Fault detection method for power distribution overhead line

Abstract

The invention relates to the technical field of electric power, and discloses a power distribution overhead line fault detection method, which comprises the following steps: acquiring a target signal set, decomposing a target signal by using wavelet transform to obtain a multi-layer high-frequency detail coefficient and a low-frequency approximation coefficient, and decomposing the low-frequency approximation coefficient by using empirical mode decomposition to obtain a multi-layer high-frequency detail coefficient and a multi-layer low-frequency detail coefficient; the method comprises the steps of obtaining a plurality of intrinsic mode functions, then constructing a multi-dimensional feature vector corresponding to a target signal, and finally inputting each multi-dimensional feature vector corresponding to each signal in a target signal set into a preset line fault prediction model to obtain a fault classification detection result. According to the method, the signal is processed by fusing wavelet transform and empirical mode decomposition, so that the fault characteristics in the non-stationary signal can be extracted more effectively, the hidden fault in the power distribution overhead line can be detected quickly and accurately, the serious line fault risk caused by the hidden fault is effectively reduced, and the stable operation of a power system is ensured.

Description

Technical Field

[0001] The present invention relates to the field of electric power technology, and in particular to a method for detecting faults in overhead distribution lines. Background Art

[0002] In power systems, line fault detection is critical to ensuring the safe and stable operation of the entire system. As the power grid continues to expand, its complexity is also increasing. This is especially true for long-distance overhead distribution lines, where the hidden and complex nature of line faults is becoming increasingly pronounced. Overhead distribution lines are directly exposed to the external environment and are highly susceptible to climatic conditions (such as lightning strikes, freezing, and wind deflection) as well as external interference (such as tree contact and foreign object intrusion), making them prone to a variety of hidden faults. These faults often manifest as atypical, transient signal characteristics, such as localized arcing and intermittent contact faults.

[0003] The non-stationary, transient, and low-energy nature of hidden fault signals presents numerous challenges for existing fault analysis methods. Currently, most fault analysis methods rely on linear analysis tools (such as Fourier transforms) or employ simple feature extraction techniques. While these methods are effective for stationary signals, they have limited processing capabilities for non-stationary signals, making it difficult to accurately extract key fault features from these signals.

[0004] What is more serious is that if hidden faults are not effectively repaired for a long time, they are very likely to gradually evolve into serious line faults, which will pose a major threat to the stability of the power system and may even cause large-scale power outages, causing huge losses to social production and life.

[0005] Therefore, developing a technical solution that can quickly process non-stationary signals and accurately extract hidden fault characteristics is crucial to ensuring the intelligent operation of power systems. Summary of the Invention

[0006] Based on this, it is necessary to address the above problems and propose a fault detection method for distribution overhead lines, which can more effectively extract fault features in non-stationary signals, thereby quickly and accurately detecting hidden faults in distribution overhead lines, effectively reducing the risk of serious line failures caused by hidden faults, and ensuring the stable operation of the power system.

[0007] To achieve the above object, the present invention provides, in a first aspect, a method for detecting a fault in a distribution overhead line, the method comprising: Acquire a target signal set, wherein the target signal set includes three phase current signals, three phase voltage signals, and an electric field strength signal of a distribution overhead line; Decomposing the target signal using wavelet transform to obtain multiple layers of high-frequency detail coefficients and low-frequency approximation coefficients, wherein the target signal is any one signal in the target signal set; Decomposing the low-frequency approximation coefficients using empirical mode decomposition to obtain a plurality of intrinsic mode functions; Determine the mutual information between each intrinsic mode function and each layer's high-frequency detail coefficient, and take the intrinsic mode function corresponding to the mutual information greater than the mutual information threshold as the high-frequency intrinsic mode function; Constructing a multidimensional feature vector corresponding to the target signal based on all high-frequency intrinsic mode functions and all high-frequency detail coefficients; Each multidimensional feature vector corresponding to each signal in the target signal set is input into a preset line fault prediction model to obtain a fault classification detection result.

[0008] Optionally, the use of wavelet transform to decompose the target signal to obtain multiple layers of high-frequency detail coefficients and low-frequency approximation coefficients includes: The target signal is iteratively decomposed using the wavelet transform until the number of decomposition layers of the iterative decomposition is equal to the preset number of decomposition layers, and multiple layers of high-frequency detail coefficients corresponding to all iterative decompositions and the low-frequency approximation coefficients corresponding to the iterative decomposition are obtained.

[0009] Optionally, the method further includes: Using short-time Fourier transform, transforming the target signal to obtain a time-frequency spectrum; Performing frequency extraction on the time-frequency spectrum to obtain a signal dominant frequency; The preset number of decomposition layers is determined according to the dominant frequency of the signal.

[0010] Optionally, determining a preset number of decomposition layers according to the dominant frequency of the signal includes: Using the formula Determining the preset number of decomposition layers; in, is the preset decomposition level, is the floor symbol, is the logarithm to base 2, is the signal sampling frequency of the target signal, is the dominant frequency of the signal.

[0011] Optionally, the low-frequency approximation coefficients are decomposed using empirical mode decomposition to obtain a plurality of intrinsic mode functions, including: Using the empirical mode decomposition, the low-frequency approximation coefficients are iteratively decomposed until the energy of the residual component obtained by the iterative decomposition is less than or equal to the energy threshold, and multiple initial intrinsic mode functions corresponding to all iterative decompositions are obtained, and the preset initial intrinsic mode functions are all used as intrinsic mode functions.

[0012] Optionally, the method further includes: determining the energy of the target signal; A preset percentage of the energy of the target signal is used as the energy threshold.

[0013] Optionally, before inputting each multidimensional feature vector corresponding to each signal in the target signal set into a preset line fault prediction model to obtain a fault classification detection result, the method further includes: Using short-time Fourier transform, transforming the target signal to obtain a time-frequency spectrum; Arrange all high-frequency detail coefficients in sequence to obtain a coefficient vector; Determine the energy of each eigenmode function and arrange the energies of all eigenmode functions in sequence to obtain an energy vector; Constructing a three-dimensional tensor corresponding to the target signal according to the time-frequency spectrum, the coefficient vector, and the energy vector; The step of inputting each multidimensional feature vector corresponding to each signal in the target signal set into a preset line fault prediction model to obtain a fault classification detection result includes: Each multidimensional feature vector and each three-dimensional tensor corresponding to each signal in the target signal set is input into the preset line fault prediction model to obtain the fault classification detection result.

[0014] Optionally, the method further includes: Determine the current amplitude, voltage drop amplitude, harmonic content and fault time based on the three phase current signals and three phase voltage signals; Determine a fault type detection result based on the current amplitude, the voltage drop amplitude, the harmonic content, and the fault time using a preset fuzzy rule; A target fault classification detection result is determined according to the fault type detection result and the fault classification detection result.

[0015] Optionally, determining the mutual information between each intrinsic mode function and each layer of high-frequency detail coefficients includes: Determining a joint probability distribution between each intrinsic mode function and each layer of high-frequency detail coefficients, and determining a marginal probability distribution of each intrinsic mode function, and determining a marginal probability distribution of each layer of high-frequency detail coefficients; The mutual information between each eigenmode function and each layer of high-frequency detail coefficients is determined according to the joint probability distribution between each eigenmode function and each layer of high-frequency detail coefficients, the marginal probability distribution of each eigenmode function, and the marginal probability distribution of each layer of high-frequency detail coefficients.

[0016] Optionally, acquiring the target signal set includes: Acquiring three initial phase current signals, three initial phase voltage signals, and an initial electric field strength signal collected by an electromagnetic mutual inductor, wherein the electromagnetic mutual inductor has been installed on the power distribution overhead line; Performing Gaussian kernel denoising processing on each initial phase current signal, each initial phase voltage signal and the initial electric field strength signal to obtain three phase current signals, three phase voltage signals and an electric field strength signal; The three phase current signals, the three phase voltage signals and the electric field strength signal are all taken as elements to form the target signal set.

[0017] To achieve the above object, the present invention provides, in a second aspect, a fault detection device for a power distribution overhead line, the device comprising: An acquisition module is used to acquire a target signal set, wherein the target signal set includes three phase current signals, three phase voltage signals and an electric field strength signal of the distribution overhead line; A first decomposition module is configured to decompose a target signal using a wavelet transform to obtain multiple layers of high-frequency detail coefficients and low-frequency approximation coefficients, wherein the target signal is any one signal in the target signal set; A second decomposition module is used to decompose the low-frequency approximation coefficients using empirical mode decomposition to obtain a plurality of intrinsic mode functions; a determination module, configured to determine the mutual information between each intrinsic mode function and each layer's high-frequency detail coefficient, and to use the intrinsic mode function corresponding to the mutual information greater than a mutual information threshold as the high-frequency intrinsic mode function; A construction module, configured to construct a multidimensional feature vector corresponding to the target signal based on all high-frequency intrinsic mode functions and all high-frequency detail coefficients; The prediction module is used to input each multidimensional feature vector corresponding to each signal in the target signal set into a preset line fault prediction model to obtain a fault classification detection result.

[0018] To achieve the above-mentioned object, the present invention provides, in a third aspect, a computer-readable storage medium storing a computer program, wherein when the computer program is executed by a processor, the processor executes the method as described in any one of the first aspects.

[0019] To achieve the above-mentioned objectives, the present invention provides a computer device in a fourth aspect, comprising a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor executes the method as described in any one of the first aspects.

[0020] The embodiment of the present invention has the following beneficial effects: the above method obtains a target signal set, which includes three phase current signals, three phase voltage signals and electric field strength signals of the distribution overhead line, uses wavelet transform to decompose the target signal to obtain multiple layers of high-frequency detail coefficients and low-frequency approximation coefficients, wherein the target signal is any signal in the target signal set, and uses empirical mode decomposition to decompose the low-frequency approximation coefficient to obtain multiple intrinsic mode functions, and then determines the mutual information between each intrinsic mode function and each layer of high-frequency detail coefficients, and uses the intrinsic mode function corresponding to the mutual information greater than the mutual information threshold as the high-frequency intrinsic mode. function, and then construct the multidimensional feature vector corresponding to the target signal according to all high-frequency intrinsic mode functions and all high-frequency detail coefficients. Finally, the multidimensional feature vectors corresponding to each signal in the target signal set are input into the preset line fault prediction model to obtain the fault classification detection result; that is, by fusing wavelet transform and empirical mode decomposition to process the signal, it is possible to more effectively extract the fault characteristics in the non-stationary signal, thereby quickly and accurately detecting the hidden faults in the distribution overhead lines, providing reliable support for the intelligent operation of the power system, effectively reducing the risk of serious line failures caused by hidden faults, and ensuring the stable operation of the power system. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. 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 paying any creative work.

[0022] in: Figure 1 A schematic diagram of a method for detecting a fault in an overhead power distribution line according to an embodiment of the present application; Figure 2 This is a schematic diagram of a fault detection device for a power distribution overhead line according to an embodiment of the present application; Figure 3 1 is a diagram of the internal structure of a computer device in some embodiments. DETAILED DESCRIPTION

[0023] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0024] In power systems, line fault detection is critical to ensuring the safe and stable operation of the entire system. As the power grid continues to expand, its complexity is also increasing. This is especially true for long-distance overhead distribution lines, where the hidden and complex nature of line faults is becoming increasingly pronounced. Overhead distribution lines are directly exposed to the external environment and are highly susceptible to climatic conditions (such as lightning strikes, freezing, and wind deflection) as well as external interference (such as tree contact and foreign object intrusion), making them prone to a variety of hidden faults. These faults often manifest as atypical, transient signal characteristics, such as localized arcing and intermittent contact faults.

[0025] The non-stationary, transient, and low-energy nature of hidden fault signals presents numerous challenges for existing fault analysis methods. Currently, most fault analysis methods rely on linear analysis tools (such as Fourier transforms) or employ simple feature extraction techniques. While these methods are effective for stationary signals, they have limited processing capabilities for non-stationary signals, making it difficult to accurately extract key fault features from these signals.

[0026] What is more serious is that if hidden faults are not effectively repaired for a long time, they are very likely to gradually evolve into serious line faults, which will pose a major threat to the stability of the power system and may even cause large-scale power outages, causing huge losses to social production and life.

[0027] Therefore, developing a technical solution that can quickly process non-stationary signals and accurately extract hidden fault characteristics is crucial to ensuring the intelligent operation of power systems.

[0028] In response to the above problems, the present application proposes a fault detection method for distribution overhead lines, which can more effectively extract fault features in non-stationary signals, thereby quickly and accurately detecting hidden faults in distribution overhead lines, effectively reducing the risk of serious line failures caused by hidden faults, and ensuring the stable operation of the power system. The specific implementation principle will be described in detail in the following embodiments.

[0029] In a first aspect, the present application provides a method for detecting faults in a power distribution overhead line.

[0030] See also Figure 1, is a schematic diagram of a method for detecting a fault in a power distribution overhead line according to an embodiment of the present application, the method comprising: Step 110: Acquire a target signal set, where the target signal set includes three phase current signals, three phase voltage signals, and an electric field strength signal of the distribution overhead line.

[0031] The overhead distribution lines here refer to the lines that require fault detection.

[0032] Regarding the signal collection method, in some embodiments, a signal collection device can be installed on the distribution overhead line to collect the three-phase current signals, three-phase voltage signals and electric field strength signals of the distribution overhead line through the signal collection device; wherein, the signal collection device can be configured with multiple high-speed collection channels, that is, each signal collection can be configured with a high-speed collection channel.

[0033] It can be understood that by configuring multiple high-speed acquisition channels, it is possible to achieve simultaneous high-speed acquisition of signals between different channels, thereby ensuring the timing consistency of the acquired signals and avoiding time offset errors between signals of different channels due to different timing of the acquired signals, thereby affecting subsequent detection results.

[0034] Furthermore, in some embodiments, the signal acquisition device also supports configurations of multiple sampling frequencies and multiple resolutions, which can be adjusted according to actual needs; for example, after a fault is detected, it can be adjusted to a high-frequency mode for signal acquisition, and after no fault is detected, it can be adjusted to a low-frequency mode or a normal mode for signal acquisition.

[0035] Step 120: Decompose the target signal using wavelet transform to obtain multiple layers of high-frequency detail coefficients and low-frequency approximation coefficients, where the target signal is any signal in the target signal set.

[0036] Among them, wavelet transform is WT, which is Wavelet Transform.

[0037] Regarding the decomposition method of multiple layers of high-frequency detail coefficients, in some embodiments, wavelet transform may be used to sequentially decompose the target signal to obtain multiple layers of high-frequency detail coefficients obtained by decomposition.

[0038] Regarding the decomposition method of the low-frequency approximation coefficients, in some embodiments, wavelet transform can be used to decompose the target signal in sequence to obtain multiple layers of low-frequency approximation coefficients, and then, among the multiple layers of low-frequency approximation coefficients, any layer of low-frequency approximation coefficients can be used as the low-frequency approximation coefficients obtained by decomposition.

[0039] Step 130: Use empirical mode decomposition to decompose the low-frequency approximation coefficients to obtain multiple intrinsic mode functions.

[0040] Among them, empirical mode decomposition is EMD, which is Empirical Mode Decomposition.

[0041] It should be noted that the present application decomposes the low-frequency approximate coefficients obtained by wavelet transform decomposition through empirical mode decomposition, which can finely separate and focus on the fault characteristics of hidden faults to improve subsequent detection results.

[0042] Regarding the decomposition method of multiple intrinsic mode functions, in some embodiments, empirical mode decomposition is used to sequentially decompose low-frequency approximation coefficients to obtain multiple intrinsic mode functions obtained by decomposition.

[0043] Step 140: Determine the mutual information between each IMF and each layer of high-frequency detail coefficients, and use the IMF corresponding to the mutual information greater than the mutual information threshold as the high-frequency IMF.

[0044] The mutual information threshold may be obtained and pre-set by an operator based on a large amount of experience, experiments or statistics. Of course, it may also be set by the operator based on actual needs.

[0045] It should be noted that there is a mutual information between each layer of high-frequency detail coefficients and the nth intrinsic mode function, that is, for the nth intrinsic mode function, there are multiple mutual information corresponding to it. Among the multiple mutual information, as long as there is at least one mutual information greater than the mutual information threshold, the nth intrinsic mode function will be regarded as the high-frequency intrinsic mode function.

[0046] Step 150: Construct a multi-dimensional feature vector corresponding to the target signal based on all high-frequency intrinsic mode functions and all high-frequency detail coefficients.

[0047] Regarding the construction method of multidimensional feature vectors, in some embodiments, all high-frequency intrinsic mode functions and all high-frequency detail coefficients can be directly used as multidimensional feature vectors, that is, all high-frequency intrinsic mode functions and all high-frequency detail coefficients can be arranged in sequence to form a multidimensional feature vector.

[0048] Regarding the method of constructing a multidimensional feature vector, in other embodiments, all high-frequency intrinsic mode functions and all high-frequency detail coefficients can be used as elements to form a high-frequency feature set, and then the weight, skewness, kurtosis and approximate entropy of each data in the high-frequency feature set are determined. Finally, a multidimensional feature vector is constructed based on the weight, skewness, kurtosis and approximate entropy of all data in the high-frequency feature set.

[0049] Furthermore, in some embodiments, the weight of each data in the high-frequency feature set can be determined using the formula Determine the weight of each data in the high-frequency feature set; is the weight of the nth data in the high-frequency feature set, is the nth data in the high-frequency feature set, is the L2 normal form symbol, is the total number of data in the high-frequency feature set.

[0050] Furthermore, for a multidimensional feature vector constructed based on the weights, skewness, kurtosis, and approximate entropy of all data in the high-frequency feature set, in some embodiments, the expression of the constructed multidimensional feature vector is: ;in, is a multidimensional feature vector, is the weight of the nth data in the high-frequency feature set, is the skewness of the nth data in the high-frequency feature set, is the nth data in the high-frequency feature set, is the kurtosis of the nth data in the high-frequency feature set, is the approximate entropy of the nth data in the high-frequency feature set, is the total number of data in the high-frequency feature set.

[0051] It should be noted that this application constructs a multi-dimensional feature vector corresponding to the target signal by adopting weights, skewness, kurtosis and approximate entropy, which can accurately extract key fault features from complex hidden fault signals to effectively improve the accuracy of subsequent model fault detection.

[0052] Step 160: Input each multidimensional feature vector corresponding to each signal in the target signal set into a preset line fault prediction model to obtain a fault classification detection result.

[0053] The preset line fault prediction model herein refers to a pre-trained model used to predict and output fault classification detection results based on the multi-dimensional feature vectors corresponding to the input signals.

[0054] It should be noted that each multidimensional feature vector corresponding to each signal in the target signal set can be obtained by repeating steps 120 to 150 multiple times, which will not be described in detail here.

[0055] Regarding the method for obtaining the preset line fault prediction model, in some embodiments, the multidimensional feature vectors corresponding to each signal in a large number of historical signal sets and the corresponding fault classification labels can be obtained, and then the multidimensional feature vectors corresponding to each signal in the large number of historical signal sets and the corresponding fault classification labels are sequentially input into the initial line fault prediction model for training. After training to a certain extent, a trained preset line fault prediction model is obtained; wherein, the fault classification label can be used as the true value of the initial line fault prediction model, that is, by comparing the true value with the fault classification detection result output during the training process, it can be determined whether the initial line fault prediction model is well trained and meets the expected requirements.

[0056] Regarding the type of initial line fault prediction model, in some embodiments, a CNN-LSTM-Transformer model can be selected as the initial line fault prediction model; wherein CNN stands for Convolutional Neural Network, and LSTM stands for Long Short-Term Memory.

[0057] In the embodiment of the present application, by fusing wavelet transform and empirical mode decomposition to process the signal, the fault characteristics in the non-stationary signal can be more effectively extracted, thereby quickly and accurately detecting the hidden faults in the distribution overhead lines, providing reliable support for the intelligent operation of the power system, effectively reducing the risk of serious line failures caused by hidden faults, and ensuring the stable operation of the power system.

[0058] In addition, in addition to the above-mentioned effects of being able to quickly and accurately detect hidden faults, reduce the risk of serious line failures, and ensure stable operation of the power system, the distribution overhead line fault detection method proposed in this application also has the following advantages: Guaranteeing timing consistency: By configuring multiple high-speed acquisition channels in the signal acquisition device, simultaneous high-speed acquisition of signals between different channels is achieved, ensuring the timing consistency of the acquired signals, avoiding time offset errors caused by different signal timings, and thus providing a reliable data basis for subsequent accurate fault detection; Flexible sampling configuration: The signal acquisition device supports multiple sampling frequencies and multiple resolutions, and can be adjusted according to actual needs. For example, after a fault is detected, it is adjusted to a high-frequency mode for signal acquisition, which can capture the fault signal characteristics in more detail, and adjusted to a low-frequency or normal mode when there is no fault, which not only meets the acquisition needs in different scenarios, but also saves acquisition resources and storage space to a certain extent; Refined separation of fault characteristics: The low-frequency approximate coefficients obtained by wavelet transform decomposition are decomposed using empirical mode decomposition, which can refine Separating and focusing on the fault characteristics of hidden faults. Through this multi-level decomposition and combination processing method, more subtle and critical fault information can be mined from complex signals, improving the ability to extract hidden fault characteristics and helping to more accurately identify the fault type and location; accurately screening high-frequency intrinsic mode functions: by determining the mutual information between each intrinsic mode function and each layer of high-frequency detail coefficients, and taking the intrinsic mode function corresponding to the mutual information greater than the mutual information threshold as the high-frequency intrinsic mode function, this screening method can effectively find the parts with a strong correlation with the high-frequency detail coefficients from many intrinsic mode functions, further highlighting the features related to the fault, removing irrelevant or interfering information, and improving the quality and effectiveness of the features; improving fault detection accuracy: using statistics such as weight, skewness, kurtosis and approximate entropy to construct multidimensional feature vectors. These statistics can describe the characteristics of the data from different angles and accurately extract key fault characteristics from complex hidden fault signals, thereby effectively improving the accuracy of subsequent model fault detection and reducing the possibility of misjudgment and omission. Select an advanced model architecture: The CNN-LSTM-Transformer model is selected as the initial line fault prediction model. This model combines the advantages of convolutional neural networks (CNN) in feature extraction, long short-term memory networks (LSTM) in time series processing, and Transformer in capturing long-distance dependencies. It can more effectively process information in multi-dimensional feature vectors, further improving the accuracy and reliability of fault classification and detection.

[0059] In a feasible implementation, step 120 in the above embodiment uses wavelet transform to decompose the target signal to obtain multiple layers of high-frequency detail coefficients and low-frequency approximation coefficients, including: using wavelet transform to iteratively decompose the target signal until the number of decomposition layers of the iterative decomposition is equal to the preset number of decomposition layers, and obtain multiple layers of high-frequency detail coefficients corresponding to all iterative decompositions, and low-frequency approximation coefficients corresponding to the iterative decomposition.

[0060] The preset number of decomposition layers may be obtained and pre-set by the operator based on a large amount of experience, experiments or statistics, and of course, may also be set by the operator based on actual needs.

[0061] Regarding the value of the preset number of decomposition layers, in some embodiments, the present application may preferably set the preset number of decomposition layers to 6.

[0062] For the iterative decomposition method of multiple layers of high-frequency detail coefficients, in some embodiments, assuming that there are six layers of high-frequency detail coefficients, wavelet transform can be used to decompose the target signal to obtain the first layer of high-frequency detail coefficients, decompose the first layer of high-frequency detail coefficients to obtain the second layer of high-frequency detail coefficients, decompose the second layer of high-frequency detail coefficients to obtain the third layer of high-frequency detail coefficients, decompose the third layer of high-frequency detail coefficients to obtain the fourth layer of high-frequency detail coefficients, decompose the fourth layer of high-frequency detail coefficients to obtain the fifth layer of high-frequency detail coefficients, decompose the fifth layer of high-frequency detail coefficients to obtain the sixth layer of high-frequency detail coefficients, so as to obtain six layers of high-frequency detail coefficients corresponding to all iterative decompositions.

[0063] Regarding the iterative decomposition method of low-frequency approximation coefficients, in some embodiments, assuming that there are six layers of high-frequency detail coefficients, wavelet transform can be used to decompose the target signal to obtain the first layer of low-frequency approximation coefficients, decompose the first layer of low-frequency approximation coefficients to obtain the second layer of low-frequency approximation coefficients, decompose the second layer of low-frequency approximation coefficients to obtain the third layer of low-frequency approximation coefficients, decompose the third layer of low-frequency approximation coefficients to obtain the fourth layer of low-frequency approximation coefficients, decompose the fourth layer of low-frequency approximation coefficients to obtain the fifth layer of low-frequency approximation coefficients, decompose the fifth layer of low-frequency approximation coefficients to obtain the sixth layer of low-frequency approximation coefficients, and use the sixth layer of low-frequency approximation coefficients as the low-frequency approximation coefficients corresponding to the iterative decomposition.

[0064] In the embodiment of the present application, by iteratively decomposing the target signal, multiple layers of high-frequency detail coefficients and low-frequency approximation coefficients can be obtained more accurately, providing a more reliable basis for subsequent fault feature extraction and detection.

[0065] It can be understood that accurately obtaining multi-level coefficients: By iteratively decomposing until the preset number of decomposition levels is reached, the high-frequency detail coefficients and final low-frequency approximation coefficients of the target signal at different decomposition levels can be systematically obtained. This multi-level decomposition method helps to more comprehensively understand the characteristics of the signal and provide rich information for subsequent fault feature extraction; adapting to different fault characteristics: different faults may exhibit different characteristics at different decomposition levels. By presetting the number of decomposition levels, the decomposition depth can be adjusted according to actual needs to better adapt to the feature extraction requirements of different types of hidden faults and improve the accuracy of fault detection; improving the quality of feature extraction: the iterative decomposition method can gradually refine the signal, so that the obtained high-frequency detail coefficients and low-frequency approximation coefficients can more accurately reflect the essential characteristics of the signal, which helps to more effectively extract key fault features in subsequent steps, reduce the interference of irrelevant information or noise, and improve the quality of fault feature extraction; enhancing the versatility of the method: the preset number of decomposition levels can be set according to actual conditions, making the method universal and flexible. Whether facing faults of different types and severities, or distribution overhead lines under different operating environments, the fault detection effect can be optimized by adjusting the preset number of decomposition levels.

[0066] In a feasible implementation, the method in the above embodiment also includes: using short-time Fourier transform to transform the target signal to obtain a time-frequency spectrum; performing frequency extraction on the time-frequency spectrum to obtain the dominant frequency of the signal; and determining a preset number of decomposition layers based on the dominant frequency of the signal.

[0067] Regarding the method for determining the preset number of decomposition layers, in some embodiments, the signal acquisition frequency of the target signal may be obtained, and the preset number of decomposition layers may be determined based on the signal acquisition frequency and the signal dominant frequency.

[0068] In the embodiment of the present application, by dynamically determining the preset number of decomposition layers according to the dominant frequency of the signal, the adaptability and accuracy of fault detection can be improved.

[0069] It can be understood that, the adaptability is enhanced: different distribution overhead line fault signals have different dominant frequencies. The time-frequency spectrum is obtained by short-time Fourier transform and the dominant frequency of the signal is extracted. Then, the preset decomposition layer number is determined according to the dominant frequency, so that the method can be adaptively adjusted according to the characteristics of different fault signals, thereby enhancing the adaptability of the fault detection method to different types of faults; the accuracy is improved: the preset decomposition layer number directly affects the effect of fault feature extraction. If the decomposition layer number is set unreasonably, it may cause key fault features to be ignored or excessive interference from irrelevant information. The preset decomposition layer number is dynamically determined according to the dominant frequency of the signal, which can more accurately obtain high-frequency detail coefficients and low-frequency approximation coefficients related to the fault, thereby more effectively extracting key fault features and improving the accuracy of fault detection; resource utilization is optimized: dynamic determination of the preset decomposition layer number can avoid over-decomposition or under-decomposition. Over-decomposition will increase computational complexity and time cost, while under-decomposition may not fully extract fault features. By reasonably determining the decomposition layer number according to the dominant frequency of the signal, the utilization of computing resources can be optimized and the detection efficiency can be improved under the premise of ensuring the accuracy of fault detection.

[0070] In a feasible implementation, the method of determining the preset number of decomposition layers according to the dominant frequency of the signal in the above embodiment includes: Using the formula Determine the preset number of decomposition layers; in, To preset the number of decomposition layers, is the floor symbol, is the logarithm to base 2, is the signal sampling frequency of the target signal, is the dominant frequency of the signal.

[0071] In the embodiment of the present application, by using a formula to determine the preset number of decomposition layers according to the dominant frequency of the signal, a scientific and reasonable determination of the number of decomposition layers is achieved, thereby improving the accuracy and efficiency of fault detection.

[0072] It can be understood that scientifically determining the number of decomposition layers: calculating the preset number of decomposition layers through a formula avoids the subjectivity and arbitrariness of manually setting the number of decomposition layers, making the determination of the number of decomposition layers more scientific and reasonable. The formula comprehensively considers the signal sampling frequency and the signal dominant frequency. These two factors are key factors affecting the signal decomposition effect. Therefore, the decomposition layer number calculated by this formula can better adapt to the signal characteristics; improving fault detection accuracy: a reasonable number of decomposition layers can ensure that fault features are fully extracted during the signal decomposition process, while avoiding interference from irrelevant information or noise. The decomposition layer number determined by the formula can more accurately obtain high-frequency detail coefficients and low-frequency approximation coefficients related to the fault, thereby more effectively extracting key fault features and improving the accuracy of fault detection; optimizing computing resource utilization: dynamically determining the preset number of decomposition layers can avoid over-decomposition or under-decomposition, thereby optimizing the utilization of computing resources. Over-decomposition will increase computational complexity and time cost, while under-decomposition may not be able to fully extract fault features. Reasonably determining the number of decomposition layers through a formula can improve detection efficiency and reduce computing costs while ensuring fault detection accuracy.

[0073] In a feasible implementation, step 130 in the above embodiment uses empirical mode decomposition to decompose the low-frequency approximation coefficients to obtain multiple intrinsic mode functions, including: using empirical mode decomposition to iteratively decompose the low-frequency approximation coefficients until the energy of the residual component obtained by the iterative decomposition is less than or equal to the energy threshold, obtaining multiple initial intrinsic mode functions corresponding to all iterative decompositions, and using the previously preset initial intrinsic mode functions as the intrinsic mode functions.

[0074] The energy threshold and the specific number of the preset initial eigenmode functions can be obtained and pre-set by the operator based on a large amount of experience, experiments or statistics. Of course, they can also be set by the operator according to actual needs.

[0075] Regarding the value of the energy threshold, in some embodiments, the present application may preferably set the energy threshold to 0.7.

[0076] Regarding the specific number of the first preset initial eigenmode functions, in some embodiments, the present application may preferably set the first preset initial eigenmode functions to the first three initial eigenmode functions.

[0077] For the iterative decomposition method of multiple initial intrinsic mode functions, in some embodiments, assuming that the energy of the residual component obtained after six decompositions is less than or equal to the energy threshold, empirical mode decomposition can be used to decompose the target signal to obtain the first initial intrinsic mode function and the first residual component, decompose the first residual component to obtain the second initial intrinsic mode function and the second residual component, decompose the second residual component to obtain the third initial intrinsic mode function and the third residual component, decompose the third residual component to obtain the fourth initial intrinsic mode function and the fourth residual component, decompose the fourth residual component to obtain the fifth initial intrinsic mode function and the fifth residual component, decompose the fifth residual component to obtain the sixth initial intrinsic mode function and the sixth residual component, at this time, the sixth residual component is less than or equal to the energy threshold, and the multiple initial intrinsic mode functions are six initial intrinsic mode functions.

[0078] It should be noted that if the first three initial eigenmode functions are set as eigenmode functions, and the energy of the remaining components obtained after decomposition twice can be less than or equal to the energy threshold, then the number of eigenmode functions is only two, and no further decomposition is required.

[0079] In the embodiment of the present application, the intrinsic mode function is obtained by iterative decomposition by setting an energy threshold, which effectively avoids noise interference caused by over-decomposition and improves the accuracy and reliability of fault feature extraction.

[0080] It can be understood that over-decomposition is avoided: in the actual signal decomposition process, if the decomposition is excessive, a large amount of noise and irrelevant information will be introduced. These noise and irrelevant information will interfere with the extraction of fault features and reduce the accuracy of fault detection. By setting an energy threshold, the decomposition is stopped when the energy of the residual component obtained by iterative decomposition is less than or equal to the threshold. This can effectively avoid over-decomposition and reduce the impact of noise and irrelevant information on fault feature extraction; improve feature extraction accuracy: since over-decomposition is avoided, the obtained intrinsic mode function can more accurately reflect the essential characteristics related to the fault in the signal. In this way, when constructing multi-dimensional feature vectors and performing fault detection in the subsequent process, it can be based on more accurate and relevant feature information, thereby improving the accuracy of fault feature extraction and thus improving the reliability of the entire fault detection method; enhance the robustness of the method: in the actual distribution overhead line fault detection scenario, the signal is often affected by various noises and interferences. The method of setting an energy threshold for iterative decomposition can resist these noises and interferences to a certain extent, enhance the adaptability and robustness of the method to different environmental conditions, and ensure that fault features can be effectively extracted under different conditions and accurate fault detection can be achieved.

[0081] In a feasible implementation, the method in the above embodiment further includes: determining the energy of the target signal; and using a preset percentage of the energy of the target signal as an energy threshold.

[0082] The preset percentage may be obtained and set in advance by the operator based on a large amount of experience, experiments or statistics. Of course, it may also be set by the operator based on actual needs.

[0083] Regarding the value of the preset percentage, in some embodiments, the present application may preferably set the preset percentage to 5%.

[0084] In the embodiment of the present application, by dynamically determining the energy threshold, the adaptability and accuracy of fault detection can be improved.

[0085] It can be understood that by determining the energy of the target signal and using a preset percentage of the target signal energy as the energy threshold, the energy threshold can be dynamically adjusted according to the actual situation of the target signal. This method avoids the inadaptability that may be caused by a fixed energy threshold, because different target signals may have different energy levels. The dynamically determined energy threshold can better adapt to various signal conditions, thereby more accurately determining the intrinsic mode function in the empirical mode decomposition process, improving the accuracy and reliability of fault feature extraction, and thereby improving the performance of the entire fault detection method.

[0086] In a feasible implementation, in step 160 of the above embodiment, before inputting each multidimensional feature vector corresponding to each signal in the target signal set into a preset line fault prediction model to obtain a fault classification detection result, the method further includes: using a short-time Fourier transform to transform the target signal to obtain a time-frequency spectrum; arranging all high-frequency detail coefficients in sequence to obtain a coefficient vector; determining the energy of each intrinsic mode function, and arranging the energies of all intrinsic mode functions in sequence to obtain an energy vector; and constructing a three-dimensional tensor corresponding to the target signal based on the time-frequency spectrum, the coefficient vector and the energy vector.

[0087] Step 160 in the above embodiment, inputting each multidimensional feature vector corresponding to each signal in the target signal set into a preset line fault prediction model to obtain a fault classification detection result, includes: inputting each multidimensional feature vector and each three-dimensional tensor corresponding to each signal in the target signal set into a preset line fault prediction model to obtain a fault classification detection result.

[0088] Regarding the construction of a three-dimensional tensor, in some embodiments, the product of the time-frequency spectrum, the coefficient vector, and the energy vector may be used as the three-dimensional tensor.

[0089] In the embodiment of the present application, by constructing a three-dimensional tensor and combining it with a multi-dimensional feature vector for fault detection, the accuracy and comprehensiveness of fault classification detection are improved.

[0090] It can be understood that providing richer fault information: by using short-time Fourier transform to obtain a time-frequency spectrum, arranging all high-frequency detail coefficients into coefficient vectors, and determining the energy of each intrinsic mode function and arranging them into energy vectors, and finally constructing a three-dimensional tensor, this three-dimensional tensor integrates the information of the signal in multiple dimensions such as time domain, frequency domain and energy distribution, providing a richer and more comprehensive feature representation for fault detection, which helps to capture fault characteristics more accurately; enhancing the model's ability to identify faults: the multidimensional feature vectors corresponding to each signal in the target signal set and the constructed three-dimensional tensor are input into the preset line fault prediction model together. The multidimensional feature vector mainly describes the fault characteristics from the perspective of the statistical characteristics of the signal, while the three-dimensional tensor It supplements information from the perspective of time-frequency energy distribution. The combination of the two enables the model to learn and understand fault characteristics from different angles and levels, thereby greatly enhancing the model's ability to identify various types of faults and improving the accuracy of fault classification detection; adapting to complex and changeable fault conditions: In the actual operation of distribution overhead lines, the fault conditions are complex and diverse. By introducing a three-dimensional tensor, a feature representation method that contains multi-dimensional information, the fault detection method can better adapt to these complex and changeable fault conditions. Regardless of the type of hidden fault, this method can rely on input data that integrates multi-dimensional information to more effectively extract key fault characteristics, achieve accurate fault classification detection, and ensure the stable operation of the power system.

[0091] In a feasible implementation, the method in the above embodiment also includes: determining the current amplitude, voltage drop amplitude, harmonic content and fault time based on three phase current signals and three phase voltage signals; using preset fuzzy rules to determine the fault type detection result based on the current amplitude, voltage drop amplitude, harmonic content and fault time; determining the target fault classification detection result based on the fault type detection result and the fault classification detection result.

[0092] The preset fuzzy rules may be obtained and pre-set by the operator based on a large amount of experience, experiments or statistics. Of course, they may also be set by the operator based on actual needs.

[0093] Regarding the method for determining the preset fuzzy rules, in some embodiments, the preset fuzzy rules can be determined based on historical fault data and expert experience.

[0094] Regarding the method of determining the target fault classification detection result, in some embodiments, it can be determined whether the fault type detection result and the fault classification detection result are consistent. If they are consistent, the fault classification detection result is used as the target fault classification detection result. If they are inconsistent, the method of the present application is re-executed for re-detection.

[0095] Furthermore, in some embodiments, if the number of re-detections reaches a preset number and the fault type detection result is still inconsistent with the fault classification detection result, the fault classification detection result obtained for the last time is used as the target fault classification detection result; of course, in other embodiments, if the number of re-detections reaches a preset number and the fault type detection result is still inconsistent with the fault classification detection result, the first probability of the fault type detection result is determined, and the second probability of the fault classification detection result is determined, and the target fault classification detection result is determined by comparing the first probability and the second probability, that is, when the first probability is greater than the second probability, the fault type detection result is used as the target fault classification detection result, and when the first probability is less than or equal to the second probability, the fault classification detection result is used as the target fault classification detection result.

[0096] In the embodiment of the present application, the accuracy and reliability of fault detection are improved by combining the fault type detection results based on phase current and phase voltage signals and the fault classification detection results based on multi-dimensional feature vectors.

[0097] It can be understood that complementary fault information is provided: according to the three-phase current signals and the three-phase voltage signals, the current amplitude, voltage drop amplitude, harmonic content and fault time and other characteristics are determined, and the preset fuzzy rules are used to determine the fault type detection result. This method provides fault information from the perspective of the basic electrical characteristics of current and voltage, while the fault classification detection result obtained based on the multi-dimensional feature vector in the previous step is the fault information extracted from the perspective of complex characteristics such as non-stationarity and transientness of the signal. The combination of the two provides complementary fault information, which helps to understand the fault situation more comprehensively; enhance the accuracy of fault detection: since the fault type detection result and the fault classification detection result are obtained from different perspectives, they can be compared with each other. Mutual verification: when the two are consistent, the type of fault can be more determined, which improves the accuracy of fault detection. When the two are inconsistent, further analysis and judgment can be made by re-testing or comparing the probabilities of the two detection results, avoiding misjudgment or missed judgment that may occur in a single detection method, thereby improving the reliability of fault detection; adapting to complex and changeable fault conditions: in the actual operation of distribution overhead lines, the fault conditions are complex and diverse. By combining the detection results of two different methods, the fault detection method can better adapt to these complex and changeable fault conditions. Whether it is a simple fault or a complex hidden fault, this method can more accurately determine the fault type by comprehensively analyzing multiple information to ensure the stable operation of the power system.

[0098] In a feasible implementation, step 140 in the above embodiment, determining the mutual information between each eigenmode function and each layer of high-frequency detail coefficients, includes: determining the joint probability distribution between each eigenmode function and each layer of high-frequency detail coefficients, and determining the marginal probability distribution of each eigenmode function, and determining the marginal probability distribution of each layer of high-frequency detail coefficients; determining the mutual information between each eigenmode function and each layer of high-frequency detail coefficients based on the joint probability distribution between each eigenmode function and each layer of high-frequency detail coefficients, the marginal probability distribution of each eigenmode function, and the marginal probability distribution of each layer of high-frequency detail coefficients.

[0099] In the embodiment of the present application, by accurately calculating the mutual information, the intrinsic mode functions that are strongly correlated with the high-frequency detail coefficients are more accurately screened out, thereby improving the quality and effectiveness of fault feature extraction.

[0100] It can be understood that the correlation can be accurately quantified: by determining the joint probability distribution and marginal probability distribution between each intrinsic mode function and each layer of high-frequency detail coefficients, and calculating the mutual information based on this, the correlation between the intrinsic mode functions and the high-frequency detail coefficients can be accurately quantified. This quantification method avoids the uncertainty caused by subjective judgment and empirical estimation, making the correlation assessment more objective and accurate; improving the quality of feature extraction: mutual information, as an indicator of the degree of correlation between two variables, can effectively screen out those with strong correlation with the high-frequency detail coefficients from many intrinsic mode functions. By using the intrinsic mode functions greater than the mutual information threshold as high-frequency intrinsic mode functions, irrelevant or interfering information can be removed, further highlighting the features related to the fault, thereby improving the quality of fault feature extraction; enhancing the accuracy of fault detection: because the selected high-frequency intrinsic mode functions have a stronger correlation with the high-frequency detail coefficients, they can better reflect the essential characteristics of the fault. Therefore, when constructing multidimensional feature vectors and performing fault detection in the future, more accurate and relevant feature information can be used, thereby improving the accuracy of fault detection and reducing the possibility of misjudgment and missed detection.

[0101] In a feasible implementation, step 110 in the above embodiment, obtaining the target signal set, includes: obtaining three initial phase current signals, three initial phase voltage signals and an initial electric field strength signal collected by an electromagnetic transformer, wherein the electromagnetic transformer has been installed on the distribution overhead line; performing Gaussian kernel denoising processing on each initial phase current signal, each initial phase voltage signal and initial electric field strength signal to obtain three phase current signals, three phase voltage signals and an electric field strength signal; and taking the three phase current signals, three phase voltage signals and the electric field strength signal as elements to form the target signal set.

[0102] In the embodiment of the present application, the quality of signal acquisition and the accuracy of fault detection are improved by optimizing the signal acquisition device and denoising processing.

[0103] It can be understood that frequency band compatibility and fast response: using electromagnetic mutual inductor as signal acquisition device, different turns ratios can be designed for high-frequency and low-frequency signals respectively to achieve frequency band compatibility. At the same time, the use of nanocrystalline soft magnetic materials to construct the core components of the sensor reduces the hysteresis effect and improves the response speed to sudden faults. This helps to more accurately capture fault signals in the distribution overhead lines, especially hidden fault signals, and provides a reliable data basis for subsequent fault detection; high-sensitivity signal capture: integrating high-sensitivity Hall elements to assist in capturing electric field changes. Hall elements are highly sensitive to electric field changes and can more accurately reflect changes in electric field strength in distribution overhead lines, thereby facilitating more accurate detection. Fault; Effective denoising processing: Gaussian kernel denoising processing is performed on each initial phase current signal, each initial phase voltage signal and initial electric field strength signal. Gaussian kernel denoising processing can effectively remove noise interference in the signal, improve the signal-to-noise ratio of the signal, and make subsequent fault feature extraction and detection more accurate and reliable; Improve fault detection accuracy: By optimizing the signal acquisition device and denoising processing, higher quality three-phase current signals, three-phase voltage signals and electric field strength signals are obtained, which can more accurately reflect the actual operating status of the distribution overhead line. These signals are used as elements to form a target signal set, and subsequent fault detection processing is carried out, which helps to improve the accuracy of fault detection and reduce the possibility of misjudgment and missed judgment.

[0104] In a second aspect, the present application provides a fault detection device for a power distribution overhead line.

[0105] See also Figure 2 , is a schematic diagram of a fault detection device for a power distribution overhead line according to an embodiment of the present application, wherein the device 210 includes: An acquisition module 211 is configured to acquire a target signal set, the target signal set including three phase current signals, three phase voltage signals, and an electric field strength signal of a distribution overhead line; A first decomposition module 212 is configured to decompose a target signal using wavelet transform to obtain multiple layers of high-frequency detail coefficients and low-frequency approximation coefficients, wherein the target signal is any one signal in the target signal set; A second decomposition module 213 is configured to decompose the low-frequency approximation coefficients using empirical mode decomposition to obtain a plurality of intrinsic mode functions; A determination module 214 is configured to determine the mutual information between each IMF and each layer's high-frequency detail coefficients, and to use the IMF corresponding to the mutual information greater than a mutual information threshold as the high-frequency IMF; A construction module 215 is used to construct a multi-dimensional feature vector corresponding to the target signal based on all high-frequency intrinsic mode functions and all high-frequency detail coefficients; The prediction module 216 is configured to input each multidimensional feature vector corresponding to each signal in the target signal set into a preset line fault prediction model to obtain a fault classification detection result.

[0106] In the embodiment of the present application, the relevant contents of the acquisition module 211, the first decomposition module 212, the second decomposition module 213, the determination module 214, the construction module 215 and the prediction module 216 can be referred to. Figure 1 The contents of the illustrated embodiments are not described in detail here.

[0107] It should be noted that the device 210 of the present application also includes some other modules. It can be understood that the method of the present application and the device 210 have a one-to-one correspondence. Therefore, the other modules of the device 210 of the present application are the contents corresponding to the method of the present application in the above-mentioned embodiment.

[0108] In the embodiment of the present application, by fusing wavelet transform and empirical mode decomposition to process the signal, the fault characteristics in the non-stationary signal can be more effectively extracted, thereby quickly and accurately detecting the hidden faults in the distribution overhead lines, providing reliable support for the intelligent operation of the power system, effectively reducing the risk of serious line failures caused by hidden faults, and ensuring the stable operation of the power system.

[0109] In addition, in addition to the above-mentioned effects of being able to quickly and accurately detect hidden faults, reduce the risk of serious line failures, and ensure stable operation of the power system, the distribution overhead line fault detection device proposed in this application also has the following advantages: Guaranteeing timing consistency: By configuring multiple high-speed acquisition channels in the signal acquisition device, simultaneous high-speed acquisition of signals between different channels is achieved, ensuring the timing consistency of the acquired signals, avoiding time offset errors caused by different signal timings, and thus providing a reliable data basis for subsequent accurate fault detection; Flexible sampling configuration: The signal acquisition device supports multiple sampling frequencies and multiple resolutions, and can be adjusted according to actual needs. For example, after a fault is detected, it is adjusted to a high-frequency mode for signal acquisition, which can capture the fault signal characteristics in more detail, and adjusted to a low-frequency or normal mode when there is no fault, which not only meets the acquisition needs in different scenarios, but also saves acquisition resources and storage space to a certain extent; Refined separation of fault characteristics: The low-frequency approximate coefficients obtained by wavelet transform decomposition are decomposed using empirical mode decomposition, which can refine Separating and focusing on the fault characteristics of hidden faults. Through this multi-level decomposition and combination processing method, more subtle and critical fault information can be mined from complex signals, improving the ability to extract hidden fault characteristics and helping to more accurately identify the fault type and location; accurately screening high-frequency intrinsic mode functions: by determining the mutual information between each intrinsic mode function and each layer of high-frequency detail coefficients, and taking the intrinsic mode function corresponding to the mutual information greater than the mutual information threshold as the high-frequency intrinsic mode function, this screening method can effectively find the parts with a strong correlation with the high-frequency detail coefficients from many intrinsic mode functions, further highlighting the features related to the fault, removing irrelevant or interfering information, and improving the quality and effectiveness of the features; improving fault detection accuracy: using statistics such as weight, skewness, kurtosis and approximate entropy to construct multidimensional feature vectors. These statistics can describe the characteristics of the data from different angles and accurately extract key fault characteristics from complex hidden fault signals, thereby effectively improving the accuracy of subsequent model fault detection and reducing the possibility of misjudgment and omission. Select an advanced model architecture: The CNN-LSTM-Transformer model is selected as the initial line fault prediction model. This model combines the advantages of convolutional neural networks (CNN) in feature extraction, long short-term memory networks (LSTM) in time series processing, and Transformer in capturing long-distance dependencies. It can more effectively process information in multi-dimensional feature vectors, further improving the accuracy and reliability of fault classification and detection.

[0110] In a third aspect, the present application further provides a computer-readable storage medium storing a computer program. When the computer program is executed by a processor, the processor executes a method for detecting a fault of a distribution overhead line in the above method embodiment.

[0111] In a fourth aspect, the present application further provides a computer device including a memory and a processor, wherein the memory stores a computer program. When the computer program is executed by the processor, the processor executes a fault detection method for a distribution overhead line in the above method embodiment.

[0112] Figure 3 The internal structure diagram of the computer device in some embodiments is shown. The computer device can be a terminal, a server, or a gateway. Figure 3 As shown, the computer device includes a processor, a memory, and a network interface connected via a system bus.

[0113] The memory includes a non-volatile storage medium and an internal memory. The non-volatile storage medium of the computer device stores an operating system and may also store a computer program. When the computer program is executed by the processor, the processor can implement the various steps in the above method embodiment. The internal memory may also store a computer program. When the computer program is executed by the processor, the processor can implement the various steps in the above method embodiment. It will be understood by those skilled in the art that Figure 3 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than shown in the figure, or combine certain components, or have a different component arrangement.

[0114] Those skilled in the art will understand that all or part of the processes in the above-mentioned embodiment methods can be implemented by instructing related hardware through a computer program. The program can be stored in a non-volatile computer-readable storage medium. When the program is executed, it can include the processes of the embodiments of the above-mentioned methods.

[0115] Among them, any reference to memory, storage, database or other media used in the various embodiments provided in this application may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).

[0116] The technical features of the above embodiments can be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0117] The above-described embodiments merely represent several implementation methods of the present application. While the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the present invention. It should be noted that a person of ordinary skill in the art may make various modifications and improvements without departing from the spirit of the present invention, all of which fall within the scope of protection of the present invention. Therefore, the scope of protection of the present invention shall be determined by the appended claims.

Claims

1. A method for detecting faults in a power distribution overhead line, characterized in that: The method comprises: Acquire a target signal set, wherein the target signal set includes three phase current signals, three phase voltage signals, and an electric field strength signal of a distribution overhead line; Decomposing the target signal using wavelet transform to obtain multiple layers of high-frequency detail coefficients and low-frequency approximation coefficients, wherein the target signal is any one signal in the target signal set; Decomposing the low-frequency approximation coefficients using empirical mode decomposition to obtain a plurality of intrinsic mode functions; Determine the mutual information between each intrinsic mode function and each layer's high-frequency detail coefficient, and take the intrinsic mode function corresponding to the mutual information greater than the mutual information threshold as the high-frequency intrinsic mode function; Constructing a multidimensional feature vector corresponding to the target signal based on all high-frequency intrinsic mode functions and all high-frequency detail coefficients; Each multidimensional feature vector corresponding to each signal in the target signal set is input into a preset line fault prediction model to obtain a fault classification detection result.

2. The method according to claim 1, characterized in that The target signal is decomposed by using wavelet transform to obtain multiple layers of high-frequency detail coefficients and low-frequency approximation coefficients, including: The target signal is iteratively decomposed using the wavelet transform until the number of decomposition layers of the iterative decomposition is equal to the preset number of decomposition layers, and multiple layers of high-frequency detail coefficients corresponding to all iterative decompositions and the low-frequency approximation coefficients corresponding to the iterative decomposition are obtained.

3. The method according to claim 2, characterized in that The method further comprises: Using short-time Fourier transform, transforming the target signal to obtain a time-frequency spectrum; Performing frequency extraction on the time-frequency spectrum to obtain a signal dominant frequency; The preset number of decomposition layers is determined according to the dominant frequency of the signal.

4. The method according to claim 3, characterized in that The step of determining the preset number of decomposition layers according to the dominant frequency of the signal includes: Using the formula Determining the preset number of decomposition layers; in, is the preset decomposition level, is the floor symbol, is the logarithm to base 2, is the signal sampling frequency of the target signal, is the dominant frequency of the signal.

5. The method according to claim 1, wherein The low-frequency approximation coefficients are decomposed using empirical mode decomposition to obtain a plurality of intrinsic mode functions, including: Using the empirical mode decomposition, the low-frequency approximation coefficients are iteratively decomposed until the energy of the residual component obtained by the iterative decomposition is less than or equal to the energy threshold, and multiple initial intrinsic mode functions corresponding to all iterative decompositions are obtained, and the preset initial intrinsic mode functions are all used as intrinsic mode functions.

6. The method according to claim 5, characterized in that The method further comprises: determining the energy of the target signal; A preset percentage of the energy of the target signal is used as the energy threshold.

7. The method according to claim 1, characterized in that Before inputting each multidimensional feature vector corresponding to each signal in the target signal set into a preset line fault prediction model to obtain a fault classification detection result, the method further includes: Using short-time Fourier transform, transforming the target signal to obtain a time-frequency spectrum; Arrange all high-frequency detail coefficients in sequence to obtain a coefficient vector; Determine the energy of each eigenmode function and arrange the energies of all eigenmode functions in sequence to obtain an energy vector; Constructing a three-dimensional tensor corresponding to the target signal according to the time-frequency spectrum, the coefficient vector, and the energy vector; The step of inputting each multidimensional feature vector corresponding to each signal in the target signal set into a preset line fault prediction model to obtain a fault classification detection result includes: Each multidimensional feature vector and each three-dimensional tensor corresponding to each signal in the target signal set is input into the preset line fault prediction model to obtain the fault classification detection result.

8. The method according to claim 1, characterized in that The method further comprises: Determine the current amplitude, voltage drop amplitude, harmonic content and fault time based on the three phase current signals and three phase voltage signals; Determine a fault type detection result based on the current amplitude, the voltage drop amplitude, the harmonic content, and the fault time using a preset fuzzy rule; A target fault classification detection result is determined according to the fault type detection result and the fault classification detection result.

9. The method according to claim 1, characterized in that Determining the mutual information between each intrinsic mode function and each layer of high-frequency detail coefficients includes: Determining a joint probability distribution between each intrinsic mode function and each layer of high-frequency detail coefficients, and determining a marginal probability distribution of each intrinsic mode function, and determining a marginal probability distribution of each layer of high-frequency detail coefficients; The mutual information between each eigenmode function and each layer of high-frequency detail coefficients is determined according to the joint probability distribution between each eigenmode function and each layer of high-frequency detail coefficients, the marginal probability distribution of each eigenmode function, and the marginal probability distribution of each layer of high-frequency detail coefficients.

10. The method according to claim 1, characterized in that The acquiring of the target signal set includes: Acquiring three initial phase current signals, three initial phase voltage signals, and an initial electric field strength signal collected by an electromagnetic mutual inductor, wherein the electromagnetic mutual inductor has been installed on the power distribution overhead line; Performing Gaussian kernel denoising processing on each initial phase current signal, each initial phase voltage signal and the initial electric field strength signal to obtain three phase current signals, three phase voltage signals and an electric field strength signal; The three phase current signals, the three phase voltage signals and the electric field strength signal are all taken as elements to form the target signal set.

Images