A method for evaluating the insulation performance of power cables

Abstract

The present invention relates to the technical field of insulation performance evaluation, and particularly to a method for evaluating the insulation performance of power cables. The method includes: adopting a measurement method based on time-frequency coupling enhancement and nonlinear adaptive filtering to collect the electrical parameters of the power cable in real time; preprocessing the electrical parameters of the power cable to obtain the preprocessed electrical parameters; and performing fusion processing on the preprocessed electrical parameters to obtain the fused electrical parameters; using an evaluation algorithm based on a hybrid neural network and fuzzy logic to evaluate the performance of the fused electrical parameters, so as to obtain the comprehensive health evaluation value of the power cable. The technical problem of inaccurate collection and processing of the electrical parameters of the power cable is solved.

Description

Technical Field

[0001] The present invention relates to the technical field of insulation performance evaluation, and in particular to a method for evaluating the insulation performance of a power cable. Background Art

[0002] The continuous development of power systems and the expansion of power grids are placing higher demands on the operational monitoring and maintenance management of power cables. Power cables play a vital role in modern power systems, being widely used in power transmission, distribution, and equipment connection. However, because power cables are often buried underground or hidden within buildings and exposed to various environmental factors, cable faults can be difficult to detect in a timely manner. Cable faults not only affect the stability and safety of power systems but can also cause equipment damage and economic losses. Therefore, how to monitor the electrical status of power cables in real time and identify potential faults early on has become a critical issue that needs to be addressed in the power industry.

[0003] Traditional power cable monitoring methods rely primarily on regular manual inspections and traditional electrical testing equipment, but these methods often fail to meet the requirements of real-time, comprehensive, and high-precision monitoring. With the continuous advancement of intelligent sensing, wireless communication, and data processing technologies, sensor network-based power cable monitoring systems are becoming an effective solution. These systems enable real-time data acquisition and fault diagnosis of power cables, and have broad application prospects, particularly in the monitoring of high-voltage power cables.

[0004] However, the existing power cable insulation performance evaluation method has the following technical problems: the collection and processing of the electrical parameters of the power cable are not accurate enough. Summary of the Invention

[0005] The present invention provides a method for evaluating the insulation performance of a power cable, so as to solve the technical problem of inaccurate collection and processing of electrical parameters of the power cable.

[0006] A method for evaluating the insulation performance of a power cable according to the present invention specifically includes the following technical solutions:

[0007] A method for evaluating the insulation performance of a power cable comprises the following steps:

[0008] S1. Using a measurement method based on time-frequency coupling enhancement and nonlinear adaptive filtering, collect electrical parameters of the power cable in real time; preprocess the electrical parameters of the power cable to obtain preprocessed electrical parameters; and fuse the preprocessed electrical parameters to obtain fused electrical parameters;

[0009] S2. The fused electrical parameters are evaluated using an evaluation algorithm based on a hybrid neural network and fuzzy logic to obtain a comprehensive health assessment value of the power cable.

[0010] Preferably, the S1 specifically includes:

[0011] In the measurement method based on time-frequency coupling enhancement and nonlinear adaptive filtering, sensors are deployed to collect signals in real time, and the time-frequency coupling analysis method is used to perform time-frequency coupling conversion on the collected signals to obtain the converted signals.

[0012] Preferably, the S1 specifically includes:

[0013] In the measurement method based on time-frequency coupling enhancement and nonlinear adaptive filtering, a nonlinear adaptive filtering algorithm is introduced to filter the converted signals. The nonlinear adaptive filtering algorithm dynamically adjusts the filtering parameters according to the changes in the converted signals to remove the noise of the converted signals and obtain the filtered signals.

[0014] Preferably, the S1 specifically includes:

[0015] The specific calculation formula of the filtered signal is:

[0016] ,

[0017] where is the filtered signal; is the window size of the adaptive filter; is the weighting coefficient of the adaptive filter at time and frequency ; is the input signal at time and frequency , that is, the converted signal at time and frequency ; is the nonlinear enhancement coefficient; is the number of historical time steps used in the nonlinear enhancement; is the converted signal at time and frequency .

[0018] Preferably, the S1 specifically includes:

[0019] In the measurement method based on time-frequency coupling enhancement and nonlinear adaptive filtering, after the converted signals are filtered, the filtered signals are enhanced to obtain the enhanced signals.

[0020] Preferably, the S1 specifically includes:

[0021] In the measurement method based on time-frequency coupling enhancement and nonlinear adaptive filtering, a time-frequency domain joint modeling method is introduced based on the enhanced signal. The time-frequency domain joint modeling method combines time-domain and frequency-domain information and extracts the electrical parameters of the power cable by performing integral calculations on the enhanced signal.

[0022] Preferably, the S1 specifically includes:

[0023] The implementation formula of the time-frequency domain joint modeling method is:

[0024] ,

[0025] where, are the electrical parameters of the power cable; is the enhanced signal; is the time-domain weighting coefficient; is the frequency-domain weighting coefficient; is the upper limit of the frequency part integral; is the upper limit of the time part integral.

[0026] Preferably, the S2 specifically includes:

[0027] In the implementation process of the evaluation algorithm based on the hybrid neural network and fuzzy logic, a neural network model is designed and trained. The neural network model consists of an input layer, a hidden layer, and an output layer; the input layer receives the fused electrical parameters; the hidden layer learns the relationships between different fused electrical parameters through multi-layer nonlinear transformations; and the output layer generates a preliminary health evaluation value.

[0028] Preferably, the S2 specifically includes:

[0029] In the implementation process of the evaluation algorithm based on the hybrid neural network and fuzzy logic, by introducing fuzzy rules and fuzzy membership functions, the health state of the power cable is inferred, a fuzzy value is output, and the fuzzy value is defuzzified to obtain a health evaluation value; the comprehensive health evaluation value of the power cable is obtained by weighted fusion of the preliminary health evaluation value in the output layer of the neural network model and the health evaluation value.

[0030] The beneficial effects of the technical solution of the present invention are:

[0031] 1. Under a multi-channel sensor network, using a measurement method based on time-frequency coupling enhancement and non-linear adaptive filtering to collect the electrical parameters of power cables in real time, better revealing the frequency-domain characteristics of signals, effectively solving the problem that traditional time-domain analysis methods cannot capture high-frequency signals. At the same time, through the adjustment of non-linear enhancement coefficients and adaptive filters, it can adaptively denoise and enhance electrical characteristics in dynamically changing signals, effectively eliminating noise from environmental interference and enhancing useful information in the signals. Finally, introducing composite signal enhancement processing to perform multiple enhancements and suppressions on the converted signals, improving the quality of the converted signals and the extraction accuracy of electrical parameters.

[0032] 2. Introducing an evaluation algorithm based on a hybrid neural network and fuzzy logic, by combining the deep learning ability of the neural network and the non-linear reasoning advantages of the fuzzy logic system, making the health state evaluation of power cables more robust. The training of the neural network model can learn the complex non-linear relationships between electrical parameters, while fuzzy logic reasoning can effectively handle the uncertainty of data, providing more accurate results for health evaluation. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] Figure 1 Flowchart of a method for evaluating the insulation performance of a power cable according to the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0034] In order to further elaborate on the technical means and effects adopted by the present invention to achieve the intended invention purpose, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0035] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the technical field to which the present invention belongs.

[0036] The following specifically describes the specific solution of a method for evaluating the insulation performance of a power cable provided by the present invention with reference to the accompanying drawings. Figure 1 It shows a flowchart of a method for evaluating the insulation performance of a power cable provided by an embodiment of the present invention, and the method includes the following steps:

[0037] S1. Adopting a measurement method based on time-frequency coupling enhancement and non-linear adaptive filtering to collect the electrical parameters of the power cable in real time; preprocessing the electrical parameters of the power cable to obtain preprocessed electrical parameters; and performing fusion processing on the preprocessed electrical parameters to obtain fused electrical parameters;

[0038] To achieve real-time monitoring of power cables, a multi-channel sensor network is designed. The multi-channel sensor network collects electrical parameters of power cables in real time through multi-channel configuration. The specific design includes arranging sensors, such as current sensors, voltage sensors, temperature and humidity sensors, dielectric loss sensors, etc., at key positions of power cables (such as joints, cable middles, access ends, corners, and parts under large loads) according to the layout of power cables. The key positions of electrical cables are determined by the expert experience method. In the deployed multi-channel sensor network, a measurement method based on time-frequency coupling enhancement and nonlinear adaptive filtering is used to collect electrical parameters of power cables in real time, including resistance, leakage current, dielectric loss, voltage, temperature, humidity, etc. The measurement method based on time-frequency coupling enhancement and nonlinear adaptive filtering converts signals through time-frequency coupling analysis method and introduces multiple enhancement processes to further extract electrical parameters of power cables. The specific implementation process is as follows:

[0039] First, start measuring electrical parameters of power cables from signal acquisition. By deploying sensors, such as dielectric loss sensors, current sensors, voltage sensors, temperature sensors, and humidity sensors, etc., signals are collected in real time at key positions of electrical cables ; denotes the signal collected by the th sensor at time , where

[0040] is a time variable. To better capture the frequency domain characteristics in the collected signal , a time-frequency coupling analysis method, such as wavelet transform and short-time Fourier transform, is used to perform time-frequency coupling conversion on the collected signal to obtain the converted signal ; the signal is transformed from the time domain to the time-frequency domain, thus retaining both the time information and frequency information of the signal, providing rich time-frequency information for subsequent signal enhancement and extraction of electrical parameters. For example, the signal is represented in the time-frequency domain as , where

[0041] is a frequency variable.

[0042] ,

[0043] ,

[0044] ,

[0045] wherein, is the filtered signal; is the window size of the adaptive filter, determined according to the specific scenario; is the weighting coefficient of the adaptive filter at time and frequency , representing the contribution degree of the input signal to the filtered signal ; is the input signal at time and frequency , i.e., the converted signal at time and frequency ; is the non - linear enhancement coefficient, representing the degree of non - linear enhancement of the converted signal, and is a dynamic adjustment factor that adjusts the intensity of non - linear enhancement according to the signal characteristics; is the number of historical time steps used in non - linear enhancement, controlling the historical range of non - linear enhancement, determined according to the specific scenario; is the converted signal at time and frequency ; is the weighting coefficient of the adaptive filter at the previous time and frequency ; is the adaptive step - size parameter, used to control the rate of weight update, determining the sensitivity of the adaptive filter to adjust weights, determined according to the expert experience method; is the desired output signal, which is the target output of the adaptive filter, compared with the actual filtered signal, and calculated through the existing theoretical model; is the adjustment factor, determining the influence degree of the signal energy of the converted signal on the non - linear enhancement coefficient , used to control the contribution of the fluctuation of the signal energy of the converted signal to non - linear enhancement, determined according to the expert experience method; is the adjustment parameter, used to control the response speed of the signal energy of the converted signal to the non - linear enhancement coefficient , determined through the experimental method; is the signal energy of the converted signal; is the length of the time window, representing the time period used to calculate the signal energy of the converted signal, determined according to the expert experience method.

[0046] After the converted signal is filtered, the filtered signal is further enhanced through a composite signal enhancement algorithm. The specific composite signal enhancement formula is as follows:

[0047] ,

[0048] where, is the enhanced signal; is the exponential enhancement coefficient, which is used to control the enhancement intensity of the filtered signal and is obtained through the experimental method; is the power coefficient of the exponent, which is used to control the relationship between the enhancement amplitude and the size of the filtered signal and is determined through the experimental method; is the suppression coefficient, which determines the suppression effect when the filtered signal is too strong and is determined according to the expert experience method; is the exponential coefficient in the suppression term, which determines the suppression sensitivity and is determined through the experimental method; is the power exponent of the suppression term, which is used to control the non-linear influence of the suppression intensity of the filtered signal and is determined through the experimental method; is the logarithmic enhancement coefficient, which is used to control the enhancement amplitude of the low-intensity filtered signal and is determined according to the expert experience method;

[0049] After the composite signal enhancement processing, the obtained enhanced signal , contains multi-dimensional electrical characteristics of the power cable. In order to extract the electrical parameters of the power cable from the enhanced signal, a time-frequency domain joint modeling method is introduced. The time-frequency domain joint modeling method combines time-domain and frequency-domain information, and through the integral calculation of the enhanced signal , extracts the electrical parameters of the power cable, such as resistance, dielectric loss, load current, etc. The implementation formula of the time-frequency domain joint modeling method is:

[0050] ,

[0051] where, is the electrical parameter of the power cable; is the time-domain weighting coefficient, which reflects the influence of the power of the frequency component on the electrical parameter of the power cable and is determined through the experimental method; is the frequency-domain weighting coefficient, which is used to control the influence of the cumulative effect of the enhanced signal in a specific frequency band on the electrical parameter of the power cable and is obtained through the experimental method; is the upper limit of the frequency part integral, which is determined according to the expert experience method; is the upper limit of the time part integral, which is determined according to the expert experience method.

[0052] Further, preprocess the electrical parameters of the power cable, such as smoothing, calibration, normalization, etc., to obtain the preprocessed electrical parameters. The preprocessing process uses existing technologies well-known to those skilled in the art and will not be elaborated here.

[0053] Finally, fuse the preprocessed electrical parameters using existing data fusion algorithms (such as weighted average method, multi-sensor data fusion algorithm, Kalman filter, etc.) to obtain the fused electrical parameters. 。

[0054] S2. Use an evaluation algorithm based on a hybrid neural network and fuzzy logic to evaluate the performance of the fused electrical parameters, and obtain the comprehensive health evaluation value of the power cable.

[0055] The evaluation algorithm based on a hybrid neural network and fuzzy logic combines the deep learning ability of the neural network and the non-linear reasoning advantage of the fuzzy logic system, can handle the complex relationships between various electrical parameters involved in the power cable monitoring process, and effectively cope with the uncertainty and fuzziness of the data. The specific implementation process is as follows:

[0056] First, design a neural network model. The neural network model, such as a feedforward neural network, consists of an input layer, a hidden layer, and an output layer. The input layer receives the fused electrical parameters. Each of its elements corresponds to a fused electrical parameter, such as resistance, dielectric loss, etc.; the hidden layer learns the potential relationships of the fused electrical parameters through multiple non-linear transformations; the output layer generates a preliminary health evaluation value. The preliminary health evaluation value is a quantitative indicator of the health status of the power cable. The specific number of transformation layers in the hidden layer of multiple non-linear transformations is determined according to the specific scenario, and the multiple non-linear transformations are implemented by selecting existing activation functions according to the expert experience method.

[0057] Further, use the existing backpropagation algorithm to adjust the weights and biases of the neural network model to achieve model training, so as to obtain a model that can accurately reflect the insulation health status of the power cable.

[0058] Further, introduce the existing fuzzy logic reasoning technology. The fuzzy logic reasoning technology reasons about the health status of the power cable by introducing fuzzy rules and fuzzy membership functions, and processes the fuzzy relationships between the fused electrical parameters. The input of the fuzzy logic reasoning process includes the preliminary health evaluation value generated by the output layer of the neural network model. And the fused electrical parameters, and define appropriate fuzzy membership functions according to the specific scenario to convert the input of the fuzzy logic reasoning process into fuzzy values. The fuzzy logic reasoning process combines fuzzy rules for decision-making and outputs fuzzy values.

[0059] The fuzzy membership function is as follows:

[0060] ,

[0061] ,

[0062] Among them, is the fused electrical parameter of any power cable, and represent the minimum and maximum values of the fused electrical parameters; and are the membership degrees of the low value and the high value respectively.

[0063] According to the membership degrees calculated by the fuzzy membership function, fuzzy rules are defined. For example: if the resistance is low and the current is low, the power cable is healthy; if the resistance is high and the current is high, the power cable is aged.

[0064] Through the above existing fuzzy logic inference technology, a fuzzy value is output, and the output fuzzy value is defuzzified by an existing defuzzification method such as the weighted average method to obtain a health assessment value .

[0065] Finally, the preliminary health assessment value in the output layer of the neural network model and the health assessment value of fuzzy logic inference

[0066] are weighted and fused to obtain the comprehensive health assessment value of the power cable. The goal of this weighted fusion process is to combine the advantages of the two methods to obtain a more accurate cable health assessment value.

[0067] In summary, a method for evaluating the insulation performance of a power cable is completed.

[0068] The sequence of the invention embodiments is only for description and does not represent the superiority or inferiority of the embodiments. The processes depicted in the drawings do not necessarily require the specific order or sequential order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0069] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that: they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements on some of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included in the protection scope of the present invention.

Claims

1. A method for evaluating the insulation performance of a power cable, characterized in that, It includes the following steps: S1. Adopt a measurement method based on time-frequency coupling enhancement and nonlinear adaptive filtering to collect the electrical parameters of the power cable in real time; In the measurement method based on time-frequency coupling enhancement and nonlinear adaptive filtering, deploy sensors to collect signals in real time, perform time-frequency coupling conversion on the collected signals to obtain the converted signals, perform filtering processing on the converted signals to obtain the filtered signals, perform enhancement processing on the filtered signals to obtain the enhanced signals; based on the enhanced signals, introduce a time-frequency domain joint modeling method, which combines time-domain and frequency-domain information, and extracts the electrical parameters of the power cable by performing integral calculation on the enhanced signals. The implementation formula of the time-frequency domain joint modeling method is: ， Among them, are the electrical parameters of the power cable; is the enhanced signal; is the time-domain weighting coefficient; is the frequency-domain weighting coefficient; is the upper limit of the frequency part integral; is the upper limit of the time part integral; Preprocess the electrical parameters of the power cable to obtain the preprocessed electrical parameters; and perform fusion processing on the preprocessed electrical parameters to obtain the fused electrical parameters; S2. Use an evaluation algorithm based on a hybrid neural network and fuzzy logic to evaluate the performance of the fused electrical parameters to obtain the comprehensive health evaluation value of the power cable.

2. The method for evaluating the insulation performance of a power cable according to claim 1, characterized in that The S1 specifically includes: In the measurement method based on time-frequency coupling enhancement and nonlinear adaptive filtering, introduce a nonlinear adaptive filtering algorithm to perform filtering processing on the converted signals. The nonlinear adaptive filtering algorithm dynamically adjusts the filtering parameters according to the changes in the converted signals to remove the noise of the converted signals and obtain the filtered signals.

3. The method for evaluating the insulation performance of a power cable according to claim 2, wherein The S1 specifically includes: The specific calculation formula of the filtered signal is: ， wherein, is the filtered signal; is the window size of the adaptive filter; is the weighted coefficient of the adaptive filter at time and frequency ; is the input signal at time and frequency , that is, the converted signal at time and frequency ; is the nonlinear enhancement coefficient; is the number of historical time steps used in the nonlinear enhancement; is the converted signal at time and frequency .

4. A method for evaluating the insulation performance of a power cable according to claim 1, characterized in that, The S2 specifically includes: In the implementation process of the evaluation algorithm based on a hybrid neural network and fuzzy logic, design and train a neural network model, which consists of an input layer, a hidden layer, and an output layer; the input layer receives the fused electrical parameters; the hidden layer learns the relationships between different fused electrical parameters through multi-layer nonlinear transformations; the output layer generates a preliminary health evaluation value.

5. The method for evaluating the insulation performance of a power cable according to claim 4, characterized in that, The S2 specifically includes: In the implementation process of the evaluation algorithm based on a hybrid neural network and fuzzy logic, introduce fuzzy rules and fuzzy membership functions to infer the health state of the power cable, output a fuzzy value, and perform defuzzification processing on the fuzzy value to obtain the health evaluation value; the comprehensive health evaluation value of the power cable is obtained by weighted fusion of the preliminary health evaluation value in the output layer of the neural network model and the health evaluation value.

Images