TW202303319A - Anomaly detection device - Google Patents

Abstract

An anomaly detection device according to an aspect is an anomaly detection device that detects an anomaly in a rolling system for rolling a material through a predetermined step, the device including: a pre-step specification judging unit that judges whether or not the material before the predetermined step meets predetermined specifications; a post-step specification judging unit that judges whether or not the material after the predetermined step meets predetermined specifications; an intervening operation amount extracting unit that extracts an operation amount in which an operator intervened in the predetermined step; a judgment unit that judges that there was an anomaly in the predetermined step if at least the post-step specification judging unit judges that the predetermined specifications are met and if the operation amount extracted by the intervening operation amount extracting unit exceeds a predetermined operation amount; and an output unit that outputs a result of judgment made by the judgment unit.

Description

Anomaly detection device

The present invention relates to an abnormality detection device for detecting abnormality of a rolling system.

For example, a rolling system is used in a rolling process (process) for rolling steel and the like. This rolling system has a hot rolling line (line) and a cold rolling line, and the raw material (rolled material: material) is processed in a plurality of lines. Or rolling is carried out separately in the process.

In addition, the calendering process is abnormal due to the deterioration of the equipment over time, which may affect the quality of the calendered material.

For example, when the coolant balance of the calender roll is broken due to an abnormal valve, etc., there will be an abnormal shape such as bending or off-centering of the rolled material. Abnormal mobility.

In addition, the operator (operator) monitors the state of the material or the process data (process data) during the operation, and in order to finish the material to an appropriate quality accuracy, he intervenes in the rolling conditions set by the computer in advance. Control, and manual operation to adjust the rolling conditions.

In this case, even if the quality and accuracy of the material can be maintained through the operator's technical proficiency (skill) or technical knowledge (know-how), it is still impossible to know whether the calendering process has occurred. abnormal. Generally speaking, in order to prevent equipment abnormalities that are likely to have a significant impact on operations, it is an issue for the entire equipment industry to detect abnormalities in the process at an early stage.

Also, in the generally widely used method of statistical process management, abnormalities are detected based on the results of the rolling process such as the accuracy of the rolled material or the stability of the rolling, and as long as there is an abnormality in the rolling results, the It is judged that an abnormality has occurred in the rolling process that also includes equipment abnormality.

Also, regarding the abnormal diagnosis of the rolling process, there is a method based on a model base, which is disclosed in Patent Document 1. A basic method for utilizing data is disclosed in Patent Document 2, for example.

In the abnormality diagnosis method disclosed in Patent Document 1, the actual performance values of a plurality of variables in the manufacturing process obtained during normal operation are used to create a plurality of mathematical expressions for predicting the state of the product under manufacture. model (sub-model), and use the scenario inference list (scenario inference list) to deduce the abnormal cause.

Here, regarding each equipment constituting the manufacturing process, it refers to the actual performance of materials before manufacturing, the setting of equipment, the actual performance of equipment, the actual performance of manual operation intervention, the actual status of intermediate products, and the actual product performance after manufacturing. Correlations are made into submodels expressed mathematically. In addition, a rule (rule) used when estimating the cause of an abnormal state that has occurred in the manufacturing process is defined in advance as a situation inference table.

Then, the presence or absence of abnormality is detected for each sub-model, and the cause of the abnormality in the post-manufacture product is estimated using the situational inference table.

For example, the main factor (upstream side factor) of the abnormal state of the manufacturing process in equipment i is "equipment performance", and the subfactor (downstream side factor) of the abnormal state of the manufacturing process in equipment k Sub) In the case of "manual intervention results", it is presumed that "an equipment abnormality occurs in equipment i, and the operator operates equipment k to make up for the equipment abnormality, but a mistake has occurred" as the cause of the abnormal state of the manufacturing process. .

However, when an abnormality has occurred in the actual manufacturing results, this diagnostic method searches for the cause of the abnormality, equipment, settings, and human intervention, as its causal relationship. That is, this diagnosis method does not include the case where whether there is an abnormality in the rolling process including the equipment is detected regardless of whether there is an abnormality in the actual manufacturing results.

In addition, Patent Document 2 discloses an abnormality diagnosis method, which can specifically clarify the causal relationship between the cause of the abnormality and the abnormal state for the method and device for diagnosing the abnormal state of the equipment and product process in the rolling equipment and the like.

In addition, Patent Document 2 has disclosed a method of discarding explanatory variables, which is a method for diagnosing the abnormal state of the product process by performing linear multivariate analysis, and using a neural network or a genetic algorithm to Among the methods of identifying the causal relationship between the cause of the abnormality and the abnormal state, etc., there is a method of discarding unnecessary explanatory variables that become external disturbances.

However, these diagnostic methods are also diagnostic methods for finding the cause of abnormalities in the rolling process including equipment abnormalities for abnormalities found as a result of rolling.

[Prior Art Literature]

[Patent Document]

Patent Document 1: Japanese Patent No. 6662222

Patent Document 2: Japanese Patent No. 3892614

As described above, there has been conventionally proposed a method of finding abnormalities in the rolling process of equipment, etc. considered to be the cause when abnormalities have occurred in rolling results such as product quality and rolling stability. However, it is impossible to detect the abnormality of the rolling process when the rolling result is normally maintained by the operator intervening in the control by manual operation and making up for the abnormality of the rolling process.

In the actual manufacturing process, most of the operators etc. intervene to adjust the calendering process in order to obtain the desired calendering result.

The present invention was developed to solve the above-mentioned problems, and its object is to provide an abnormality detection device that can detect abnormalities that have occurred in the predetermined process even if the material after the predetermined process of processing materials meets the predetermined specifications.

An abnormality detection device according to an aspect of the present invention detects an abnormality in a rolling system that rolls materials through a predetermined process, and the abnormality detection device includes: a pre-process specification determination unit that determines whether the material before the predetermined process satisfies Predetermined specifications; post-process specification determination section, which determines whether the material after the aforementioned predetermined process meets the predetermined specifications; intervention operation amount extraction section, which extracts the operation amount that the operator has intervened in the aforementioned predetermined process; determination section, is At least when the specification judging unit judges that the predetermined specification is met after the aforementioned process, and the operation amount extracted by the intervention operation amount extraction unit exceeds the previously determined operation amount, it is determined that an abnormality has occurred in the aforementioned predetermined process; and The output unit outputs the result determined by the determination unit.

In addition, in the abnormality detection device according to an aspect of the present invention, it is preferable that the determination unit determines whether an abnormality has occurred based on the specification of the material or a standard that differs for each operator.

In addition, in the abnormality detection device according to an aspect of the present invention, it is preferable that the determination unit determines whether an abnormality has occurred using a non-statistical method.

According to the present invention, even if the material after the predetermined process of processing the material satisfies the predetermined specification, it is possible to detect an abnormality that has occurred in the predetermined process.

1: Process control system

2: Hot calendering production line (calendering system)

3: Cold calendering production line (calendering system)

4: Abnormal detection device

10: Control network

20: heating furnace

21: Rough calender

22: Cutting machine

23: Fine calender

24: cooling device

25:Coiler

26-1~26-4: Sensors

27: The first operation panel

28: First control device

30: Uncoiler

31: Entrance side shear

32: Welding machine

33: Looper

34:Calender

35: Export side shear

36: tension reel

37: Sensor

38: The second operation panel

39: Second control device

40:Production Information Memory Department

41: Performance memory department

42:Operation memory unit

43: Unanimous Judgment Department

44: Pre-process specification judgment department

45: Post-process specification judgment department

46: Intervention operation volume extraction part

47: Judgment Department

48: output part

FIG. 1 is a diagram showing a configuration example of a process control system including a rolling system and an abnormality detection device for detecting an abnormality of the rolling system.

Fig. 2 is a diagram showing a configuration example of an abnormality detection device according to one embodiment.

Fig. 3 is a flowchart (flowchart) showing an example of the operation of the abnormality detection device of one embodiment.

FIG. 4 is a series of diagrams showing specific examples of a process in which an abnormality is detected by an abnormality detection device, deterioration of affected equipment, and manual intervention by an operator.

FIG. 5 is a series of diagrams showing specific examples of a process in which an abnormality is detected by an abnormality detection device, deterioration of affected equipment, and manual intervention by an operator.

Hereinafter, one embodiment of an abnormality detection device for detecting abnormality in a rolling system will be described using drawings. FIG. 1 is a diagram showing a configuration example of a process control system 1 including a rolling system and an abnormality detection device for detecting an abnormality in the rolling system.

As shown in FIG. 1, the process control system 1 is composed of, for example, a hot calendering production line 2 and a cold calendering production line 3, which are two calendering systems, respectively connected to an abnormality detection device 4 through a control network 10, and by The material (rolled material) is sequentially rolled by a plurality of rolling steps.

The control network 10 is, for example, a LAN (Local Area Network; Local Area Network) or the like, and can also be configured to include a control LAN and an information system LAN.

The hot calendering production line 2, for example, has a heating furnace (RF: Reheating Furnace) 20, a rough calender (RM: Roughing mill) 21, a shearing machine (CS: Crop Shear) 22, and a fine calender (FM: Finishing mill) 23. Cooling device (ROT: Run Out Table) 24, and coiler (DC: Down Coiler) 25. Moreover, the hot calendering production line 2 is, for example, equipped with sensors (sensors) 26-1 to 26-4, and is controlled by the first control device 28 through the operator operating the first operation panel 27 to form a hot calendering production line. 2 is composed of each part.

The hot calendering production line 2 is a rough calender 21 that performs rough calendering on the slab output from the heating furnace 20 , and feeds the calendered material cut by the cutter 22 to the fine calender 23 . Then, the hot calendering production line 2 uses the finishing calender 23 to further calender the roughly calendered calendered material to a predetermined specification, and is coiled by the coiler 25 after being cooled by the cooling device 24 . In this way, the hot calendering production line 2 includes a plurality of processes to perform predetermined processing on the calendered material.

In addition, the sensor 26-1, for example, is arranged at the exit side of the rough calender 21, detects the rolling performance data of the rolled material rolled by the rough calender 21, and outputs it to the first controller. Control device 28. For example, the sensor 26-1 detects various actual values in the rolling process, so that the first control device 28 can obtain the length of the rolled material in each of a plurality of positions different in the rolling direction (traveling direction of the rolled material) , thickness, width and temperature.

The sensor 26 - 2 is arranged at the inlet side of the finishing calender 23 , detects the rolling performance data of the calendered material fed to the finishing calender 23 , and outputs it to the first control device 28 . For example, the sensor 26-2 detects various performance values relative to the rolled material, so that the first control device 28 can obtain the length, thickness, width and temperature of the rolled material in each of a plurality of positions with different rolling directions. wait.

The sensor 26 - 3 is arranged at the exit side of the finishing calender 23 , detects the rolling performance data of the calendered material rolled by the finishing calender 23 , and outputs it to the first control device 28 . For example, the sensor 26-3 detects various performance values in the rolling process, so that the first control device 28 can obtain the length, thickness, width, temperature, etc. of the rolled material in each of a plurality of positions with different rolling directions. .

The sensor 26 - 4 is arranged at the outlet side of the cooling device 24 , detects the rolling performance data of the rolled material cooled by the cooling device 24 , and outputs it to the first control device 28 . For example, the sensor 26-4 detects various performance values after the cooling process, so that the first control device 28 can obtain the length, thickness, width, temperature, etc. of the rolled material in each of a plurality of positions with different rolling directions. .

Then, the first control device 28 sends the actual rolling performance data detected by the sensors 26 - 1 to 26 - 4 to the abnormality detection device 4 through the control network 10 . In addition, the hot rolling production line 2 is provided with a plurality of other sensors not shown, and sends various data detected by the sensors to the abnormality detection device 4 respectively.

The cold rolling production line 3, for example, has a pay-off reel 30, an entry side shear 31, a welding machine 32, a looper 33, a calender 34, an exit side shear 35, a tension reel ( tension reel) 36, a plurality of sensors 37, a second operation panel 38 and a second control device 39. Then, the cold rolling line 3 is configured such that each part constituting the cold rolling line 3 is controlled by the second control device 39 by the operator operating the second operation panel 38 .

The cold calendering production line 3 uses the tail end of the steel coil (coil) of the rolled material coiled by the coiler 25 of the hot calendering production line 2 as the front end, and further processes the rolled material rolled by the hot calendering production line 2 calendering.

More specifically, the uncoiler 30 takes the tail end of the steel coil coiled by the coiler 25 as the front end, and guides the rolled material toward the inlet side shearer 31 . The entrance side shears 31 and the exit side shears 35 are used to pinch and pass the calendered material through front and rear pinch rolls (pinch rolls) (not shown) in order to adjust the front and rear ends of the calendered material, and According to the control of the second control device 39, the rolled material is cut off.

The welding machine 32 has a function of connecting steel coils by welding so that the rolling process performed on a plurality of steel coils can be continuously performed. The looper 33 stores the calendered material, and supplies the calendered material to the calender 34 in a fixed manner.

The calender 34 is for further calendering the calendered material. The tension drum 36 is used to take up the calendered material calendered by the calender 34 . In this way, the cold rolling production line 3 includes a plurality of processes to perform predetermined processing on the rolled material.

A plurality of sensors 37 are arranged around a plurality of stands arranged in the calender 34 , for example, detect the thickness of the calendered material, and output to the second control device 39 . In addition, the sensor 37 can also detect various performance characteristics in the rolling process, so that the second control device 39 can The length, width, temperature, and the like of the rolled material at each of a plurality of positions with different rolling directions can be obtained.

Then, the second control device 39 sends each rolling performance data detected by each sensor 37 to the abnormality detection device 4 through the control network 10 . Also, the cold rolling line 3 is provided with a plurality of other sensors not shown, and sends various data detected by the sensors to the abnormality detection device 4 respectively.

In this way, the process control system 1 uses the hot rolling production line 2 and the cold rolling production line 3 to respectively roll the rolling materials, and the abnormality detection device 4 receives actual performance time series data in each rolling process. Also, the abnormality detection device 4 receives information indicating that the operator has operated the first operation panel 27 and the second operation panel 38 . In other words, the abnormality detection device 4 collects the amount of operations that the operator has performed on the hot rolling line 2 and the cold rolling line 3 .

Next, a configuration example of the abnormality detection device 4 will be described in detail. FIG. 2 is a diagram showing a configuration example of an abnormality detection device 4 according to one embodiment. The abnormality detection device 4 is, for example, a computer (computer) equipped with a CPU, and has a production information storage unit 40, an actual performance storage unit 41, an operation amount storage unit 42, a coincidence determination unit 43, a pre-process specification determination unit 44, and a post-processing A specification determination unit 45 , an intervention operation amount extraction unit 46 , a determination unit 47 , and an output unit 48 .

The production information storage unit 40 stores the production information (production conditions, etc.) of the rolled materials in the hot rolling production line 2 and the cold rolling production line 3 in advance. The production information is, for example, pre-set computer data including the steel type and product size of the materials put into the hot rolling production line 2 and the cold rolling production line 3 .

The actual performance memory unit 41 stores the actual performance time series data sent by the hot rolling production line 2 and the cold rolling production line 3 . Time-series data of actual performance detects hot rolling for each sensor The results of the rolled materials processed in the production line 2 and the cold rolling production line 3 in each process, etc. For example, actual performance time series data are data related to the quality accuracy of rolled materials, the stability of rolling phenomena, and quality issues other than precision.

The operation amount memory unit 42 memorizes information (time series) indicating manual operations (lever operations, gain adjustments, etc.) performed by the operator on the first operation panel 27 and the second operation panel 38 based on his/her own judgment. material). In other words, the abnormality detection device 4 is capable of collecting the respective operation amounts that the operator has operated on the hot rolling line 2 and the cold rolling line 3 .

The coincidence determination unit 43 obtains the information indicating that the operator has entered the process start operation on the hot rolling production line 2 or the cold rolling production line 3 through the first operation panel 27 or the second operation panel 38, and reads the information from the production information storage unit 40 The memorized production information. In addition, the coincidence determination unit 43 determines that the process executed in each process of the hot rolling line 2 or the cold rolling line 3 is consistent with the read production information.

Next, the coincidence determination unit 43 specifically specifies the manufacturing process executed in each process of the hot rolling production line 2 or the cold rolling production line 3 from among the plurality of production conditions stored in the production information storage unit 40 . Then, the coincidence determination unit 43 outputs the reference specification that should exist before a predetermined process to the pre-process specification determination unit 44 . Also, the coincidence determination unit 43 outputs the standard specification that should exist after a predetermined process to the post-process specification determination unit 45 . Also, the coincidence determination unit 43 outputs to the determination unit 47 an operation amount (reference operation amount: for example, a threshold value) that the operator can usually intervene in a predetermined process.

The pre-process specification judging unit 44 compares the reference specification before the scheduled process input from the coincidence judging unit 43 with the actual performance time-series data before the scheduled process read from the actual performance storage unit 41, and judges that the pre-planned process Whether the calendered material meets the predetermined specifications. In other words In other words, the pre-process specification determination unit 44 quantifies the accuracy of the rolled material before a predetermined process and monitors changes. Then, the pre-process specification determination unit 44 outputs the determination result to the determination unit 47 .

The post-process specification determination unit 45 compares the standard specification after the scheduled process input from the coincidence determination unit 43 with the actual performance time-series data after the scheduled process read from the actual performance storage unit 41, and determines that the post-scheduled process Whether the calendered material meets the predetermined specifications. In other words, the post-process specification determination unit 45 quantifies the accuracy of the rolled material after a predetermined process and monitors changes. Then, the post-process specification determination unit 45 outputs the determination result to the determination unit 47 .

The intervention operation amount extracting unit 46 reads information (time series data) indicating that the operator has manually operated the first operation panel 27 and the second operation panel 38 by his/her own judgment from the operation amount storage unit 42, and respectively The intervention operation amount indicating the operation amount manually intervened by the operator in each predetermined process in the hot rolling line 2 and the cold rolling line 3 is extracted and output to the determination unit 47 . In other words, the intervention operation amount extracting unit 46 extracts the operator's operation amount, quantifies it, and monitors the change.

The judging unit 47 determines at least after the process that the specification judging unit 45 satisfies the predetermined specification, and the operation amount (intervention operation amount) extracted by the intervention operation amount extraction unit 46 exceeds a previously determined operation amount (reference operation amount: for example, a temporary limit value), it is determined that an abnormality has occurred in the predetermined process. Then, the determination unit 47 outputs the determined result to the output unit 48 .

Furthermore, the determination unit 47 can also determine whether to use the rolling material specifications input through the first operation panel 27 or the second operation panel 38, or different criteria (such as specific conditions) for each operator, for example. An exception has occurred. This is because it is considered that the amount of intervention will vary according to the specifications of each rolled material or each operator. In addition, the determination unit 47 may use non-statistical techniques such as clustering to determine whether or not an abnormality has occurred.

In this way, the judging unit 47 judges whether the operator's operation is conducive to maintaining the accuracy of the rolled material based on the operator's operation amount, and judges the operator's operation amount and the predetermined process by using logic and threshold values. The relationship between the accuracy of the rolled material after.

The output unit 48 outputs the determination result input from the determination unit 47 . For example, the output unit 48 may output by displaying the determination result or process change input from the determination unit 47 on a display or the like, or may output by sound or light.

In other words, even if the production conditions are the same and the material accuracy (rolled material accuracy) after the process satisfies the predetermined specification, the abnormality detection device 4 deviates from the reference past data and the like in the operation amount of the operator in the process. In the case of , it is judged as an abnormality of the process.

FIG. 3 is a flow chart showing an example of the operation of the abnormality detection device 4 of one embodiment. As shown in FIG. 3 , in step 100 (S100), the abnormality detection device 4 specifically specifies the hot rolling production line 2 or the cold rolling production line by the coincidence determination unit 43 from among a plurality of production conditions stored in the production information storage unit 40. 3 Processes performed in each process.

In step 102 (S102), the pre-process specification determination unit 44 compares the reference specification before the scheduled process input from the coincidence determination unit 43 with the actual performance time series before the scheduled process read from the actual performance storage unit 41. data, and determine whether the calendered material before the predetermined process meets the predetermined specifications.

In step 104 (S104), the post-process specification determination unit 45 compares the standard specification after the planned process input from the coincidence determination unit 43 with the actual performance time series after the scheduled process read from the actual performance storage unit 41. data, and determine whether the calendered material after the predetermined process meets the predetermined specifications.

In step 106 (S106), the intervention operation amount extracting unit 46 reads information indicating that the operator has manually operated the first operation panel 27 and the second operation panel 38 from the operation amount storage unit 42 ( Time-series data), and respectively extract the intervention operation amount representing the operation amount that the operator has manually intervened in each predetermined process in the hot calendering production line 2 and the cold calendering production line 3 .

In step 108 (S108), the judging unit 47 determines at least after the process that the specification judging unit 45 satisfies the predetermined specification, and the operation amount (intervention operation amount) extracted by the intervention operation amount extraction unit 46 exceeds the previously determined operation amount. In the case of a certain amount (reference operation amount: for example, a threshold value), it is determined that an abnormality has occurred in the predetermined process.

In step 110 ( S110 ), the output unit 48 outputs the determination result input from the determination unit 47 .

Next, a specific example of the operation of the abnormality detection device 4 will be described by taking, as an example, bending during rolling of materials in the hot rolling line 2 .

First, the coincidence determination unit 43 uses production information (steel type, product size, etc.) layer number. At this time, all materials are assigned layer numbers, and the same layer number is regarded as the same production condition.

The pre-process specification judgment unit 44 takes the actual performance time-series data of the material before the process (quality accuracy, instability of rolling phenomenon, quality problems other than accuracy, etc.) and monitor changes in the amount of curvature of the raw material at the exit side of rough rolling. The pre-process specification judging unit 44 uses, as a monitoring method, a method of judging data exceeding a threshold value of 1σ obtained, for example, from an X-s management chart as abnormal.

The post-process specification judgment unit 45 takes the actual performance time-series data of the post-process material (quality accuracy, instability of rolling phenomenon, quality problems other than precision, etc.) and monitor changes in the amount of bending of the material at the exit side of finish rolling. The post-process specification judging unit 45 uses, as a monitoring method, a method of judging data exceeding a threshold value of 1σ obtained, for example, from an X-s management chart as abnormal.

The intervention operation amount extraction unit 46 takes the time-series data of the operator's operation amount (lever, gain, etc.) It is extracted and quantified, and changes in the operating volume are monitored. The intervention manipulation amount extracting unit 46 uses, as a monitoring method, a method of judging data exceeding a threshold value of 2σ obtained, for example, from an X-s management chart as abnormal.

The judging unit 47 judges, for example, the presence or absence of a process change in the finish rolling or rough rolling and an abnormality in the process as follows. In the process variation, there are left-right imbalance caused by the gap of the roll (roll), and left-right imbalance caused by the gap of the pressing device.

Then, the judging unit 47 judges the relationship between the operation amount of the operator and the precision of the material after the process by using the logic and the threshold value, and detects the change of the process.

For example, the judging unit 47 may have a normal accuracy with no change in the bending amount of the material at the exit side of the rough rolling, and a normal accuracy with no change in the bending amount of the material at the exit side of the finish rolling, and the operator When the change in the amount of operation is large, it is determined that the process change has been absorbed by the operator's operation (operation).

In addition, the judging unit 47 may have a change in the amount of curvature of the material at the exit side of the rough rolling, and have no change in the amount of curvature of the material at the exit side of the finish rolling, which is normal accuracy, and operate If the operator's operation amount changes greatly, it is determined that the change has occurred in both the process and the previous process process, or in either one of the processes, and the process change has been absorbed by the operator's operation.

Specifically, the judging unit 47, for example, judges that the finishing rolling process after rolling the coil is valid when the operation amount of the operator exceeds the threshold value of 2σ and the accuracy of the material after the process is within 1σ. Possibility of abnormal process.

The result determined by the determination unit 47 is displayed on, for example, a display or the like by the output unit 48 .

Next, specific examples of the process (controlled object) in which the abnormality detection device 4 detects an abnormality and the manual intervention by the operator (intervention method) will be described. The series of FIGS. 4 and 5 show specific examples of the abnormality detection process (controlled object) detected by the abnormality detection device 4 , affected equipment degradation, and manual intervention by the operator (intervention method).

In the strip leveler intermesh (cold rolling/process), the operator performs the operation of increasing the amount of intermesh under the condition that the pushing pressure of the leveler has changed.

Also, in the shape control (cold rolling) of the cold-rolled steel sheet, the operator manually performs the bender/leveling by turning off the shape control when the thickness of the roll has changed. Machine (leveling) intervention operation, or target shape change.

In addition, in the determination of the completion of coiling for the tension reel <cold rolling/process>, the operator increases the speed lead rate (speed) when the wear coefficient of the mandrel has deteriorated (decreased in roughness). lead rate), or an operation that delays the winding completion timing.

In addition, in the tension control of the tension device (bridle) (without tension meter (tension meter)) (cold rolling/process), the operator is tied to the GD2 variable of the tension device/gear In the case of deterioration (influence due to deterioration/insufficiency of grease), in order to avoid loosening of the tensioner/gear, increase the tension setting and provide torque to the motor.

Also, in the hot rolling CTC control <hot rolling>, the operator switches to the fixed warehouse when the on/off speed of the spray valve has deteriorated. (bank) setting, or the operation of making the valve delay speed (valve delay speed) setting match the actual state.

Also, in the position control of the coil car (coil car) on the entrance side/exit side (cold rolling/process), the operator is in the situation that the oil pressure source has decreased and the friction coefficient of the rail has changed, because It becomes impossible to stop at the predetermined position, so fine adjustment is performed by inching operation.

In addition, in the plate thickness measurement (hot rolling/cold rolling) by the plate thickness gauge, the operator performs the target plate thickness under the condition that the line source of the plate thickness gauge is lowered and the X-line axis is shifted. Additional compensation (offset), or the operation of additional compensation for the plate thickness variation target (usually 0) controlled by AGC.

Also, in the speed control of the calender (hot calendering/cold calendering), when the operator has changed the resonance frequency due to the deterioration of the calender housing (housing), rollers, gears, and motors, Perform operations that have avoided the resonance velocity area, or change the schedule of the steel plate produced.

In addition, in the loop car (cold rolling/process), the operator is in the situation where the tilting of the loop car shell and the failure of the swing mechanism (snaking, or the vibration of the plate becomes larger) have occurred. Next, set the synchronous position on the short end side so that the looper does not reach the long end side.

Also, in the control of the wiper roll (cold rolling/process), the operator increases the size of the wiper roll when the surface of the wiper roll (rubber roll) has deteriorated. Operation of pushing amount.

Also, in the process line (process line) in the straight line of the furnace plate (snake), the operator is based on the characteristics of the furnace, or when the heating device has deteriorated and changed, and additionally set the compensation amount so that the actual performance becomes Operation of the desired tension.

Also, in the elongation rate of SPM (skin pass mill), the operator did not reach the target set by L2 (predetermined level 2) due to the deterioration and change of the rigidity and reduction system of the skin pass mill. In the case of load, perform the operation to change the setting of the target value of SPM load.

As explained above, since the abnormality detection device 4 judges at least after the process that the standard determination unit 45 satisfies the predetermined specification, and the operation amount extracted by the intervention operation amount extraction unit 46 exceeds the previously determined operation amount, it is judged as An abnormality has occurred in the predetermined process, so even if the material after the predetermined process of processing the material satisfies the predetermined specification, the abnormality that has occurred in the predetermined process can still be detected.

Furthermore, each of the functions of the abnormality detection device 4 has a part or all of which can be implemented by hardware such as PLD (Programmable Logic Device; Programmable Logic Device) or FPGA (Field Programmable Gate Array; Field Programmable Gate Array). Body (hardware), and can also constitute a program (program) executed by a processor (processor) such as a CPU.

4: Abnormal detection device

40:Production Information Memory Department

41: Performance memory department

42:Operation memory unit

43: Unanimous Judgment Department

44: Pre-process specification judgment department

45: Post-process specification judgment department

46: Intervention operation volume extraction part

47: Judgment Department

48: output part

Claims (3)

An abnormality detection device for detecting abnormality of a calendering system for calendering materials through a predetermined process, the abnormality detection device having: The pre-process specification judgment department is to judge whether the material before the aforementioned predetermined process meets the predetermined specification; The post-process specification determination section is to determine whether the material after the aforementioned predetermined process meets the predetermined specification; Intervention operation amount extracting part is to extract the operation amount that the operator has intervened in the aforementioned predetermined process; The judging unit judges that at least after the aforementioned process, the specification judging unit satisfies the predetermined specification, and when the manipulation amount extracted by the intervention manipulating amount extracting unit exceeds the previously determined manipulation amount, it is judged that the predetermined process has been completed. an exception occurs; and The output unit outputs the result determined by the determination unit. The abnormality detection device according to claim 1, wherein the determination unit determines whether the abnormality has occurred based on the specification of the material or a different standard for each operator. The abnormality detection device according to claim 1 or 2, wherein the determination unit uses a non-statistical method to determine whether an abnormality has occurred.

Images