CN111914732A - Insulation fault diagnosis method, device, equipment and computer readable storage medium - Google Patents
Insulation fault diagnosis method, device, equipment and computer readable storage medium Download PDFInfo
- Publication number
- CN111914732A CN111914732A CN202010738600.3A CN202010738600A CN111914732A CN 111914732 A CN111914732 A CN 111914732A CN 202010738600 A CN202010738600 A CN 202010738600A CN 111914732 A CN111914732 A CN 111914732A
- Authority
- CN
- China
- Prior art keywords
- fault
- insulation fault
- target
- neural network
- insulation
- 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
- 238000009413 insulation Methods 0.000 title claims abstract description 186
- 238000003745 diagnosis Methods 0.000 title claims abstract description 81
- 238000000034 method Methods 0.000 title claims abstract description 66
- 238000013528 artificial neural network Methods 0.000 claims abstract description 91
- 238000000605 extraction Methods 0.000 claims abstract description 23
- 238000012549 training Methods 0.000 claims abstract description 19
- 238000000354 decomposition reaction Methods 0.000 claims description 15
- 238000004088 simulation Methods 0.000 claims description 15
- 238000012795 verification Methods 0.000 claims description 13
- 230000015654 memory Effects 0.000 claims description 12
- 239000000126 substance Substances 0.000 claims 1
- 230000008569 process Effects 0.000 abstract description 15
- 238000010586 diagram Methods 0.000 description 11
- 230000006870 function Effects 0.000 description 7
- 230000002068 genetic effect Effects 0.000 description 6
- 238000004364 calculation method Methods 0.000 description 5
- 238000012545 processing Methods 0.000 description 5
- 230000008901 benefit Effects 0.000 description 4
- 238000004422 calculation algorithm Methods 0.000 description 4
- 230000001172 regenerating effect Effects 0.000 description 4
- 230000002457 bidirectional effect Effects 0.000 description 3
- 238000004891 communication Methods 0.000 description 3
- 230000008878 coupling Effects 0.000 description 3
- 238000010168 coupling process Methods 0.000 description 3
- 238000005859 coupling reaction Methods 0.000 description 3
- 230000000694 effects Effects 0.000 description 3
- 238000011156 evaluation Methods 0.000 description 3
- 210000004205 output neuron Anatomy 0.000 description 3
- 230000002035 prolonged effect Effects 0.000 description 3
- 230000005540 biological transmission Effects 0.000 description 2
- 238000004134 energy conservation Methods 0.000 description 2
- 230000007613 environmental effect Effects 0.000 description 2
- 238000007667 floating Methods 0.000 description 2
- 230000035772 mutation Effects 0.000 description 2
- 238000005457 optimization Methods 0.000 description 2
- 230000006978 adaptation Effects 0.000 description 1
- 230000032683 aging Effects 0.000 description 1
- 239000003990 capacitor Substances 0.000 description 1
- 230000015556 catabolic process Effects 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 238000005265 energy consumption Methods 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 230000002427 irreversible effect Effects 0.000 description 1
- 239000004973 liquid crystal related substance Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 230000000644 propagated effect Effects 0.000 description 1
- 230000002441 reversible effect Effects 0.000 description 1
- 239000002699 waste material Substances 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/12—Classification; Matching
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR 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
- G06N3/086—Learning methods using evolutionary algorithms, e.g. genetic algorithms or genetic programming
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/08—Feature extraction
- G06F2218/10—Feature extraction by analysing the shape of a waveform, e.g. extracting parameters relating to peaks
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- Computing Systems (AREA)
- Software Systems (AREA)
- Mathematical Physics (AREA)
- Molecular Biology (AREA)
- General Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physiology (AREA)
- Evolutionary Biology (AREA)
- Signal Processing (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Testing Of Short-Circuits, Discontinuities, Leakage, Or Incorrect Line Connections (AREA)
Abstract
The application provides an insulation fault diagnosis method, an insulation fault diagnosis device, an insulation fault diagnosis equipment and a computer readable storage medium. The method comprises the steps of obtaining an alternating current waveform of a device to be detected; carrying out feature extraction on the alternating current waveform to obtain a target fault feature quantity; inputting the target fault characteristic quantity into a preset neural network, wherein the preset neural network is obtained through fault characteristic quantity and insulation fault type training; the target insulation fault type output by the preset neural network is obtained, so that the current waveform can be collected in the normal operation process of the device to be detected, the accurate target insulation fault type is obtained according to the input target fault characteristic quantity, fault diagnosis is simple to realize, and the accuracy of a diagnosis result is improved.
Description
Technical Field
The present application relates to the field of fault detection, and in particular, to an insulation fault diagnosis method, apparatus, device, and computer-readable storage medium.
Background
The urban rail transit has the advantages of safety, comfort, large passenger capacity, high running speed, energy conservation, environmental protection and the like, so that the urban rail transit becomes the first choice for people to go out daily, and along with the wide use of the urban rail transit, the important points are to improve the safety of urban rail transit vehicles and reduce the energy consumption of the urban rail transit vehicles.
In order to improve the energy utilization rate, the train adopts an energy feedback power supply device to supply power for the whole train, thereby realizing the bidirectional flow of energy and achieving the effect of feedback and reuse of regenerative braking energy of the train. However, the power of the power supply device is high, and insulation failure is easy to occur, so that potential safety hazard is caused. In the prior art, whether insulation faults occur in devices in the energy feeding and supplying device is generally judged by detecting whether current exists between insulation devices.
However, in the prior art, the insulation fault diagnosis is complex to realize, and the accuracy of the diagnosis result is low.
Disclosure of Invention
The application provides an insulation fault diagnosis method, device, equipment and a computer readable storage medium, thereby solving the technical problems of complex implementation of insulation fault diagnosis and low accuracy of diagnosis results in the prior art.
In a first aspect, the present application provides an insulation fault diagnosis method, including:
acquiring an alternating current waveform of a device to be detected;
carrying out feature extraction on the alternating current waveform to obtain a target fault feature quantity;
inputting the target fault characteristic quantity into a preset neural network, wherein the preset neural network is obtained through fault characteristic quantity and insulation fault type training;
and acquiring the target insulation fault type output by the preset neural network.
Here, the embodiment of the present application can collect a current waveform in the normal operation process of the device to be detected, so that a target fault characteristic quantity obtained by performing characteristic extraction on the current waveform is input to a trained preset neural network based on the collected current waveform to obtain a target insulation fault type, and whether the device to be detected has a fault and a fault type can be determined, and the fault can be found in time, so that the fault can be processed, and the operation life of the device to be detected is prolonged, and further, since the neural network is obtained by training the fault characteristic quantity and the insulation fault type, an accurate target insulation fault type can be obtained for the input target fault characteristic quantity, so that the accuracy of a diagnosis result is improved, in addition, the embodiment of the present application can determine the insulation fault of the device only by obtaining the alternating current waveform of the device, and the fault diagnosis process is simple, is suitable for application.
Optionally, the performing feature extraction on the alternating current waveform to obtain a target fault feature quantity includes:
decomposing the alternating current waveform based on three-layer wavelet packet decomposition;
and calculating wavelet packet norm entropy of the decomposed waveform, and obtaining the target fault characteristic quantity according to the norm entropy.
Here, the embodiment of the application adopts a wavelet packet decomposition and norm entropy calculation mode to preprocess an alternating current waveform, so as to extract fault characteristic information, wherein the wavelet packet processing can perform fine analysis on a signal, and perform fine decomposition on a high-frequency part and a low-frequency part of the signal, so that accurate signal characteristics can be extracted, and then, an accurate target fault characteristic quantity is obtained, and the accuracy of a diagnosis result is further improved.
Optionally, the neural network includes an input layer, an output layer, and a hidden layer, where the number of nodes in the input layer is determined by the number of the fault feature quantities input at a single time, and the number of nodes in the output layer is determined by the number of the insulation fault types.
Here, the neural network may be determined according to actual conditions, and the embodiment of the present application is not particularly limited thereto.
Optionally, before the inputting the target fault characteristic quantity into a preset neural network, the method further includes:
transmitting the fault characteristic quantity from the input layer and the hidden layer to the output layer to obtain an insulation fault type corresponding to the fault characteristic quantity;
determining an error according to the obtained insulation fault type and the expected insulation fault type;
if the error meets a preset condition, judging that the training of the preset neural network is finished;
if the error does not meet the preset condition, transmitting the error from the output layer and the hidden layer to the input layer, and re-executing the step of transmitting the fault characteristic quantity from the input layer and the hidden layer to the output layer.
Before the target fault characteristic quantity is input into a preset neural network, the neural network is trained, the fault characteristic quantity is input into the neural network for forward propagation, the error between the type of the insulation fault output from the neural network and the type of the expected insulation fault is obtained after the forward propagation, then the weight threshold value of the neural network is corrected by the backward propagation in the neural network, and the operation is circulated for many times until the conditions are met, so that the trained accurate neural network can be obtained, and the accuracy of the diagnosis result is further improved.
Optionally, after the obtaining of the target insulation fault type output by the preset neural network, the method further includes:
carrying out simulation verification on the target insulation fault type;
and determining the diagnosis accuracy of the target insulation fault type according to the simulation verification result.
After the diagnosis result is output, the target insulation fault type of the diagnosis result is subjected to simulation verification to obtain the diagnosis accuracy, the insulation fault condition of the equipment to be detected can be better mastered, and meanwhile, the insulation diagnosis method of the embodiment of the application can be adjusted and optimized, so that the accuracy of the diagnosis result is further improved.
Optionally, the target insulation fault category includes:
one or more of a single phase-to-ground insulation fault, a phase-to-phase insulation fault, a two phase-to-ground insulation fault, and a three phase-to-ground insulation fault.
Here, the target insulation fault type may be determined according to actual conditions, and this is not particularly limited by the embodiment of the present application.
Optionally, the alternating current waveform is a three-phase alternating current waveform.
In a second aspect, an embodiment of the present application provides an insulation fault diagnosis apparatus, including:
the first acquisition module is used for acquiring the alternating current waveform of the device to be detected;
the extraction module is used for carrying out feature extraction on the alternating current waveform to obtain target fault feature quantity;
the input module is used for inputting the target fault characteristic quantity into a preset neural network, wherein the preset neural network is obtained through fault characteristic quantity and insulation fault type training;
and the second acquisition module is used for acquiring the target insulation fault type output by the preset neural network.
Optionally, the extraction module is specifically configured to:
decomposing the alternating current waveform based on three-layer wavelet packet decomposition;
and calculating wavelet packet norm entropy of the decomposed waveform, and obtaining the target fault characteristic quantity according to the norm entropy.
Optionally, the neural network includes an input layer, an output layer, and a hidden layer, where the number of nodes in the input layer is determined by the number of the fault feature quantities input at a single time, and the number of nodes in the output layer is determined by the number of the insulation fault types.
Optionally, the apparatus further comprises:
the training module is used for transmitting the fault characteristic quantity from the input layer and the hidden layer to the output layer before the target fault characteristic quantity is input into a preset neural network by the input module, so that an insulation fault type corresponding to the fault characteristic quantity is obtained;
determining an error according to the obtained insulation fault type and the expected insulation fault type;
if the error meets a preset condition, judging that the training of the preset neural network is finished;
if the error does not meet the preset condition, transmitting the error from the output layer and the hidden layer to the input layer, and re-executing the step of transmitting the fault characteristic quantity from the input layer and the hidden layer to the output layer.
Optionally, the apparatus further comprises:
and the simulation module is used for performing simulation verification on the target insulation fault type after the target insulation fault type output by the preset neural network is obtained, and determining the diagnosis accuracy.
Optionally, the target insulation fault category includes:
single phase-to-ground insulation faults, phase-to-phase insulation faults, two phase-to-ground insulation faults, and three phase-to-ground insulation faults.
Optionally, the alternating current waveform is a three-phase alternating current waveform.
In a third aspect, an embodiment of the present application provides an insulation fault diagnosis apparatus, including: at least one processor; and a memory communicatively coupled to the at least one processor; wherein the memory stores instructions executable by the at least one processor to enable the at least one processor to perform the insulation fault diagnosis method of the first aspect or the alternatives of the first aspect.
In a fourth aspect, embodiments of the present application provide a computer-readable storage medium, in which computer-executable instructions are stored, and when the computer-executable instructions are executed by a processor, the computer-readable storage medium is configured to implement the insulation fault diagnosis method according to the first aspect or the alternatives of the first aspect.
The method, the device, the equipment and the computer readable storage medium for diagnosing the insulation fault, which are provided by the embodiment of the application, collect a current waveform in the normal operation process of a device to be detected, input a target fault characteristic quantity obtained after characteristic extraction is carried out on the current waveform to a trained preset neural network based on the collected current waveform to obtain a target insulation fault type, determine whether the device to be detected has a fault and a fault type, find the fault in time, further process the fault, prolong the service life of the device to be detected, and obtain an accurate target insulation fault type according to the input target fault characteristic quantity and the insulation fault type due to the fact that the neural network is obtained through the training of the fault characteristic quantity and the insulation fault type, so that the accuracy of a diagnosis result is improved, in addition, the embodiment of the application can determine the insulation fault of the device only by obtaining the alternating current waveform of the device, the fault diagnosis process is simple and is suitable for application.
Drawings
The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present disclosure and together with the description, serve to explain the principles of the disclosure.
Fig. 1 is a schematic diagram of an insulation fault diagnosis and evaluation system according to an embodiment of the present disclosure
Fig. 2 is a flowchart of an insulation fault diagnosis method according to an embodiment of the present application;
fig. 3 is an exploded schematic view of a three-layer wavelet packet according to an embodiment of the present application;
fig. 4 is a schematic circuit diagram of a device to be tested according to an embodiment of the present disclosure;
fig. 5 is a schematic flowchart of another insulation fault diagnosis method provided in an embodiment of the present application;
fig. 6 is a schematic structural diagram of a neural network provided in an embodiment of the present application;
fig. 7 is a schematic flowchart of another insulation fault diagnosis method according to an embodiment of the present application;
fig. 8 is a schematic structural diagram of an insulation fault diagnosis apparatus according to an embodiment of the present application;
fig. 9 is a schematic structural diagram of an insulation fault diagnosis apparatus according to an embodiment of the present application.
With the foregoing drawings in mind, certain embodiments of the disclosure have been shown and described in more detail below. These drawings and written description are not intended to limit the scope of the disclosed concepts in any way, but rather to illustrate the concepts of the disclosure to those skilled in the art by reference to specific embodiments.
Detailed Description
Reference will now be made in detail to the exemplary embodiments, examples of which are illustrated in the accompanying drawings. When the following description refers to the accompanying drawings, like numbers in different drawings represent the same or similar elements unless otherwise indicated. The implementations described in the exemplary embodiments below are not intended to represent all implementations consistent with the present disclosure. Rather, they are merely examples of apparatus and methods consistent with certain aspects of the present disclosure, as detailed in the appended claims.
The terms "first," "second," "third," and "fourth," if any, in the description and claims of this application and the above-described figures are used for distinguishing between similar elements and not necessarily for describing a particular sequential or chronological order. It is to be understood that the data so used is interchangeable under appropriate circumstances such that the embodiments of the application described herein are capable of operation in sequences other than those illustrated or described herein. Furthermore, the terms "comprises," "comprising," and "having," and any variations thereof, are intended to cover a non-exclusive inclusion, such that a process, method, system, article, or apparatus that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed, but may include other steps or elements not expressly listed or inherent to such process, method, article, or apparatus
Urban rail transit is a preferred scheme for solving the problem of increasingly serious urban congestion due to the advantages of safety, comfort, large passenger capacity, high running speed, energy conservation, environmental protection and the like, so that urban rail transit is developed and constructed in many places at present. In most traction power supply systems at present, a diode rectifier is adopted as a power supply device, energy can only flow in a single direction, regenerative braking energy of a vehicle can only be consumed through a braking resistor, huge waste of energy is caused, the additional equipped braking resistor can also increase cost and vehicle weight, and the problems of tunnel temperature rise and the like are caused. The energy feedback power supply device can realize bidirectional flow of energy, the voltage at the direct current side is adjustable, and the feedback reutilization of the regenerative braking energy of the train can be realized.
In order to improve the energy utilization rate, the train adopts the energy-feedback power supply device to supply power to the whole train, thereby realizing the bidirectional flow of energy and achieving the effect of feedback and reutilization of regenerative braking energy of the train, the energy-feedback device has the advantage of excellent performance because the energy-feedback device has precise devices, but the more precise devices have smaller tolerance to faults, the insulation faults are taken as a common fault type, are particularly easy to occur in the application scene of high voltage and large current, the rated power of the energy-feedback device is very large, the insulation faults are easy to occur, once the energy-feedback device is subjected to insulation aging and even insulation breakdown, the irreversible damage can be caused to the energy-feedback device, and the safety and the service life of the train are influenced, so that the energy-feedback device is necessary for providing a more stable and good operation environment and carrying out real-time monitoring and real-time diagnosis on the insulation faults.
However, in the prior art, whether a device in the feedable power supply device generates an insulation fault is generally judged by detecting whether a current exists between the insulation devices, so that the method is complex to implement, the misjudgment rate of the insulation fault is low, and the accuracy of a diagnosis result is low.
In order to solve the above problems, embodiments of the present application provide an insulation fault diagnosis method, apparatus, device, and computer-readable storage medium, in which a current waveform is collected during normal operation of a device to be detected, so that a target fault characteristic quantity obtained by extracting characteristics of the current waveform is input to a trained preset neural network based on the collected current waveform, a target insulation fault type is obtained, and it is possible to determine whether a fault and a fault type occur in the device to be detected, find a fault in time, process the fault, and improve the operation life of the device to be detected.
Optionally, fig. 1 is a schematic diagram of an insulation fault diagnosis and evaluation system according to an embodiment of the present application. In fig. 1, the above-described architecture includes at least one of a receiving device 101, a processor 102, and a display device 103.
It is to be understood that the illustrated structure of the embodiment of the present application does not constitute a specific limitation to the insulation fault diagnosis evaluation architecture. In other possible embodiments of the present application, the foregoing architecture may include more or less components than those shown in the drawings, or combine some components, or split some components, or arrange different components, which may be determined according to practical application scenarios, and is not limited herein. The components shown in fig. 2 may be implemented in hardware, software, or a combination of software and hardware.
In a specific implementation process, the receiving device 101 may be an input/output interface or a communication interface.
The processor 102 can collect current waveforms in the normal operation process of the device to be detected through the receiving device 101, so that based on the collected current waveforms, target fault characteristic quantities obtained after characteristic extraction is carried out on the current waveforms are input to a trained preset neural network to obtain target insulation fault types, whether the device to be detected has faults or not and fault types can be determined, faults can be found in time, the faults are further processed, and the operation life of the device to be detected is prolonged.
The display device 103 may be used to display the above results and the like.
The display device may also be a touch display screen for receiving user instructions while displaying the above-mentioned content to enable interaction with a user.
It should be understood that the processor may be implemented by reading instructions in the memory and executing the instructions, or may be implemented by a chip circuit.
In addition, the network architecture and the service scenario described in the embodiment of the present application are for more clearly illustrating the technical solution of the embodiment of the present application, and do not constitute a limitation to the technical solution provided in the embodiment of the present application, and it can be known by a person skilled in the art that along with the evolution of the network architecture and the appearance of a new service scenario, the technical solution provided in the embodiment of the present application is also applicable to similar technical problems.
The technical scheme of the application is described in detail by combining specific embodiments as follows:
fig. 2 is a flowchart of an insulation fault diagnosis method according to an embodiment of the present application. The execution subject of the embodiment of the present application may be the processor 102 in fig. 1, and the specific execution subject may be determined according to an actual application scenario. As shown in fig. 2, the method comprises the steps of:
s201: and acquiring the alternating current waveform of the device to be detected.
Here, the device to be detected may be an energy-feeding device in the vehicle for supplying power.
Optionally, the alternating current waveform is a three-phase alternating current waveform.
Optionally, the method for acquiring the ac current waveform may be through a current sensor inside the device to be detected, or through an ac current oscilloscope.
S202: and (4) carrying out feature extraction on the alternating current waveform to obtain a target fault feature quantity.
Optionally, the performing feature extraction on the alternating current waveform to obtain a target fault feature quantity includes: decomposing the alternating current waveform based on three-layer wavelet packet decomposition; and calculating the wavelet packet norm entropy of the decomposed waveform, and obtaining the target fault characteristic quantity according to the norm entropy.
Fig. 3 is a schematic diagram of a three-layer wavelet packet decomposition provided in an embodiment of the present application, where S denotes a signal to be decomposed, i.e., an alternating current waveform, a denotes a low-frequency component, and D denotes a high-frequency component. As shown in FIG. 3, each layer of wavelet packet decomposition decomposes each component of the layer into two wavelets of low frequency and high frequency, and the number of wavelets of the j-th layer is 2j. After three layers of wavelet packet decomposition, the decomposition has the following relational expression:
S＝A(3,0)+D(3,1)+A(3,2)+D(3,3)+A(3,4)+D(3,5)+A(3,6)+D(3,7)
here, the embodiment of the application adopts a wavelet packet decomposition and norm entropy calculation mode to preprocess an alternating current waveform, so as to extract fault characteristic information, wherein the wavelet packet processing can perform fine analysis on a signal, and perform fine decomposition on a high-frequency part and a low-frequency part of the signal, so that accurate signal characteristics can be extracted, a target fault characteristic quantity is extracted through the norm entropy calculation, a network structure is simplified through the wavelet packet decomposition and norm entropy calculation, an accurate target fault characteristic quantity is obtained, and the accuracy of a diagnosis result is further improved.
S203: and inputting the target fault characteristic quantity into a preset neural network.
The preset neural network is obtained through fault characteristic quantity and insulation fault type training.
Optionally, the neural network includes an input layer, an output layer, and a hidden layer, the number of nodes in the input layer is determined by the number of single-input fault features, and the number of nodes in the output layer is determined by the number of insulation fault types, where the structure of the neural network is determined as the input layer, the output layer, and the hidden layer, the number of nodes in the input layer is determined by the number of single-input fault features, and the number of nodes in the output layer is determined by the number of insulation fault types, so that the fault features obtained in the embodiment of the present application and the insulation fault types to be diagnosed are combined to the input and output of the neural network, which facilitates the diagnosis of the insulation fault types, and improves the accuracy of the diagnosis result.
S204: and acquiring the target insulation fault type output by the preset neural network.
Optionally, the target insulation fault category includes: one or more of a single phase-to-ground insulation fault, a phase-to-phase insulation fault, a two phase-to-ground insulation fault, and a three phase-to-ground insulation fault.
Exemplarily, fig. 4 is a circuit schematic diagram of a device to be detected provided by an embodiment of the present application, as shown in fig. 4, a three-phase circuit is taken as an example, N is power in a power grid, the circuit includes resistors R1, R2, and R3 and inductors L1, L2, and L3, each resistor is connected in series with an inductor and is input to a direct current side through nodes A, B and C, the direct current side of the circuit includes switching tubes G1, G2, G3, G4, G5, and G6, phase a is connected with a series branch of switching tubes G1 and G2, phase B is connected with a series branch of switching tubes G3 and G4, phase C is connected with a series branch of switching tubes G5 and G6, and an output end further includes a capacitor C1 connected in parallel with the circuit. Optionally, the single phase-to-ground insulation fault includes an a phase-to-ground insulation fault, a B phase-to-ground insulation fault, and a C phase insulation fault, optionally, the phase-to-phase insulation fault includes an AB phase-to-phase insulation fault, an AC phase-to-phase insulation fault, and a BC phase-to-phase insulation fault, optionally, the two phase-to-ground insulation faults include an AB phase-to-ground insulation fault, an AC phase-to-ground insulation fault, and a BC phase-to-ground insulation fault, optionally, the three phase-to-ground insulation faults include an ABC three-phase ground. The target insulation fault types are divided into single phase-to-ground insulation faults, phase-to-phase insulation faults, two phase-to-ground insulation faults and three phase-to-ground insulation faults, and the classification of the target insulation fault types can better analyze the insulation faults, facilitate the analysis of the position and the reason of the insulation faults, and further improve the accuracy of diagnosis results.
The current waveform can be collected in the normal operation process of the device to be detected, so that faults can be timely obtained, the faults can be conveniently processed, the service life of the device to be detected is prolonged, the target fault characteristic quantity obtained after characteristic extraction is carried out on the current waveform after the current waveform is collected is input into a trained preset neural network, the target insulation fault type is obtained, whether the device to be detected breaks down or not can be determined, the fault type can be obtained, accurate output results can be generated for input by the trained neural network, accurate target insulation fault types can be obtained for the input target fault characteristic quantity, fault diagnosis is simple to achieve, and the accuracy of diagnosis results is improved.
Optionally, the neural network includes an input layer, an output layer, and a hidden layer, the number of nodes in the input layer is determined by the number of feature quantities of a single-time input fault, and the number of nodes in the output layer is determined by the number of types of insulation faults. Before the target fault characteristic quantity is input into the preset neural network, the neural network may be trained, fig. 5 is a schematic flow chart of another insulation fault diagnosis method provided in the embodiment of the present application, and as shown in fig. 5, the method includes:
s501: and acquiring the alternating current waveform of the device to be detected.
S502: and (4) carrying out feature extraction on the alternating current waveform to obtain a target fault feature quantity.
The steps S501 and S502 are the same as the steps S201 and S202, and are not described herein again.
S503: and transmitting the fault characteristic quantity from the input layer and the hidden layer to the output layer to obtain the insulation fault type corresponding to the fault characteristic quantity.
S504: and determining an error according to the obtained insulation fault type and the expected insulation fault type.
S505: if the error meets a preset condition, judging that the training of the preset neural network is finished; and if the error does not meet the preset condition, transmitting the error from the output layer and the hidden layer to the input layer, and re-executing the step of transmitting the fault characteristic quantity from the input layer and the hidden layer to the output layer.
Here, the neural network includes an input layer, an output layer, and a hidden layer, the number of nodes in the input layer is determined by the number of single-input fault feature quantities, and the number of nodes in the output layer is determined by the number of types of insulation faults. Exemplarily, the target fault characteristic quantity after three-phase alternating current wavelet packet decomposition and norm entropy processing as shown in fig. 3 is 24, so that the number of input layer nodes is 24.
Optionally, the number of output layer nodes is determined by the number of binary encoding bits of the insulation fault type. If 10 cases are included, the output result needs 4-bit binary number representation, and the judgment bit representing the rectification or inversion state is added, so that 5-bit binary number is needed in total, and the number of the output nodes is 5.
Optionally, the number of nodes of the hidden layer is obtained based on the number of nodes of the input layer and the output layer.
Fig. 6 is a schematic structural diagram of a preset neural network provided in an embodiment of the present application, and as shown in fig. 6, the preset neural network includes an input layer, a hidden layer, and an output layer, a forward flow direction of a signal is transmitted from the input layer and the hidden layer to the output layer, an error is determined according to an obtained insulation fault type and an expected insulation fault type, and then the error is propagated in a reverse direction to adjust a weight threshold of the neural network.
Optionally, the preset condition that the error meets may be that the error is smaller than the maximum preset error, and it can be understood that the maximum preset error may be determined according to an actual situation, and the present application is not particularly limited.
In addition, before the neural network is trained, the weight threshold of the neural network can be optimized through a genetic algorithm, so that the accurate weight threshold can be obtained more easily in the process of training the neural network.
Optionally, the weight threshold of the neural network is optimized through a genetic algorithm in the following manner: when the genetic algorithm is applied, parameters to be optimized, namely weights of the preset neural network in the application, need to be encoded by using floating point numbers, and the length of encoding is the number of the weights to be optimized. The preference of the genetic network is judged based on the fitness, the fitness function adopted by the method is the sum of squares of errors between the fault types obtained by training individuals in the population through the neural network and the actual fault types, and then the reciprocal is taken to obtain the fitness F, and the calculation formula is as follows:
in the above formula, n is the number of output neurons of the neural network, yiIs the desired output of the ith output neuron, oiK is the coefficient for the actual output of the ith output neuron.
The main method for realizing the selection operation is a fitness proportion method, the larger the individual fitness is, the larger the probability of being selected is, and the probability of each individual i being selected is as follows:
fi＝kFi
in the above formula FiIs the fitness of an individual i, k is a coefficient, and n is a population individualAnd (4) counting.
Kth individual a in floating point number interleaving operationkAnd the first individual alWhen the j bit is crossed, the implementation method comprises the following steps:
akj＝akj(1-b)+aljb
alj＝alj(1-b)+akjb
in the above formula, b is a random number taken from [0,1 ].
If the j-th bit of the ith individual is selected to have variation, the specific implementation method is as follows:
aij+(aij-ama)x*f(g)(r≥0.5)
in the above formula amaxAnd aminAre respectively an individual aijUpper and lower limits of (f), (g) r2(1-g/Gmax)，r、r2Are all taken from [0,1]]G is the current iteration number, GmaxIs the maximum number of evolutions.
After each time of selective cross mutation operation, the mutation individuals are screened through a fitness function, individuals with good fitness are reserved, and individuals with poor fitness are eliminated, so that a newly generated group inherits good information of a previous generation, and the operation of optimizing the group is completed. Through repeated population optimization until the new population meets the requirements set in advance, the accurate weight threshold value can be obtained more easily in the process of training the neural network. After the structure of the neural network is determined, an initial weight threshold value of the neural network is randomly generated, then the initial weight threshold value of the neural network is coded by using a genetic algorithm, then the coded parameters are processed by utilizing selection, crossing and variation operations in genetics, and the advantages and disadvantages of the variation individuals are screened by a fitness function, so that the population fitness performance of each generation is better than that of the previous generation. Through the optimization of the secondary population, the new population meets the requirement of advance fitness or reaches the maximum genetic algebra. At this time, the optimal individual in the population is decoded and used as an initial weight threshold value when the neural network is trained. The neural network trained by the method has high diagnostic performance.
S506: and inputting the target fault characteristic quantity into a preset neural network.
S507: and acquiring the target insulation fault type output by the preset neural network.
Steps S506 and S507 are the same as the implementation of steps S203 and S204, and are not described herein again.
According to the method and the device, before the target fault characteristic quantity is input into the preset neural network, the neural network is trained, the fault characteristic quantity is input into the neural network to be transmitted in the forward direction, the error between the type of the insulation fault output from the neural network and the type of the expected insulation fault is obtained after the forward transmission, then the weight threshold value of the neural network is corrected through the backward transmission in the neural network, and the operation is repeated for many times until the condition is met, so that the trained accurate neural network can be obtained, and the accuracy of the diagnosis result is further improved.
In addition, after acquiring a target insulation fault type output by a preset neural network, the embodiment of the present application also considers performing simulation verification on the target insulation fault type, and fig. 7 is a flowchart of another insulation fault diagnosis method provided in the embodiment of the present application, and as shown in fig. 7, the method includes the following steps:
s701: and acquiring the alternating current waveform of the device to be detected.
S702: carrying out feature extraction on the AC waveform to obtain target fault feature quantity
S703: and inputting the target fault characteristic quantity into a preset neural network.
S704: and acquiring the target insulation fault type output by the preset neural network.
The steps S701 to S704 are the same as the steps S201 to S204, and are not described herein again.
S705: carrying out simulation verification on the target insulation fault type; and determining the diagnosis accuracy of the target insulation fault type according to the simulation verification result.
Optionally, the target insulation fault type is subjected to simulation verification, and compared with an expected diagnosis insulation fault type, so as to obtain a diagnosis accuracy.
Optionally, if the diagnosis accuracy is smaller than the preset diagnosis accuracy threshold, the neural network continues to be trained. It is understood that the threshold value of the diagnosis accuracy can be determined according to actual conditions, and the application is not particularly limited.
After the diagnosis result is output, the target insulation fault type of the diagnosis result is subjected to simulation verification to obtain the diagnosis accuracy, the insulation fault condition of the equipment to be detected can be better mastered, and meanwhile, the insulation diagnosis method of the embodiment of the application can be adjusted and optimized, so that the accuracy of the diagnosis result is further improved.
Fig. 8 is a schematic structural diagram of an insulation fault diagnosis apparatus according to an embodiment of the present application, and as shown in fig. 8, the apparatus according to the embodiment of the present application includes: a first acquisition module 801, an extraction module 802, an input module 803, and a second acquisition module 804. The insulation fault diagnosis device here may be the processor 102 itself described above, or a chip or an integrated circuit that realizes the functions of the processor 102. It should be noted here that the division of the first obtaining module 801, the extracting module 802, the inputting module 803, and the second obtaining module 804 is only a division of logic functions, and the two may be integrated or independent physically.
The first obtaining module 801 is configured to obtain an alternating current waveform of a device to be detected;
an extraction module 802, configured to perform feature extraction on an alternating current waveform to obtain a target fault feature quantity;
an input module 803, configured to input the target fault feature quantity into a preset neural network, where the preset neural network is obtained through fault feature quantity and insulation fault category training;
a second obtaining module 804, configured to obtain a target insulation fault type output by the preset neural network.
Optionally, the extraction module 802 is specifically configured to decompose the ac current waveform based on three-layer wavelet packet decomposition;
and calculating the wavelet packet norm entropy of the decomposed waveform, and obtaining the target fault characteristic quantity according to the norm entropy.
Optionally, the neural network includes an input layer, an output layer, and a hidden layer, the number of nodes in the input layer is determined by the number of feature quantities of a single-time input fault, and the number of nodes in the output layer is determined by the number of types of insulation faults.
Optionally, the apparatus further comprises:
the training module 805 is configured to transmit the fault feature quantity from the input layer and the hidden layer to the output layer before the input module inputs the target fault feature quantity into the preset neural network, so as to obtain an insulation fault type corresponding to the fault feature quantity; determining an error according to the obtained insulation fault type and the expected insulation fault type; if the error meets a preset condition, judging that the training of the preset neural network is finished; and if the error does not meet the preset condition, transmitting the error from the output layer and the hidden layer to the input layer, and re-executing the step of transmitting the fault characteristic quantity from the input layer and the hidden layer to the output layer.
Optionally, the apparatus further comprises:
the simulation module 806 is configured to perform simulation verification on a target insulation fault type output by the preset neural network after the target insulation fault type is obtained; and determining the diagnosis accuracy of the target insulation fault type according to the simulation verification result.
Optionally, the target insulation fault category includes:
single phase-to-ground insulation faults, phase-to-phase insulation faults, two phase-to-ground insulation faults, and three phase-to-ground insulation faults.
Optionally, the alternating current waveform is a three-phase alternating current waveform.
Fig. 9 is a schematic structural diagram of an insulation fault diagnosis apparatus according to an embodiment of the present application. The components shown herein, their connections and relationships, and their functions, are meant to be exemplary only, and are not limiting to the implementations of the present application described and/or claimed herein.
As shown in fig. 9, the insulation fault diagnosis apparatus includes: a processor 901 and a memory 902, which are connected to each other using different buses, and may be mounted on a common motherboard or in other manners as needed. The processor 901 may process instructions executed within the insulation fault diagnosis device, including instructions of graphical information stored in or on a memory for display on an external input/output apparatus (such as a display device coupled to an interface). In other embodiments, multiple processors and/or multiple buses may be used, along with multiple memories and multiple memories, as desired. Fig. 9 illustrates an example of a processor 901.
The memory 902, which is a non-transitory computer-readable storage medium, may be used to store non-transitory software programs, non-transitory computer-executable programs, and modules, such as program instructions/modules corresponding to the method of the insulation fault diagnosis apparatus in the embodiment of the present application (for example, the first acquisition module 801, the extraction module 802, and the input module 803 shown in fig. 8). The processor 901 executes various functional applications of the server and data processing, i.e., a method of implementing the insulation fault diagnosis apparatus in the above-described method embodiments, by executing non-transitory software programs, instructions, and modules stored in the memory 902.
The insulation fault diagnosis apparatus may further include: an input device 903 and an output device 904. The processor 901, the memory 902, the input device 903 and the output device 904 may be connected by a bus or other means, and fig. 9 illustrates the connection by a bus as an example.
The input device 903 may receive input numeric or character information and generate key signal inputs related to user settings and function control of the insulation fault diagnosis apparatus, such as a touch screen, a keypad, a mouse, or a plurality of mouse buttons, a trackball, a joystick, and the like. The output device 904 may be an output apparatus such as a display apparatus of the insulation fault diagnosis apparatus. The display device may include, but is not limited to, a Liquid Crystal Display (LCD), a Light Emitting Diode (LED) display, and a plasma display. In some implementations, the display device can be a touch screen.
The insulation fault diagnosis device according to the embodiment of the present application may be configured to execute the technical solutions in the method embodiments of the present application, and the implementation principle and the technical effect are similar, which are not described herein again.
An embodiment of the present application further provides a computer-readable storage medium, in which computer-executable instructions are stored, and when the computer-executable instructions are executed by a processor, the computer-executable instructions are used to implement any one of the insulation fault diagnosis methods described above.
In the several embodiments provided in the present application, it should be understood that the disclosed system, apparatus and method may be implemented in other manners. For example, the above-described apparatus embodiments are merely illustrative, and for example, the division of the units is only one logical division, and other divisions may be realized in practice, for example, a plurality of units or components may be combined or integrated into another system, or some features may be omitted, or not executed. In addition, the shown or discussed mutual coupling or direct coupling or communication connection may be an indirect coupling or communication connection through some interfaces, devices or units, and may be in an electrical, mechanical or other form.
In addition, functional units in the embodiments of the present application may be integrated into one processing unit, or each unit may exist alone physically, or two or more units are integrated into one unit. The integrated unit can be realized in a form of hardware, and can also be realized in a form of a software functional unit.
Other embodiments of the disclosure will be apparent to those skilled in the art from consideration of the specification and practice of the disclosure disclosed herein. This application is intended to cover any variations, uses, or adaptations of the disclosure following, in general, the principles of the disclosure and including such departures from the present disclosure as come within known or customary practice within the art to which the disclosure pertains. It is intended that the specification and examples be considered as exemplary only, with a true scope and spirit of the disclosure being indicated by the following claims.
It will be understood that the present disclosure is not limited to the precise arrangements described above and shown in the drawings and that various modifications and changes may be made without departing from the scope thereof. The scope of the present disclosure is limited only by the appended claims.
Claims (10)
1. An insulation fault diagnosis method characterized by comprising:
acquiring an alternating current waveform of a device to be detected;
carrying out feature extraction on the alternating current waveform to obtain a target fault feature quantity;
inputting the target fault characteristic quantity into a preset neural network, wherein the preset neural network is obtained through fault characteristic quantity and insulation fault type training;
and acquiring the target insulation fault type output by the preset neural network.
2. The method of claim 1, wherein the performing feature extraction on the alternating current waveform to obtain a target fault feature quantity comprises:
decomposing the alternating current waveform based on three-layer wavelet packet decomposition;
and calculating wavelet packet norm entropy of the decomposed waveform, and obtaining the target fault characteristic quantity according to the norm entropy.
3. The method according to claim 1, wherein the neural network comprises an input layer, an output layer and an implicit layer, the number of nodes of the input layer is determined by the number of the fault characteristic quantities input at a time, and the number of nodes of the output layer is determined by the number of the insulation fault types.
4. The method according to claim 3, before the inputting the target failure characteristic quantity into a preset neural network, further comprising:
transmitting the fault characteristic quantity from the input layer and the hidden layer to the output layer to obtain an insulation fault type corresponding to the fault characteristic quantity;
determining an error according to the obtained insulation fault type and the expected insulation fault type;
if the error meets a preset condition, judging that the training of the preset neural network is finished;
if the error does not meet the preset condition, transmitting the error from the output layer and the hidden layer to the input layer, and re-executing the step of transmitting the fault characteristic quantity from the input layer and the hidden layer to the output layer.
5. The method of claim 1, further comprising, after said obtaining a target insulation fault category for said preset neural network output:
carrying out simulation verification on the target insulation fault type;
and determining the diagnosis accuracy of the target insulation fault type according to the simulation verification result.
6. The method according to any one of claims 1 to 5, wherein the target insulation fault category comprises one or more of a single phase-to-ground insulation fault, a phase-to-phase insulation fault, a two phase-to-ground insulation fault, and a three phase-to-ground insulation fault.
7. The method of any of claims 1-5, wherein the AC current waveform is a three-phase AC current waveform.
8. An insulation fault diagnosis apparatus characterized by comprising:
the first acquisition module is used for acquiring the alternating current waveform of the device to be detected;
the extraction module is used for carrying out feature extraction on the alternating current waveform to obtain target fault feature quantity;
the input module is used for inputting the target fault characteristic quantity into a preset neural network, wherein the preset neural network is obtained through fault characteristic quantity and insulation fault type training;
and the second acquisition module is used for acquiring the target insulation fault type output by the preset neural network.
9. An insulation fault diagnosis apparatus characterized by comprising:
at least one processor; and
a memory communicatively coupled to the at least one processor; wherein the content of the first and second substances,
the memory stores instructions executable by the at least one processor to enable the at least one processor to perform the method of any one of claims 1 to 7.
10. A computer-readable storage medium having stored therein computer-executable instructions for implementing the insulation fault diagnosis method according to any one of claims 1 to 7 when executed by a processor.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010738600.3A CN111914732A (en) | 2020-07-28 | 2020-07-28 | Insulation fault diagnosis method, device, equipment and computer readable storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010738600.3A CN111914732A (en) | 2020-07-28 | 2020-07-28 | Insulation fault diagnosis method, device, equipment and computer readable storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN111914732A true CN111914732A (en) | 2020-11-10 |
Family
ID=73286443
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010738600.3A Pending CN111914732A (en) | 2020-07-28 | 2020-07-28 | Insulation fault diagnosis method, device, equipment and computer readable storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111914732A (en) |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107153735A (en) * | 2017-04-28 | 2017-09-12 | 无锡开放大学 | Motor driven systems PWM inverter method for diagnosing faults |
| CN108181562A (en) * | 2018-01-18 | 2018-06-19 | 福州大学 | Insulator breakdown diagnostic device and method based on Study On Reliability Estimation Method For Cold Standby Systems |
| CN108279364A (en) * | 2018-01-30 | 2018-07-13 | 福州大学 | Wire selection method for power distribution network single phase earthing failure based on convolutional neural networks |
| US20190242936A1 (en) * | 2018-02-05 | 2019-08-08 | Wuhan University | Fault diagnosis method for series hybrid electric vehicle ac/dc converter |
-
2020
- 2020-07-28 CN CN202010738600.3A patent/CN111914732A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107153735A (en) * | 2017-04-28 | 2017-09-12 | 无锡开放大学 | Motor driven systems PWM inverter method for diagnosing faults |
| CN108181562A (en) * | 2018-01-18 | 2018-06-19 | 福州大学 | Insulator breakdown diagnostic device and method based on Study On Reliability Estimation Method For Cold Standby Systems |
| CN108279364A (en) * | 2018-01-30 | 2018-07-13 | 福州大学 | Wire selection method for power distribution network single phase earthing failure based on convolutional neural networks |
| US20190242936A1 (en) * | 2018-02-05 | 2019-08-08 | Wuhan University | Fault diagnosis method for series hybrid electric vehicle ac/dc converter |
Non-Patent Citations (2)
| Title |
|---|
| 曹会阳: "《能馈供电装置故障诊断技术研究》", 《优秀硕士论文》 * |
| 郑宏兴等编著，华中科技大学出版社: "《小波包分解》", 《MATLAB 5.X工具箱使用技巧与实例》, 31 December 2001 (2001-12-31), pages 151 - 155 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102496069B (en) | Cable multimode safe operation evaluation method based on fuzzy analytic hierarchy process (FAHP) | |
| CN103995237A (en) | Satellite power supply system online fault diagnosis method | |
| CN106202886A (en) | Track circuit red band Fault Locating Method based on fuzzy coarse central Yu decision tree | |
| CN102520341B (en) | Analog circuit fault diagnosis method based on Bayes-KFCM (Kernelized Fuzzy C-Means) algorithm | |
| Rafique et al. | End to end machine learning for fault detection and classification in power transmission lines | |
| CN102722471B (en) | Fuzzy relation matrix generating method based on comprehensive correlation matrix | |
| CN101819253A (en) | Probabilistic neural network-based tolerance-circuit fault diagnosis method | |
| CN103345384B (en) | A kind of change neighborhood multiple goal test assignment dispatching method based on decomposition and platform | |
| CN106250934A (en) | Defect data classification method and device | |
| CN106372724A (en) | Artificial neural network algorithm | |
| CN110991472B (en) | Method for diagnosing minor faults of high-speed train traction system | |
| CN104037760B (en) | Anticipated accident selection method for electric power system with uncertain injection power | |
| CN104077483A (en) | Determining method of overall influence degrees of failure modes and failure causes of urban rail vehicle components | |
| CN104318485A (en) | Power transmission line fault identification method based on nerve network and fuzzy logic | |
| CN106208043A (en) | A kind of low and medium voltage distribution network key power consumption link discrimination method | |
| CN105572492A (en) | City rail train auxiliary inverter fault diagnosis device | |
| CN105117847A (en) | Method for evaluating transformer failure importance | |
| CN113093058A (en) | NPC three-level inverter open-circuit fault diagnosis method | |
| CN113203914A (en) | Underground cable early fault detection and identification method based on DAE-CNN | |
| CN113610119B (en) | Method for identifying power transmission line development faults based on convolutional neural network | |
| CN111914732A (en) | Insulation fault diagnosis method, device, equipment and computer readable storage medium | |
| CN111126750B (en) | Key unit determination method and device and storage medium | |
| CN106443394A (en) | Local discharge detection method for hydraulic turbine generator based on association probability wavelet transformation | |
| CN113722239B (en) | Airborne embedded software quality detection method, device, medium and electronic equipment | |
| CN105572493A (en) | Vehicle-mounted-ground auxiliary inverter remote fault diagnosis system |
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 |