Cable fault identification method, system and device based on deep learning, and medium

Through deep learning technology, the MSPSO optimization algorithm and deep belief network model are used to solve the problem of cumbersome and low efficiency in cable insulation fault identification, and the rapid and accurate identification of cable faults is achieved, and the identification efficiency and accuracy are improved.

CN118981709APending Publication Date: 2024-11-19STATE GRID BEIJING ELECTRIC POWER CO +1
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411058689.3
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-02
Publication Date
2024-11-19

AI Technical Summary

Technical Problem

The prior art has problems such as cumbersome process, low efficiency and low recognition accuracy in identifying cable insulation faults, and it is difficult to effectively identify cable faults, especially in the case of insulation aging and partial discharge.

Method used

A deep learning-based method is adopted to obtain cable measurement data, perform continuous wavelet transform feature extraction, and use the MSPSO optimization algorithm to train the deep belief network model to optimize the cable insulation fault recognition model, and improve the recognition accuracy and efficiency.

Benefits of technology

It realizes rapid classification and identification of cable faults, improves identification accuracy and efficiency, simplifies the operation process, and ensures the training speed and stable convergence of the model.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118981709A_ABST
    Figure CN118981709A_ABST
Patent Text Reader

Abstract

The invention discloses a cable fault identification method, system and device based on deep learning, and a medium. The method comprises the steps of obtaining measurement data of a cable; performing feature extraction on the measurement data of the cable; the extracted feature data is input into the trained cable insulation fault identification model to identify the fault of the cable, and the cable insulation fault identification model performs optimization training based on an MSPSO optimization algorithm, the method, system, device and medium can improve the accuracy and efficiency of cable fault identification, and the operation is relatively simple.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention belongs to the technical field of fault identification and relates to a cable fault identification method, system, equipment and medium based on deep learning. Background Art

[0002] Cross-linked polyethylene (XLPE) cables are known for their excellent electrical performance, which is reflected in their simple accessory configuration, low sensitivity to environmental changes, high rated operating temperature and reliable durability. A large amount of statistical data analysis shows that the vast majority of XLPE cable failures are caused by insulation failures, and about 80% of power accidents are related to insulation aging. The cable insulation layer is frequently subjected to high currents and operating overvoltages, which not only aggravates the aging of the cable, but also shortens its service life. Under certain specific conditions, partial discharge may also occur, which may trigger a series of emergencies and pose a potential threat to the stable operation of the power system. The four common failure modes of power cables are as follows:

[0003] Ordinary power cable (Type I): Before using the power cable, each layer of material must be stripped separately according to the construction regulations to avoid short circuit caused by the copper conductor being too close to the ground. In addition, terminal connectors must be added to increase the surface distance to prevent high-voltage creepage on the cable surface. Its wavy structure means that surface discharge accidents can be effectively avoided.

[0004] Short outer semiconductive layer (Type II): The power cable terminal joints must be installed according to the manufacturer's instructions. If the construction specifications are not followed, the stress cone inside the terminal joint cannot be located on the outer semiconductive layer, resulting in uneven dielectric constant and electric field distribution inside the terminal joint, which leads to partial discharge.

[0005] Insulation layer with impurities (Class III): Due to the influence of natural dust in the construction environment or during the installation of the power cable terminal connector, some impurities may remain on the insulation layer. According to the power cable construction regulations, the insulation layer must be wiped in a fixed direction with a special cleaning cloth to prevent the impurities remaining on the insulation layer from affecting the electric field distribution.

[0006] Damaged insulation layer (Type IV): During the installation, maintenance, disassembly and operation of power cables, the insulation layer may be damaged due to external factors. Digging with mechanical tools or being bitten by animals during construction may cause serious damage to the insulation of power cables. After the insulation layer is damaged, even if the power is not cut off immediately, water and gas will penetrate into the insulation layer during long-term operation, eventually causing insulation breakdown and failure.

[0007] In the identification of cable insulation faults, domestic and foreign scholars have achieved many results, which are mainly divided into three categories: cable fault identification based on machine learning, cable fault identification based on multivariate statistical theory, and cable fault identification based on deep learning. The above research has a certain effect in the identification of cable insulation faults, but in general, there are still problems such as cumbersome processes, low efficiency, and low identification accuracy. Summary of the invention

[0008] The purpose of the present invention is to overcome the shortcomings of the above-mentioned prior art and provide a cable fault identification method, system, device and medium based on deep learning. The method, system, device and medium can improve the accuracy and efficiency of cable fault identification and are relatively simple to operate.

[0009] In order to achieve the above object, the present invention adopts the following technical scheme:

[0010] In one aspect of the present invention, the cable fault identification method based on deep learning of the present invention comprises:

[0011] Get the measurement data of the cable;

[0012] Performing feature extraction on the measurement data of the cable;

[0013] The extracted feature data is input into a trained cable insulation fault identification model to identify cable faults, wherein the cable insulation fault identification model is optimized and trained based on the MSPSO optimization algorithm.

[0014] The cable fault identification method based on deep learning described in the present invention is further improved in that:

[0015] Furthermore, the process of extracting features from the measurement data of the cable is as follows:

[0016] The measured data of the cable is subjected to feature extraction using continuous wavelet transform.

[0017] Furthermore, before inputting the extracted feature data into the trained cable insulation fault identification model, the method further includes:

[0018] Construct a cable insulation fault identification model;

[0019] Obtain cable insulation fault data samples;

[0020] Extracting cable insulation fault data features from cable insulation fault data samples;

[0021] The extracted cable insulation fault data features are used to construct training sets and test sets;

[0022] The MSPSO optimization algorithm is adopted to train and test the cable insulation fault identification model using a training set and a test set to optimize the parameters of the cable insulation fault identification model and obtain the trained cable insulation fault identification model.

[0023] Furthermore, the process of constructing the cable insulation fault identification model is as follows:

[0024] Based on the deep belief network, the cable insulation fault identification model is constructed.

[0025] Furthermore, the process of extracting the cable insulation fault data features from the cable insulation fault data samples is as follows:

[0026] Continuous wavelet transform is used to extract cable insulation fault data features from cable insulation fault data samples.

[0027] In a second aspect of the present invention, the cable fault identification system based on deep learning of the present invention comprises:

[0028] A first acquisition module, used to acquire measurement data of the cable;

[0029] A first extraction module, used for extracting features from the measurement data of the cable;

[0030] The identification module is used to input the extracted feature data into a trained cable insulation fault identification model to identify the cable fault, wherein the cable insulation fault identification model is optimized and trained based on the MSPSO optimization algorithm.

[0031] The cable fault identification system based on deep learning described in the present invention is further improved in that:

[0032] Furthermore, the process of extracting features from the measurement data of the cable is as follows:

[0033] The measured data of the cable is subjected to feature extraction using continuous wavelet transform.

[0034] Furthermore, it also includes:

[0035] The first building module is used to build a cable insulation fault identification model;

[0036] The second acquisition module is used to acquire cable insulation fault data samples;

[0037] A second extraction module is used to extract cable insulation fault data features from the cable insulation fault data sample;

[0038] The second construction module is used to construct a training set and a test set based on the extracted cable insulation fault data features;

[0039] The training module is used to adopt the MSPSO optimization algorithm, use the training set and the test set to train and test the cable insulation fault identification model, so as to optimize the parameters of the cable insulation fault identification model and obtain the trained cable insulation fault identification model.

[0040] In the third aspect of the present invention, the computer device of the present invention includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the steps of the cable fault identification method based on deep learning are implemented.

[0041] In a fourth aspect of the present invention, the computer-readable storage medium of the present invention stores a computer program, and when the computer program is executed by a processor, the steps of the cable fault identification method based on deep learning are implemented.

[0042] The present invention has the following beneficial effects:

[0043] During specific operation, the cable fault identification method, system, device and medium based on deep learning described in the present invention input the extracted feature data into the trained cable insulation fault identification model to identify the cable fault, wherein the cable insulation fault identification model is optimized and trained based on the MSPSO optimization algorithm, which can ensure the training speed and stable convergence of the cable insulation fault identification model, while improving the recognition accuracy of the model, and realizing rapid classification and identification of cable faults. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] The accompanying drawings constituting a part of the present invention are used to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the accompanying drawings:

[0045] Figure 1 Flowchart for optimization of deep belief network parameters;

[0046] Figure 2 is a flow chart of the method of the present invention; DETAILED DESCRIPTION

[0047] The present invention will be described in detail below with reference to the accompanying drawings and in combination with embodiments. It should be noted that the embodiments and features in the embodiments of the present application can be combined with each other without conflict.

[0048] The following detailed description is an exemplary description, which is intended to provide further detailed description of the present invention. Unless otherwise specified, all technical terms used in the present invention have the same meaning as those generally understood by those skilled in the art to which the present application belongs. The terms used in the present invention are only for describing specific embodiments, and are not intended to limit exemplary embodiments according to the present invention.

[0049] Embodiment 1

[0050] Figure 2 The cable fault identification method based on deep learning of the present invention comprises the following steps:

[0051] 1) Obtain the measurement data of the cable;

[0052] 2) extracting features from the measured data of the cable using continuous wavelet transform;

[0053] 3) Inputting the extracted feature data into the trained cable insulation fault identification model to identify the fault type of the cable, wherein the cable insulation fault identification model is optimized and trained based on the MSPSO optimization algorithm.

[0054] The process of optimizing and training the cable insulation fault identification model based on the MSPSO optimization algorithm in step 3) is as follows:

[0055] 31) Based on the deep belief network (DBN), a cable insulation fault identification model is constructed;

[0056] 32) Obtain cable insulation fault data samples;

[0057] 33) Using continuous wavelet transform to extract cable insulation fault data features from cable insulation fault data samples;

[0058] 34) constructing a training set and a test set based on the extracted cable insulation fault data features;

[0059] 35) The cable insulation fault identification model is trained and tested using a training set and a test set by using an MSPSO optimization algorithm to optimize the parameters of the cable insulation fault identification model and obtain a trained cable insulation fault identification model.

[0060] In addition, it should be noted that the particle swarm optimization algorithm is a random search strategy derived from group behavior. The algorithm involves two important groups: one is the group that records the current position of the particle (called pbest), and the other is the group that stores the historical optimal position of the particle. The pbest group serves as a potential solution in the search space, while the latter guides the direction update of the former during the optimization process. In the particle swarm optimization framework, each particle has two characteristics, namely the velocity vector V and the position vector X. The movement rate of the particle in the search space will be dynamically adjusted synchronously based on the past performance of itself and its neighboring particles. Specifically, the speed and position updates of the particles follow specific mathematical rules:

[0061] v id (t+1)=ω×v id (t)+c1×r1×[p id (t)-x id (t)]+c2×r2×[p gd (t)-x id (t)] (1)

[0062] x id (t+1)=x id (t)+v id (t+1) (2)

[0063] Among them, c1 and c2 are acceleration coefficients reflecting the weight of the random acceleration term that pulls each particle to the pbest and gbest positions respectively, r1 and r2 represent two random numbers uniformly distributed in the range of (0, 1), and the characteristics of r1 and r2 are similar to the temperature parameters in simulated annealing (SA), and ω is the inertia weight used to balance the global search and the local search. Generally speaking, a large inertia weight is conducive to global optimization, while a small inertia weight is often conducive to local optimization. The position of the i-th particle can be represented by a D-dimensional vector as X i =[x i1 , x i2 , …, x ij , …, x iD ], where x ij ∈[xmax min , represents the position of the i-th particle in the j-th dimension, and the corresponding velocity is V i =[v i1 , v i2 , …, v ij , …, v iD ], where v ij ∈[vmax min To reduce the possibility of particles leaving the search space, the best previous position of the ith particle is recorded as pbest and denoted as P i =[pi1 , p i2 ,…,p ij ,…,p iD ], and the global best position reached by the entire particle swarm so far is recorded as gbest and expressed as P g =[p g1 , p g2 ,…,p gj ,…,p gD ].

[0064] The MSPSO optimization algorithm described in the present invention has the following two differences compared with the traditional particle swarm optimization algorithm:

[0065] a1) Introduce the linear decreasing method into the speed calculation formula to adjust the inertia weight, specifically:

[0066]

[0067] Among them, w is the inertial weight, t is the current iteration number; w ini 、w end and T max are the initial inertia weight, final inertia weight and maximum number of iterations respectively.

[0068] a2) Adopting population mutation strategy, specifically: when the population converges to the local minimum, the particles in the population are close to each other, and the population loses diversity. The mutated particle group makes the particles repel each other, improves the diversity of the particle group, and makes the particle group jump to the local minimum with a higher probability. In the MSPSO algorithm, when the global optimal fitness value of the population does not change within the predetermined number of iterations, the mutation operation is adopted to randomly select half of the particles in the population for reinitialization.

[0069] In the present invention, when the multi-strategy improved particle swarm optimization (MSPSO) algorithm is used to optimize the structural DBN parameters, the DBN and the MSPSO algorithm are first initialized, and the number of hidden layers of the DBN is set to 3, corresponding to m1, m2 and m3 neurons respectively. N represents the number of particles, which is set to 10. The maximum number of iterations of the particle swarm optimization is M. The process of optimizing the DBN model based on MSPSO is as follows: Figure 1 shown.

[0070] The parameter optimization process of the DBN model is:

[0071] (1) Initialize particle swarm velocity Initialize the position of the particle swarm

[0072] (2) Calculate the recognition error rate of all particles and find the optimal particle of the particle swarm in this round, including the optimal particle that has completed the search;

[0073] (3) Use formula (3) and formula (4) to update the particle velocity and position.

[0074] (4) When MSPSO ends the optimization, one of the two conditions must be met to stop. One is that the classification error rate of the experimental data is lower than the preset value or the number of iterations reaches the preset value. Otherwise, go to step (2), increase the number of iterations, and repeat steps (3) and (4) until the judgment condition is met. The other is that the fitness satisfies the condition that the number of iterations is equal to N.

[0075] (5) Substitute the optimized parameters into the original DBN model and obtain the fault identification results through retraining and retesting of data samples.

[0076] Embodiment 2

[0077] The cable fault identification system based on deep learning of the present invention comprises:

[0078] A first acquisition module, used to acquire measurement data of the cable;

[0079] A first extraction module, used for extracting features from the measurement data of the cable;

[0080] The first building module is used to build a cable insulation fault identification model;

[0081] The second acquisition module is used to acquire cable insulation fault data samples;

[0082] A second extraction module is used to extract cable insulation fault data features from the cable insulation fault data sample;

[0083] The second construction module is used to construct a training set and a test set based on the extracted cable insulation fault data features;

[0084] A training module, used for training and testing the cable insulation fault identification model using a training set and a test set by using an MSPSO optimization algorithm, so as to optimize the parameters of the cable insulation fault identification model and obtain a trained cable insulation fault identification model;

[0085] The identification module is used to input the extracted feature data into a trained cable insulation fault identification model to identify the cable fault, wherein the cable insulation fault identification model is optimized and trained based on the MSPSO optimization algorithm.

[0086] As an implementation manner of the present invention, the process of extracting features from the measurement data of the cable is: extracting features from the measurement data of the cable using continuous wavelet transform.

[0087] The division of modules in the embodiments of the present application is schematic and is only a logical function division. There may be other division methods in actual implementation. In addition, each functional module in each embodiment of the present application may be integrated into a processor, or may exist physically separately, or two or more modules may be integrated into one module. The above-mentioned integrated modules may be implemented in the form of hardware or in the form of software functional modules.

[0088] Embodiment 3

[0089] A computer device comprises a memory, a processor and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the steps of implementing the cable fault identification method based on deep learning include: constructing a cable insulation fault identification model; obtaining cable insulation fault data samples; extracting cable insulation fault data features from the cable insulation fault data samples; constructing a training set and a test set with the extracted cable insulation fault data features; using an MSPSO optimization algorithm, using the training set and the test set to train and test the cable insulation fault identification model, so as to optimize the parameters of the cable insulation fault identification model, and obtaining the trained cable insulation fault identification model. model; obtain the measurement data of the cable; extract the features of the measurement data of the cable; input the extracted feature data into the trained cable insulation fault identification model to identify the cable fault, wherein the cable insulation fault identification model is optimized and trained based on the MSPSO optimization algorithm, wherein the memory may include a memory, such as a high-speed random access memory, and may also include a non-volatile memory, such as at least one disk memory, etc.; the processor, the network interface, and the memory are interconnected through an internal bus, and the internal bus may be an industrial standard architecture bus, a peripheral component interconnection standard bus, an extended industrial standard architecture bus, etc., and the bus may be divided into an address bus, a data bus, a control bus, etc. The memory is used to store programs, and specifically, the program may include a program code, and the program code includes computer operation instructions. The memory may include a memory and a non-volatile memory, and provide instructions and data to the processor.

[0090] Embodiment 4

[0091] A computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of implementing the cable fault identification method based on deep learning include: constructing a cable insulation fault identification model; obtaining a cable insulation fault data sample; extracting cable insulation fault data features from the cable insulation fault data sample; constructing a training set and a test set with the extracted cable insulation fault data features; using an MSPSO optimization algorithm, training and testing the cable insulation fault identification model with the training set and the test set to optimize the parameters of the cable insulation fault identification model, and obtaining a trained cable insulation fault identification model; obtaining cable measurement data; extracting features from the cable measurement data; inputting the extracted feature data into the trained cable insulation fault identification model to identify the cable fault, wherein the cable insulation fault identification model is optimized and trained based on the MSPSO optimization algorithm, and specifically, the computer-readable storage medium includes, but is not limited to, for example, a volatile memory and / or a non-volatile memory. The volatile memory may include a random access memory (RAM) and / or a cache memory, etc. The non-volatile memory may include a read-only memory (ROM), a hard disk, a flash memory, an optical disk, a magnetic disk, and the like.

[0092] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented in one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that include computer-usable program code.

[0093] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0094] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0095] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0096] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit the present invention. Although the present invention has been described in detail with reference to the above embodiments, it should be understood by those skilled in the art that the specific implementation modes of the present invention can still be modified or replaced by equivalents without departing from the spirit and scope of the present invention.

[0097] The description should all be included in the protection scope of the claims of the present invention.

Claims

1. A cable fault identification method based on deep learning, It is characterized in that include: Get the measurement data of the cable; Performing feature extraction on the measurement data of the cable; The extracted feature data is input into a trained cable insulation fault identification model to identify cable faults, wherein the cable insulation fault identification model is optimized and trained based on the MSPSO optimization algorithm.

2. The cable fault identification method based on deep learning according to claim 1, It is characterized in that The process of extracting features from the measurement data of the cable is as follows: The measured data of the cable is subjected to feature extraction using continuous wavelet transform.

3. The cable fault identification method based on deep learning according to claim 1, It is characterized in that Before the extracted feature data is input into the trained cable insulation fault identification model, the method further includes: Construct a cable insulation fault identification model; Obtain cable insulation fault data samples; Extracting cable insulation fault data features from cable insulation fault data samples; The extracted cable insulation fault data features are used to construct training sets and test sets; The MSPSO optimization algorithm is adopted to train and test the cable insulation fault identification model using a training set and a test set to optimize the parameters of the cable insulation fault identification model and obtain the trained cable insulation fault identification model.

4. The cable fault identification method based on deep learning according to claim 1, It is characterized in that The process of constructing the cable insulation fault identification model is as follows: Based on the deep belief network, the cable insulation fault identification model is constructed.

5. The cable fault identification method based on deep learning according to claim 1, It is characterized in that The process of extracting cable insulation fault data features from cable insulation fault data samples is as follows: Continuous wavelet transform is used to extract cable insulation fault data features from cable insulation fault data samples.

6. A cable fault identification system based on deep learning, It is characterized in that include: A first acquisition module, used to acquire measurement data of the cable; A first extraction module, used for extracting features from the measurement data of the cable; The identification module is used to input the extracted feature data into a trained cable insulation fault identification model to identify the cable fault, wherein the cable insulation fault identification model is optimized and trained based on the MSPSO optimization algorithm.

7. The cable fault identification system based on deep learning according to claim 6, It is characterized in that The process of extracting features from the measurement data of the cable is as follows: The measured data of the cable is subjected to feature extraction using continuous wavelet transform.

8. The cable fault identification system based on deep learning according to claim 6, It is characterized in that Also includes: The first building module is used to build a cable insulation fault identification model; The second acquisition module is used to acquire cable insulation fault data samples; A second extraction module is used to extract cable insulation fault data features from the cable insulation fault data sample; The second construction module is used to construct a training set and a test set based on the extracted cable insulation fault data features; The training module is used to adopt the MSPSO optimization algorithm, use the training set and the test set to train and test the cable insulation fault identification model, so as to optimize the parameters of the cable insulation fault identification model and obtain the trained cable insulation fault identification model.

9. A computer device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, It is characterized in that When the processor executes the computer program, the steps of the cable fault identification method based on deep learning as described in any one of claims 1 to 5 are implemented.

10. A computer-readable storage medium storing a computer program. It is characterized in that When the computer program is executed by a processor, the steps of the cable fault identification method based on deep learning as described in any one of claims 1 to 5 are implemented.