CN111881999A - Water service pipeline leakage detection method and system based on deep convolutional neural network - Google Patents
Water service pipeline leakage detection method and system based on deep convolutional neural network Download PDFInfo
- Publication number
- CN111881999A CN111881999A CN202010770571.9A CN202010770571A CN111881999A CN 111881999 A CN111881999 A CN 111881999A CN 202010770571 A CN202010770571 A CN 202010770571A CN 111881999 A CN111881999 A CN 111881999A
- Authority
- CN
- China
- Prior art keywords
- data
- neural network
- convolutional neural
- pipeline
- deep convolutional
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2413—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on distances to training or reference patterns
- G06F18/24133—Distances to prototypes
- G06F18/24137—Distances to cluster centroïds
- G06F18/2414—Smoothing the distance, e.g. radial basis function networks [RBFN]
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/06—Energy or water supply
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
- Y02A20/00—Water conservation; Efficient water supply; Efficient water use
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Health & Medical Sciences (AREA)
- Data Mining & Analysis (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Business, Economics & Management (AREA)
- Evolutionary Computation (AREA)
- General Health & Medical Sciences (AREA)
- Software Systems (AREA)
- Biomedical Technology (AREA)
- Molecular Biology (AREA)
- Mathematical Physics (AREA)
- Computational Linguistics (AREA)
- Biophysics (AREA)
- Computing Systems (AREA)
- Economics (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Evolutionary Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Bioinformatics & Computational Biology (AREA)
- Public Health (AREA)
- Water Supply & Treatment (AREA)
- Human Resources & Organizations (AREA)
- Marketing (AREA)
- Primary Health Care (AREA)
- Strategic Management (AREA)
- Tourism & Hospitality (AREA)
- General Business, Economics & Management (AREA)
- Examining Or Testing Airtightness (AREA)
Abstract
The invention discloses a water service pipeline leakage detection method and system based on a deep convolutional neural network, wherein the method comprises the steps of establishing and training a convolutional neural network model in advance according to pipeline data acquired by intelligent pipe fittings of all position areas in a water service pipe network water supply system, the trained deep convolutional neural network model comprises a plurality of leakage data label indexes, and each leakage data label index comprises a leakage point type and a leakage point position; real-time pipeline data acquired by intelligent pipes in each position area in a water service pipe network water supply system are acquired in real time, the real-time pipeline data are preprocessed, the preprocessed real-time pipeline data are sent into a trained deep convolution neural network model, and a prediction result is obtained through the deep convolution neural network model; and finally, comparing the prediction result with the leakage data label index of the corresponding leakage point type to judge the position of the leakage point, thereby being capable of quickly and accurately positioning the leakage point of the water service pipeline in the water service pipe network water supply system.
Description
Technical Field
The invention relates to the technical field of pipeline leakage detection, in particular to a water service pipeline leakage detection method and system based on a deep convolutional neural network.
Background
The city water service pipe network water supply system is an important component of city infrastructure, is responsible for the work of transmitting water sources, is a material foundation on which the city lives and develops, and is called a life line of the city. The water supply system for city water service pipe network consists of water source, pump station, water supply tank, water supply pipe network, water pipe valve, water using department, etc.
Due to the shortage of water resources caused by the rapid development of cities, the urban water service pipe network water supply system must be operated at full load with long time and high precision in the peak period of water use, so that the leakage phenomenon of water service pipelines can be caused, the waste of water resources is caused, and the serious people can cause the unstable water supply of the urban water service pipe network system.
In the prior art, the problems of difficult prediction, difficult leakage positioning and labor consumption exist in the leakage monitoring of the water service pipeline. The technical problem which needs to be solved urgently is how to quickly and accurately position the leakage position of the water service pipeline in the urban water service pipe network water supply system.
Disclosure of Invention
The invention mainly aims to provide a water service pipeline leakage detection method and system based on a deep convolutional neural network, and aims to solve the problems of difficult prediction, difficult leakage positioning and high labor consumption in the leakage monitoring of a water service pipeline in a current urban water service pipe network water supply system.
In order to achieve the purpose, the invention provides a water service pipeline leakage detection method and a system based on a deep convolutional neural network, wherein the method comprises the following steps:
acquiring pipeline data acquired by intelligent pipes in each position area in a water service pipe network water supply system in advance, wherein the intelligent pipes are correspondingly arranged on water service pipes in different position areas in the urban water service pipe network;
preprocessing pre-collected pipeline data;
inputting the preprocessed pipeline data into a deep convolutional neural network model for training so as to obtain a trained deep convolutional neural network model, wherein the trained deep convolutional neural network model comprises a plurality of leakage data label indexes, and each leakage data label index comprises a leakage point type and a leakage point position;
real-time pipeline data acquired by intelligent pipe fittings in each position area in a water service pipe network water supply system are acquired in real time, the real-time pipeline data are preprocessed, the preprocessed real-time pipeline data are sent into the trained deep convolutional neural network model, and a prediction result is obtained through the trained deep convolutional neural network model;
and comparing the prediction result with the leakage data label index of the corresponding leakage point type to judge the position of the leakage point.
Preferably, the step of preprocessing the pre-collected pipeline data specifically includes:
dividing pre-collected pipeline data into training data samples and testing data samples;
respectively carrying out normalization processing on the training data samples and the test data samples;
correspondingly, the step of inputting the preprocessed pipeline data into the deep convolutional neural network model for training to obtain the trained deep convolutional neural network model specifically includes:
inputting the training data sample after normalization processing into a deep convolution neural network model for training;
and testing the trained deep convolutional neural network model by using the test data sample after normalization processing, and storing the trained deep convolutional neural network model after the test is successful.
Preferably, the pipeline data includes flow data, temperature data, and pressure data;
correspondingly, the step of dividing the pre-collected pipeline data into training data samples and testing data samples comprises:
dividing the pre-collected pipeline data according to types, labeling each piece of pipeline data with different labels according to different types, and dividing the labeled pipeline data into training data samples and testing data samples;
the type of the pipeline data comprises normal data and leakage data, and the format of one piece of pipeline data acquired in real time each time is { a1 a2 … aj; b1 b2 … bj; c1 c2 … cj; j represents the pipeline data collected by the jth intelligent pipe fitting; a represents flow data, b represents temperature data, and c represents pressure data;
and constructing leakage data label indexes based on the labels of the pipeline data of which the types are leakage data, wherein each leakage data label index comprises a leakage point type and a leakage point position.
Preferably, the step of respectively performing normalization processing on the training data sample and the test data sample specifically includes:
randomly selecting the same number of pieces of data after different types of pipeline data are disordered to construct a training data sample;
scrambling each group of the selected data with the same number in each unselected pipeline data to construct a test data sample;
respectively carrying out normalization processing on the training data samples and the test data samples through the following formulas:
wherein x is a normalization processing result, x is each numerical value of each piece of data in the training data sample or the test data sample, and xmaxIs the largest value in the sample, xminIs the smallest value in the sample.
Preferably, the step of inputting the training data sample after the normalization processing into the deep convolutional neural network model for training specifically includes:
substep S1, inputting the training data samples after the normalization processing into the deep convolutional neural network model in batches for calculation, and obtaining a feedforward calculation result y, y ═ y (1), y (2).. y (f).. y (k) }, wherein y (k) is the probability that the current pipeline data belongs to the kth class predicted by the Softmax classifier; the Softmax classifier is divided into k classes, wherein k represents belonging to the kth class;
a substep S2, calculating a cross entropy error by the deep convolutional neural network model according to the feedforward calculation result, the actual output result and the label result;
substep S3, updating weight bias according to the cross entropy error;
correspondingly, the step of testing the trained deep convolutional neural network model by using the test data sample after the normalization processing, and storing the trained deep convolutional neural network model after the test is successful specifically comprises:
substep S4, performing iterative training for n times according to the substeps S1 to S3, and inputting normalized test data samples after each training is completed to obtain an output result; comparing the output result with a label corresponding to the test data sample to obtain an accuracy rate; if the accuracy meets the requirement, storing the structural parameters, the weight matrix and the bias parameters of the deep convolutional neural network model after training is finished; otherwise, adjusting the parameters of the deep convolutional neural network model or remanufacturing the sample and continuing training until the requirements are met.
Preferably, the real-time pipeline data is preprocessed in a manner of normalizing the real-time pipeline data.
Preferably, the step of sending the preprocessed real-time pipeline data into the trained deep convolutional neural network model and obtaining the prediction result through the trained deep convolutional neural network model specifically includes:
sending the real-time pipeline data after normalization processing into the trained deep convolutional neural network model for calculation to obtain a result sequence, and taking the result sequence as a prediction result;
and analyzing the result sequence, finding out the value 1 assigned to the bit with the largest value in the result sequence, and assigning 0 to other bits to obtain a new result sequence.
Preferably, the step of comparing the prediction result with the missing data tag index and determining the number and position of the missing points specifically includes:
searching a target mapping corresponding to the new result sequence from the leakage data tag index;
selecting the target mapping;
and judging the type and the position of the leakage point of the current pipeline data according to the type of the leakage point contained in the target mapping.
In addition, in order to achieve the purpose, the invention also provides a water service pipeline leakage detection system based on the deep convolutional neural network, which is applied to an urban water service pipe network, wherein intelligent pipe fittings are correspondingly arranged on water service pipelines in different position areas in the urban water service pipe network; the system comprises a data processing center and a water service pipeline leakage detection program based on the deep convolutional neural network, wherein the water service pipeline leakage detection program based on the deep convolutional neural network is stored on the memory and can run on the data processing center, the data processing center is used for receiving pipeline data transmitted by each intelligent pipe fitting, and the water service pipeline leakage detection program based on the deep convolutional neural network is configured to realize the steps of the water service pipeline leakage detection method based on the deep convolutional neural network.
In addition, in order to achieve the above object, the present invention further provides a computer readable storage medium, where a water service pipe leakage detection program based on a deep convolutional neural network is stored, and when executed by a processor, the water service pipe leakage detection program based on the deep convolutional neural network implements the steps of the water service pipe leakage detection method based on the deep convolutional neural network as described above.
The invention discloses a water service pipeline leakage detection method based on a deep convolutional neural network, which is characterized in that a convolutional neural network model is established and trained in advance according to pipeline data acquired by intelligent pipe fittings of all position areas in a water service pipe network water supply system, the trained convolutional neural network model comprises a plurality of leakage data label indexes, and each leakage data label index comprises a leakage point type and a leakage point position; then real-time pipeline data acquired by intelligent pipe fittings in each position area in the water service pipe network water supply system are acquired in real time, the real-time pipeline data are preprocessed, the preprocessed real-time pipeline data are sent into a trained convolutional neural network model, and a prediction result is obtained through a deep convolutional neural network model; and finally, comparing the prediction result with the leakage data label index of the corresponding leakage point type to judge the position of the leakage point. And then can be fast and accurately carry out the leakage point location to the water utilities pipeline in the water utilities pipe network water supply system.
Drawings
FIG. 1 is a schematic structural diagram of a water service pipeline leakage detection system based on a deep convolutional neural network in a hardware operating environment according to an embodiment of the present invention;
FIG. 2 is a partial schematic structural view of a main pipe segment of the water service pipe of the present invention;
FIG. 3 is a partial schematic view of the smart pipe of the present invention;
fig. 4 is a schematic flowchart of an embodiment of a water service pipeline leakage detection method based on a deep convolutional neural network according to the present invention.
The implementation, functional features and advantages of the objects of the present invention will be further explained with reference to the accompanying drawings.
Detailed Description
It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.
Referring to fig. 1, fig. 1 is a schematic structural diagram of a water service pipeline leakage detection system based on a deep convolutional neural network in a hardware operating environment according to an embodiment of the present invention.
As shown in fig. 1, the system may include: a data processing center 1001, such as a CPU; a communication bus 1002, a user interface 1003, a network interface 1004, and a memory 1005. Wherein a communication bus 1002 is used to enable connective communication between these components. The user interface 1003 may include a Display screen (Display), an input unit such as a Keyboard (Keyboard), and the optional user interface 1003 may also include a standard wired interface, a wireless interface. The network interface 1004 may optionally include a standard wired interface, a wireless interface (e.g., WI-FI interface). The memory 1005 may be a high speed RAM memory. The memory 1005 may alternatively be a storage device separate from the processor 1001.
The system is applied to an urban water service pipe network, intelligent pipes 1006 are correspondingly arranged on water service pipes corresponding to different position areas in the urban water service pipe network, and the data processing center 1001 is used for receiving pipe data transmitted by the intelligent pipes 1006.
Referring to fig. 2, a schematic structural diagram of a main pipe segment of a water service pipeline is shown, and fig. 3 is a schematic structural diagram of a smart pipe, which can be as shown in fig. 3, using a detection bracket as a carrier, wherein the detection bracket includes a temperature detection sensor, a pressure detection sensor and an ultrasonic transducer. The temperature detection sensor is used for detecting the temperature data of the water service pipeline, the pressure detection sensor is used for detecting the pressure data of the water service pipeline, and the ultrasonic transducer is used for detecting the flow data of the water service pipeline.
Those skilled in the art will appreciate that the configuration shown in FIG. 1 does not constitute a limitation of the water service pipe leakage detection system based on the deep convolutional neural network of the present invention, and may include more or fewer components than those shown, or some components in combination, or a different arrangement of components.
As shown in fig. 1, the memory 1005, which is a kind of computer storage medium, may include therein an operating system, a data storage module, a network communication module, a user interface module, and a water pipe leakage detection program based on a deep convolutional neural network. The data processing center 1001 is used to execute the steps of the water service pipeline leakage detection method based on the deep convolutional neural network.
Based on the hardware structure, the embodiment of the water service pipeline leakage detection method based on the deep convolutional neural network is provided.
Referring to fig. 4, fig. 4 is a schematic flowchart illustrating an embodiment of a water service pipeline leakage detection method based on a deep convolutional neural network according to the present invention.
In this embodiment, the method includes the steps of:
step S10: acquiring pipeline data acquired by intelligent pipes in each position area in a water service pipe network water supply system in advance, wherein the intelligent pipes are correspondingly arranged on water service pipelines corresponding to different position areas in the urban water service pipe network;
it should be noted that the execution subject of this embodiment is a data processing center, and the pipeline data includes flow data, temperature data, and pressure data; referring to fig. 3, the smart pipe of fig. 3 includes a temperature detecting sensor, a pressure detecting sensor, and an ultrasonic transducer. The temperature detection sensor is used for detecting the temperature data of the water service pipeline, the pressure detection sensor is used for detecting the pressure data of the water service pipeline, and the ultrasonic transducer is used for detecting the flow data of the water service pipeline.
Step S20: preprocessing pre-collected pipeline data;
in the specific implementation, the step firstly divides the pre-collected pipeline data into a training data sample and a testing data sample; then, the training data samples and the test data samples are respectively subjected to normalization processing.
Specifically, the step of dividing the pre-collected pipeline data into training data samples and testing data samples further includes:
sub-step Z1: dividing the pre-collected pipeline data according to types, labeling each piece of pipeline data with different labels according to different types, and dividing the labeled pipeline data into training data samples and testing data samples;
sub-step Z2: the types of the pipeline data comprise normal data and leakage data, and the format of one piece of pipeline data acquired in real time each time is { a1 a2 … aj; b1 b2 … bj; c1 c2 … cj; j represents the pipeline data collected by the jth intelligent pipe fitting; a represents flow data, b represents temperature data, and c represents pressure data;
sub-step Z3: and constructing leakage data label indexes based on the labels of the pipeline data of which the types are leakage data, wherein each leakage data label index comprises a leakage point type and a leakage point position.
Specifically, in the step, a leakage data tag index is constructed for the tag, and the tag comprises the type of a leakage point and the position of the leakage point of the leakage data;
{000 … 001: { leak type: normal; leakage point three-dimensional coordinate position: do not }
000 … 010: { leak type: point I is missed; leakage point three-dimensional coordinate position: (x1, y1, z1) }
000 … 100: { leak type: point II is omitted; leakage point three-dimensional coordinate position: (x2, y2, z 2).
100 … 000: { leak type: point leakage of q, m, n.. times; leakage point three-dimensional coordinate position: (xq, yq, zq), (xm, ym, zm).. and (xn, yn, zn) } };
wherein the number is 000. . . 001: { leak type: normal; leakage point position: none, 000 … 001 is an index, { leakage type: normal; leakage point position: none is the mapping corresponding to 000 … 001.
Correspondingly, the step of respectively performing normalization processing on the training data sample and the test data sample specifically includes the following substeps:
a substep A1 of randomly selecting the same number of data after different types of pipeline data are disordered to construct training data samples; in the specific implementation, in each type of data, the data with the same number is randomly selected after being scrambled to construct a training sample input _ train;
a substep A2, scrambling each group of the unselected pipeline data with the same number of selected data to construct a test data sample; in the specific implementation, the data with the same number is selected from each group of the rest data of each type and is disordered to construct a test sample input _ test;
the training data sample and test data sample formats are as follows:
training data samples
input_train＝{a11 a12…a1j；b11 b12…b1j；c11 c12…c1j；
a21 a22…a2j；b21 b22…b2j；c21 c22…b2j；
…
ai1 ai2…bij；bi1 bi2…aij；ci1 ci2…cij}；
Wherein i in aij represents i pieces of data, and j represents flow data acquired by the jth pipe fitting; i in bij represents i data, and j represents temperature data acquired by the jth pipe fitting; i in cij represents i pieces of data, and j represents pressure data acquired by the jth pipe fitting;
test data samples
input_test＝{d11 d12…d1j；e11 e12…e1j；f11 f12…f1j；
d21 d22…d2j；e21 e22…e2j；f21 f22…f2j；
…
di1 di2…dij,ei1 ei2…eij,fi1 fi2…fij,}；
Wherein i in dij represents i pieces of data, and j represents flow data acquired by the jth pipe fitting; i in eij represents i pieces of data, and j represents temperature data acquired by the jth pipe fitting; i in fij represents i pieces of data, and j represents pressure data collected by the jth pipe fitting.
Substep A3, normalizing the training data samples and the test data samples separately by:
wherein x is a normalization processing result, x is each numerical value of each piece of data in the training data sample or the test data sample, and xmaxIs the largest value in the sample, xminIs the smallest value in the sample.
Step S30: inputting the preprocessed pipeline data into a deep convolutional neural network model for training so as to obtain a trained deep convolutional neural network model, wherein the trained deep convolutional neural network model comprises a plurality of leakage data label indexes, and each leakage data label index comprises a leakage point type and a leakage point position;
in specific implementation, the training data samples after normalization processing are input into a deep convolutional neural network model for training; and testing the trained deep convolutional neural network model by using the test data sample after normalization processing, and storing the trained deep convolutional neural network model after the test is successful.
The step of inputting the training data sample after the normalization processing into the deep convolutional neural network model for training specifically includes:
substep S1, inputting the training data samples after the normalization processing into the deep convolutional neural network model in batches for calculation, and obtaining a feedforward calculation result y, y ═ y (1), y (2).. y (f).. y (k) }, wherein y (k) is the probability that the current pipeline data belongs to the kth class predicted by the Softmax classifier; the Softmax classifier is divided into k classes, wherein k represents belonging to the kth class;
it should be noted that the deep convolutional neural network model mainly includes an input layer, an output layer, a convolutional layer, an application layer and a full connection layer.
The construction of the deep convolutional neural network model comprises the following small steps:
small step 1 build up of convolutional layer:
conv is the operation of convolution with the result that,
ulrespectively representing the input and output of the first convolution layer, i represents the ith data;
an nth convolution kernel representing an l-th layer; n is the number of convolution kernels, and bn represents the bias of the nth convolution kernel; f () is an activation function.
Small step 2 convolution window:
with the size of the convolution operation input and output windows varying as
Where W is the input length/width (the input length/width is not necessarily equal), F is the size of the convolution kernel, P is the number of edge complements, and S is the step size (number of convolution intervals).
And 3, constructing a pooling layer:
formula of pooling layer
ul＝Maxpooling vl
Wherein v islIs the input for the first pooling layer, Maxpooling is the maximum pooling operation, ulIs the output of the pooling layer;
and 4, constructing a Softmax classifier, wherein a cost function calculation formula of the Softmax classifier is as follows:
k is the number to be classified and the number of neurons in the Softmax layer; a iskIndicating that the output of the connectivity layer is also the input to the Softmax classifier, y (k) is the probability that the Softmax classifier predicted this piece of data to belong to the kth class; the classifier finally needs to be classified into K classes, where K denotes belonging to the kth class.
And 5, constructing a full connection layer.
A substep S2, calculating a cross entropy error by the deep convolutional neural network model according to the feedforward calculation result, the actual output result and the label result; in a specific implementation, the error value, i.e. the cross entropy error E, is calculated by the following formula:
u is the desired output, a is the neuron actual output, and n represents the total number of samples;
substep S3, updating weight bias according to the cross entropy error; in specific implementation, the weight and bias of each neuron are derived from the cross entropy error according to a gradient descent method, and the weight and bias are updated in the direction of negative gradient:
wherein eta is the learning rate, l is the training frequency,
represents the jth neuron weight of the ith neuron matrix in the ith training,
representing the shared bias of the ith neuron matrix in the ith training;
correspondingly, the step of testing the trained deep convolutional neural network model by using the test data sample after the normalization processing, and storing the trained deep convolutional neural network model after the test is successful specifically comprises:
substep S4, performing iterative training for n times according to the substeps S1 to S3, and inputting normalized test data samples after each training is completed to obtain an output result; comparing the output result with a label corresponding to the test data sample to obtain an accuracy rate; if the accuracy meets the requirement, storing the structural parameters, the weight matrix and the bias parameters of the deep convolutional neural network model after training is finished; otherwise, adjusting the parameters of the deep convolutional neural network model or remanufacturing the sample and continuing training until the requirements are met.
Step S40: real-time pipeline data acquired by intelligent pipe fittings in each position area in a water service pipe network water supply system are acquired in real time, the real-time pipeline data are preprocessed, the preprocessed real-time pipeline data are sent into the trained deep convolutional neural network model, and a prediction result is obtained through the trained deep convolutional neural network model;
in a specific implementation, the step is used for carrying out normalization processing on the real-time pipeline data;
sending the real-time pipeline data after normalization processing into the trained deep convolutional neural network model for calculation to obtain a result sequence, and taking the result sequence as a prediction result;
and analyzing the result sequence, finding out the value 1 assigned to the bit with the largest value in the result sequence, and assigning 0 to other bits to obtain a new result sequence.
Step S50: and comparing the prediction result with the leakage data label index of the corresponding leakage point type to judge the position of the leakage point.
In a specific implementation, in this step, a target mapping corresponding to the new result sequence is first searched from the missing data tag index; then selecting the target mapping; and finally, judging the type and the position of the leakage point of the current pipeline data according to the type of the leakage point contained in the target mapping.
In the embodiment, a deep convolutional neural network model is established and trained in advance according to pipeline data acquired by intelligent pipe fittings in each position area in a water service pipe network water supply system, the trained deep convolutional neural network model comprises a plurality of leakage data label indexes, and each leakage data label index comprises a leakage point type and a leakage point position; then real-time pipeline data acquired by intelligent pipe fittings in each position area in the water service pipe network water supply system are acquired in real time, the real-time pipeline data are preprocessed, the preprocessed real-time pipeline data are sent into a trained deep convolution neural network model, and a prediction result is obtained through the deep convolution neural network model; and finally, comparing the prediction result with the leakage data label index of the corresponding leakage point type to judge the position of the leakage point. And then can be fast and accurately carry out the leakage point location to the water utilities pipeline in the water utilities pipe network water supply system.
In addition, in order to achieve the above object, the present invention further provides a computer readable storage medium, where a water service pipe leakage detection program based on a deep convolutional neural network is stored, and when executed by a processor, the water service pipe leakage detection program based on the deep convolutional neural network implements the steps of the water service pipe leakage detection method based on the deep convolutional neural network as described above.
It should be noted that, in this document, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or system that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or system. Without further limitation, an element defined by the phrase "comprising an … …" does not exclude the presence of other like elements in a process, method, article, or system that comprises the element.
The above-mentioned serial numbers of the embodiments of the present invention are merely for description and do not represent the merits of the embodiments.
Through the above description of the embodiments, those skilled in the art will clearly understand that the method of the above embodiments can be implemented by software plus a necessary general hardware platform, and certainly can also be implemented by hardware, but in many cases, the former is a better implementation manner. Based on such understanding, the technical solutions of the present invention may be embodied in the form of a software product, which is stored in a storage medium (such as ROM/RAM, magnetic disk, optical disk) and includes instructions for enabling a terminal device (such as a mobile phone, a computer, a server, an air conditioner, or a network device) to execute the method according to the embodiments of the present invention.
The above description is only a preferred embodiment of the present invention, and not intended to limit the scope of the present invention, and all modifications of equivalent structures and equivalent processes, which are made by using the contents of the present specification and the accompanying drawings, or directly or indirectly applied to other related technical fields, are included in the scope of the present invention.
Claims (10)
1. A water service pipeline leakage detection method based on a deep convolutional neural network is characterized by comprising the following steps:
acquiring pipeline data acquired by intelligent pipes in each position area in a water service pipe network water supply system in advance, wherein the intelligent pipes are correspondingly arranged on water service pipes in different position areas in the urban water service pipe network;
preprocessing pre-collected pipeline data;
inputting the preprocessed pipeline data into a deep convolutional neural network model for training so as to obtain a trained deep convolutional neural network model, wherein the trained deep convolutional neural network model comprises a plurality of leakage data label indexes, and each leakage data label index comprises a leakage point type and a leakage point position;
real-time pipeline data acquired by intelligent pipe fittings in each position area in a water service pipe network water supply system are acquired in real time, the real-time pipeline data are preprocessed, the preprocessed real-time pipeline data are sent into the trained deep convolutional neural network model, and a prediction result is obtained through the trained deep convolutional neural network model;
and comparing the prediction result with the leakage data label index of the corresponding leakage point type to judge the position of the leakage point.
2. The method of claim 1, wherein the step of preprocessing the pre-collected pipeline data comprises:
dividing pre-collected pipeline data into training data samples and testing data samples;
respectively carrying out normalization processing on the training data samples and the test data samples;
correspondingly, the step of inputting the preprocessed pipeline data into the deep convolutional neural network model for training to obtain the trained deep convolutional neural network model specifically includes:
inputting the training data sample after normalization processing into a deep convolution neural network model for training;
and testing the trained deep convolutional neural network model by using the test data sample after normalization processing, and storing the trained deep convolutional neural network model after the test is successful.
3. The method of claim 2, wherein the pipeline data includes flow data, temperature data, and pressure data;
correspondingly, the step of dividing the pre-collected pipeline data into training data samples and testing data samples comprises:
dividing the pre-collected pipeline data according to types, labeling each piece of pipeline data with different labels according to different types, and dividing the labeled pipeline data into training data samples and testing data samples;
the type of the pipeline data comprises normal data and leakage data, and the format of one piece of pipeline data acquired in real time each time is { a1 a2 … aj; b1 b2 … bj; c1 c2 … cj; j represents the pipeline data collected by the jth intelligent pipe fitting; a represents flow data, b represents temperature data, and c represents pressure data;
and constructing leakage data label indexes based on the labels of the pipeline data of which the types are leakage data, wherein each leakage data label index comprises a leakage point type and a leakage point position.
4. The method of claim 2, wherein the step of normalizing the training data samples and the test data samples separately comprises:
randomly selecting the same number of pieces of data after different types of pipeline data are disordered to construct a training data sample;
scrambling each group of the selected data with the same number in each unselected pipeline data to construct a test data sample;
respectively carrying out normalization processing on the training data samples and the test data samples through the following formulas:
wherein x is a normalization processing result, x is each numerical value of each piece of data in the training data sample or the test data sample, and xmaxIs the largest value in the sample, xminIs the smallest value in the sample.
5. The method according to claim 2, wherein the step of inputting the training data samples after the normalization processing into the deep convolutional neural network model for training specifically comprises:
substep S1, inputting the training data samples after the normalization processing into the deep convolutional neural network model in batches for calculation, and obtaining a feedforward calculation result y, y ═ y (1), y (2).. y (f).. y (k) }, wherein y (k) is the probability that the current pipeline data belongs to the kth class predicted by the Softmax classifier; the Softmax classifier is divided into k classes, wherein k represents belonging to the kth class;
a substep S2, calculating a cross entropy error by the deep convolutional neural network model according to the feedforward calculation result, the actual output result and the label result;
substep S3, updating weight bias according to the cross entropy error;
correspondingly, the step of testing the trained deep convolutional neural network model by using the test data sample after the normalization processing, and storing the trained deep convolutional neural network model after the test is successful specifically comprises:
substep S4, performing iterative training for n times according to the substeps S1 to S3, and inputting normalized test data samples after each training is completed to obtain an output result; comparing the output result with a label corresponding to the test data sample to obtain an accuracy rate; if the accuracy meets the requirement, storing the structural parameters, the weight matrix and the bias parameters of the deep convolutional neural network model after training is finished; otherwise, adjusting the parameters of the deep convolutional neural network model or remanufacturing the sample and continuing training until the requirements are met.
6. The method of claim 2, wherein the pre-processing of the real-time pipeline data is by normalizing the real-time pipeline data.
7. The method of claim 6, wherein the step of sending the pre-processed real-time pipeline data into the trained deep convolutional neural network model and obtaining the prediction result through the trained deep convolutional neural network model specifically comprises:
sending the real-time pipeline data after normalization processing into the trained deep convolutional neural network model for calculation to obtain a result sequence, and taking the result sequence as a prediction result;
and analyzing the result sequence, finding out the value 1 assigned to the bit with the largest value in the result sequence, and assigning 0 to other bits to obtain a new result sequence.
8. The method according to claim 7, wherein the step of comparing the prediction result with the leakage data tag index to determine the number and location of leakage points comprises:
searching a target mapping corresponding to the new result sequence from the leakage data tag index;
selecting the target mapping;
and judging the type and the position of the leakage point of the current pipeline data according to the type of the leakage point contained in the target mapping.
9. A water service pipeline leakage detection system based on a deep convolutional neural network is characterized in that the system is applied to an urban water service pipe network, and intelligent pipe fittings are correspondingly arranged on water service pipelines in different position areas in the urban water service pipe network; the system comprises a data processing center and a water service pipeline leakage detection program based on the deep convolutional neural network, wherein the water service pipeline leakage detection program is stored on a memory and can run on the data processing center, the data processing center is used for receiving pipeline data transmitted by each intelligent pipe fitting, and the water service pipeline leakage detection program based on the deep convolutional neural network is configured to realize the steps of the water service pipeline leakage detection method based on the deep convolutional neural network according to any one of claims 1 to 8.
10. A computer-readable storage medium, wherein the computer-readable storage medium stores thereon a deep convolutional neural network-based water service pipe leakage detection program, and when the deep convolutional neural network-based water service pipe leakage detection program is executed by a processor, the steps of the deep convolutional neural network-based water service pipe leakage detection method according to any one of claims 1 to 8 are implemented.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010770571.9A CN111881999A (en) | 2020-08-04 | 2020-08-04 | Water service pipeline leakage detection method and system based on deep convolutional neural network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010770571.9A CN111881999A (en) | 2020-08-04 | 2020-08-04 | Water service pipeline leakage detection method and system based on deep convolutional neural network |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN111881999A true CN111881999A (en) | 2020-11-03 |
Family
ID=73210143
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010770571.9A Pending CN111881999A (en) | 2020-08-04 | 2020-08-04 | Water service pipeline leakage detection method and system based on deep convolutional neural network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111881999A (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112446115A (en) * | 2020-12-10 | 2021-03-05 | 大唐东北电力试验研究院有限公司 | Furnace water-cooled wall abnormity identification algorithm based on deep convolutional neural network |
| CN113112127A (en) * | 2021-03-22 | 2021-07-13 | 浙江和达科技股份有限公司 | Metering management method and system based on artificial intelligence |
| CN113112126A (en) * | 2021-03-22 | 2021-07-13 | 浙江和达科技股份有限公司 | Water supply pipe leakage early warning method and system based on artificial intelligence |
| CN113434989A (en) * | 2021-06-28 | 2021-09-24 | 山东大学 | Pipe network leakage amount prediction method and system based on attention mechanism and LSTM |
| CN113588179A (en) * | 2021-06-24 | 2021-11-02 | 武汉众智鸿图科技有限公司 | Water supply pipe network leakage detection method and system |
| CN116624786A (en) * | 2023-05-15 | 2023-08-22 | 西安航天自动化股份有限公司 | Pipeline leakage prediction method and storage medium thereof |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20180232601A1 (en) * | 2017-02-16 | 2018-08-16 | Mitsubishi Electric Research Laboratories, Inc. | Deep Active Learning Method for Civil Infrastructure Defect Detection |
| CN109242049A (en) * | 2018-11-21 | 2019-01-18 | 安徽建筑大学 | Water supply pipe network multipoint leakage positioning method and device based on convolutional neural network |
| CN109555979A (en) * | 2018-12-10 | 2019-04-02 | 清华大学 | A kind of water supply network leakage monitoring method |
| CN109578816A (en) * | 2018-12-14 | 2019-04-05 | 中冶京诚工程技术有限公司 | Water supply pipe network pipe burst detection method and device, and control method and device |
| CN110440148A (en) * | 2019-07-15 | 2019-11-12 | 清华大学 | A kind of leakage loss acoustical signal classifying identification method, apparatus and system |
-
2020
- 2020-08-04 CN CN202010770571.9A patent/CN111881999A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20180232601A1 (en) * | 2017-02-16 | 2018-08-16 | Mitsubishi Electric Research Laboratories, Inc. | Deep Active Learning Method for Civil Infrastructure Defect Detection |
| CN109242049A (en) * | 2018-11-21 | 2019-01-18 | 安徽建筑大学 | Water supply pipe network multipoint leakage positioning method and device based on convolutional neural network |
| CN109555979A (en) * | 2018-12-10 | 2019-04-02 | 清华大学 | A kind of water supply network leakage monitoring method |
| CN109578816A (en) * | 2018-12-14 | 2019-04-05 | 中冶京诚工程技术有限公司 | Water supply pipe network pipe burst detection method and device, and control method and device |
| CN110440148A (en) * | 2019-07-15 | 2019-11-12 | 清华大学 | A kind of leakage loss acoustical signal classifying identification method, apparatus and system |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112446115A (en) * | 2020-12-10 | 2021-03-05 | 大唐东北电力试验研究院有限公司 | Furnace water-cooled wall abnormity identification algorithm based on deep convolutional neural network |
| CN113112127A (en) * | 2021-03-22 | 2021-07-13 | 浙江和达科技股份有限公司 | Metering management method and system based on artificial intelligence |
| CN113112126A (en) * | 2021-03-22 | 2021-07-13 | 浙江和达科技股份有限公司 | Water supply pipe leakage early warning method and system based on artificial intelligence |
| CN113112126B (en) * | 2021-03-22 | 2024-05-03 | 浙江和达科技股份有限公司 | Artificial intelligence-based water supply pipe leakage early warning method and system |
| CN113588179A (en) * | 2021-06-24 | 2021-11-02 | 武汉众智鸿图科技有限公司 | Water supply pipe network leakage detection method and system |
| CN113588179B (en) * | 2021-06-24 | 2023-11-21 | 武汉众智鸿图科技有限公司 | Water supply network leakage detection method and system |
| CN113434989A (en) * | 2021-06-28 | 2021-09-24 | 山东大学 | Pipe network leakage amount prediction method and system based on attention mechanism and LSTM |
| CN116624786A (en) * | 2023-05-15 | 2023-08-22 | 西安航天自动化股份有限公司 | Pipeline leakage prediction method and storage medium thereof |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111881999A (en) | Water service pipeline leakage detection method and system based on deep convolutional neural network | |
| US11669990B2 (en) | Object area measurement method, electronic device and storage medium | |
| US20230140289A1 (en) | Traffic accident prediction systems and methods | |
| CN106471851B (en) | Learning model based device positioning | |
| US11886779B2 (en) | Accelerated simulation setup process using prior knowledge extraction for problem matching | |
| CN112036169B (en) | Event recognition model optimization method, device, equipment and readable storage medium | |
| CN112465001B (en) | Classification method and device based on logistic regression | |
| CN112906375A (en) | Text data labeling method, device, equipment and storage medium | |
| EP3686812A1 (en) | System and method for context-based training of a machine learning model | |
| CN117708720B (en) | Equipment fault diagnosis system based on knowledge graph | |
| CN110599200A (en) | Detection method, system, medium and device for false address of OTA hotel | |
| CN112966957A (en) | Data link abnormity positioning method and device, electronic equipment and storage medium | |
| CN112052670A (en) | Address text word segmentation method and device, computer equipment and storage medium | |
| CN111782973A (en) | Point-of-interest state prediction method, device, electronic device and storage medium | |
| CN113807692B (en) | Quality detection method, device and equipment for logistics network points and storage medium | |
| CN108711150B (en) | End-to-end pavement crack detection and identification method based on PCA | |
| CN115482436B (en) | Image screening model training method, device and image screening method | |
| CN117521313A (en) | Method and apparatus for leak detection of pipeline networks | |
| Wang et al. | Software defect fault intelligent location and identification method based on data mining | |
| CN117689868A (en) | Target detection method, device and computer readable storage medium | |
| CN118333356B (en) | Service work order sorting finishing system and method | |
| Adha et al. | Automated building classification framework using convolutional neural network | |
| US20240037424A1 (en) | Methods and internet of things (iot) systems for managing operation progresses of smart gas pipeline networks | |
| CN109725121B (en) | Method and system for generating sampling task | |
| Snider et al. | A flexible, leak crew focused localization model using a maximum coverage search area algorithm |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| WD01 | Invention patent application deemed withdrawn after publication | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20201103 |