CN120506803A

CN120506803A - Direct-current submerged arc furnace and detection system

Info

Publication number: CN120506803A
Application number: CN202510678347.XA
Authority: CN
Inventors: 黄江成; 李志成; 刘飞; 任成文
Original assignee: Fengzhen Huaxing Chemical Co ltd
Current assignee: Fengzhen Huaxing Chemical Co ltd
Priority date: 2025-05-23
Filing date: 2025-05-23
Publication date: 2025-08-19
Anticipated expiration: 2045-05-23
Also published as: CN120506803B

Abstract

The invention relates to the technical field of chemical detection data processing, and provides a direct-current submerged arc furnace and a detection system, aiming at solving the problems that the existing direct-current submerged arc furnace cannot detect chemical substances in the furnace in real time, so that the state of equipment cannot be predicted and early-warned in time. The invention collects the concentration data and other basic data of chemical substances in the furnace in real time through the data acquisition unit, the detection unit carries out pretreatment and feature extraction on the data, the data processing unit constructs a knowledge base, the fault data is called to generate chemical detection substances, the chemical substances in the furnace are detected, and real-time electric equipment adjustment maintenance data is generated according to the detection result, so that the running state of the equipment is optimized. The invention realizes the real-time detection of chemical substances in the direct-current submerged arc furnace, improves the pre-judging and early-warning capabilities of the equipment state, and enhances the pertinence of fault treatment.

Description

Direct-current submerged arc furnace and detection system

Technical Field

The invention relates to the technical field of chemical detection data processing, in particular to a direct-current submerged arc furnace and a detection system.

Background

The direct current ore arc furnace is a smelting device which is heated by direct current, and the working principle of the direct current ore arc furnace is that electric energy is converted into heat energy, and ore or metal raw materials are melted through the high temperature effect of an electric arc. Compared with an alternating current arc furnace, the direct current ore arc furnace adopts direct current in a power supply mode, and the direct current power supply mode ensures that the electric arc is more stable, thereby being beneficial to controlling the smelting process and improving the smelting efficiency. The direct current arc furnace is widely applied to the fields of ore smelting, iron and steel smelting and the like, and is one of essential equipment in the modern metallurgical industry.

The existing direct-current submerged arc furnace is powered by an alternating-current power supply generally, so that the problem of low electric energy efficiency exists, meanwhile, because electric arc fluctuation generated by alternating current is large, a cooling device, a detection device and a monitoring device are additionally arranged for dissipating heat generated by electric arcs, the direct-current submerged arc furnace is monitored and regulated through a plurality of device equipment, and stable operation of the direct-current submerged arc furnace is realized. However, in the actual use process of the existing direct-current submerged arc furnace, chemical substances in the direct-current submerged arc furnace cannot be detected, so that the conditions of all equipment of the direct-current submerged arc furnace cannot be timely predicted and early-warned.

Disclosure of Invention

Aiming at the defects of the prior art, the invention provides a direct-current submerged arc furnace and a detection system, which solve the problems that the prior direct-current submerged arc furnace cannot detect chemical substances in the direct-current submerged arc furnace in the actual use process, so that the conditions of all equipment of the direct-current submerged arc furnace cannot be timely predicted and early-warned.

In order to solve the technical problems, the specific technical scheme of the invention is as follows:

In a first aspect, the present invention provides a direct current submerged arc furnace, comprising:

the data acquisition unit is used for acquiring basic data of the direct-current submerged arc furnace;

The detection unit is used for preprocessing the basic data of the direct-current submerged arc furnace to obtain a basic data characteristic set of the direct-current submerged arc furnace;

The state detection unit is used for establishing an operation state detection model of the direct-current submerged arc furnace based on the basic data characteristic set of the direct-current submerged arc furnace, extracting characteristics of the direct-current submerged arc furnace acquired in real time and inputting the characteristics into the operation state detection model of the direct-current submerged arc furnace to generate a direct-current submerged arc furnace state detection result, wherein the direct-current submerged arc furnace state detection result comprises a direct-current submerged arc furnace state normal state, a direct-current submerged arc furnace state early-warning state and a direct-current submerged arc furnace state fault state;

The early warning unit is used for triggering the early warning process of the direct-current submerged arc furnace when the direct-current submerged arc furnace state detection result is the direct-current submerged arc furnace state early warning state, and triggering the fault detection process of the direct-current submerged arc furnace when the direct-current submerged arc furnace state detection result is the direct-current submerged arc furnace state fault state;

The data processing unit is used for acquiring the running history data of the direct-current submerged arc furnace, constructing a direct-current submerged arc furnace knowledge base, calling fault data corresponding to the state fault state of the direct-current submerged arc furnace, obtaining real-time fault data, carrying out data feature extraction on the real-time fault data, matching the obtained real-time fault feature data with the direct-current submerged arc furnace knowledge base, obtaining real-time fault feature associated data, wherein the real-time fault feature associated data comprises historical chemical substance detection data and historical electrical equipment adjustment maintenance data, generating chemical detection substances based on the historical chemical substance detection data, detecting the direct-current submerged arc furnace by using the chemical detection substances, obtaining a chemical detection result of the direct-current submerged arc furnace, matching the chemical detection result of the direct-current submerged arc furnace with the historical electrical equipment adjustment maintenance data, obtaining real-time electrical equipment adjustment maintenance data, and sending the real-time electrical equipment adjustment maintenance data to the equipment end of the direct-current submerged arc furnace.

Further, the direct-current submerged arc furnace, the data acquisition unit, provided by the invention, is further used for:

The data acquisition unit initiates a data acquisition request to a sensor in the direct-current submerged arc furnace, the request is used for acquiring real-time data of chemical substances in the furnace, the request message body carries the address of the data acquisition unit, the address of the sensor, the time range of the request data and the request data type, the request data type comprises temperature, pressure and chemical component concentration, and the sensor responds to the request and returns the original detection data of the direct-current submerged arc furnace.

Further, the direct-current submerged arc furnace, the detection unit, provided by the invention, is further used for:

The detection unit receives the original detection data from the sensor, removes repeated, invalid or format error data, calibrates the cleaned data, and marks the temperature data, the pressure data and the chemical component concentration data according to the characteristics of the detection data.

Further, the direct-current submerged arc furnace, the data processing unit, provided by the invention, is further used for:

Collecting chemical substance data in the furnace, and identifying the types, the concentrations and the change trend of the chemical substances in the furnace through a preset algorithm model;

And receiving and setting a threshold value, constructing an abnormality detection model, detecting the types, the concentrations and the change trend of the chemical substances in the furnace by using the abnormality detection model to obtain an abnormality detection result of the chemical substances in the furnace, and alarming data exceeding a normal range to prompt the existing equipment faults or potential safety hazards.

Based on the historical data and the current data, predicting future trends of the states of chemical substances in the furnace and the running conditions of the equipment.

the method comprises the steps of calling historical chemical substance detection data from a direct-current submerged arc furnace knowledge base, and preprocessing the obtained historical chemical substance detection data to obtain preprocessed historical chemical substance detection data;

carrying out data analysis on the preprocessed historical chemical substance detection data by adopting a preset algorithm model, and identifying chemical substances in the direct-current submerged arc furnace and characteristics of the direct-current submerged arc furnace;

And matching in a direct-current submerged arc furnace knowledge base according to chemical substances in the direct-current submerged arc furnace and characteristics of the direct-current submerged arc furnace to obtain corresponding chemical detection substances.

The method comprises the steps of applying chemical detection substances to a to-be-detected part of a direct-current submerged arc furnace, wherein the to-be-detected part of the direct-current submerged arc furnace comprises a furnace interior of the direct-current submerged arc furnace and a furnace wall of the direct-current submerged arc furnace, and collecting reaction data of the chemical detection substances and chemical substances in the direct-current submerged arc furnace and reaction data of the chemical detection substances and the chemical substances in the direct-current submerged arc furnace;

Carrying out data characteristic extraction on the reaction data of the chemical detection substances and the chemical substances in the direct-current submerged arc furnace to obtain the types of the chemical substances in the direct-current submerged arc furnace, the concentration of the chemical substances in the direct-current submerged arc furnace and the color data of the chemical substances in the direct-current submerged arc furnace, and carrying out comparative analysis on the types of the chemical substances in the direct-current submerged arc furnace, the concentration of the chemical substances in the direct-current submerged arc furnace and the color data of the chemical substances in the direct-current submerged arc furnace and a preset chemical detection standard to obtain a chemical detection result;

and acquiring operation data after the equipment end of the direct-current submerged arc furnace executes real-time electric equipment adjustment maintenance data, obtaining adjusted operation data, comparing the operation data of the adjusted operation data with preset direct-current submerged arc furnace standard data, optimizing the real-time electric equipment adjustment maintenance data by using a distillation algorithm if the adjusted operation data is not in the preset direct-current submerged arc furnace standard data, and sending the optimized real-time electric equipment adjustment maintenance data to the equipment end of the direct-current submerged arc furnace.

The invention provides a direct-current submerged arc furnace detection system, which is applied to a direct-current submerged arc furnace and is characterized by comprising a server end, a data acquisition end and a direct-current submerged arc furnace equipment end, wherein the server end is respectively in communication connection with the data acquisition end and the direct-current submerged arc furnace equipment end;

the server side comprises:

Further, the direct current submerged arc furnace detection system of the invention is characterized in that the data acquisition unit is further used for:

Further, the detection system of the direct-current submerged arc furnace, provided by the invention, is further provided with a detection unit for:

Further, the direct-current submerged arc furnace detection system provided by the invention is characterized in that the data processing unit is further used for:

The invention has the beneficial effects that:

The invention realizes the real-time detection of chemical substances in the direct-current submerged arc furnace, wherein the existing direct-current submerged arc furnace cannot directly detect the chemical substances in the furnace, so that the chemical environment in the furnace cannot be accurately mastered. The invention can collect and process the concentration data of chemical substances in the furnace in real time, including temperature, pressure, chemical component concentration and the like through the data acquisition unit and the detection unit. The chemical substances in the furnace are detected in real time, so that operators can know the conditions in the furnace in time, and a data basis is provided for follow-up fault pre-judgment and early warning.

The invention can improve the pre-judging and pre-warning capability of the equipment state, namely, the existing direct-current submerged arc furnace can not pre-judge and pre-warn the equipment state because the chemical substances in the furnace can not be detected. According to the invention, the state detection unit is used for establishing and operating the operation state detection model based on the chemical substance data in the furnace and other basic data acquired in real time, so that the operation state of the direct-current submerged arc furnace can be accurately identified. When the equipment state is abnormal or is about to be abnormal, the early warning unit can trigger the early warning process in time to remind operators to take measures, so that the occurrence or expansion of faults is effectively avoided.

The invention enhances the pertinence of equipment fault treatment, namely the existing direct-current submerged arc furnace often lacks pertinence in fault treatment, so that the treatment efficiency is low. According to the invention, a direct-current submerged arc furnace knowledge base is constructed through the data processing unit, and historical fault data and adjustment maintenance data are stored. When the equipment fails, the data processing unit can call related knowledge, generate real-time electric equipment adjustment maintenance data and optimize the running state of the equipment. The fault processing mode based on the historical data and the real-time data ensures that the processing measures are more targeted and effective, and improves the reliability of the equipment.

The invention improves the production efficiency and reduces the maintenance cost, and the existing direct current submerged arc furnace usually needs to be shut down for maintenance when the fault occurs because the equipment state cannot be prejudged and early warned, so that the production efficiency is low. By means of a real-time monitoring and early warning mechanism, measures can be taken in time before the abnormal state of the equipment occurs, and unnecessary downtime is avoided. Meanwhile, through targeted fault treatment measures, the occurrence frequency and maintenance cost of equipment faults are reduced, and the overall production efficiency is improved.

The invention introduces advanced data detection and analysis technology in the field of direct current submerged arc furnaces, and promotes the technical innovation in the field. Through real-time monitoring and data analysis, powerful support is provided for the optimal design and intelligent control of the direct-current submerged arc furnace, and the upgrading and development of the whole industry are promoted.

In summary, the invention has the advantages of realizing the real-time detection of chemical substances in the direct-current submerged arc furnace, improving the pre-judging and early-warning capability of the equipment state, enhancing the pertinence and effectiveness of the equipment fault treatment, improving the production efficiency, reducing the maintenance cost, promoting the technical innovation and the industrial upgrading, and the like. The beneficial effects not only solve the problems existing in the existing direct-current submerged arc furnace, but also provide more reliable and efficient technical support for the application of the direct-current submerged arc furnace in the metallurgical industry.

Drawings

In order to more clearly illustrate the technical solution of the present invention, the drawings that are needed in the embodiments will be briefly described below, and it will be obvious to those skilled in the art that other drawings can be obtained from the drawings without inventive effort.

Fig. 1 is a schematic diagram of a functional module of a direct current submerged arc furnace according to an embodiment of the present invention.

Detailed Description

In order to make the objects, technical solutions and advantages of the present invention more apparent, the technical solutions of the present invention will be clearly and completely described below with reference to specific embodiments of the present invention and corresponding drawings. It will be apparent that the described embodiments are only some, but not all, embodiments of the invention. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention. The following describes in detail the technical solutions provided by the embodiments of the present invention with reference to the accompanying drawings.

For a better understanding of the present invention, the present invention is described in further detail below.

The invention realizes the real-time detection of chemical substances in the direct-current submerged arc furnace by the following steps:

And the data acquisition unit initiates a data acquisition request to a sensor in the direct-current submerged arc furnace to request acquisition of real-time data of chemical substances in the furnace. The request message body carries the data acquisition unit address, the sensor address, the time range of the requested data, and the type of the requested data (including temperature, pressure, chemical concentration, etc.). And the sensor responds to the request and returns the original detection data of the direct-current submerged arc furnace.

The detection unit is used for receiving the original detection data from the sensor, removing repeated, invalid or format error data, calibrating the cleaned data and ensuring the accuracy and consistency of the data.

And labeling the temperature data, the pressure data and the chemical component concentration data according to the characteristics of the detection data, so that the subsequent data processing and analysis are convenient.

And the data characteristic extraction, namely the detection unit performs data characteristic extraction on the preprocessed data to obtain a direct-current submerged arc furnace basic data characteristic set, wherein the data characteristic set comprises chemical substance concentration characteristics and the like.

The data processing unit can collect the data of the chemical substances in the furnace and identify the types, the concentrations and the change trend of the chemical substances in the furnace through a preset algorithm model. An abnormal detection model is built, the types, the concentrations and the change trends of chemical substances in the furnace are detected, and an alarm is given when the data exceeds a normal range, so that the existing equipment faults or potential safety hazards are prompted. Through the steps, the method can realize the real-time detection of chemical substances in the direct-current submerged arc furnace, and provides accurate data support for the subsequent fault pre-judgment, early warning and processing.

The early warning unit does not directly establish an operation state detection model based on the chemical substance data in the furnace acquired in real time, but the state detection unit is responsible for the task. The following is a logical reasoning process:

The state detection unit is responsible for establishing and operating an operation state detection model of the direct-current submerged arc furnace based on a basic data characteristic group (comprising electric parameter characteristics, temperature characteristics, arc stability characteristics, chemical substance concentration characteristics and the like) of the direct-current submerged arc furnace.

The data acquisition unit acquires concentration data of chemical substances in the furnace and other basic data (such as electrical parameter data, temperature data and the like) in real time. The detection unit performs pretreatment and feature extraction on the data to obtain real-time direct-current submerged arc furnace feature data.

The method comprises the steps of establishing an operation state detection model, wherein the state detection unit uses a direct-current submerged arc furnace basic data feature set in a historical data set and a corresponding operation state label, and trains through a machine learning or deep learning algorithm (such as a support vector machine, a random forest, a neural network and the like) to obtain the operation state detection model. The model can identify the normal state, the early warning state and the fault state of the direct current submerged arc furnace.

And (3) detecting the real-time state, namely inputting the real-time collected direct-current submerged arc furnace characteristic data into an operation state detection model. And the model generates a state detection result of the direct-current submerged arc furnace according to the input characteristic data.

Triggering the early warning unit, namely triggering the early warning unit when the state detection result output by the state detection unit is an early warning state or a fault state. The early warning unit triggers a corresponding early warning process or fault detection process according to the early warning or fault state, such as sending an early warning signal, recording early warning information, starting a fault detection program and the like.

Therefore, the early warning unit does not directly establish an operation state detection model, but relies on the model established by the state detection unit through historical data and real-time data to identify the operation state of the direct-current submerged arc furnace, and triggers corresponding early warning or fault processing flows when needed.

The system comprises a data acquisition unit, a control unit and a control unit, wherein the data acquisition unit is used for acquiring direct-current submerged arc furnace basic data, and the direct-current submerged arc furnace basic data comprise electrical parameter data, temperature data, arc stability data, chemical substance concentration data, cooling system data and equipment state data;

Data acquisition unit of direct-current submerged arc furnace

Basic data range of data acquisition the data acquisition unit is used for acquiring basic data of the direct-current submerged arc furnace, wherein the data comprise, but are not limited to, electrical parameter data (such as current, voltage, power factor and the like), temperature data (such as temperature in the furnace, cooling water temperature and the like), arc stability data (such as arc length, arc voltage fluctuation and the like), chemical substance concentration data (such as gas component concentration in the furnace, melt component concentration and the like), cooling system data (such as cooling water flow rate, pressure and the like) and equipment state data (such as electrode position, furnace door opening and closing state and the like).

In order to accurately acquire the above data, the data acquisition unit needs to interact with various types of sensors. For example, electrical parameter data may be obtained through a current transformer, a voltage sensor, temperature data may be obtained through a thermocouple or an infrared thermometer, arc stability data may be obtained through an arc sensor or an image recognition technique, chemical concentration data may be obtained through a gas analyzer or a spectrometer, cooling system data may be obtained through a flow meter, a pressure sensor, equipment state data may be obtained through a limit switch, a proximity switch, etc. The layout of the sensor should fully consider the environment in the furnace and the measurement requirement.

The data acquisition unit needs to establish stable communication connection with the sensor, and can adopt standard industrial communication protocols (such as Modbus, profibus and the like) or custom protocols for data transmission. Meanwhile, in order to cope with the data format and transmission rate difference of different sensors, the data acquisition unit is provided with a data format conversion and buffering mechanism.

In the data acquisition process, the data is incomplete or wrong due to the factors of sensor faults, communication interference and the like. Therefore, the data acquisition unit detects the integrity of the data through an internal preset data check and error processing mechanism, and the preset data check and error processing mechanism improves the accuracy of the data through a redundant sensor, data comparison verification and other modes.

The data acquisition unit is used as a data source of the detection system and needs to be closely cooperated with other units (such as a detection unit, a state detection unit, an early warning unit and a data processing unit). For example, the detection unit needs to acquire the preprocessed data from the data acquisition unit for feature extraction, the state detection unit needs to establish and update an operation state detection model based on the data provided by the data acquisition unit, and the early warning unit and the data processing unit need to perform fault early warning and adjustment maintenance decision according to the data provided by the data acquisition unit.

The whole detection system forms a closed loop feedback mechanism from data acquisition to state detection, early warning and processing. The data acquisition unit is used as a basis to realize the data input of the subsequent steps, the detection unit and the state detection unit process and analyze the data, a basis is provided for early warning and processing, and the early warning unit and the data processing unit take corresponding measures according to the analysis result to realize the safe and stable operation of the direct-current submerged arc furnace.

The detection unit is used for preprocessing the direct-current submerged arc furnace basic data to obtain preprocessed direct-current submerged arc furnace basic data, and extracting data characteristics of the preprocessed direct-current submerged arc furnace basic data to obtain a direct-current submerged arc furnace basic data characteristic set, wherein the direct-current submerged arc furnace basic data characteristic set comprises electrical parameter characteristics, temperature characteristics, electric arc stability characteristics, chemical substance concentration characteristics, cooling system characteristics and equipment state characteristics;

The detection unit firstly receives the basic data from the data acquisition unit. The data cleaning process includes removing duplicate data, processing missing values (e.g., filling in by interpolation), identifying and correcting outliers (e.g., thresholding based on statistical methods or domain knowledge).

And (3) data calibration, namely aiming at the precision difference of different sensors, performing data calibration, and ensuring that all data are compared under the same dimension and standard. For example, temperature data is subjected to temperature compensation, and electrical parameter data is subjected to zero drift correction.

Data normalization/standardization, namely, in order to eliminate dimensional differences among different features, improving the efficiency of subsequent feature extraction and model training, and carrying out normalization or standardization treatment on the data.

The data normalization/standardization is not only helpful to improve the training efficiency of the model, but also can enhance the generalization capability of the model.

Based on the domain knowledge and the statistical analysis, characteristics with obvious influence on the running state of the submerged arc furnace are selected from the preprocessed data. For example, current, voltage, power factor are selected as electrical parameter characteristics, furnace temperature, cooling water temperature as temperature characteristics, arc length, arc voltage fluctuation as arc stability characteristics, etc.

Feature dimension reduction, namely, for high-dimensional data, adopting methods such as Principal Component Analysis (PCA), linear Discriminant Analysis (LDA) and the like to perform feature dimension reduction, reducing calculation complexity and simultaneously keeping key information.

Feature engineering, namely constructing new features according to specific application scenes. For example, the arc power fluctuation rate is calculated as a new feature of arc stability, or the heat flow distribution in the furnace is calculated from the temperature gradient, etc.

The detection unit serves as a bridge connecting the data acquisition unit and a subsequent analysis unit (such as a state detection unit) and plays a role in going up and down. By preprocessing the original data and extracting the characteristics, data support is provided for subsequent state monitoring, fault early warning and the like.

After the data acquisition unit provides the original data, the detection unit firstly carries out preprocessing, and then the detection unit carries out feature extraction to convert the high-dimensional and complex original data into low-dimensional and easy-to-analyze feature vectors. These feature vectors are then input into a state detection unit for establishing and updating an operating state detection model to realize real-time state monitoring and fault early warning of the submerged arc furnace. The whole process forms a closed loop feedback mechanism from data acquisition to state monitoring.

The state detection unit is used for establishing an operation state detection model of the direct-current submerged arc furnace based on the basic data characteristic set of the direct-current submerged arc furnace, wherein the operation state detection model of the direct-current submerged arc furnace is used for identifying the operation state of the direct-current submerged arc furnace, extracting characteristics of the direct-current submerged arc furnace acquired in real time to obtain real-time direct-current submerged arc furnace characteristic data, inputting the real-time direct-current submerged arc furnace characteristic data into the operation state detection model of the direct-current submerged arc furnace, and generating a direct-current submerged arc furnace state detection result by the operation state detection model of the direct-current submerged arc furnace, wherein the direct-current submerged arc furnace state detection result comprises a direct-current submerged arc furnace state normal state, a direct-current submerged arc furnace state early-warning state and a direct-current submerged arc furnace state fault state;

And selecting a proper machine learning or deep learning algorithm as a basis of an operation state detection model according to the operation characteristics and monitoring requirements of the direct-current submerged arc furnace. For example, support Vector Machines (SVMs), random forests, neural networks, etc. algorithms may be selected.

And (3) before the model is built, further feature selection, feature transformation or feature construction is carried out on the basic data feature set of the direct current submerged arc furnace so as to extract the most useful information for identifying the running state.

And training the model by using the direct-current submerged arc furnace basic data feature set and the corresponding running state label in the historical data set to train the selected algorithm so as to obtain a running state detection model. In the training process, the method can be used for optimizing model parameters by adopting methods such as cross verification, grid search and the like, so that the model performance is improved.

Preprocessing and extracting features of the direct-current submerged arc furnace data acquired in real time to obtain real-time direct-current submerged arc furnace feature data.

And (3) inputting the real-time direct-current submerged arc furnace characteristic data into an operation state detection model, and generating a direct-current submerged arc furnace state detection result according to the input characteristic data by the model. The detection result can comprise the states of normal state, early warning or fault and the like of the direct current submerged arc furnace.

And (3) processing and feeding back results, namely processing and analyzing a state detection result generated by the model, such as triggering a corresponding alarm mechanism or maintenance measures according to early warning or fault states. Meanwhile, the detection result is fed back to a system user or an operator so that the user or the operator can know the running state of the direct-current submerged arc furnace in time.

And in a normal state, when all operating parameters of the direct current submerged arc furnace are in a normal range and no abnormal fluctuation or trend exists, judging the direct current submerged arc furnace to be in the normal state.

And the early warning state is judged to be the early warning state when some operation parameters of the direct-current submerged arc furnace have slight abnormality or trend change but do not reach the fault threshold. The early warning state aims to discover potential problems in advance so as to take preventive measures.

And in the fault state, when one or more operating parameters of the direct-current submerged arc furnace exceed a normal range or obvious abnormal fluctuation or trend occurs, judging the direct-current submerged arc furnace as the fault state. The fault condition requires immediate maintenance or shutdown measures to avoid more serious consequences.

The normal state is an ideal state of the operation of the direct current submerged arc furnace and is also a target pursued by a state detection system.

The early warning state is one of important functions of the state detection system, and the occurrence and influence of faults can be reduced by finding potential problems in advance.

The fault state is the least unexpected condition in the running of the direct-current submerged arc furnace, but the state detection system can timely find and alarm, thereby being beneficial to reducing the loss caused by the fault.

The state detection unit is used as a core part of the running state detection system of the direct-current submerged arc furnace and is responsible for establishing and running a running state detection model so as to realize real-time and accurate identification of the running state of the direct-current submerged arc furnace.

Firstly, basic data of the direct-current submerged arc furnace are acquired and preprocessed through a data acquisition unit and a detection unit, and characteristic data are extracted.

Then, the feature data is input into an operation state detection model in the state detection unit, and the model generates a state detection result from the input data.

And finally, processing and analyzing the state detection result, and adopting corresponding measures or feeding back to a user to ensure the safe and stable operation of the direct-current submerged arc furnace.

The whole flow forms a closed loop system from data acquisition to state detection, and the running state of the direct-current submerged arc furnace is accurately identified in real time, so that powerful guarantee is provided for production efficiency and safety.

The early warning unit is used for triggering the early warning process of the direct-current submerged arc furnace when the direct-current submerged arc furnace state detection result is the direct-current submerged arc furnace state early warning state or the direct-current submerged arc furnace state fault state, and triggering the direct-current submerged arc furnace fault detection process when the direct-current submerged arc furnace state detection result is the direct-current submerged arc furnace state fault state;

According to the operation characteristics and the safety standard of the direct-current submerged arc furnace, the specific conditions for triggering the early warning state are preset. These conditions may include abnormal fluctuations in certain critical operating parameters, outside of a preset range or trending changes, and the like.

And the early warning process is triggered by the early warning unit immediately when the state detection result of the direct current submerged arc furnace output by the state detection unit is an early warning state. The early warning process can comprise sending early warning signals to an operator or a control system, recording relevant information such as early warning time, early warning reasons and the like, and taking corresponding precautions according to early warning levels.

The early warning signal is transmitted to an operator or a control system in various modes, such as sound alarm, light indication, short message notification, system popup window and the like, so that the operator can timely receive and respond to the early warning information.

The setting of the early warning condition is the basis of the operation of the early warning unit, and the reasonable early warning condition can ensure that the early warning process is triggered in time when the potential problem occurs to the equipment. The triggering of the early warning process is the core function of the early warning unit, and through timely and accurately transmitting early warning information, operators can be prompted to take preventive measures, and equipment faults are prevented from being generated or expanded. The transmission mode of the early warning signal is selected according to the actual use scene and the requirement so as to ensure that the early warning information can be rapidly and accurately transmitted to related personnel.

And setting fault conditions, namely presetting specific conditions for triggering the fault state according to the operation characteristics and the safety standard of the direct-current submerged arc furnace. These conditions are typically more stringent than pre-warning conditions, such as critical operating parameters severely out of normal ranges, obvious signs of failure of the equipment, etc.

And the fault detection flow is triggered by the early warning unit immediately when the state detection result of the direct current submerged arc furnace output by the state detection unit is a fault state. The fault detection flow may include stopping the operation of the device, sending a fault alarm signal, starting a fault troubleshooting program, recording relevant information such as fault time, fault type, etc., and taking corresponding maintenance or repair measures according to the fault level.

The fault alarm and processing means that the fault alarm signal should be transmitted to the operator or the control system in a significant manner, such as emergency stop indication, audible and visual alarm, system emergency popup window, etc., so as to ensure that the operator can immediately recognize the severity of the fault and take corresponding processing measures.

The setting of fault conditions is a key to ensure that the equipment can be shut down in time and trigger a fault handling process when serious problems occur. The triggering of the fault detection flow is the core function of the early warning unit when the equipment fails, and the fault information can be rapidly and accurately transmitted, so that operators can be prompted to timely take maintenance or repair measures, and the loss caused by the fault is reduced.

The early warning unit is used as part of a direct-current submerged arc furnace running state detection and early warning system and is responsible for identifying early warning and fault states of the direct-current submerged arc furnace and triggering corresponding processing flows.

Firstly, a state detection unit detects the running state of the direct-current submerged arc furnace in real time and outputs a state detection result.

And then, the early warning unit recognizes the early warning or fault state of the direct-current submerged arc furnace according to the state detection result and triggers a corresponding early warning or fault detection flow.

And then, the early warning or fault detection flow ensures the safe and stable operation of the equipment by sending alarm signals, recording related information, taking preventive measures or maintenance measures and the like.

Finally, the efficiency and accuracy of early warning and fault processing are continuously improved through the means of flow connection, information sharing, flow optimization and the like.

The whole flow forms a closed loop system from state detection to early warning or fault processing, and provides powerful guarantee for the safe and stable operation of equipment by accurately identifying and processing the operation state of the direct-current submerged arc furnace in real time.

The system comprises a data processing unit, a direct-current ore-smelting furnace operation history data processing unit and a direct-current ore-smelting furnace operation history data processing unit, wherein the data processing unit is used for obtaining direct-current ore-smelting furnace operation history data, constructing a direct-current ore-smelting furnace knowledge base, calling fault data corresponding to a direct-current ore-smelting furnace state fault state, taking the fault data corresponding to the direct-current ore-smelting furnace state fault state as real-time fault data, carrying out data characteristic extraction on the real-time fault data to obtain real-time fault characteristic data, matching the real-time fault characteristic data in the direct-current ore-smelting furnace knowledge base to obtain real-time fault characteristic associated data, generating chemical detection substances based on the historical chemical substance detection data, detecting the direct-current ore-smelting furnace by using the chemical detection substances, obtaining a direct-current ore-smelting furnace chemical detection result, matching the direct-smelting furnace chemical detection result with the historical electric equipment adjustment maintenance data, obtaining real-time electric equipment adjustment maintenance data, sending the real-time electric equipment adjustment maintenance data to an equipment end, acquiring the real-time fault characteristic data after the real-time electric equipment adjustment maintenance data, obtaining the adjusted operation data, comparing the operation data of the adjusted operation data with preset direct-current ore-smelting furnace operation data with preset direct-smelting furnace operation data, and carrying out the preset direct-smelting furnace operation data to be optimized after the electric equipment adjustment data is carried out the real-time adjustment according to the real-time electric equipment adjustment data, and carrying out optimization after the electric equipment adjustment data is carried out on the electric equipment adjustment by the electric equipment adjustment data.

The data processing unit is connected with a control system or a sensor network of the direct-current submerged arc furnace to acquire operation history data of the direct-current submerged arc furnace in real time or periodically, including but not limited to temperature, current, voltage, chemical substance content, equipment operation state and the like.

The knowledge base construction, namely, based on the acquired operation history data, the data processing unit constructs a direct-current submerged arc furnace knowledge base which comprises normal operation data, history fault data, fault processing schemes, chemical substance detection data, electric equipment adjustment maintenance data and the like, and data support is provided for subsequent fault analysis and processing.

And (3) fault data calling, namely when the direct-current submerged arc furnace fails, the data processing unit calls fault data corresponding to the fault state from the knowledge base, wherein the fault data comprise historical fault cases, fault phenomena, fault reasons and the like.

And data feature extraction, namely performing data feature extraction, such as statistical analysis, trend analysis, spectrum analysis and the like, on the retrieved fault data to obtain real-time fault feature data for subsequent fault matching and processing.

And (3) fault feature matching, namely matching the real-time fault feature data in a direct-current submerged arc furnace knowledge base, and searching similar historical fault cases to obtain real-time fault feature associated data, wherein the real-time fault feature associated data comprises historical chemical substance detection data and historical electrical equipment adjustment maintenance data.

Chemical detection substance generation, namely generating a chemical detection substance list or formula based on historical chemical detection data, wherein the chemical detection substance list or formula is used for carrying out chemical detection on the direct-current submerged arc furnace so as to verify or determine the cause of the fault.

And obtaining chemical detection results, namely detecting the direct-current submerged arc furnace by using the generated chemical detection substances to obtain chemical detection results of the direct-current submerged arc furnace, such as chemical substance content, reaction products and the like.

The fault feature matching is a bridge connecting the current fault and the historical fault case, and a similar fault processing scheme or experience can be found through matching. The generation and detection of chemical detection substances is an important means of verifying the cause of a fault, especially when the fault is chemical-related.

And matching the chemical detection result of the direct-current submerged arc furnace with historical electrical equipment adjustment maintenance data, and searching similar adjustment maintenance cases to obtain real-time electrical equipment adjustment maintenance data.

And sending the adjustment maintenance data, namely sending the adjustment maintenance data of the real-time electrical equipment to the equipment end of the direct-current submerged arc furnace, and guiding the equipment to carry out corresponding adjustment or maintenance.

And acquiring the operation data after the adjustment of the real-time electrical equipment at the equipment end of the direct-current submerged arc furnace to obtain the adjusted operation data.

And (3) comparing the adjusted operation data with preset direct-current submerged arc furnace standard data, and if the adjusted operation data is not in a preset range, optimizing the adjustment maintenance data of the real-time electric equipment by using a distillation algorithm so as to improve the accuracy and the effectiveness of adjustment maintenance.

And (3) transmitting optimized real-time electric equipment adjustment and maintenance data to a direct-current submerged arc furnace equipment end to guide the equipment to carry out adjustment or maintenance again.

Matching of adjustment maintenance data is key to achieving pertinence of adjustment or maintenance measures, and similar adjustment maintenance experience or scheme can be found through matching. The transmission and execution of the adjustment maintenance data are actual operation links of fault processing, and accuracy of the data and timeliness of execution need to be ensured. The collection and comparison of the operation data after adjustment are important bases for verifying the adjustment maintenance effect, and the effectiveness of the adjustment maintenance measures can be evaluated through comparison.

The application of the distillation algorithm is an effective means for optimizing the adjustment maintenance data, and can improve the accuracy and efficiency of the adjustment maintenance.

The data processing unit is used as a core part of the direct-current submerged arc furnace data processing and fault processing system and is responsible for acquiring and processing the operation history data of the direct-current submerged arc furnace, constructing a knowledge base, and calling related data to perform fault analysis, detection and processing in the fault state, so that powerful support is provided for the stable operation of the direct-current submerged arc furnace.

Firstly, the data processing unit acquires the operation history data of the direct-current submerged arc furnace in real time or periodically in a data acquisition mode, and builds a knowledge base.

Then, in the fault state, the data processing unit retrieves fault data from the knowledge base and performs data feature extraction to obtain real-time fault feature data.

And then, the data processing unit matches the real-time fault characteristic data in a knowledge base to obtain real-time fault characteristic associated data, generates chemical detection substances based on historical chemical substance detection data, and detects the direct-current submerged arc furnace.

And then, the data processing unit matches the chemical detection result with the historical electric equipment adjustment maintenance data to obtain real-time electric equipment adjustment maintenance data, and sends the real-time electric equipment adjustment maintenance data to the equipment end of the direct-current submerged arc furnace for execution.

And finally, the data processing unit acquires the adjusted operation data, compares the operation data with preset direct-current submerged arc furnace standard data, optimizes the real-time electric equipment adjustment maintenance data by using a distillation algorithm if the operation data does not meet the expectations, and sends the operation data to the equipment end for re-execution.

The whole flow forms a closed loop system from data acquisition to fault processing, and provides powerful guarantee for the stable operation of equipment by accurately processing and analyzing the operation data of the direct-current submerged arc furnace in real time. Meanwhile, maintenance data are continuously optimized and adjusted, so that the efficiency and accuracy of fault processing are improved.

Specifically, the direct-current submerged arc furnace, the data acquisition unit is further used for:

And the data acquisition unit address is used for identifying the data acquisition unit initiating the request so that the sensor or other parts of the system can identify the source of the request.

Sensor address-explicitly specifying the sensor that needs to respond to the request, and causing data to be retrieved from the correct sensor.

The time range of the requested data is a time range of designating the data time period required to be acquired, which can be real-time data or historical data.

Request data type-data type explicitly required to be collected, including but not limited to temperature, pressure, chemical composition concentration, etc., to meet data requirements in different scenarios.

After the sensor receives the data acquisition request, corresponding data acquisition tasks are executed according to the information in the request message body.

After the sensor collects the data, the data is packaged into a response message, and the response message is returned to the data acquisition unit which initiates the request.

The returned raw test data includes, but is not limited to, temperature values, pressure values, chemical component concentration values, etc., and is specifically determined based on the type of data requested.

Specifically, the direct-current submerged arc furnace, the detection unit is further used for:

The detection unit is responsible for receiving raw detection data from the sensor in real time or periodically. The types of data received include, but are not limited to, temperature data, pressure data, chemical composition concentration data, and the like.

The detection unit is internally provided with a data cleaning module which is responsible for removing repeated data, invalid data and data with wrong format in the received original detection data. Repeated data is caused by repeated sending of a sensor or data transmission errors, invalid data is caused by sensor faults or environment interference, and data with wrong format is caused by coding errors or decoding errors in the data transmission process. The data cleaning module automatically identifies and removes the bad data through a preset algorithm or rule, and ensures the accuracy of subsequent processing.

The detection unit is provided with a data calibration module which is responsible for calibrating the cleaned data.

The calibration process includes linear calibration, nonlinear calibration, temperature compensation, pressure compensation, etc., and is specifically determined based on the characteristics of the sensor and the nature of the sensed data.

The data calibration module processes the cleaned data through preset calibration parameters or algorithms, and ensures the accuracy and consistency of the data.

The detection unit is internally provided with a data labeling module which is responsible for labeling the calibrated data.

Labeling includes, but is not limited to, the type of data (e.g., temperature, pressure, chemical concentration, etc.), acquisition time, sensor number, etc.

The data marking module automatically adds marking information to the calibrated data through a preset marking rule or algorithm, so that the subsequent data analysis and processing are facilitated.

Specifically, the direct-current submerged arc furnace, the data processing unit is further used for:

The data processing unit is responsible for collecting raw data of chemical substances in the furnace in real time or periodically, including but not limited to temperature, pressure, chemical component concentration and the like.

The data processing unit is internally provided with a preset algorithm model, and the model can identify the types, the concentrations and the change trends of chemical substances in the furnace based on methods such as machine learning, deep learning or statistical analysis.

The algorithm model learns the characteristic mode of the chemical substances through training historical data, so that the newly acquired data can be accurately identified.

By identifying the types, the concentrations and the change trend of the chemical substances, the chemical environment in the furnace can be known in time, and a basis is provided for subsequent abnormality detection and prediction.

The data processing unit is provided with a threshold setting module which is responsible for receiving and setting the normal range threshold of the chemical species, concentration and change trend in the furnace.

Based on the set threshold value, the data processing unit builds an abnormality detection model capable of monitoring the state of the chemical substances in the furnace in real time and comparing with the threshold value.

When detecting that the state of chemical substances in the furnace exceeds the normal range, the abnormal detection model triggers an alarm mechanism to prompt the existing equipment failure or potential safety hazard.

The data processing unit is internally provided with a prediction model, and the model predicts future trends of chemical substance states and equipment running conditions in the furnace based on historical data and current data by using methods such as time sequence analysis, machine learning or deep learning. The predictive model can take into account various factors such as changes in furnace temperature, pressure, chemical concentration, and equipment run time, load, etc., to improve the accuracy of the predictions. The prediction result can be displayed in the forms of a chart, a report and the like, so that operators can intuitively know the state of chemical substances in the furnace and the future trend of the running condition of equipment.

The data processing unit firstly collects the original data of chemical substances in the furnace, and then identifies the chemical substances through a preset algorithm model to obtain the types, the concentrations and the change trend of the chemical substances. Based on the set threshold value, the data processing unit builds an abnormality detection model, monitors the states of chemical substances in the furnace in real time, and timely discovers and alarms the abnormal states. The data processing unit predicts future trends of chemical substance states and equipment running conditions in the furnace based on historical data and current data through a prediction model, and provides decision support for operators.

In summary, the data processing unit and the data processing method of the direct-current submerged arc furnace realize comprehensive monitoring of the chemical substance state and the equipment operation condition in the direct-current submerged arc furnace through the steps of collecting, identifying, detecting and predicting the chemical substance data.

And carrying out data characteristic extraction on the reaction data of the chemical detection substances and the chemical substances in the direct-current submerged arc furnace to obtain the types of the chemical substances in the direct-current submerged arc furnace, the concentration of the chemical substances in the direct-current submerged arc furnace and the color data of the chemical substances in the direct-current submerged arc furnace, and carrying out comparison analysis on the types of the chemical substances in the direct-current submerged arc furnace, the concentration of the chemical substances in the direct-current submerged arc furnace and the color data of the chemical substances in the direct-current submerged arc furnace and a preset chemical detection standard to obtain a chemical detection result.

The data processing unit can retrieve historical chemical substance detection data from a direct current submerged arc furnace knowledge base. The data in the knowledge base includes past detection records, chemical species, concentrations, colors, reaction characteristics, and the like. And performing pretreatment operations such as cleaning, denoising, normalization and the like on the collected historical chemical substance detection data to eliminate abnormal values, missing values or noise in the data, and obtaining the pretreated historical chemical substance detection data.

The data processing unit adopts a preset algorithm model, such as a machine learning algorithm, a statistical analysis method or a pattern recognition technology, to perform data analysis on the preprocessed historical chemical substance detection data.

Through data analysis, the chemical species and concentration distribution in the direct-current submerged arc furnace and the operation characteristics of the direct-current submerged arc furnace, such as temperature change, pressure fluctuation and the like, are identified.

And according to the identified chemical substances in the direct-current submerged arc furnace and the characteristics of the direct-current submerged arc furnace, the data processing unit matches in a direct-current submerged arc furnace knowledge base and selects or prepares corresponding chemical detection substances. The selection or preparation of the chemical detection substance needs to consider the reaction characteristics, detection sensitivity, safety and other factors of the chemical detection substance and the target chemical substance.

Matching and preparation of chemical detection substances are key steps of chemical detection, and directly affect the accuracy and reliability of detection. The matching process in the knowledge base needs to be based on accurate chemical and characteristic information to ensure that the selected or prepared chemical detection substances are applicable.

The prepared chemical detection substances are applied to the parts to be detected of the direct current submerged arc furnace, including the furnace interior, the furnace wall and the like.

The data processing unit acquires the reaction data of the chemical detection substances and the chemical substances in the direct-current submerged arc furnace in real time or periodically, wherein the reaction data comprise the types, the concentrations, the colors, the reaction time and the like of reaction products.

The data processing unit is internally provided with a data characteristic extraction module, and performs characteristic extraction on the collected reaction data of the chemical detection substances and the chemical substances in the direct-current submerged arc furnace. And obtaining data such as the types, the concentrations and the colors of chemical substances in the direct-current submerged arc furnace through feature extraction. And comparing and analyzing the extracted data with a preset chemical detection standard to obtain a chemical detection result, wherein the chemical detection result comprises whether the chemical substance is abnormal, whether the concentration exceeds the standard, whether the color is changed and the like.

Data feature extraction is one of the key steps in chemical detection, which requires extraction of useful information from complex reaction data. The chemical detection standard needs to be set based on factors such as industry specifications, equipment requirements, safety standards and the like so as to ensure the accuracy and reliability of a detection result.

The data processing unit firstly calls historical chemical substance detection data and performs preprocessing to provide accurate basic data for subsequent data analysis. And carrying out data analysis on the preprocessed data through a preset algorithm model, and accurately identifying chemical substances and characteristics in the direct-current submerged arc furnace. And matching in a knowledge base according to the identified chemical substances and the features, and selecting or preparing the applicable chemical detection substances. The chemical detection substances are applied to the part to be detected, and reaction data are collected in real time or periodically, so that timeliness and accuracy of the data are ensured. And extracting features of the reaction data, and comparing and analyzing the features with a preset standard to obtain an accurate chemical detection result.

In summary, the data processing unit and the chemical detection method of the direct-current submerged arc furnace realize comprehensive detection of chemical substances in the direct-current submerged arc furnace through the steps of historical data acquisition, data analysis, chemical detection substance preparation, application and reaction data acquisition, data characteristic extraction and the like.

The invention provides a direct-current submerged arc furnace detection system, which is applied to a direct-current submerged arc furnace and comprises a server end, a data acquisition end and a direct-current submerged arc furnace equipment end, wherein the server end is respectively in communication connection with the data acquisition end and the direct-current submerged arc furnace equipment end;

the server side comprises:

The system comprises a data acquisition unit, a control unit and a control unit, wherein the data acquisition unit acquires direct-current submerged arc furnace basic data, and the direct-current submerged arc furnace basic data comprises electrical parameter data, temperature data, arc stability data, chemical substance concentration data, cooling system data and equipment state data;

The detection unit is used for preprocessing direct-current submerged arc furnace basic data to obtain preprocessed direct-current submerged arc furnace basic data, and extracting data characteristics of the preprocessed direct-current submerged arc furnace basic data to obtain a direct-current submerged arc furnace basic data characteristic set, wherein the direct-current submerged arc furnace basic data characteristic set comprises electrical parameter characteristics, temperature characteristics, electric arc stability characteristics, chemical substance concentration characteristics, cooling system characteristics and equipment state characteristics;

A data processing unit for obtaining operation history data of the direct-current submerged arc furnace, constructing a knowledge base of the direct-current submerged arc furnace, calling fault data corresponding to the fault state of the direct-current submerged arc furnace, taking the fault data corresponding to the fault state of the direct-current submerged arc furnace as real-time fault data, carrying out data feature extraction on the real-time fault data to obtain real-time fault feature data, matching the real-time fault feature data in the knowledge base of the direct-current submerged arc furnace to obtain real-time fault feature related data, wherein the real-time fault feature related data comprises historical chemical substance detection data and historical electrical equipment adjustment maintenance data, generating chemical detection substances based on the historical chemical substance detection data, detecting the direct-current submerged arc furnace by using the chemical detection substances to obtain chemical detection results of the direct-current submerged arc furnace, matching a chemical detection result of the direct-current submerged arc furnace with historical electrical equipment adjustment maintenance data to obtain real-time electrical equipment adjustment maintenance data, sending the real-time electrical equipment adjustment maintenance data to equipment ends of the direct-current submerged arc furnace, collecting operation data after the equipment ends of the direct-current submerged arc furnace execute the real-time electrical equipment adjustment maintenance data to obtain adjusted operation data, comparing the operation data of the adjusted operation data with preset direct-current submerged arc furnace standard data, optimizing the real-time electrical equipment adjustment maintenance data by using a distillation algorithm if the adjusted operation data is not in the preset direct-current submerged arc furnace standard data, and sending the optimized real-time electrical equipment adjustment maintenance data to the equipment ends of the direct-current submerged arc furnace

Specifically, the direct-current submerged arc furnace detection system provided by the invention is characterized in that the data acquisition unit is also used for:

Specifically, the detection system of the direct-current submerged arc furnace, provided by the invention, is characterized in that the detection unit is also used for:

Specifically, the direct-current submerged arc furnace detection system provided by the invention is characterized in that the data processing unit is also used for:

The technical scheme of the invention effectively solves the problems that the existing direct current submerged arc furnace cannot detect chemical substances in the furnace in the actual use process, and further cannot timely pre-judge and pre-warn the states of all equipment.

The existing direct-current submerged arc furnace lacks means for directly detecting chemical substances in the furnace in real time, so that the chemical environment in the furnace cannot be accurately mastered. The chemical substances in the furnace cannot be detected, so that the working state of the equipment cannot be predicted according to the change of the chemical substances, and the early warning can not be timely performed when the state of the equipment is abnormal.

The technical scheme of the invention comprises the following steps:

The data acquisition unit acquires various basic data including chemical substance concentration data by initiating a data acquisition request to a sensor in the direct-current submerged arc furnace. The detection unit performs pretreatment and feature extraction on the acquired basic data, including chemical substance concentration features and the like. Noise and invalid data are removed through data cleaning, calibration and feature extraction, key features are extracted, and high-quality data are provided for subsequent state detection. The state detection unit establishes and operates an operation state detection model based on the extracted characteristic data, and identifies the operation state of the direct-current submerged arc furnace. The model can automatically learn the relation between the characteristics such as the concentration of chemical substances in the furnace and the running state of equipment through a machine learning or deep learning algorithm, thereby realizing accurate identification of the state.

When the state detection unit recognizes the early warning or fault state, the early warning unit triggers a corresponding early warning or fault detection flow. By monitoring key indexes such as chemical substance concentration in the furnace in real time, the early warning unit can timely send out early warning before the abnormal state of the equipment occurs, so that enough time is provided for operators to take measures, and faults are prevented from occurring or expanding.

Constructing a direct current submerged arc furnace knowledge base, storing historical chemical substance detection data and the like. When the equipment fails, the failure data is called to generate chemical detection substances, and the chemical substances in the furnace are detected. And generating real-time electric equipment adjustment maintenance data according to the detection result and the historical data, and optimizing the running state of the equipment. The construction of the knowledge base provides rich historical data support for data analysis. Through the application of chemical detection substances, the types and the concentrations of the chemical substances in the furnace can be directly detected, so that the state of the equipment can be accurately judged. The generated adjustment maintenance data can solve the equipment problem in a targeted manner, and the reliability and stability of the equipment are improved.

According to the technical scheme, through the close collaboration of a plurality of links such as data acquisition, detection, state identification, early warning and data processing, the real-time detection of chemical substances in the direct-current submerged arc furnace and the accurate pre-judgment of the equipment state are realized. The scheme not only solves the problem that the existing direct-current submerged arc furnace cannot detect chemical substances in the furnace, but also realizes safe and stable operation of equipment through early warning and adjustment of maintenance flow, effectively improves production efficiency and reduces maintenance cost.

The invention relates to a model and a construction step, a usage and model parameters:

Running state detection model:

The construction steps are as follows:

And collecting basic data of the direct-current submerged arc furnace, including electrical parameters, temperature, arc stability, chemical substance concentration, cooling system data and equipment state data.

And (3) preprocessing data, namely cleaning, calibrating and extracting features of the collected data to form a feature data set.

Model selection, namely selecting a proper machine learning or deep learning algorithm (such as a support vector machine, a random forest, a neural network and the like) as a model base.

Feature engineering, further performing feature selection, transformation or construction on the feature data set to extract information most useful for operating state identification.

Model training, namely training a model by using characteristic data in a historical data set and corresponding running state labels, and optimizing model parameters by means of cross verification, grid search and the like.

The running state detection model is used for identifying the running state of the direct current submerged arc furnace in real time, and comprises a normal state, an early warning state and a fault state.

Running state detection model parameters, namely specific parameters (such as the number of layers of a neural network, the number of nodes, the learning rate and the like) of a machine learning or deep learning algorithm. Parameters used in feature selection, transformation, or construction processes.

An abnormality detection model construction step:

and collecting data of the types, the concentrations and the change trend of chemical substances in the furnace.

Setting a threshold value, namely setting a normal range threshold value of chemical substance data in the furnace according to historical data and actual operation experience.

And constructing an abnormality detection model based on the set threshold value.

The abnormality detection model is used for monitoring the state of chemical substances in the furnace in real time, comparing the state with a threshold value, and triggering an alarm when the data exceeds a normal range to prompt the existence of equipment faults or potential safety hazards.

Model parameters of the abnormality detection model are normal range thresholds of the furnace chemical data.

A prediction model construction step:

data collection, namely collecting historical data and current data, including changes of temperature, pressure and chemical component concentration in the furnace, equipment running time, load and the like.

Model selection, namely selecting a proper time sequence analysis, machine learning or deep learning algorithm as a prediction model basis.

Model training, namely training a prediction model by using historical data and current data, and optimizing model parameters to improve prediction accuracy.

The prediction model is used for predicting future trends of chemical substance states and equipment running conditions in the furnace and provides decision support for operators.

Predictive model parameters-specific parameters of time series analysis, machine learning or deep learning algorithms.

Historical data and current data for training the model.

When optimizing real-time electrical equipment tuning maintenance data using a distillation algorithm, the specific steps can be summarized as follows:

Determining a knowledge distillation target, namely determining a distillation algorithm target, namely improving the accuracy and the effectiveness of real-time electric equipment adjustment maintenance data, and enabling the real-time electric equipment adjustment maintenance data to be more close to preset direct-current submerged arc furnace standard data.

And constructing a teacher model, namely selecting or constructing a teacher model with excellent performance. In the present scenario, the teacher model may be a model that adjusts maintenance data and the running state of the direct-current submerged arc furnace based on the historical optimal electrical equipment, and it can accurately predict and adjust the equipment state. The teacher model should have high accuracy and generalization capability and can be used as a reference standard for adjusting and maintaining data optimization of the real-time electrical equipment.

And preparing a student model, namely, a real-time electric equipment adjustment maintenance data model which needs to be optimized.

Parameters of the initial chemo model, which are derived from preliminary regulatory maintenance data or from a simple rule-based model.

Data preparation, namely preparing training data for distillation, wherein the training data comprises historical optimal electric equipment adjustment maintenance data, corresponding direct-current submerged arc furnace running state data and current real-time electric equipment adjustment maintenance data.

Knowledge distillation process:

and in the prediction stage, firstly, the teacher model predicts the training data to obtain a prediction result of the teacher model.

And in the learning stage, the student model predicts the training data and compares the prediction result with the prediction result of the teacher model.

And calculating the difference between the student model prediction and the teacher model prediction, namely distilling loss. Meanwhile, standard loss can be calculated by combining the real label of the original data.

Back propagation and optimization by back propagation algorithm, parameters of the student model are optimized together according to distillation loss and standard loss.

Model evaluation and adjustment:

During the distillation process, the performance of the student model is periodically evaluated to ensure that the optimization direction is correct.

And according to the evaluation result, adjusting the super parameters in the distillation process, such as the weight, the learning rate and the like of the distillation loss and the standard loss.

And outputting the optimized model, namely stopping the distillation process when the performance of the student model reaches the expected performance, and outputting the optimized real-time electric equipment adjustment maintenance data model.

And (3) deploying and applying, namely deploying the optimized model to a direct-current submerged arc furnace equipment end for guiding actual electric equipment adjustment and maintenance work.

Through the steps, the distillation algorithm can effectively transfer the knowledge of the teacher model to the student model, and the accuracy and the effectiveness of the adjustment and maintenance data of the real-time electrical equipment are improved, so that the running state of the direct-current submerged arc furnace is optimized.

Claims

1. A direct current submerged arc furnace, comprising:

2. The direct current submerged arc furnace of claim 1, further comprising a data acquisition unit for:

3. The direct current submerged arc furnace of claim 1, further comprising a detection unit for:

4. The direct current submerged arc furnace of claim 1, further comprising a data processing unit that:

5. The direct current submerged arc furnace of claim 1, further comprising a data processing unit that:

6. The direct-current submerged arc furnace detection system is applied to the direct-current submerged arc furnace according to any one of claims 1-5, and is characterized by comprising a server end, a data acquisition end and a direct-current submerged arc furnace equipment end, wherein the server end is respectively in communication connection with the data acquisition end and the direct-current submerged arc furnace equipment end;

the server side comprises:

7. The direct current submerged arc furnace detection system of claim 6, further comprising a data acquisition unit that is further configured to:

8. The direct current submerged arc furnace detection system of claim 6, further comprising a detection unit configured to:

9. The direct current submerged arc furnace detection system of claim 6, further comprising a data processing unit that:

10. The direct current submerged arc furnace detection system of claim 6, further comprising a data processing unit that: