Submarine cable grounding fault distance measurement method and system of high-voltage alternating-current submarine cable system, and medium

Abstract

The invention discloses a submarine cable grounding fault distance measurement method and system for a high-voltage alternating-current submarine cable system and a medium. The method comprises the following steps of S1, assembling a zero-sequence voltage transformer, a power frequency voltage sensor and a power frequency current sensor on sheaths at the two ends of the high-voltage alternating-current submarine cable system; the zero-sequence voltage transformer extracts zero-sequence voltage on the submarine cable sheath before and after the fault, and judges a fault phase according to the zero-sequence voltage temporary rising condition before and after the fault; a power frequency voltage sensor and a power frequency current sensor collect fault traveling wave signals at the fault moment of a fault phase, and the fault traveling wave signals are analyzed through a wavelet decomposition method and a wavelet transformation method to obtain fault distance measurement characteristic quantities about a time domain and a frequency domain; according to the method, fault phase selection is completed through data collection and calculation, the corresponding relation between the fault characteristic quantity and the fault distance is obtained through data training learning and curve fitting, fault distance measurement is completed, and the method has the advantage of being good in accuracy.

Description

Technical Field

[0001] The present invention relates to the field of power systems, and in particular to a method, system and medium for measuring grounding faults in a high-voltage AC submarine cable system. Background Art

[0002] With the rapid growth of marine resource development and offshore power transmission needs, the operational reliability of submarine cables, as the core carrier of cross-sea power transmission, offshore wind power grid connection, and island power supply, has attracted much attention. However, the complex marine environment will cause mechanical damage to submarine cables. The submarine cables will generate heat during operation, and the damaged parts may crack due to aging. The seawater and armor provide multiple return paths for the current, causing grounding faults in the submarine cable system.

[0003] The laying environment of submarine cables is more complex and extreme than that of terrestrial cables. Submarine cables face difficulties in repair when faults occur. Specialized equipment and personnel are usually required to conduct long-term fault inspections and construction work. Large-scale and prolonged power outages can lead to serious economic losses. Therefore, research on fault location of submarine cables can timely detect submarine cable grounding faults and ensure safer and more stable offshore power transmission. Summary of the Invention

[0004] The present invention aims to provide a method, system, and medium for measuring ground fault distance in a high-voltage AC submarine cable system. This method collects and calculates data to select the fault phase, and utilizes data training and curve fitting to determine the relationship between fault characteristics and fault distance, thereby achieving high accuracy in fault distance measurement.

[0005] The technical solution of the present invention is a method for measuring the ground fault of a submarine cable in a high-voltage AC submarine cable system, which is carried out according to the following steps:

[0006] Step S1: Installing a zero-sequence voltage transformer, a power frequency voltage sensor, and a power frequency current sensor on the sheaths at both ends of the high-voltage AC submarine cable system; the zero-sequence voltage transformer extracts the zero-sequence voltage on the submarine cable sheath before and after the fault, and determines the fault phase by the temporary rise of the zero-sequence voltage before and after the fault; the power frequency voltage sensor and the power frequency current sensor collect the fault traveling wave signal of the fault phase at the moment of the fault, and analyze the fault traveling wave signal using wavelet decomposition and wavelet transform methods to obtain fault location feature quantities in the time domain and frequency domain;

[0007] Step S2: constructing a ranging feature vector group using the fault ranging feature quantity, dividing the ranging feature vector group into training samples and test samples as sample data, and performing normalization processing on the training samples and the test samples;

[0008] Step S3: constructing a wavelet neural network and determining the number of nodes in the input layer, hidden layer and output layer of the wavelet neural network;

[0009] Step S4: Input the training samples and test samples processed in step S2 into the wavelet neural network for training. During the training process, the wavelet neural network performs forward propagation and backward propagation until the training samples meet the end training conditions, and the training is terminated to obtain a fitting curve of the relationship between the fault feature quantity and the fault distance;

[0010] Step S5: Determine the distance between the fault location of the high-voltage AC submarine cable system and the onshore sensors at both ends by fitting a curve of the relationship between the fault characteristic and the fault distance.

[0011] In the above-mentioned method for measuring the ground fault location of a high-voltage AC submarine cable system, the fault state is that the high-voltage AC submarine cable system is directly grounded and short-circuited.

[0012] In the aforementioned method for measuring the ground fault location of a submarine cable in a high-voltage AC submarine cable system, the calculation of the zero-sequence voltage swell is shown in the following formula:

[0013]

[0014] Where, is the voltage swell, is the zero-sequence voltage before the fault, is the zero-sequence voltage during the fault period.

[0015] In the aforementioned method for measuring the grounding fault location of a submarine cable in a high-voltage AC submarine cable system, the fault location characteristic quantities include the transient voltage, transient current, Lipschitz index of the fault traveling wave signal, the mutation time of the first three modulus maxima of the fault traveling wave signal, and the fault signal parameters after the fault occurs.

[0016] In the aforementioned method for measuring the ground fault location of a submarine cable in a high-voltage AC submarine cable system, the Lipschitz index of the fault traveling wave signal is obtained by wavelet transform. The relationship between the attenuation characteristics of the wavelet transform coefficient of the fault traveling wave signal and the Lipschitz index is shown in the following formula:

[0017] |WT x (t,s)|≤A s S a ;

[0018] Where W is the maximum value of the wavelet coefficient modulus, T x (t,s) is the signal T at time t transformed to scale s x , t is time, S is the wavelet transform scale, a is the scaling factor, As is the Li index, and the Li index is calculated as follows:

[0019]

[0020] Among them, when calculating the Li index, n in the formula approaches infinity, f(x k ,α) is the system function, x kis the k-th iteration value;

[0021] The mutation time of the first three modulus maxima of the fault traveling wave signal is obtained by wavelet transform and is calculated as follows:

[0022]

[0023]

[0024] Where t1 is the arrival time of the fault wave head, t2 is the arrival time of the reflected wave at the fault point, t3 is the arrival time of the reflected wave at the other end, v is the wave speed, L f is the fault distance, and L is the line length.

[0025] In the aforementioned method for measuring the ground fault location of a submarine cable in a high-voltage AC submarine cable system, the fault signal parameters include the modulus maximum, the stretch factor, and the mother wavelet of the fault traveling wave signal. The fault traveling wave signal is decomposed into multiple scales using the wavelet decomposition method. For a continuous fault traveling wave signal f(t), its calculation is shown in the following formula:

[0026]

[0027] Where f(t) is the fault traveling wave signal to be analyzed, is the mother wavelet function, a is the scaling factor, b is the translation parameter, is the wavelet coefficient.

[0028] In the aforementioned method for measuring submarine cable ground fault location in a high-voltage AC submarine cable system, sample data is transferred from the input layer to the hidden layer during the forward propagation of the wavelet neural network. At this time, the data of the jth node in the hidden layer is calculated by the following formula:

[0029]

[0030] Where x k is the kth sample data of the input sample data, ω kj is the link weight of the hidden layer node; the calculated result S j The wavelet basis is scaled and changed at the hidden layer nodes. The change calculation formula is as follows:

[0031]

[0032] h j =φ[(S j -b j ) / a j ];

[0033] Where φ(x) is the wavelet base, b j is the smoothing factor of the basis function, aj is the scaling factor of the basis function, h j is the output data of the jth node in the hidden layer; h j Entering the output layer, after calculation, it is output from the tth node of the output layer. The calculation formula is as follows:

[0034]

[0035] Where, ω jt is the link weight of the output layer, is the activation function.

[0036] In the aforementioned method for measuring ground faults in a high-voltage AC submarine cable system, the backpropagation process is to modify the wavelet neural network parameters and update the smoothing factor, scaling factor, and two link weights of the hidden layer and the output layer of the wavelet basis function. The modification formula is shown below:

[0037]

[0038] Where E is the error function, μ is the learning rate, ω is t is the corrected weight, ω t-1 is the weight of the previous item.

[0039] The system of the above-mentioned method for measuring the ground fault location of a submarine cable in a high-voltage AC submarine cable system comprises a data acquisition and processing module, a data analysis module, a feature vector construction module and a fault location module connected in sequence;

[0040] The data acquisition and processing module is used to collect the zero-sequence voltage and fault traveling wave signals under the fault state extracted by the power frequency current sensor, power frequency voltage sensor and zero-sequence voltage transformer, and perform data initialization processing;

[0041] The data analysis module is used to combine wavelet decomposition method and wavelet transform method to analyze the fault traveling wave signal to obtain the fault distance feature value;

[0042] The feature vector construction module is used to construct a ranging feature vector group;

[0043] The fault distance measurement module is used to measure the distance to the fault location by fitting a curve between the fault characteristic quantity and the fault distance.

[0044] A computer-readable storage medium stores computer-executable instructions. When the computer-executable instructions are executed, the method for measuring grounding faults in a high-voltage AC submarine cable system is implemented.

[0045] Compared with the existing technology, the method for measuring the grounding fault of a high-voltage AC submarine cable system provided by the present invention realizes the accurate judgment of the fault phase and the effective collection and analysis of the fault traveling wave signal by equipping a variety of sensors on the sheaths at both ends of the submarine cable; uses the wavelet decomposition method and the wavelet transform method to extract comprehensive time domain and frequency domain fault ranging characteristics from the fault traveling wave signal, providing rich and accurate data support for subsequent fault ranging; by constructing and training a wavelet neural network, a fitting curve between the fault characteristic quantity and the fault distance is established, which can intuitively and accurately reflect the correlation between the fault signal and the distance of each fault phase, thereby realizing the high-voltage AC submarine cable system. Accurately judge the distance between the system fault location and the onshore sensors at both ends; first, the present invention realizes the accurate selection of the fault phase through the zero-sequence voltage temporary rise, avoiding blind investigation; secondly, the fault signal feature vector extracted by the wavelet function correlation method covers multiple aspects of information such as transient voltage, transient current, Li index, modulus maximum mutation time and fault signal parameters, which describes the fault characteristics more comprehensively and greatly improves the accuracy of subsequent neural network training and learning; based on the learning of the distance measurement feature vector group by the wavelet neural network, the obtained fitting curve can more accurately reflect the relationship between the fault feature quantity and the fault distance, significantly improving the accuracy of submarine cable grounding fault ranging. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] Figure 1 Schematic diagram of the flow of a method for measuring ground faults in a high-voltage AC submarine cable system according to the present invention;

[0047] Figure 2 Schematic diagram of the high-voltage AC submarine cable system in Example 1 of the present invention;

[0048] Figure 3 This is a schematic diagram of the sensor arrangement positions of the high-voltage AC submarine cable system in Example 1 of the present invention;

[0049] Figure 4 1 is a schematic diagram of a submarine cable grounding fault circuit of a high-voltage AC submarine cable system in Example 1 of the present invention;

[0050] Figure 5 This is a topological structure diagram of the wavelet neural network training process in Example 1 of the present invention.

[0051] Figure 6 It is a schematic diagram of the framework of the submarine cable grounding fault distance measurement system of the high-voltage AC submarine cable system of the present invention. DETAILED DESCRIPTION

[0052] The present invention will be further described below with reference to the accompanying drawings and examples, but they are not intended to limit the present invention.

[0053] Example 1: A method for measuring the ground fault location of a submarine cable in a high-voltage AC submarine cable system, as shown in the attached Figure 1As shown, follow the steps below:

[0054] Step S1: As shown in the attached Figure 2 -Attached Figure 4 As shown in the figure, the high-voltage AC submarine cable system is a double-end grounded 110kV high-voltage AC submarine cable grounding system. Zero-sequence voltage transformers (VTA, VTB, VTC), power frequency voltage sensors (PTA, PTB, PTC), and power frequency current sensors (CTA, CTB, CTC) are pre-installed on the submarine cable sheath. The zero-sequence voltage on the submarine cable sheath before and after the high-voltage AC submarine cable system is directly grounded and short-circuited is extracted through the zero-sequence voltage transformer. The number 0 on the dial represents the normal state, the number 1 represents a single-phase grounding fault, and single-phase refers to any one of the three phases ABC being grounded. The number 4 represents a two-phase grounding fault, and two-phase refers to any two of the three phases ABC being grounded. The number 7 represents a three-phase grounding fault, and three-phase refers to the three phases ABC. The fault phase is determined by the temporary rise of the zero-sequence voltage before and after the fault. The calculation is shown in the following formula:

[0055]

[0056] Where, is the voltage swell, is the zero-sequence voltage before the fault, is the zero-sequence voltage during the fault period;

[0057] The power frequency voltage sensor and the power frequency current sensor collect the fault traveling wave signal at the moment of the fault phase fault. The fault traveling wave signal is analyzed by wavelet decomposition and wavelet transform methods to obtain fault location feature quantities in the time domain and frequency domain. The fault location feature quantities include the transient voltage and transient current after the fault occurs, the Lipschitz index of the fault traveling wave signal, the mutation time of the first three modulus maxima of the fault traveling wave signal, and the fault signal parameters.

[0058] The Lipschitz index of the fault traveling wave signal is obtained by wavelet transform method. The relationship between the attenuation characteristics of the wavelet transform coefficient of the fault traveling wave signal and the Lipschitz index is shown in the following formula:

[0059] |WT x (t,s)|≤A s S a ;

[0060] Where W is the maximum value of the wavelet coefficient modulus, T x (t,s) is the signal T at time t transformed to scale s x , t is time, S is the wavelet transform scale, a is the expansion factor, As is the Li index, and Li index A S The calculation is as follows:

[0061]

[0062] Among them, when calculating the Li index, n in the formula approaches infinity, f(x k ,α) is the system function, x k is the k-th iteration value;

[0063] The mutation time of the first three modulus maxima of the fault traveling wave signal is obtained by wavelet transform and is calculated as follows:

[0064]

[0065] Where t1 is the arrival time of the fault wave head, t2 is the arrival time of the reflected wave at the fault point, t3 is the arrival time of the reflected wave at the other end, v is the wave speed, L f is the fault distance, and L is the line length.

[0066] The fault signal parameters include the modulus maximum, the stretch factor, and the mother wavelet of the fault traveling wave signal, which are obtained by multi-scale decomposition of the fault traveling wave signal using the wavelet decomposition method. For the continuous fault traveling wave signal f(t), the calculation is shown in the following formula:

[0067]

[0068] Where f(t) is the fault traveling wave signal to be analyzed, is the mother wavelet function, a is the scaling factor, b is the translation parameter, is the wavelet coefficient.

[0069] Step S2: Use the fault distance feature quantity to construct a distance feature vector group containing spatial domain functions and time domain functions. Use the distance feature vector group as sample data to divide it into training samples and test samples. The training samples and test samples are normalized so that the input sample values ​​are mapped within the range of [0, 1]. The calculation formula is as follows:

[0070]

[0071] Among them, x i is the original sample data, x min and x min are the minimum and maximum values ​​of the original samples to be normalized in the wavelet neural network, y i and the input sample value after normalization of the wavelet neural network, i (i=1,2,…) is the number of samples;

[0072] Step S3: As shown in the attached Figure 5As shown, a wavelet neural network is constructed (a wavelet neural network is an algorithm that combines wavelet analysis with artificial neural networks, and uses the two to process nonlinear, non-stationary signals or complex data. It includes an input layer, a hidden layer, and an output layer. The input layer is used to receive original data, and the hidden layer has multiple nodes that are analyzed and output by wavelet basis functions. The output layer linearly combines the results of the hidden layer and outputs predicted values ​​or classification results). The number of nodes in the input layer, hidden layer, and output layer of the wavelet neural network are determined to be n, i, and mm respectively according to the sample capacity.

[0073] Step S4: Input the training samples and test samples processed in step S2 into the wavelet neural network for training. During the training process, the wavelet neural network performs forward propagation and backward propagation until the training samples meet the end training conditions, and the training is terminated to obtain a fitting curve of the relationship between the fault feature quantity and the fault distance;

[0074] During the forward propagation of the wavelet neural network, the sample data is transferred from the input layer to the hidden layer. At this time, the data of the jth node in the hidden layer is calculated by the following formula:

[0075]

[0076] Where x k is the kth sample data of the input sample data, ω kj is the link weight of the hidden layer node; the calculated result S j The wavelet basis is scaled and changed at the hidden layer nodes. The change calculation formula is as follows:

[0077]

[0078] h j =φ[(S j -b j ) / a j ];

[0079] Where φ(x) is the wavelet base, b j is the smoothing factor of the basis function, a j is the scaling factor of the basis function, h j is the output data of the jth node in the hidden layer; h j Entering the output layer, after calculation, it is output from the tth node of the output layer. The calculation formula is as follows:

[0080]

[0081] Where, ω jt is the link weight of the output layer, is the activation function.

[0082] The back propagation process is to modify the wavelet neural network parameters and update the smoothing factor, scaling factor, and two link weights of the hidden layer and the output layer of the wavelet basis function. The modification formula is shown below:

[0083]

[0084] Where E is the error function, μ is the learning rate, ω is t is the corrected weight, ω t-1 is the weight of the previous item.

[0085] The condition that the training sample meets the end of training refers to setting the output error when setting the wavelet neural network parameters. When the output error value of the wavelet neural network is less than the preset value during the training process, the training mark is completed, the training completed data is output, and the fitting curve of the relationship between the fault feature quantity and the fault distance is obtained.

[0086] Step S5: Determine the distance between the fault location of the high-voltage AC submarine cable system and the onshore sensors at both ends by fitting a curve of the relationship between the fault characteristic and the fault distance.

[0087] To better enable relevant personnel in this technical field to fully understand and implement the present invention, the following is a specific implementation case applied to submarine cable ground fault distance measurement in a high-voltage AC submarine cable system. Combining the technical solution in the specification with actual scenarios, the application process and effects of the present invention are demonstrated in detail: Zero-sequence voltage transformers (VTA, VTB, VTC) are respectively installed on the sheaths at both ends of the submarine cable (near the landing section) to collect the zero-sequence voltage of the three-phase sheaths A, B, and C. When the zero-sequence voltage swell of a phase is greater than a threshold, the phase is determined to be a fault phase; the power frequency voltage sensor and the power frequency current sensor respectively collect the voltage and current traveling waves of the fault phase, and the sampling frequency is set to 10MHz;

[0088] The transient signal peak within 0-10ms after the fault is intercepted, the fault signal singularity is calculated through wavelet transform, the wavelet coefficient modulus maximum W at the specified scale S is taken, and the modulus maximum attenuation rate is further obtained. The Lipschitz index is extracted using the formula to obtain the degree of fault signal mutation; the arrival time of the first three wave heads of the fault traveling wave is extracted; the traveling wave signal is decomposed into multiple layers using wavelet decomposition to extract the modulus maximum, expansion factor and mother wavelet phase information at each scale.

[0089] The characteristic quantities of the traveling wave signal of the same fault are integrated into a multi-dimensional feature vector, which is divided into training samples and test samples in a ratio of 8:2 and then normalized using minimum and maximum normalization.

[0090] The number of nodes in the input layer, hidden layer, and output layer of the wavelet neural network are set to n = 10 (corresponding to the dimension of the feature vector), i = 7 (the activation function uses the Morlet wavelet basis function), and m = 1 (the output fault distance, and the activation function is a linear function). The hidden layer weights are randomly initialized to [-0.1, 0.1], and the initial values ​​of the wavelet basis smoothing factor and the scaling factor are both set to 1.

[0091] The wavelet neural network calculates the hidden layer output through forward propagation. The backward propagation sets the learning rate μ = 0.01 and the number of iterations to 500. The smoothing factor, scaling factor, and two link weights of the wavelet basis function are updated. When the test sample error is less than 10 meters or the number of iterations reaches the upper limit, the training is stopped and a fitting curve of the relationship between the fault feature quantity and the fault distance is generated.

[0092] A phase A grounding fault was set 5 km away from end A of the submarine cable. The collected feature vectors were normalized and input into the wavelet neural network, outputting fault = 4.98 km, with an error of 0.4%. The fault location was displayed on the host computer interface, with the distances to ends A and B marked, and an alarm was triggered to alert the operation and maintenance personnel.

[0093] Example 2: High-voltage AC submarine cable system submarine cable ground fault location system, as shown in the attached Figure 6 As shown, a method for measuring ground fault location of a submarine cable in a high-voltage AC submarine cable system includes a data acquisition and processing module, a data analysis module, a feature vector construction module, and a fault location module connected in sequence;

[0094] The data acquisition and processing module is used to collect the zero-sequence voltage and fault traveling wave signals under the fault state extracted by the power frequency current sensor, the power frequency voltage sensor and the zero-sequence voltage transformer, and perform data initialization processing; the data acquisition and processing module is an acquisition device, a signal conditioning circuit and a controller. The acquisition device includes but is not limited to a PLC and a single-chip microcomputer, which converts the acquired signal into a digital signal that can be processed by the microcontroller. The signal conditioning circuit amplifies (gain adjustment), filters (noise removal), isolates (anti-interference), linearizes (compensates for nonlinear errors), and performs analog-to-digital conversion (ADC) on the acquired signal. The controller can be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), off-the-shelf programmable gate arrays (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components, etc.

[0095] The data analysis module is an embedded processing system based on digital signal processing, or an embedded processing system based on a field programmable gate array, which integrates algorithms including wavelet decomposition and wavelet transform to analyze the fault traveling wave signal to obtain the fault distance feature value;

[0096] The feature vector construction module is an embedded processing system based on digital signal processing, or an embedded processing system based on a field programmable gate array. The fault ranging feature vector group is constructed using the fault ranging feature quantity. The ranging feature vector group is divided into training samples and test samples as sample data, and the training samples and the test samples are normalized.

[0097] The fault distance measurement module is an embedded processing system based on digital signal processing, or an embedded processing system based on field programmable gate array, which integrates a wavelet neural network and measures the fault location distance by fitting a curve between the fault feature quantity and the fault distance.

[0098] Embodiment 3: A computer-readable storage medium stores computer-executable instructions, which, when executed, implement a method for measuring grounding faults in a high-voltage AC submarine cable system.

[0099] In this embodiment, it should be understood that the disclosed system can be implemented in other ways. For example, the system embodiment described above is merely illustrative. For example, the module division is merely a logical functional division. In actual implementation, other division methods may be used. For example, multiple modules or components may be combined or integrated into another system, or some features may be ignored or not implemented. In addition, the coupling or direct coupling or communication connection shown or discussed may be an indirect coupling or communication connection through some interface, device or module, and may be electrical or other forms.

[0100] The modules described as separate components may or may not be physically separate, and the components shown as modules may or may not be physical modules, that is, they may be located in one place or distributed across multiple network modules. Some or all of the modules may be selected to achieve the purpose of the present embodiment according to actual needs.

[0101] In addition, the functional modules in various embodiments of the present invention may be integrated into a single processing module, or each module may exist physically separately, or two or more modules may be integrated into a single module. The aforementioned integrated modules may be implemented in the form of hardware or software functional modules.

[0102] If the integrated module is implemented in the form of a software functional module and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, server or network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), mobile hard disk, magnetic disk or optical disk, etc. Various media that can store program codes.

[0103] In summary, the method for measuring the distance between the grounding cable of a high-voltage AC submarine cable system provided by the present invention realizes the rapid phase selection of the fault phase (based on the calculation of the zero-sequence voltage swell) and the multi-scale analysis of the fault traveling wave signal (combined with the wavelet decomposition and transformation method, the extracted feature quantities include transient voltage and current, Li index, modulus maximum mutation time and other multi-dimensional information, breaking through the limitations of traditional single physical quantity analysis and providing a basis for the refined description of fault characteristics; combining the time-frequency localization advantage of wavelet analysis with the nonlinear fitting ability of neural networks, and constructing a high-resolution fault feature quantity and distance through the parameter optimization of forward propagation and backward propagation. Precision mapping relationship, compared with the traditional traveling wave ranging algorithm, this model has stronger robustness to complex submarine cable environments (such as parameter frequency variation, multipath reflection), and significantly improves the ranging accuracy; by directly correlating the fault characteristics and distance through the fitting curve, the distance between the fault point and the onshore sensors at both ends can be quickly determined, which greatly shortens the submarine cable fault investigation time and maintenance cost, ensures the stability of the offshore power transmission network, and reduces industrial losses; for the extreme conditions of strong corrosion, high water pressure and dynamic load of submarine cable laying, the present invention ensures data reliability in harsh environments through sensor optimization layout and signal denoising algorithm (wavelet transform filtering), which is suitable for various scenarios such as cross-sea power transmission and offshore wind power grid connection.

Claims

1. A method for measuring ground fault location in a high-voltage AC submarine cable system, characterized by: Follow these steps: Step S1: Installing a zero-sequence voltage transformer, a power frequency voltage sensor, and a power frequency current sensor on the sheaths at both ends of the high-voltage AC submarine cable system; the zero-sequence voltage transformer extracts the zero-sequence voltage on the submarine cable sheath before and after the fault, and determines the fault phase by the temporary rise of the zero-sequence voltage before and after the fault; the power frequency voltage sensor and the power frequency current sensor collect the fault traveling wave signal of the fault phase at the moment of the fault, and analyze the fault traveling wave signal using wavelet decomposition and wavelet transform methods to obtain fault location feature quantities in the time domain and frequency domain; Step S2: constructing a ranging feature vector group using the fault ranging feature quantity, dividing the ranging feature vector group into training samples and test samples as sample data, and performing normalization processing on the training samples and the test samples; Step S3: constructing a wavelet neural network and determining the number of nodes in the input layer, hidden layer and output layer of the wavelet neural network; Step S4: Input the training samples and test samples processed in step S2 into the wavelet neural network for training. During the training process, the wavelet neural network performs forward propagation and backward propagation until the training samples meet the end training conditions, and the training is terminated to obtain a fitting curve of the relationship between the fault feature quantity and the fault distance; Step S5: Determine the distance between the fault location of the high-voltage AC submarine cable system and the onshore sensors at both ends by fitting a curve of the relationship between the fault characteristic and the fault distance.

2. The method for measuring ground fault location of a high voltage AC submarine cable system according to claim 1, characterized in that: The fault state is that the high-voltage AC submarine cable system is directly grounded and short-circuited.

3. The method for measuring ground fault location of a high voltage AC submarine cable system according to claim 1, characterized in that: The calculation of the zero-sequence voltage swell is shown in the following formula: Where, is the voltage swell, is the zero-sequence voltage before the fault, is the zero-sequence voltage during the fault period.

4. The method for measuring ground fault location of a high voltage AC submarine cable system according to claim 1, characterized in that: The fault location characteristic quantities include transient voltage, transient current, Lipschitz index of the fault traveling wave signal, mutation time of the first three modulus maxima of the fault traveling wave signal and fault signal parameters after the fault occurs.

5. The method for measuring ground fault location of a high voltage AC submarine cable system according to claim 4, characterized in that: The Lipschitz index of the fault traveling wave signal is obtained by wavelet transform method. The relationship between the attenuation characteristics of the wavelet transform coefficient of the fault traveling wave signal and the Lipschitz index is shown in the following formula: |WT x (t,s)|≤A s S a ; Where W is the maximum value of the wavelet coefficient modulus, T x (t,s) is the signal T at time t transformed to scale s x , t is time, S is the wavelet transform scale, a is the scaling factor, A S is the Lee index, Lee index A S The calculation is as follows: Among them, when calculating the Li index, n in the formula approaches infinity, f(x k ,α) is the system function, x k is the k-th iteration value; The mutation time of the first three modulus maxima of the fault traveling wave signal is obtained by wavelet transform and is calculated as follows: Where t1 is the arrival time of the fault wave head, t2 is the arrival time of the reflected wave at the fault point, t3 is the arrival time of the reflected wave at the other end, v is the wave speed, L f is the fault distance, and L is the line length.

6. The method for measuring ground fault location of a high voltage AC submarine cable system according to claim 4, characterized in that: The fault signal parameters include the modulus maximum, the stretch factor, and the mother wavelet of the fault traveling wave signal, which are obtained by multi-scale decomposition of the fault traveling wave signal using the wavelet decomposition method. For the continuous fault traveling wave signal f(t), the calculation is shown in the following formula: Where f(t) is the fault traveling wave signal to be analyzed, is the mother wavelet function, a is the scaling factor, b is the translation parameter, is the wavelet coefficient.

7. The method for measuring ground fault location of a high voltage AC submarine cable system according to claim 1, characterized in that: During the forward propagation of the wavelet neural network, the sample data is transferred from the input layer to the hidden layer. At this time, the data of the jth node in the hidden layer is calculated by the following formula: Where x k is the kth sample data of the input sample data, ω kj is the link weight of the hidden layer node; the calculated result S j The wavelet basis is scaled and changed at the hidden layer nodes. The change calculation formula is as follows: h j =φ[(S j -b j ) / a j ]; Where φ(x) is the wavelet base, b j is the smoothing factor of the basis function, a j is the scaling factor of the basis function, h j is the output data of the jth node in the hidden layer; h j Entering the output layer, after calculation, it is output from the tth node of the output layer. The calculation formula is as follows: Where, ω jt is the link weight of the output layer, is the activation function.

8. The method for measuring ground fault location of a high voltage AC submarine cable system according to claim 7, characterized in that: The back propagation process is to modify the wavelet neural network parameters and update the smoothing factor, scaling factor, and two link weights of the hidden layer and the output layer of the wavelet basis function. The modification formula is shown below: Where E is the error function, μ is the learning rate, ω is t is the corrected weight, ω t-1 is the weight of the previous item.

9. The method for measuring the ground fault location of a high-voltage AC submarine cable system according to any one of claims 1 to 8, characterized in that: The system includes a data acquisition and processing module, a data analysis module, a feature vector construction module and a fault location module connected in sequence; The data acquisition and processing module is used to collect the zero-sequence voltage and fault traveling wave signals under the fault state extracted by the power frequency current sensor, power frequency voltage sensor and zero-sequence voltage transformer, and perform data initialization processing; The data analysis module is used to combine wavelet decomposition method and wavelet transform method to analyze the fault traveling wave signal to obtain the fault distance feature value; The feature vector construction module is used to construct a ranging feature vector group; The fault distance measurement module is used to integrate the wavelet neural network and measure the fault position distance by fitting the relationship curve between the fault characteristic quantity and the fault distance.

10. A computer-readable storage medium storing computer-executable instructions, wherein when the computer-executable instructions are executed, the method for measuring ground fault location of a submarine cable in a high-voltage AC submarine cable system according to any one of claims 1 to 8 is implemented.

Images