Cable joint state detection model training method, cable joint state detection method, device, equipment, medium and program
Through particle swarm optimization algorithm, the neural network model is optimized and the cable high-order harmonic characteristics are extracted, the accuracy of cable joint state detection is improved, the problem of low detection accuracy in the existing technology is solved, and the reliability of cable joint state detection is enhanced.
Patent Information
- Application Number
- CN202411054556.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-02
- Publication Date
- 2024-11-22
AI Technical Summary
The detection accuracy of existing cable connector status detection is low, resulting in inaccurate detection of cable connector status, which may cause safety hazards such as cable fires.
The network model parameters of the training neural network model are optimized by using particle swarm optimization algorithm (PSO) to obtain the target network model parameters, and the cable joint state detection model is trained based on the cable joint state sample data to extract the high-order cable of the induced current data. Harmonic features, input to the detection model to output the status detection result.
The detection accuracy of the cable joint status detection model is improved, the accuracy of the cable joint status detection is enhanced, and the risk of cable joint failure is reduced.
Smart Images

Figure CN119005278A_ABST
Abstract
Description
Technical Field
[0001] The embodiments of the present invention relate to the field of power grid detection technology, and in particular to a cable joint state detection model training method and a cable joint state detection method, device, equipment, medium and program. Background Art
[0002] With the continuous construction of urbanization and distribution networks, the popularity of power grid cabling is increasing, and there are more and more cables laid in cable trenches or directly buried. Due to the needs of power supply and operation and maintenance, there are more cable joints (can be referred to as cable heads for short) to meet the requirements of power supply safety and operation and maintenance. Busy urban areas and industrially intensive areas are becoming cable-intensive areas. Widely distributed cable joints are also the weakest link in the safe operation of the power system. This is mainly due to the influence of unfavorable factors such as the on-site environment during production and the limitations of the production process, which causes some potential hidden dangers in the crimping quality of the joints. Along with the aging of insulation, the corrosion of cable joints by rain, various acids and salts and other corrosive gases and liquids, the contact resistance of the cable joints is too large. In the long-term operation process, it is easy to cause the cable head to overheat and break down the insulation, resulting in the safety hazard of fire in the cable trench or directly buried cable. Summary of the invention
[0003] The embodiments of the present invention provide a cable joint state detection model training method and a cable joint state detection method, device, equipment, medium and program, which can improve the detection accuracy of the cable joint state detection model and thus improve the accuracy of the cable joint state detection.
[0004] According to one aspect of the present invention, a cable joint state detection model training method is provided, comprising:
[0005] Obtain network model parameters of the neural network model to be trained;
[0006] The network model parameters of the neural network model to be trained are optimized by using a particle swarm optimization (PSO) optimization algorithm to obtain target network model parameters;
[0007] The network model parameters of the neural network model to be trained are updated according to the target network model parameters to obtain a cable joint state detection model to be trained;
[0008] The cable joint state detection model to be trained is trained according to the cable joint state sample data to obtain a target cable joint state detection model;
[0009] Wherein, the target cable joint state detection model is used to detect the cable joint state.
[0010] According to another aspect of the present invention, there is provided a method for detecting a cable connector state, comprising:
[0011] Acquire the current collected data of the cable joint to be detected; wherein the current collected data includes key state quantity data and induced current data of the cable joint to be detected;
[0012] Extracting cable high-order harmonic characteristics of the induced current data of the cable joint to be detected;
[0013] Inputting the key state quantity data of the cable joint to be detected and the cable high-order harmonic characteristics into the target cable joint state detection model;
[0014] Outputting the state detection result of the cable joint to be detected through the target cable joint state detection model;
[0015] The target cable joint state detection model is obtained by training the cable joint state detection model training method described in any embodiment of the present invention.
[0016] According to another aspect of the present invention, there is provided a cable joint state detection model training device, comprising:
[0017] A network model parameter acquisition module is used to obtain network model parameters of the neural network model to be trained;
[0018] A network model parameter optimization module is used to optimize the network model parameters of the neural network model to be trained by using a PSO optimization algorithm to obtain target network model parameters;
[0019] A network model parameter updating module, used for updating the network model parameters of the neural network model to be trained according to the target network model parameters, so as to obtain a cable joint state detection model to be trained;
[0020] A detection model training module is used to train the cable joint state detection model to be trained according to the cable joint state sample data to obtain a target cable joint state detection model;
[0021] Wherein, the target cable joint state detection model is used to detect the cable joint state.
[0022] According to another aspect of the present invention, there is provided a cable connector status detection device, comprising:
[0023] A current acquisition data acquisition module is used to acquire the current acquisition data of the cable joint to be detected; wherein the current acquisition data includes key state quantity data and induced current data of the cable joint to be detected;
[0024] A cable high-order harmonic feature extraction module, used to extract the cable high-order harmonic features of the induced current data of the cable joint to be detected;
[0025] A data feature input module, used to input the key state quantity data of the cable joint to be detected and the cable high-order harmonic characteristics into a target cable joint state detection model;
[0026] A state detection result output module, used for outputting the state detection result of the cable joint to be detected through the target cable joint state detection model;
[0027] The target cable joint state detection model is obtained by training the cable joint state detection model training method described in any embodiment of the present invention.
[0028] According to another aspect of the present invention, an electronic device is provided, the electronic device comprising:
[0029] at least one processor; and
[0030] a memory communicatively connected to the at least one processor; wherein,
[0031] The memory stores a computer program that can be executed by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the cable joint status detection model training method or the cable joint status detection method described in any embodiment of the present invention.
[0032] According to another aspect of the present invention, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the cable joint state detection model training method or the cable joint state detection method described in any embodiment of the present invention when executed.
[0033] According to another aspect of the present invention, a computer program product is provided, including a computer program, which, when executed by a processor, implements the cable joint state detection model training method or the cable joint state detection method described in any embodiment of the present invention.
[0034] The embodiment of the present invention optimizes the network model parameters of the neural network model to be trained by adopting the PSO optimization algorithm to obtain the target network model parameters, thereby updating the network model parameters of the neural network model to be trained according to the target network model parameters, obtaining the cable joint state detection model to be trained, and then training the cable joint state detection model to be trained according to the cable joint state sample data to obtain the target cable joint state detection model. After the training of the target cable joint state detection model is completed, the current collected data such as the key state quantity data and the induced current data of the cable joint to be detected can be obtained, and the cable high-order harmonic characteristics of the induced current data of the cable joint to be detected can be extracted, so that the key state quantity data and the cable high-order harmonic characteristics of the cable joint to be detected are input into the target cable joint state detection model, so as to output the state detection result of the cable joint to be detected through the target cable joint state detection model. The above technical scheme can solve the problem of low detection accuracy in the existing cable joint state detection, improve the detection accuracy of the cable joint state detection model, and thus improve the accuracy of the cable joint state detection.
[0035] It should be understood that the contents described in this section are not intended to identify the key or important features of the embodiments of the present invention, nor are they intended to limit the scope of the present invention. Other features of the present invention will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS
[0036] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.
[0037] Figure 1 It is a flow chart of a cable joint state detection model training method provided by the first embodiment of the present invention;
[0038] Figure 2 is a flow chart of a cable joint status detection model training method provided by Embodiment 2 of the present invention;
[0039] Figure 3 It is a flow chart of a cable joint state detection model training method provided by Embodiment 2 of the present invention;
[0040] Figure 4 is a flow chart of a cable connector status detection method provided by Embodiment 3 of the present invention;
[0041] Figure 5 It is a flow chart of a cable connector status detection method provided in Embodiment 3 of the present invention;
[0042] Figure 6 It is a schematic diagram of a cable joint state detection model training device provided by a fourth embodiment of the present invention;
[0043] Figure 7 is a schematic diagram of a cable connector status detection device provided by Embodiment 5 of the present invention;
[0044] Figure 8 A schematic diagram of the structure of an electronic device provided in Embodiment 6 of the present invention. DETAILED DESCRIPTION
[0045] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.
[0046] It should be noted that the terms "include", "have" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, system, product or apparatus comprising a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to these processes, methods, products or apparatus.
[0047] Embodiment 1
[0048] Figure 1 This is a flowchart of a cable joint state detection model training method provided by the first embodiment of the present invention. This embodiment is applicable to the case where the PSO optimization algorithm is used to train the cable joint state detection model. The method can be executed by a cable joint state detection model training device, which can be implemented by software and / or hardware, and can generally be integrated in an electronic device. The electronic device can be a terminal device or a server device. As long as the cable joint state detection model training method can be executed, the embodiment of the present invention does not limit the specific device type of the electronic device. Accordingly, Figure 1 As shown, the method includes the following operations:
[0049] S110, obtaining network model parameters of the neural network model to be trained.
[0050] The neural network model to be trained may be an original untrained neural network model. Optionally, the neural network model to be trained may be any type of neural network model, for example, including but not limited to convolutional neural network, multi-layer forward neural network, recursive neural network, back propagation neural network (BPNN) and feedforward neural network, etc., as long as it can be used to detect and identify the state of the cable connector, the embodiment of the present invention does not limit the specific network type of the neural network model to be trained. The network model parameters may include but are not limited to parameter types such as weights and thresholds of the neural network model to be trained.
[0051] In an embodiment of the present invention, before training the neural network model to be trained, the network model parameters of the neural network model to be trained may be first obtained to preliminarily optimize the network model parameters of the neural network model to be trained.
[0052] S120, using a PSO optimization algorithm to optimize the network model parameters of the neural network model to be trained to obtain target network model parameters.
[0053] The target network model parameters may be network model parameters obtained by optimizing the initial network model parameters of the neural network model to be trained using a PSO (Particle Swarm Optimization) optimization algorithm.
[0054] Since the error of the traditional neural network is insensitive to the changes of network model parameters such as weights during the training process, the model has many iterations, slow convergence, and is very easy to fall into the local minimum. To solve this technical problem, the embodiment of the present invention uses the PSO optimization algorithm to perform preliminary optimization on the network model parameters of the neural network model to be trained, to search for the optimal values of the network model parameters of the neural network model to be trained, such as searching for the optimal weights and thresholds, etc., to obtain the target network model parameters, thereby accelerating the convergence speed of the neural network model to be trained and improving the global search capability.
[0055] The above technical scheme, by utilizing the PSO optimization algorithm to optimize the network model parameters of the neural network model to be trained, can not only improve the training efficiency and training quality of the neural network model to be trained, but also improve the model accuracy of the target cable joint state detection model obtained by training the neural network model to be trained, thereby improving the detection accuracy of the target cable joint state detection model.
[0056] S130. Update the network model parameters of the neural network model to be trained according to the target network model parameters to obtain a cable connector state detection model to be trained.
[0057] The cable joint status detection model to be trained may be a network model that needs to be further trained using sample data to optimize network model parameters.
[0058] Correspondingly, after the target network model parameters are obtained by optimizing the PSO optimization algorithm, the network model parameters of the neural network model to be trained can be assigned according to the target network model parameters, so as to update the network model parameters of the neural network model to be trained and obtain the cable joint state detection model to be trained that needs further optimization training.
[0059] S140: training the cable joint state detection model to be trained according to the cable joint state sample data to obtain a target cable joint state detection model.
[0060] The cable joint state sample data may be sample data for training the cable joint state detection model to be trained. The target cable joint state detection model may be a model obtained after training the cable joint state detection model to be trained, and the detection accuracy meets the requirements. The target cable joint state detection model may be used to detect the cable joint state.
[0061] Accordingly, after obtaining the cable joint state detection model to be trained, the obtained cable joint state sample data can be used to train the training cable joint state detection model. After repeated iterative training and testing, a target cable joint state detection model with detection accuracy that meets the requirements can be obtained. The target cable joint state detection model can establish a logical relationship between the input characteristic parameters of the tested cable joint and the operating state of the tested cable joint. After obtaining the target cable joint state detection model, the target cable joint state detection model can be used to perform real-time and intelligent detection of the cable joint state.
[0062] The embodiment of the present invention optimizes the network model parameters of the neural network model to be trained by adopting the PSO optimization algorithm to obtain the target network model parameters, thereby updating the network model parameters of the neural network model to be trained according to the target network model parameters to obtain the cable joint state detection model to be trained, and then training the cable joint state detection model to be trained according to the cable joint state sample data to obtain the target cable joint state detection model. After the target cable joint state detection model is trained, it can be used to detect the cable joint state. The above technical scheme can solve the problem of low detection accuracy in the existing cable joint state detection, can improve the detection accuracy of the cable joint state detection model, and thus improve the accuracy of the cable joint state detection.
[0063] Embodiment 2
[0064] Figure 2is a flow chart of a cable joint status detection model training method provided by Embodiment 2 of the present invention, Figure 3 The flowchart of a cable joint state detection model training method provided by the second embodiment of the present invention is a specific embodiment based on the above embodiment. In this embodiment, multiple specific optional implementation methods are provided for optimizing the network model parameters of the neural network model to be trained by using the PSO optimization algorithm and training the cable joint state detection model to be trained according to the cable joint state sample data. Figure 2 As shown, the method of this embodiment may include:
[0065] S210, obtaining network model parameters of the neural network model to be trained.
[0066] S220. Construct an initial particle population according to the network model parameters of the neural network model to be trained.
[0067] In the embodiment of the present invention, optionally, the network model parameters of each network layer of the neural network model to be trained can be used to construct an initial particle population. Exemplarily, assuming that the neural network to be trained is a BP neural network, four network model parameters, including the weight vector W1 connecting the input layer and the hidden layer of the neural network model to be trained, the weight vector W2 connecting the hidden layer and the output layer, the threshold a of the hidden layer, and the threshold b of the output layer, can be used to construct particles, and the initial particle population can be further constructed based on the constructed particles.
[0068] For example, when constructing a particle population, the particle population size, the objective function and the search space can be set. The position and velocity of each particle are randomly initialized, and its individual fitness is calculated.
[0069] S230, calculating the fitness of each particle in the particle population and the global optimal solution of the particle population according to the fitness function.
[0070] Optionally, the fitness function of the PSO optimization algorithm can be implemented using the following formula:
[0071]
[0072] Among them, f(i) represents the fitness function of the PSO optimization algorithm, that is, the objective function, P i is the state evaluation value of the neural network model to be trained, P i m is the actual state evaluation value of the neural network model to be trained for the test sample, and N is the number of training samples of the neural network model to be trained.
[0073] That is, the PSO optimization algorithm uses the mean square error between the state evaluation value output by the neural network model to be trained and the actual evaluation value of the sample as the fitness function.
[0074] S240, updating the position and velocity of each particle, performing a crossover operation and adopting a greedy strategy to update the particle population.
[0075] For each particle, the influence of its individual optimal solution (pBest) and the global optimal solution (gBest) is calculated, and then the velocity and position of the particle are updated, a crossover operation is performed, and a greedy strategy is adopted to update the particle population.
[0076] S250, determine whether the maximum evolutionary generation of the current stage has been reached, if so, execute S260, otherwise, return to execute S230.
[0077] The PSO optimization algorithm simulates the behavior of groups of organisms such as bird flocks or fish schools, and regards the problem to be solved as a process of finding the optimal solution in a multidimensional space. Each possible solution is regarded as a particle in a multidimensional space. The position of the particle represents a feasible solution, and the velocity represents the change during the update. By assigning a random velocity vector to each particle, the particle is guided to explore, and the global optimum and local optimum are used to guide the search direction of the particle, gradually approaching the global optimal solution to achieve the optimization purpose.
[0078] Correspondingly, if it is determined that the current number of iterations reaches the maximum evolutionary generation, the subsequent process can be continued; otherwise, return to re-execute the iteration process to update the next round of population.
[0079] S260: Determine whether the stage number of the current stage is the maximum stage number. If yes, execute S270; otherwise, execute S280.
[0080] S270. Determine the target network model parameters according to the optimal particle of the current particle population.
[0081] S280, optimizing and adjusting the population size of the particle population, and updating the stage number of the current stage, and then returning to execute S230.
[0082] In order to ensure the operating efficiency of the PSO optimization algorithm, the embodiment of the present invention divides the topological structure of the particle swarm of the algorithm into several parts according to the number of iterations, so as to execute the algorithm process in stages. Specifically, the maximum number of iterations of the algorithm can be set to maxgen, and the entire optimization process can be divided into R stages, then the number of iterations of each stage is T times, that is, maxgen = R × T.
[0083] The PSO optimization algorithm is prone to fall into a local optimum in the later stage, resulting in failure to obtain a global optimal solution. To address this problem, an embodiment of the present invention provides a ladder strategy of a dynamic topology structure to improve the premature maturity problem of the PSO optimization algorithm.
[0084] The core principle of the ladder strategy of dynamic topology structure is to update the topology of PSO optimization algorithm according to the number of iterations, and adopt different population sizes under a certain number of iterations. On the one hand, it can ensure the convergence speed of the algorithm, and on the other hand, it can introduce disturbances to make the algorithm escape from the local optimum.
[0085] Specifically, Figure 3 As shown in , after each calculation to obtain the individual optimal value and the global optimal value in the particle population, it can be determined whether the current number of iterations reaches the maximum number of iterations T of the current stage. If it is determined that the current number of iterations does not reach the maximum number of iterations T of the current stage, it indicates that the current staged iterative optimization has not been completed. After updating the number of iterations and the population, it returns to the step of calculating the population adaptation. If it is determined that the current number of iterations reaches the maximum number of iterations T of the current stage, it indicates that the current staged iterative optimization is completed, and it can continue to determine whether the number of stages in the current stage is the maximum number of stages. If it is determined that the number of stages in the current stage is not the maximum number of stages, it indicates that the overall process of the algorithm has not ended, and the population size of the particle population can be optimized and adjusted. After updating the number of stages in the current stage, it returns to the operation of calculating the fitness of each particle in the particle population and the global optimal solution of the particle population according to the fitness function, and realizes the optimization update process for the next stage. If it is determined that the number of stages in the current stage is the maximum number of stages, it indicates that the overall process of the algorithm is over. At this time, the target network model parameters can be determined according to the optimal particles of the current particle population finally obtained.
[0086] That is, after the particle population optimization is completed in each stage, the topological structure of the obtained particle population can be dynamically updated. Dynamically updating the topological structure of the particle population based on the stage dimension, rather than updating the topological structure of the particle population immediately after the particle population is updated, can not only ensure the convergence speed of the algorithm, but also allow the algorithm to escape from the local optimum by introducing disturbances.
[0087] In an optional embodiment of the present invention, the optimizing and adjusting the population size of the particle population may include: determining the degree of dispersion of the particle population; reducing the size of the particle population when it is determined that the degree of dispersion of the particle population is less than or equal to a set dispersion threshold; and expanding the size of the particle population when it is determined that the degree of dispersion of the particle population is greater than the set dispersion threshold.
[0088] The dispersion degree threshold may be set according to actual needs, and the embodiment of the present invention does not limit the specific value of the dispersion degree threshold.
[0089] Optionally, the population size of the particle population can be optimized and adjusted with reference to the dispersion degree of the particle population. That is, if after a certain stage of iterative optimization is completed, it is determined that the dispersion degree of the current particle population is less than or equal to the set dispersion degree threshold, the size of the particle population can be reduced to improve the compactness of the particle population. If after a certain stage of iterative optimization is completed, it is determined that the dispersion degree of the current particle population is greater than the set dispersion degree threshold, the size of the particle population can be expanded to improve the dispersion degree of the particle population.
[0090] S290. Update the network model parameters of the neural network model to be trained according to the target network model parameters to obtain the cable connector status detection model to be trained.
[0091] S2110. Train the cable joint state detection model to be trained according to the cable joint state sample data to obtain a target cable joint state detection model.
[0092] Wherein, the target cable joint state detection model is used to detect the cable joint state. Optionally, if the cable joint state detection model to be trained adopts a BP neural network, the target cable joint state detection model obtained through the above training process can be called an improved PSO-BP neural network, that is, a dynamic topology ladder PSO algorithm is used to optimize the parameters of the neural network to form an LPSO-BPNN.
[0093] For example, Figure 3 As shown in the figure, after the optimized weights and thresholds and other target network model parameters are calculated by the PSO optimization algorithm, the parameters of the target network model are used to set the parameters of the neural network model to be trained, and the training samples are input into the neural network model to be trained to obtain the target cable joint state detection model. After the training is completed, the test samples can also be input into the trained target cable joint state detection model to compare the output results to verify the correctness of the obtained target cable joint state detection model.
[0094] In an optional embodiment of the present invention, before training the cable joint state detection model to be trained according to the cable joint state sample data, it may also include: collecting key state quantity data of the cable joint in the operating state and the induced current data of the cable joint; wherein the key state quantity data may include at least one of the operating years, average load rate, load change rate, surface temperature rise, ambient humidity, ambient hazardous gas content and partial discharge; evaluating the operating state of the cable joint in the operating state to obtain the current operating state; and using the key state quantity data and the current operating state as the state quantity sample data of the cable joint.
[0095] In order to more comprehensively complete the evaluation of the operating status of the cable joint, a variety of key state quantity data of the cable joint in normal operation and various fault states can be collected, which may include but are not limited to: operating years, average load rate, load change rate, surface temperature rise, ambient humidity, ambient hazardous gas content and partial discharge, etc. At the same time, the induced current data of the cable joint can also be collected. For example, the temperature of the cable joint surface and the ambient temperature of the cable joint can be detected by a temperature sensor; the humidity of the operating environment of the cable joint can be detected by a humidity sensor; the gas condition of the surrounding environment of the cable joint can be detected by a gas sensor, mainly including the content of gas types such as carbon monoxide, oxygen and combustible gas; the real-time current data of the cable joint can be detected by a mutual inductor and the load rate can be obtained by calculation. After collecting various key state quantity data of the cable joint in the operating state, the operating state of the cable joint in the operating state can be evaluated to obtain the current operating state. Exemplarily, the operation status of the cable joint can be divided into three states: normal state, attention state and abnormal state, and the score range of each state is defined, wherein the score range of the normal state is 80-100, the score range of the attention state is 60-80, and the score range of the abnormal state is less than 60. The scores of the current operation status of different cable joints can be used as labels, the key state quantity data and the current operation status can be used as the state quantity sample data of the cable joint, and the induced current data of the cable joint can be used as the feature quantity data to realize the supervised training of the cable joint state detection model to be trained.
[0096] In an optional embodiment of the present invention, the training of the cable joint state detection model to be trained according to the cable joint state sample data to obtain the target cable joint state detection model may include: extracting the cable higher harmonic characteristics of the induced current data of the cable joint; inputting the cable higher harmonic characteristics and the state quantity sample data of the cable joint into the cable joint state detection model to be trained for training; and when it is determined that the training of the cable joint state detection model to be trained is completed, using the trained cable joint state detection model to be trained as the target cable joint state detection model.
[0097] It is understandable that the induced current data of the cable joint includes the characteristics of the high-order harmonics of the cable. Therefore, it is necessary to extract the characteristics of the induced current data of the cable joint, and then use the extracted cable high-order harmonic characteristics and the state quantity sample data of the cable joint as input data, and input them into the cable joint state detection model to be trained for training. During the training process, the cable high-order harmonic characteristics and the state quantity sample data of the cable joint can be used for repeated iterative training until it is determined that the accuracy of the cable joint state detection model to be trained meets the requirements, or the set number of training times is reached, and the training is completed. The trained cable joint state detection model to be trained is used as the target cable joint state detection model.
[0098] In an optional embodiment of the present invention, the extraction of cable high-order harmonic characteristics of the induced current data of the cable joint may include: performing a fast Fourier transform on the induced current data of the cable joint to obtain a transformed current signal; calculating the capacitive component and resistive current of the leakage current; and calculating the amplitude of the capacitive current of the target harmonic and the amplitude of the resistive current of the target harmonic based on the transformed current signal, the capacitive component of the leakage current and the resistive current.
[0099] The transformed current signal may be a current signal obtained by performing fast Fourier transform on the induced current data of the cable joint.
[0100] In a specific example, when applying fast Fourier transform to perform harmonic analysis on induced current, since the induced current of the cable joint satisfies the Dirichlet formula, the induced current data of the cable joint can be expressed as follows through fast Fourier transform;
[0101]
[0102] Among them, i x is the converted current signal, I0 is the DC component of the induced current of the cable connector, I km is the harmonic amplitude of the induced current of the cable joint, β k is the phase angle of each harmonic of the induced current of the cable joint, k represents each harmonic, and ω represents the angle.
[0103] Furthermore, the capacitive component of the leakage current can be calculated based on the following formula:
[0104]
[0105] I ck =kωCU km
[0106] Among them, i c represents the capacitive component of the leakage current, αc Represents the phase angle of the capacitive current, I ck Indicates the capacitive current of each harmonic of the induced current of the cable joint, C represents capacitance, U represents km Indicates the voltage amplitude of each harmonic of the induced current in the cable joint.
[0107] Order I Rk is the amplitude of the kth harmonic resistive current, since the kth harmonic voltage is in phase with the kth harmonic resistive current. Therefore, further, the resistive current can be calculated based on the following formula:
[0108]
[0109] Among them, i R represents the resistive current, α k Represents the phase angle of resistive current.
[0110] Because I x =i R +i C , substitute the above formula into i x =i R +i C , and can be obtained through trigonometric functions and integral transformations;
[0111]
[0112] I Ck =I km (cosα k .sinβ k -sinα k .cosβ k )
[0113] That is, I ck As the capacitive current amplitude of the target subharmonic finally extracted, I Rk As the amplitude of the target subharmonic resistive current finally extracted. Optionally, the value of k can be 3. Optionally, the amplitude of the capacitive current of the target subharmonic finally extracted and the amplitude of the target subharmonic resistive current can be used as the cable high-order harmonic features of the induced current data of the cable connector in the sample data.
[0114] The above technical solution introduces the ladder structure of the dynamic topology structure into the optimization of the PSO algorithm, which can ensure the running speed of the PSO algorithm while avoiding the situation where the global optimal solution cannot be obtained due to the premature maturity of the PSO algorithm. The target cable joint state detection model is constructed by using the real-time monitoring information and historical data information of the cable joint through the multi-parameter fusion method of the improved neural network. At the same time, combined with Fourier transform, high-order harmonics are incorporated into the state evaluation of the cable joint, so that the detection accuracy of the trained target cable joint state detection model is further improved, thereby improving the evaluation results of the cable joint by the target cable joint state detection model. More comprehensive and accurate.
[0115] Embodiment 3
[0116] Figure 4 is a flow chart of a cable connector status detection method provided by Embodiment 3 of the present invention. Figure 5 This is a flow chart of a cable connector status detection method provided by the third embodiment of the present invention. This embodiment is applicable to the case where the cable connector status detection model trained by any of the above embodiments is used to detect the cable connector status. The method can be executed by a cable connector status detection device, which can be implemented by software and / or hardware, and can generally be integrated in an electronic device. The electronic device can be a terminal device or a server device. As long as it can execute the cable connector status detection method, the embodiment of the present invention does not limit the specific device type of the electronic device. Accordingly, Figure 4 and Figure 5 As shown, the method includes the following operations:
[0117] S310, obtaining currently collected data of the cable joint to be detected; wherein the currently collected data includes key state quantity data and induced current data of the cable joint to be detected.
[0118] The cable joint to be detected may be a cable joint that needs to be monitored in state using a target cable joint state detection model.
[0119] Exemplarily, the key state quantity data of the cable joint to be detected may include but are not limited to the operating years, average load rate, load change rate, surface temperature rise, ambient humidity, ambient hazardous gas content, and partial discharge of the cable joint to be detected. After the key state quantity data of the cable joint to be detected are collected, the operating state of the cable joint to be detected can be evaluated to obtain the current operating state, and the key state quantity data and the current operating state of the cable joint to be detected are used as the state quantity collection data of the cable joint to be detected.
[0120] For example, Figure 5As shown, if the improved PSO-BP neural network is used as the neural network model to be trained, after the improved PSO-BP neural network is trained with the cable joint state sample data to obtain the target cable joint state detection model, the key state quantity data and induced current data of the cable joint to be detected can be collected as the current collected data, so as to continue the state detection of the cable joint to be detected in real time through the target cable joint state detection model.
[0121] S320, extracting cable high-order harmonic features from the induced current data of the cable joint to be detected.
[0122] Optionally, the induced current data of the cable joint to be detected can be subjected to a fast Fourier transform to obtain a transformed current signal, and the capacitive component and resistive current of the leakage current can be further calculated, thereby obtaining the amplitude of the capacitive current of the target harmonic and the amplitude of the resistive current of the target harmonic based on the transformed current signal calculated from the induced current data of the cable joint to be detected, the capacitive component of the leakage current, and the resistive current, as the cable high-order harmonic characteristics of the induced current data of the cable joint to be detected.
[0123] S330, inputting the key state quantity data of the cable joint to be detected and the cable high-order harmonic characteristics into the target cable joint state detection model.
[0124] S340: Outputting the state detection result of the cable joint to be detected through the target cable joint state detection model.
[0125] The target cable joint state detection model is obtained by training using the cable joint state detection model training method described in any one of the above embodiments.
[0126] Correspondingly, such as Figure 5 As shown, after extracting the characteristic quantities of the cable high-order harmonic characteristics of the cable joint to be detected, the key state quantity data and the cable high-order harmonic characteristics of the cable joint to be detected can be input into the target cable joint state detection model, so as to output the state detection result of the cable joint to be detected through the target cable joint state detection model, thereby completing the evaluation of the cable joint operating status.
[0127] The above cable joint state detection method realizes a comprehensive evaluation of the cable joint operation state by introducing the target cable joint state detection model and combining multiple operation state quantities of the cable joint, thereby improving the accuracy and scientificity of the evaluation. The target cable joint state detection model can use multi-state quantity fusion to evaluate the cable joint state. The evaluation results reflect the degradation of electrical quantities and non-electrical quantities, and the evaluation results are more comprehensive and accurate. At the same time, the above cable joint state detection method can realize non-stop detection of cable heads, realize operation monitoring and management of cable joints, prevent the occurrence of operation failures of cable intermediate joints, and greatly improve the power supply reliability of the distribution network.
[0128] The embodiment of the present invention optimizes the network model parameters of the neural network model to be trained by adopting the PSO optimization algorithm to obtain the target network model parameters, thereby updating the network model parameters of the neural network model to be trained according to the target network model parameters, obtaining the cable joint state detection model to be trained, and then training the cable joint state detection model to be trained according to the cable joint state sample data to obtain the target cable joint state detection model. After the training of the target cable joint state detection model is completed, the current collected data such as the key state quantity data and the induced current data of the cable joint to be detected can be obtained, and the cable high-order harmonic characteristics of the induced current data of the cable joint to be detected can be extracted, so that the key state quantity data and the cable high-order harmonic characteristics of the cable joint to be detected are input into the target cable joint state detection model, so as to output the state detection result of the cable joint to be detected through the target cable joint state detection model. The above technical scheme can solve the problem of low detection accuracy in the existing cable joint state detection, improve the detection accuracy of the cable joint state detection model, and thus improve the accuracy of the cable joint state detection.
[0129] In the technical solution of the present disclosure, the collection, storage, use, processing, transmission, provision and disclosure of user personal information are in compliance with the relevant laws and regulations and do not violate public order and good morals.
[0130] It should be noted that any arrangement and combination of the technical features in the above embodiments also falls within the protection scope of the present invention.
[0131] Embodiment 4
[0132] Figure 6 is a schematic diagram of a cable joint state detection model training device provided by a fourth embodiment of the present invention, such as Figure 6 As shown, the device includes: a network model parameter acquisition module 410, a network model parameter optimization module 420, a network model parameter update module 430 and a detection model training module 440, wherein:
[0133] A network model parameter acquisition module 410 is used to acquire network model parameters of the neural network model to be trained;
[0134] A network model parameter optimization module 420 is used to optimize the network model parameters of the neural network model to be trained by using a PSO optimization algorithm to obtain target network model parameters;
[0135] A network model parameter updating module 430 is used to update the network model parameters of the neural network model to be trained according to the target network model parameters to obtain a cable joint state detection model to be trained;
[0136] A detection model training module 440 is used to train the cable joint state detection model to be trained according to the cable joint state sample data to obtain a target cable joint state detection model;
[0137] Wherein, the target cable joint state detection model is used to detect the cable joint state.
[0138] The embodiment of the present invention optimizes the network model parameters of the neural network model to be trained by adopting the PSO optimization algorithm to obtain the target network model parameters, thereby updating the network model parameters of the neural network model to be trained according to the target network model parameters to obtain the cable joint state detection model to be trained, and then training the cable joint state detection model to be trained according to the cable joint state sample data to obtain the target cable joint state detection model. After the target cable joint state detection model is trained, it can be used to detect the cable joint state. The above technical scheme can solve the problem of low detection accuracy in the existing cable joint state detection, can improve the detection accuracy of the cable joint state detection model, and thus improve the accuracy of the cable joint state detection.
[0139] Optionally, the network model parameter optimization module 420 is also used to: construct an initial particle population according to the network model parameters of the neural network model to be trained; calculate the fitness of each particle in the particle population and the global optimal solution of the particle population according to the fitness function; update the position and speed of each particle; perform a crossover operation and adopt a greedy strategy to update the particle population; return to execute the operation of calculating the fitness of each particle in the particle population and the global optimal solution of the particle population according to the fitness function until it is determined that the maximum evolutionary generation of the current stage is reached; when it is determined that the number of stages in the current stage is not the maximum number of stages, optimize and adjust the population size of the particle population and update the number of stages in the current stage; return to execute the operation of calculating the fitness of each particle in the particle population and the global optimal solution of the particle population according to the fitness function until it is determined that the maximum evolutionary generation of the current stage is reached; when it is determined that the number of stages in the current stage is the maximum number of stages, determine the target network model parameters according to the optimal particles of the current particle population.
[0140] Optionally, the network model parameter optimization module 420 is also used to: determine the degree of dispersion of the particle population; when it is determined that the degree of dispersion of the particle population is less than or equal to a set dispersion degree threshold, reduce the size of the particle population; when it is determined that the degree of dispersion of the particle population is greater than the set dispersion degree threshold, expand the size of the particle population.
[0141] Optionally, the above-mentioned device also includes a state sample data acquisition module, which is used to: collect key state quantity data of the cable connector in the operating state and the induced current data of the cable connector; wherein the key state quantity data includes at least one of operating years, average load rate, load change rate, surface temperature rise, ambient humidity, ambient hazardous gas content and partial discharge; evaluate the operating state of the cable connector in the operating state to obtain the current operating state; and use the key state quantity data and the current operating state as the state quantity sample data of the cable connector.
[0142] Optionally, the detection model training module 440 is also used to: extract cable high-order harmonic characteristics from the induced current data of the cable joint; input the cable high-order harmonic characteristics and the state quantity sample data of the cable joint into the cable joint state detection model to be trained for training; when it is determined that the training of the cable joint state detection model to be trained is completed, use the trained cable joint state detection model to be trained as the target cable joint state detection model.
[0143] Optionally, the detection model training module 440 is also used to: perform fast Fourier transform on the induced current data of the cable connector to obtain a transformed current signal; calculate the capacitive component and resistive current of the leakage current; and calculate the amplitude of the capacitive current of the target harmonic and the amplitude of the resistive current of the target harmonic based on the transformed current signal, the capacitive component of the leakage current and the resistive current.
[0144] The cable joint state detection model training device can execute the cable joint state detection model training method provided by any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of the execution method. For technical details not described in detail in this embodiment, please refer to the cable joint state detection model training method provided by any embodiment of the present invention.
[0145] Since the cable joint state detection model training device introduced above is a device that can execute the cable joint state detection model training method in the embodiment of the present invention, based on the cable joint state detection model training method introduced in the embodiment of the present invention, the technical personnel in this field can understand the specific implementation of the cable joint state detection model training device of this embodiment and its various variations, so how the cable joint state detection model training device implements the cable joint state detection model training method in the embodiment of the present invention is not introduced in detail here. As long as the technical personnel in this field implement the device used by the cable joint state detection model training method in the embodiment of the present invention, it belongs to the scope of protection of this application.
[0146] Embodiment 5
[0147] Figure 7 is a schematic diagram of a cable connector status detection device provided in Embodiment 5 of the present invention, such as Figure 7 As shown, the device includes: a current acquisition data acquisition module 510, a cable high-order harmonic feature extraction module 520, a data feature input module 530 and a state detection result output module 540, wherein:
[0148] The current acquisition data acquisition module 510 is used to acquire the current acquisition data of the cable joint to be detected; wherein the current acquisition data includes key state quantity data and induced current data of the cable joint to be detected;
[0149] A cable high-order harmonic feature extraction module 520, used to extract cable high-order harmonic features of the induced current data of the cable joint to be detected;
[0150] A data feature input module 530, used to input the key state quantity data of the cable joint to be detected and the cable high-order harmonic characteristics into a target cable joint state detection model;
[0151] A state detection result output module 540 is used to output the state detection result of the cable joint to be detected through the target cable joint state detection model;
[0152] The target cable joint state detection model is obtained by training the cable joint state detection model training method described in any embodiment of the present invention.
[0153] The embodiment of the present invention optimizes the network model parameters of the neural network model to be trained by adopting the PSO optimization algorithm to obtain the target network model parameters, thereby updating the network model parameters of the neural network model to be trained according to the target network model parameters, obtaining the cable joint state detection model to be trained, and then training the cable joint state detection model to be trained according to the cable joint state sample data to obtain the target cable joint state detection model. After the training of the target cable joint state detection model is completed, the current collected data such as the key state quantity data and the induced current data of the cable joint to be detected can be obtained, and the cable high-order harmonic characteristics of the induced current data of the cable joint to be detected can be extracted, so that the key state quantity data and the cable high-order harmonic characteristics of the cable joint to be detected are input into the target cable joint state detection model, so as to output the state detection result of the cable joint to be detected through the target cable joint state detection model. The above technical scheme can solve the problem of low detection accuracy in the existing cable joint state detection, improve the detection accuracy of the cable joint state detection model, and thus improve the accuracy of the cable joint state detection.
[0154] The above-mentioned cable joint state detection device can execute the cable joint state detection method provided by any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of the execution method. For technical details not described in detail in this embodiment, please refer to the cable joint state detection method provided by any embodiment of the present invention.
[0155] Since the cable joint state detection device introduced above is a device that can execute the cable joint state detection method in the embodiment of the present invention, based on the cable joint state detection method introduced in the embodiment of the present invention, the technical personnel in the field can understand the specific implementation of the cable joint state detection device in this embodiment and its various variations, so how the cable joint state detection device implements the cable joint state detection method in the embodiment of the present invention is not described in detail here. As long as the technical personnel in the field implement the device used by the cable joint state detection method in the embodiment of the present invention, it belongs to the scope of protection of this application.
[0156] Embodiment 6
[0157] Figure 8A schematic diagram of the structure of an electronic device 10 that can be used to implement an embodiment of the present invention is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices (such as helmets, glasses, watches, etc.) and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present invention described and / or required herein.
[0158] like Figure 8 As shown, the electronic device 10 includes at least one processor 11, and a memory connected to the at least one processor 11, such as a read-only memory (ROM) 12, a random access memory (RAM) 13, etc., wherein the memory stores a computer program that can be executed by at least one processor, and the processor 11 can perform various appropriate actions and processes according to the computer program stored in the read-only memory (ROM) 12 or the computer program loaded from the storage unit 18 to the random access memory (RAM) 13. In the RAM 13, various programs and data required for the operation of the electronic device 10 can also be stored. The processor 11, the ROM 12, and the RAM 13 are connected to each other through a bus 14. An input / output (I / O) interface 15 is also connected to the bus 14.
[0159] A number of components in the electronic device 10 are connected to the I / O interface 15, including: an input unit 16, such as a keyboard, a mouse, etc.; an output unit 17, such as various types of displays, speakers, etc.; a storage unit 18, such as a disk, an optical disk, etc.; and a communication unit 19, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 19 allows the electronic device 10 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.
[0160] The processor 11 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various dedicated artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, digital signal processors (DSPs), and any appropriate processors, controllers, microcontrollers, etc. The processor 11 executes the various methods and processes described above, such as a cable joint state detection model training method or a cable joint state detection method.
[0161] Optionally, the cable connector status detection model training method may include:
[0162] Obtain network model parameters of the neural network model to be trained;
[0163] The network model parameters of the neural network model to be trained are optimized by using a particle swarm optimization (PSO) optimization algorithm to obtain target network model parameters;
[0164] The network model parameters of the neural network model to be trained are updated according to the target network model parameters to obtain a cable joint state detection model to be trained;
[0165] The cable joint state detection model to be trained is trained according to the cable joint state sample data to obtain a target cable joint state detection model;
[0166] Wherein, the target cable joint state detection model is used to detect the cable joint state.
[0167] Optionally, the cable connector status detection method may include:
[0168] Acquire the current collected data of the cable joint to be detected; wherein the current collected data includes key state quantity data and induced current data of the cable joint to be detected;
[0169] Extracting cable high-order harmonic characteristics of the induced current data of the cable joint to be detected;
[0170] Inputting the key state quantity data of the cable joint to be detected and the cable high-order harmonic characteristics into the target cable joint state detection model;
[0171] Outputting the state detection result of the cable joint to be detected through the target cable joint state detection model;
[0172] The target cable joint state detection model is obtained by training the cable joint state detection model training method described in any embodiment of the present invention.
[0173] In some embodiments, the cable joint state detection model training method or the cable joint state detection method may be implemented as a computer program, which is tangibly contained in a computer-readable storage medium, such as a storage unit 18. In some embodiments, part or all of the computer program may be loaded and / or installed on the electronic device 10 via the ROM 12 and / or the communication unit 19. When the computer program is loaded into the RAM 13 and executed by the processor 11, one or more steps of the cable joint state detection model training method or the cable joint state detection method described above may be performed. Alternatively, in other embodiments, the processor 11 may be configured to execute the cable joint state detection model training method or the cable joint state detection method in any other appropriate manner (e.g., by means of firmware).
[0174] Various implementations of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), load programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various implementations can include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.
[0175] Computer programs for implementing the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, so that when the computer program is executed by the processor, the functions / operations specified in the flow chart and / or block diagram are implemented. The computer program may be executed entirely on the machine, partially on the machine, partially on the machine and partially on a remote machine as a stand-alone software package, or entirely on a remote machine or server.
[0176] In the context of the present invention, a computer-readable storage medium may be a tangible medium that may contain or store a computer program for use by or in combination with an instruction execution system, device or equipment. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices or equipment, or any suitable combination of the foregoing. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.
[0177] To provide interaction with a user, the systems and techniques described herein may be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and a pointing device (e.g., a mouse or trackball) through which the user can provide input to the electronic device. Other types of devices may also be used to provide interaction with the user; for example, the feedback provided to the user may be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user may be received in any form (including acoustic input, voice input, or tactile input).
[0178] The systems and techniques described herein may be implemented in a computing system that includes backend components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes frontend components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such backend components, middleware components, or frontend components. The components of the system may be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), a blockchain network, and the Internet.
[0179] A computing system may include a client and a server. The client and the server are generally remote from each other and usually interact through a communication network. The client and server relationship is generated by computer programs running on the corresponding computers and having a client-server relationship with each other. The server may be a cloud server, also known as a cloud computing server or cloud host, which is a host product in the cloud computing service system to solve the defects of difficult management and weak business scalability in traditional physical hosts and VPS services.
[0180] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps recorded in this disclosure can be executed in parallel, sequentially or in different orders, as long as the desired results of the technical solutions disclosed in this disclosure can be achieved, and this document does not limit this.
[0181] The above specific implementations do not constitute a limitation on the protection scope of the present disclosure. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principle of the present disclosure shall be included in the protection scope of the present disclosure.
Claims
1. A cable joint status detection model training method, It is characterized in that include: Obtain network model parameters of the neural network model to be trained; The network model parameters of the neural network model to be trained are optimized by using a particle swarm optimization (PSO) optimization algorithm to obtain target network model parameters; The network model parameters of the neural network model to be trained are updated according to the target network model parameters to obtain a cable joint state detection model to be trained; The cable joint state detection model to be trained is trained according to the cable joint state sample data to obtain a target cable joint state detection model; Wherein, the target cable joint state detection model is used to detect the cable joint state.
2. The method according to claim 1, It is characterized in that The PSO optimization algorithm is used to optimize the network model parameters of the neural network model to be trained to obtain the target network model parameters, including: Constructing an initial particle population according to the network model parameters of the neural network model to be trained; Calculating the fitness of each particle in the particle population and the global optimal solution of the particle population according to the fitness function; Update the position and velocity of each particle; Performing a crossover operation and adopting a greedy strategy to update the particle population; Return to execute the operation of calculating the fitness of each particle in the particle population and the global optimal solution of the particle population according to the fitness function, until it is determined that the maximum evolutionary generation of the current stage is reached; When it is determined that the number of stages in the current stage is not the maximum number of stages, the population size of the particle population is optimized and adjusted, and the number of stages in the current stage is updated; Return to execute the operation of calculating the fitness of each particle in the particle population and the global optimal solution of the particle population according to the fitness function, until it is determined that the maximum evolutionary generation of the current stage is reached; When it is determined that the stage number of the current stage is the maximum stage number, the target network model parameters are determined according to the optimal particles of the current particle population.
3. The method according to claim 2, It is characterized in that The optimizing and adjusting the population size of the particle population includes: determining a degree of dispersion of the particle population; When it is determined that the dispersion degree of the particle population is less than or equal to a set dispersion degree threshold, reducing the size of the particle population; When it is determined that the dispersion degree of the particle population is greater than the set dispersion degree threshold, the size of the particle population is expanded.
4. The method according to claim 1, It is characterized in that Before training the cable joint state detection model to be trained according to the cable joint state sample data, the method further includes: Collecting key state quantity data of the cable joint in the operating state and the induced current data of the cable joint; wherein the key state quantity data includes at least one of the operating years, average load rate, load change rate, surface temperature rise, ambient humidity, ambient hazardous gas content and partial discharge; Evaluate the operating state of the cable connector in the operating state to obtain the current operating state; The key state quantity data and the current operating state are used as state quantity sample data of the cable connector.
5. The method according to claim 4, It is characterized in that The step of training the cable joint state detection model to be trained according to the cable joint state sample data to obtain a target cable joint state detection model includes: Extracting cable high-order harmonic characteristics of the induced current data of the cable joint; Inputting the cable high-order harmonic characteristics and the state quantity sample data of the cable joint into the cable joint state detection model to be trained for training; When it is determined that the training of the cable joint state detection model to be trained is completed, the trained cable joint state detection model to be trained is used as the target cable joint state detection model.
6. The method according to claim 5, wherein the extracting of the cable high-order harmonic characteristics of the induced current data of the cable joint is include: Performing fast Fourier transform on the induced current data of the cable joint to obtain a transformed current signal; Calculate the capacitive component of the leakage current and the resistive current; The amplitude of the capacitive current of the target subharmonic and the amplitude of the resistive current of the target subharmonic are calculated according to the converted current signal, the capacitive component of the leakage current and the resistive current.
7. A method for detecting a cable joint state, It is characterized in that include: Acquire the current collected data of the cable joint to be detected; wherein the current collected data includes key state quantity data and induced current data of the cable joint to be detected; Extracting cable high-order harmonic characteristics of the induced current data of the cable joint to be detected; Inputting the key state quantity data of the cable joint to be detected and the cable high-order harmonic characteristics into the target cable joint state detection model; Outputting the state detection result of the cable joint to be detected through the target cable joint state detection model; Wherein, the target cable joint state detection model is obtained by training the cable joint state detection model training method according to any one of claims 1-6.
8. An electronic device, It is characterized in that The electronic device comprises: at least one processor; and a memory communicatively connected to the at least one processor; wherein, The memory stores a computer program that can be executed by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the cable joint status detection model training method described in any one of claims 1-6, or execute the cable joint status detection method described in claim 7.
9. A computer-readable storage medium, It is characterized in that The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the cable joint state detection model training method described in any one of claims 1 to 6, or to execute the cable joint state detection method described in claim 7 when executed.
10. A computer program product comprising a computer program / instructions, in, When the computer program / instructions are executed by a processor, the cable joint state detection model training method described in any one of claims 1 to 6 is implemented, or the cable joint state detection method described in claim 7 is executed.
Citation Information
Cited By
Cable
CN119650171A
Cable health condition evaluation system and method based on edge calculation
CN119830718A
