Online monitoring method and system for fault state of high-voltage parallel cable line

Abstract

The invention discloses a high-voltage parallel cable line fault state on-line monitoring method and system, and relates to the technical field of cable line monitoring, and the method comprises the steps: obtaining the electrical parameters of each monitoring point of a high-voltage parallel cable line in a preset sampling interval, and carrying out the data preprocessing of each electrical parameter based on the sampling interval, performing feature extraction on the feature data through a first detection network to obtain a first feature map, performing feature extraction on the first feature map through a second detection network to obtain a second feature map, and performing state evaluation on the second feature map by adopting a Leaky Relu activation layer to obtain a state evaluation result; and obtaining a fault state evaluation result corresponding to the high-voltage parallel cable line. The technical problems that an existing cable line fault state monitoring method mainly depends on artificial experience and simple statistical analysis, the actual fault state of a cable is difficult to reflect in real time, and the operation reliability of the cable line is reduced are solved.

Description

Technical Field

[0001] The present invention relates to the technical field of cable line monitoring, and in particular to a method and system for online monitoring of fault status of a high-voltage parallel cable line. Background Art

[0002] High-voltage parallel cables are an important part of urban power grids. Cable failure will not only cause economic losses and adverse social impacts, but also affect the safe and stable operation of urban power grids. Therefore, it is very important to monitor the fault status of high-voltage parallel cable lines in real time.

[0003] At present, the existing cable line fault status monitoring methods mainly rely on manual experience and simple statistical analysis, which is difficult to reflect the actual fault status of the cable in real time, reducing the reliability of cable line operation. Summary of the invention

[0004] The present invention provides a method and system for online monitoring of the fault status of a high-voltage parallel cable line, which solves the technical problem that the existing cable line fault status monitoring method mainly relies on manual experience and simple statistical analysis, is difficult to reflect the actual fault status of the cable in real time, and reduces the reliability of the cable line operation.

[0005] A first aspect of the present invention provides a method for online monitoring of a fault state of a high-voltage parallel cable line, comprising:

[0006] Acquire electrical quantity parameters of each monitoring point of the high-voltage parallel cable line within a preset sampling interval, and perform data preprocessing on each of the electrical quantity parameters based on the sampling interval to obtain corresponding characteristic data;

[0007] Inputting the feature data into a pre-trained line fault state assessment model, wherein the line fault state assessment model includes a first detection network, a second detection network and a Leaky Relu activation layer connected in sequence;

[0008] Performing feature extraction on the feature data through the first detection network to obtain a first feature map;

[0009] Performing feature extraction on the first feature map by using the second detection network to obtain a second feature map;

[0010] A Leaky Relu activation layer is used to perform state evaluation on the second feature map to obtain a fault state evaluation result corresponding to the high-voltage parallel cable line.

[0011] Optionally, the electrical quantity parameters include a first ground current, a second ground current, a load current, a sheath induced voltage, a fault time, a start time, and an end time, and the step of performing data preprocessing on each of the electrical quantity parameters based on the sampling interval to obtain corresponding characteristic data includes:

[0012] performing difference processing on each of the second grounding currents and the associated first grounding current to obtain a first difference;

[0013] Performing ratio processing on the first difference and the sampling interval respectively to obtain the ground current change rate corresponding to each monitoring point;

[0014] Inputting two load currents connected in parallel at the same node in each of the load currents into a preset load current imbalance function to obtain a plurality of load current imbalances;

[0015] Inputting each of the load currents into a preset phase-to-phase current imbalance function respectively to obtain the phase-to-phase current imbalance corresponding to each of the monitoring points;

[0016] Performing ratio processing on each of the second grounding currents and the associated load current respectively to obtain the grounding current load ratio corresponding to each of the monitoring points;

[0017] Each of the grounding current change rates, each of the load current imbalances, each of the phase current imbalances, each of the grounding current load ratios, each of the sheath induced voltages, each of the second grounding currents, each of the load currents, each of the fault times, each of the start times and each of the end times are used as characteristic data.

[0018] Optionally, the first detection network includes a 10×10 standard convolution layer and a data normalization layer connected in sequence, and the step of extracting features from the feature data through the first detection network to obtain a first feature map includes:

[0019] A 10×10 standard convolutional layer is used to extract features from the feature data to obtain a first data feature map;

[0020] The first data feature map is normalized by a data normalization layer to obtain a first feature map.

[0021] Optionally, the second detection network includes a first detection module, a 1×1 convolutional layer, a second detection module, a 1×1 convolutional layer and a feature fusion layer, and the step of extracting features from the first feature map through the second detection network to obtain a second feature map includes:

[0022] Using the first detection module to perform feature extraction on the first feature map to obtain a first detection feature map, wherein the first detection module includes a first extraction branch, a feature fusion layer, a batch normalization layer, and a Tanh activation layer;

[0023] Downsampling the first detection feature map through a 1×1 convolutional layer to obtain a first sampling feature map;

[0024] Performing feature extraction on the first detection feature map through the second detection module to obtain a second detection feature map, wherein the second detection module includes a second extraction branch, a 1×1 convolution layer, a feature fusion layer, a batch normalization layer and a Relu activation layer;

[0025] Performing feature extraction on the second detection feature map through a 1×1 convolutional layer to obtain a second sampling feature map;

[0026] A feature fusion layer is used to perform feature fusion on the first sampling feature map and the second sampling feature map to obtain a second feature map.

[0027] Optionally, the step of using the first detection module to extract features from the first feature map to obtain a first detection feature map includes:

[0028] Performing feature extraction on the first feature map through the first extraction branch to obtain a first residual feature map, wherein the first extraction branch includes a 1×1 convolutional layer, a 3×3 convolutional layer, a batch normalization layer, a Tanh activation layer, and a 1×1 standard convolutional layer connected in sequence;

[0029] Using a feature fusion layer to perform feature fusion on the first residual feature map and the first feature map to obtain a first fused feature map;

[0030] The first fusion feature map is nonlinearly transformed through a batch normalization layer and a Relu activation layer in sequence to obtain a first detection feature map.

[0031] Optionally, the step of extracting features from the first detection feature map by the second detection module to obtain a second detection feature map includes:

[0032] Performing feature extraction on the first detection feature map through the second extraction branch to obtain a second residual feature map, wherein the second extraction branch includes a 1×1 standard convolution layer, a batch normalization layer, a ReLU activation layer, a 3×3 convolution layer, a ReLU activation layer, and a 1×1 standard convolution layer connected in sequence;

[0033] Performing feature extraction on the first detection feature map through a 1×1 convolutional layer to obtain a third residual feature map;

[0034] Using a feature fusion layer to perform feature fusion on the second residual feature map and the third residual feature map to obtain a second fused feature map;

[0035] The second fusion feature map is nonlinearly mapped through a batch normalization layer and a Relu activation layer in sequence to obtain a second detection feature map.

[0036] A second aspect of the present invention provides an online monitoring system for fault status of a high-voltage parallel cable line, comprising:

[0037] An acquisition module is used to obtain electrical quantity parameters of each monitoring point of the high-voltage parallel cable line within a preset sampling interval, and perform data preprocessing on each of the electrical quantity parameters based on the sampling interval to obtain corresponding characteristic data;

[0038] An input module, used for inputting the feature data into a pre-trained line fault state assessment model, wherein the line fault state assessment model comprises a first detection network, a second detection network and a Leaky Relu activation layer connected in sequence;

[0039] A first extraction module, configured to extract features from the feature data through the first detection network to obtain a first feature map;

[0040] A second extraction module, configured to extract features from the first feature map through the second detection network to obtain a second feature map;

[0041] An evaluation module is used to use a Leaky Relu activation layer to perform a state evaluation on the second feature map to obtain a fault state evaluation result corresponding to the high-voltage parallel cable line.

[0042] A third aspect of the present invention provides an electronic device, 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 steps of the method for online monitoring of the fault status of a high-voltage parallel cable line as described in any one of the above items.

[0043] A fourth aspect of the present invention provides a computer-readable storage medium having a computer program stored thereon, and when the computer program is executed, the method for online monitoring of the fault status of a high-voltage parallel cable line as described in any one of the above items is implemented.

[0044] A fifth aspect of the present invention provides a computer program product, which includes a computer program stored on a non-transitory computer-readable storage medium, and the computer program includes program instructions, wherein when the program instructions are executed by a computer, the computer executes the online monitoring method for the fault status of a high-voltage parallel cable line as described in any one of the above items.

[0045] It can be seen from the above technical solutions that the present invention has the following advantages:

[0046] By obtaining the electrical quantity parameters of each monitoring point of the high-voltage parallel cable line within a preset sampling interval, and performing data preprocessing on each electrical quantity parameter, characteristic data is obtained. The characteristic data contains the status information of each monitoring point in the high-voltage parallel cable line. By inputting the characteristic data into a pre-trained line fault state assessment model, the fault state assessment result of the high-voltage parallel cable line can be obtained quickly and accurately. It overcomes the technical problem that the existing cable line fault state monitoring method mainly relies on manual experience and simple statistical analysis, which is difficult to reflect the actual fault state of the cable in real time and reduces the reliability of the cable line operation. Compared with the traditional cable line fault state monitoring method, the present invention obtains the characteristic data of the high-voltage parallel cable line and inputs it into the preset line fault state assessment model, and optimizes the internal structure of the line fault state assessment model, thereby achieving accurate assessment of the fault state assessment result of the high-voltage parallel cable line, and improving the reliability of the cable line operation. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. 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 creative labor.

[0048] Figure 1 A flowchart of a method for online monitoring of fault status of a high-voltage parallel cable line provided in Embodiment 1 of the present invention;

[0049] Figure 2 A schematic diagram of the structure of a high-voltage parallel cable circuit provided in Embodiment 1 of the present invention;

[0050] Figure 3 A flowchart of the steps of a method for online monitoring of fault status of a high-voltage parallel cable line provided in Embodiment 2 of the present invention;

[0051] Figure 4 A schematic diagram of the structure of a line fault status assessment model provided in Embodiment 2 of the present invention;

[0052] Figure 5 A schematic diagram of the structure of a first detection module provided in Embodiment 2 of the present invention;

[0053] Figure 6 A schematic diagram of the structure of a second detection module provided in Embodiment 2 of the present invention;

[0054] Figure 7 A structural block diagram of an online monitoring system for fault status of a high-voltage parallel cable line provided in Embodiment 3 of the present invention;

[0055] Figure 8 This is a structural block diagram of an electronic device provided in Embodiment 4 of the present invention. DETAILED DESCRIPTION

[0056] The embodiment of the present invention provides a method and system for online monitoring of the fault status of a high-voltage parallel cable line, which is used to solve the technical problem that the existing cable line fault status monitoring method mainly relies on manual experience and simple statistical analysis, is difficult to reflect the actual fault status of the cable in real time, and reduces the reliability of the cable line operation.

[0057] In order to make the purpose, features and advantages of the present invention more obvious and easy to understand, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the embodiments described below are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0058] See also Figure 1 , Figure 1 A flowchart of the steps of a method for online monitoring of fault status of a high-voltage parallel cable line provided in Embodiment 1 of the present invention.

[0059] The present invention provides a method for online monitoring of fault status of a high-voltage parallel cable line, comprising:

[0060] Step 101, obtaining electrical quantity parameters of each monitoring point of the high-voltage parallel cable line within a preset sampling interval, and performing data preprocessing on each electrical quantity parameter based on the sampling interval to obtain corresponding characteristic data;

[0061] In the embodiment of the present invention, the electrical quantity parameters of each monitoring point are collected by the monitoring terminal of each monitoring point according to the preset sampling interval, and the data preprocessing is performed on each electrical quantity parameter based on the sampling interval to generate corresponding characteristic data.

[0062] It should be noted that the electrical quantity parameters include but are not limited to monitoring nodes, sampling intervals, start time, fault time, end time, load current, sheath grounding current and sheath induced voltage, etc.

[0063] It should be noted that, see Figure 2As shown, the monitoring terminal includes a high-speed data acquisition module, a data analysis module, a clock module, a communication module, a main controller and a power module; wherein the clock module is used to adjust the time synchronization between different devices, the power module provides power to the monitoring terminal, the high-speed data acquisition module converts the obtained analog voltage signal into a digital signal, and the data analysis module performs post-processing on the basic parameters; the clock module provides time for the high-speed data acquisition module and is used for time synchronization between different devices. The main controller is used to perform data preprocessing on the electrical quantity parameters of the acquisition module and transmit the characteristic data to the monitoring center. The high-speed data acquisition module includes but is not limited to a sheath induction voltage sensor, a sheath current sensor and a load current sensor.

[0064] Step 102: input the feature data into a pre-trained line fault state assessment model, wherein the line fault state assessment model includes a first detection network, a second detection network, and a Leaky Relu activation layer connected in sequence;

[0065] In an embodiment of the present invention, the feature data is used as input of a pre-trained line fault state assessment model, and the obtained line fault state assessment model includes a first detection network, a second detection network and a LeakyRelu activation layer connected in sequence.

[0066] It is worth mentioning that before inputting the feature data into the pre-trained line fault state assessment model, multiple historical electrical quantity parameters of the high-voltage parallel cable line are obtained, and each historical electrical quantity parameter is pre-processed to obtain a training feature data set. The training feature data set is used to train the line fault state assessment model to obtain a trained line fault state assessment model. The training process is specifically as follows: 1. Input the training feature data set into the line fault state assessment model to obtain the line assessment result. 2. Select the standard assessment result corresponding to the line assessment result from the training feature data set. 3. Calculate the deviation between the line assessment result and the standard assessment result. 4. When the deviation is less than the preset deviation threshold, a trained line fault state assessment model is generated. 5. When the deviation is greater than or equal to the preset deviation threshold, the gradient descent method is used to adjust the parameters of the line fault state assessment model, and steps 1-3 are re-executed until the deviation is less than the preset deviation threshold.

[0067] Step 103: extracting features from the feature data through a first detection network to obtain a first feature map;

[0068] In the embodiment of the present invention, a first detection network is used to extract features from feature data to obtain a first feature map.

[0069] Step 104: extract features from the first feature map through a second detection network to obtain a second feature map;

[0070] In an embodiment of the present invention, a second detection network is used to extract features from the first feature map to obtain a second feature map.

[0071] Step 105: Use the Leaky Relu activation layer to perform state evaluation on the second feature map to obtain a fault state evaluation result corresponding to the high-voltage parallel cable line.

[0072] In an embodiment of the present invention, the state of the second feature map is evaluated through the Leaky Relu activation layer to obtain a fault state evaluation result corresponding to the high-voltage parallel cable line, wherein the fault state evaluation result includes the overall state of the line (the overall state includes normal 1, attention 2, abnormal 3, and dangerous 4) and the fault type corresponding to the fault monitoring point (the fault type includes none, single-phase short circuit to ground fault, two-phase short circuit fault, two-phase short circuit to ground fault, three-phase short circuit, etc.).

[0073] In an embodiment of the present invention, by obtaining the electrical quantity parameters of each monitoring point of the high-voltage parallel cable line within a preset sampling interval, and performing data preprocessing on each electrical quantity parameter, characteristic data is obtained, and the characteristic data includes the status information of each monitoring point in the high-voltage parallel cable line. By inputting the characteristic data into a pre-trained line fault state assessment model, the fault state assessment result of the high-voltage parallel cable line can be obtained quickly and accurately. The technical problem that the existing cable line fault state monitoring method mainly relies on manual experience and simple statistical analysis, is difficult to reflect the actual fault state of the cable in real time, and reduces the reliability of the cable line operation is overcome. Compared with the traditional cable line fault state monitoring method, the present invention obtains the characteristic data of the high-voltage parallel cable line and inputs it into the preset line fault state assessment model, and optimizes the internal structure of the line fault state assessment model, thereby achieving accurate assessment of the fault state assessment result of the high-voltage parallel cable line, and improving the reliability of the cable line operation.

[0074] See also Figure 3 , Figure 3 A flowchart of the steps of a method for online monitoring of fault status of a high-voltage parallel cable line provided in Embodiment 2 of the present invention.

[0075] The present invention provides a method for online monitoring of fault status of a high-voltage parallel cable line, comprising:

[0076] Step 201, obtaining electrical quantity parameters of each monitoring point of the high-voltage parallel cable line within a preset sampling interval, and performing data preprocessing on each electrical quantity parameter based on the sampling interval to obtain corresponding characteristic data;

[0077] Furthermore, the electrical quantity parameters include the first grounding current, the second grounding current, the load current, the sheath induced voltage, the fault time, the start time and the end time. Step 201 includes the following sub-steps:

[0078] S11, performing difference processing on each second ground current and the associated first ground current respectively to obtain a first difference;

[0079] S12, performing ratio processing on the first difference and the sampling interval respectively to obtain the ground current change rate corresponding to each monitoring point;

[0080] The first ground current refers to the ground current value at the monitoring point at the previous moment.

[0081] The second ground current refers to the ground current value of the monitoring point at the current moment.

[0082] In the embodiment of the present invention, the sampling interval, each second ground current and the associated first ground current are respectively input into a preset ground current change rate function to obtain the ground current change rate corresponding to each monitoring point.

[0083] It should be noted that the ground current change rate function is specifically:

[0084]

[0085] in, is the ground current change rate, is the first ground current, is the second ground current, is the sampling interval.

[0086] S13, respectively inputting two load currents connected in parallel at the same node in each load current into a preset load current imbalance function to obtain multiple load current imbalances;

[0087] In the embodiment of the present invention, the load current unbalance degree of two load currents connected in parallel at the same node in each load current is calculated respectively by using a preset load current unbalance degree function.

[0088] It should be noted that the load current imbalance function is specifically:

[0089]

[0090] in, is the load current imbalance, is the load current value of parallel circuit 1, is the load current value of parallel circuit 2.

[0091] It should be noted that the two load currents connected in parallel at the same node are load currents collected by two sensors with consecutive numbers.

[0092] It should be noted that the load current is the load current of the three phases of the monitoring points a, b, and c, and the load current unbalance degree is the load current unbalance degree of the three phases of the monitoring points a, b, and c.

[0093] S14, inputting each load current into a preset phase current imbalance function to obtain the phase current imbalance corresponding to each monitoring point;

[0094] In the embodiment of the present invention, the inter-phase current imbalance of each load current is calculated respectively by using a preset inter-phase current imbalance function.

[0095] It should be noted that the phase current imbalance function is specifically:

[0096]

[0097]

[0098]

[0099] in, is the phase current imbalance degree of phase a at monitoring point, is the phase current imbalance degree of phase b at monitoring point, is the phase current imbalance degree of phase c at monitoring point, is the load current of phase a at monitoring point, is the load current of phase b at the monitoring point, is the load current of phase c at the monitoring point.

[0100] S15, performing ratio processing on each second grounding current and the associated load current respectively to obtain the grounding current load ratio corresponding to each monitoring point;

[0101] In the embodiment of the present invention, the ground current load ratio between each second ground current and the associated load current is calculated respectively.

[0102] It should be noted that the ground current load ratio expression is:

[0103]

[0104] in, is the ground current load ratio, is the load current.

[0105] It should be noted that the grounding current load ratio includes the grounding current load ratio of the three phases of monitoring points a, b, and c.

[0106] S16. Use each grounding current change rate, each load current imbalance, each phase current imbalance, each grounding current load ratio, each sheath induced voltage, each second grounding current, each load current, each fault time, each start time and each end time as characteristic data.

[0107] In an embodiment of the present invention, each ground current change rate, each load current imbalance, each phase current imbalance, each ground current load ratio, each sheath induced voltage, each second ground current, each load current, each fault time, each start time and each end time are used as characteristic data.

[0108] Step 202: input the feature data into a pre-trained line fault state assessment model, wherein the line fault state assessment model includes a first detection network, a second detection network, and a Leaky Relu activation layer connected in sequence;

[0109] In an embodiment of the present invention, feature data is input into a pre-trained line fault state assessment model for state assessment, wherein the line fault state assessment model includes a first detection network, a second detection network and a Leaky Relu activation layer connected in sequence.

[0110] In another embodiment, the feature data is input into a pre-trained line fault state assessment model, wherein the line fault state assessment model may be a support vector machine or a random forest model.

[0111] It is worth mentioning that before inputting the feature data into the pre-trained line fault state assessment model for state assessment, the feature data can be pre-processed to obtain pre-processed feature data. The pre-processing specifically includes but is not limited to data standardization, calculation of co-amplification difference matrix or dimensionality reduction processing.

[0112] Step 203: extracting features from the feature data through a first detection network, wherein the first detection network includes a 10×10 standard convolutional layer and a data normalization layer connected in sequence;

[0113] In the embodiment of the present invention, refer to Figure 4 As shown, a first detection network is used to extract features from feature data, wherein the first detection network includes a 10×10 standard convolutional layer and a data normalization layer connected in sequence.

[0114] Step 204: extract features from the feature data using a 10×10 standard convolutional layer to obtain a first data feature map;

[0115] In the embodiment of the present invention, feature extraction is performed on feature data through a 10×10 standard convolution layer to obtain a first data feature map.

[0116] Step 205: normalize the first data feature map through a data normalization layer to obtain a first feature map.

[0117] In the embodiment of the present invention, the first data feature map is normalized by a data normalization layer to obtain a first feature map.

[0118] It is worth mentioning that by normalizing the first feature map of the input, the data has specific distribution characteristics, thereby improving the training efficiency and performance of the line fault status assessment model.

[0119] Step 206: extract features from the first feature map through a second detection network to obtain a second feature map;

[0120] Further, see Figure 4 As shown, the second detection network includes a first detection module, a 1×1 convolutional layer, a second detection module, a 1×1 convolutional layer and a feature fusion layer, and step 206 includes the following sub-steps:

[0121] S21, using a first detection module to perform feature extraction on the first feature map to obtain a first detection feature map, wherein the first detection module includes a first extraction branch, a feature fusion layer, a batch normalization layer and a Tanh activation layer;

[0122] Further, see Figure 5 As shown, S21 includes the following sub-steps:

[0123] S211, extracting features from the first feature map through a first extraction branch to obtain a first residual feature map, wherein the first extraction branch includes a 1×1 convolutional layer, a 3×3 convolutional layer, a batch normalization layer, a Tanh activation layer, and a 1×1 standard convolutional layer connected in sequence;

[0124] In this embodiment of the present invention, feature extraction is performed on the first feature map through a 1×1 convolution layer, a 3×3 convolution layer, a batch normalization layer, a Tanh activation layer and a 1×1 standard convolution layer in sequence to obtain a first residual feature map.

[0125] S212, using a feature fusion layer to perform feature fusion on the first residual feature map and the first feature map to obtain a first fused feature map;

[0126] In the embodiment of the present invention, the first residual feature map and the first feature map are feature fused by a feature fusion layer to obtain a first fused feature map.

[0127] S213, performing nonlinear transformation on the first fusion feature map through a batch normalization layer and a Relu activation layer in sequence to obtain a first detection feature map.

[0128] In the embodiment of the present invention, the first fusion feature map is nonlinearly transformed through a batch normalization layer and a Relu activation layer in sequence to obtain a first detection feature map.

[0129] S22, performing a downsampling operation on the first detection feature map through a 1×1 convolutional layer to obtain a first sampling feature map;

[0130] In the embodiment of the present invention, a 1×1 convolutional layer is used to perform a downsampling operation on the first detection feature map to obtain a first sampling feature map.

[0131] S23, extracting features from the first detection feature map through a second detection module to obtain a second detection feature map, wherein the second detection module includes a second extraction branch, a 1×1 convolution layer, a feature fusion layer, a batch normalization layer, and a Relu activation layer;

[0132] Further, see Figure 6 As shown, S23 includes the following sub-steps:

[0133] S231, extracting features from the first detection feature map through a second extraction branch to obtain a second residual feature map, wherein the second extraction branch includes a 1×1 standard convolution layer, a batch normalization layer, a ReLU activation layer, a 3×3 convolution layer, a ReLU activation layer, and a 1×1 standard convolution layer connected in sequence;

[0134] In this embodiment of the present invention, feature extraction is performed on the first detection feature map through a 1×1 standard convolution layer, a batch normalization layer, a ReLU activation layer, a 3×3 convolution layer, a ReLU activation layer and a 1×1 standard convolution layer in sequence to obtain a second residual feature map.

[0135] S232, extracting features from the first detection feature map through a 1×1 convolutional layer to obtain a third residual feature map;

[0136] In the embodiment of the present invention, a 1×1 convolutional layer is used to perform feature extraction on the first detection feature map to obtain a third residual feature map.

[0137] S233, using a feature fusion layer to perform feature fusion on the second residual feature map and the third residual feature map to obtain a second fused feature map;

[0138] In the embodiment of the present invention, the second residual feature map and the third residual feature map are feature fused by a feature fusion layer to obtain a second fused feature map.

[0139] S234, performing nonlinear mapping on the second fusion feature map through a batch normalization layer and a Relu activation layer in sequence to obtain a second detection feature map.

[0140] In the embodiment of the present invention, the second fusion feature map is nonlinearly mapped through a batch normalization layer and a Relu activation layer in sequence to obtain a second detection feature map.

[0141] It is worth mentioning that in the process of feature data analysis, the Relu activation layer has fast calculation speed and can alleviate the gradient disappearance problem.

[0142] S245, extracting features from the second detection feature map through a 1×1 convolutional layer to obtain a second sampling feature map;

[0143] In the embodiment of the present invention, a 1×1 convolutional layer is used to perform feature extraction on the second detection feature map to obtain a second sampling feature map.

[0144] S25. Use a feature fusion layer to fuse the first sampling feature map and the second sampling feature map to obtain a second feature map.

[0145] In the embodiment of the present invention, the first sampling feature map and the second sampling feature map are feature fused by a feature fusion layer to obtain a second feature map.

[0146] Step 207: Use the Leaky Relu activation layer to perform state evaluation on the second feature map to obtain a fault state evaluation result corresponding to the high-voltage parallel cable line.

[0147] In an embodiment of the present invention, the state of the second feature map is evaluated through the Leaky Relu activation layer to obtain a fault state evaluation result corresponding to the high-voltage parallel cable line, and the fault state evaluation result is loaded into the monitoring page for visual display.

[0148] It should be noted that the visual display of the fault status assessment result can be any one of a dynamic chart, a heat map or a visual report.

[0149] It is worth mentioning that the Leaky Relu activation layer can accurately analyze the feature data when there are some negative fluctuations or special circumstances in the feature data.

[0150] In an embodiment of the present invention, by obtaining the electrical quantity parameters of each monitoring point of the high-voltage parallel cable line within a preset sampling interval, and performing data preprocessing on each electrical quantity parameter, characteristic data is obtained, and the characteristic data includes the status information of each monitoring point in the high-voltage parallel cable line. By inputting the characteristic data into a pre-trained line fault state assessment model, the fault state assessment result of the high-voltage parallel cable line can be obtained quickly and accurately. The technical problem that the existing cable line fault state monitoring method mainly relies on manual experience and simple statistical analysis, is difficult to reflect the actual fault state of the cable in real time, and reduces the reliability of the cable line operation is overcome. Compared with the traditional cable line fault state monitoring method, the present invention obtains the characteristic data of the high-voltage parallel cable line and inputs it into the preset line fault state assessment model, and optimizes the internal structure of the line fault state assessment model, thereby achieving accurate assessment of the fault state assessment result of the high-voltage parallel cable line, and improving the reliability of the cable line operation.

[0151] See also Figure 7 , Figure 7 This is a structural block diagram of a high-voltage parallel cable line fault status online monitoring system provided in Embodiment 3 of the present invention.

[0152] The present invention provides an online monitoring system for fault status of a high-voltage parallel cable line, comprising:

[0153] The acquisition module 301 is used to obtain the electrical quantity parameters of each monitoring point of the high-voltage parallel cable line within a preset sampling interval, and perform data preprocessing on each electrical quantity parameter based on the sampling interval to obtain corresponding characteristic data;

[0154] An input module 302, used to input feature data into a pre-trained line fault state assessment model, wherein the line fault state assessment model includes a first detection network, a second detection network and a Leaky Relu activation layer connected in sequence;

[0155] A first extraction module 303, configured to extract features from feature data through a first detection network to obtain a first feature map;

[0156] A second extraction module 304 is used to extract features from the first feature map through a second detection network to obtain a second feature map;

[0157] The evaluation module 305 is used to perform state evaluation on the second feature map by using a Leaky Relu activation layer to obtain a fault state evaluation result corresponding to the high-voltage parallel cable line.

[0158] Furthermore, the electrical quantity parameters include the first grounding current, the second grounding current, the load current, the sheath induced voltage, the fault time, the start time and the end time. The acquisition module 301 includes:

[0159] A first data analysis submodule is used to perform difference processing on each second ground current and the associated first ground current to obtain a first difference;

[0160] The first difference value is respectively processed with the sampling interval to obtain the ground current change rate corresponding to each monitoring point;

[0161] The second data analysis submodule is used to input two load currents connected in parallel at the same node in each load current into a preset load current imbalance function to obtain multiple load current imbalances;

[0162] The third data analysis submodule is used to input each load current into a preset phase current imbalance function to obtain the phase current imbalance corresponding to each monitoring point;

[0163] The fourth data analysis submodule is used to perform ratio processing on each second grounding current and the associated load current to obtain the grounding current load ratio corresponding to each monitoring point;

[0164] The fifth data analysis submodule is used to use each ground current change rate, each load current imbalance, each phase current imbalance, each ground current load ratio, each sheath induced voltage, each second ground current, each load current, each fault time, each start time and each end time as characteristic data.

[0165] Furthermore, the first detection network includes a 10×10 standard convolution layer and a data normalization layer connected in sequence, and the first extraction module 303 includes:

[0166] A first extraction submodule, used to extract features from feature data using a 10×10 standard convolutional layer to obtain a first data feature map;

[0167] The second extraction submodule is used to normalize the first data feature map through a data normalization layer to obtain a first feature map.

[0168] Furthermore, the second detection network includes a first detection module, a 1×1 convolution layer, a second detection module, a 1×1 convolution layer and a feature fusion layer, and a second extraction module 304 includes:

[0169] A third extraction submodule, configured to perform feature extraction on the first feature map using the first detection module to obtain a first detection feature map, wherein the first detection module includes a first extraction branch, a feature fusion layer, a batch normalization layer, and a Tanh activation layer;

[0170] A sampling submodule, used for performing a downsampling operation on the first detection feature map through a 1×1 convolution layer to obtain a first sampling feature map;

[0171] a fourth extraction submodule, configured to perform feature extraction on the first detection feature map through a second detection module to obtain a second detection feature map, wherein the second detection module includes a second extraction branch, a 1×1 convolution layer, a feature fusion layer, a batch normalization layer, and a Relu activation layer;

[0172] a fifth extraction submodule, configured to perform feature extraction on the second detection feature map through a 1×1 convolutional layer to obtain a second sampling feature map;

[0173] The first fusion submodule is used to use a feature fusion layer to perform feature fusion on the first sampling feature map and the second sampling feature map to obtain a second feature map.

[0174] Furthermore, the third extraction submodule includes:

[0175] A first extraction unit, configured to perform feature extraction on the first feature map through a first extraction branch to obtain a first residual feature map, wherein the first extraction branch includes a 1×1 convolutional layer, a 3×3 convolutional layer, a batch normalization layer, a Tanh activation layer, and a 1×1 standard convolutional layer connected in sequence;

[0176] A first fusion unit is used to perform feature fusion on the first residual feature map and the first feature map by using a feature fusion layer to obtain a first fused feature map;

[0177] The first transformation unit is used to perform nonlinear transformation on the first fusion feature map through a batch normalization layer and a Relu activation layer in sequence to obtain a first detection feature map.

[0178] Furthermore, the fourth extraction submodule includes:

[0179] A second extraction unit is used to extract features from the first detection feature map through a second extraction branch to obtain a second residual feature map, wherein the second extraction branch includes a 1×1 standard convolution layer, a batch normalization layer, a ReLU activation layer, a 3×3 convolution layer, a ReLU activation layer, and a 1×1 standard convolution layer connected in sequence;

[0180] A third extraction unit, used for performing feature extraction on the first detection feature map through a 1×1 convolution layer to obtain a third residual feature map;

[0181] A second fusion unit is used to perform feature fusion on the second residual feature map and the third residual feature map by using a feature fusion layer to obtain a second fused feature map;

[0182] The second transformation unit is used to perform nonlinear mapping on the second fusion feature map through a batch normalization layer and a Relu activation layer in sequence to obtain a second detection feature map.

[0183] See also Figure 8, Figure 8 This is a structural block diagram of an electronic device provided in Embodiment 4 of the present invention.

[0184] An electronic device according to an embodiment of the present invention comprises: a memory 401 and a processor 402, wherein the memory 402 stores a computer program; when the computer program is executed by the processor 402, the processor 402 executes a method for online monitoring of a fault status of a high-voltage parallel cable line according to any of the above embodiments.

[0185] The memory 401 may be an electronic memory such as a flash memory, an EEPROM (electrically erasable programmable read-only memory), an EPROM, a hard disk or a ROM. The memory 401 has a storage space 403 for a program code 413 for executing any method step in the above method. For example, the storage space 403 for the program code may include individual program codes 413 for implementing the various steps in the above method, respectively. These program codes may be read from or written to one or more computer program products. These computer program products include program code carriers such as a hard disk, a compact disk (CD), a memory card or a floppy disk. The program code may be compressed, for example, in an appropriate form. When these codes are run by a computing and processing device, the computing and processing device is caused to execute the various steps in the above-described method.

[0186] Embodiment 5 of the present invention further provides a computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, the method for online monitoring of the fault status of a high-voltage parallel cable line as described in any of the above embodiments is implemented.

[0187] Embodiment 6 of the present invention further provides a computer program product, which includes a computer program stored on a non-transitory computer-readable storage medium, and the computer program includes program instructions, wherein when the program instructions are executed by a computer, the computer executes the online monitoring method for the fault status of a high-voltage parallel cable line as in any of the above embodiments.

[0188] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0189] In the several embodiments provided in the present application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of units is only a logical function division. There may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be an indirect coupling or communication connection through some interfaces, devices or units, which can be electrical, mechanical or other forms.

[0190] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0191] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0192] If the integrated unit is implemented in the form of a software functional unit 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, in essence, 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, including several instructions for a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the methods of each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc. Various media that can store program codes.

[0193] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for online monitoring of fault status of a high-voltage parallel cable line, characterized in that: include: Acquire electrical quantity parameters of each monitoring point of the high-voltage parallel cable line within a preset sampling interval, and perform data preprocessing on each of the electrical quantity parameters based on the sampling interval to obtain corresponding characteristic data; Inputting the feature data into a pre-trained line fault state assessment model, wherein the line fault state assessment model includes a first detection network, a second detection network and a Leaky Relu activation layer connected in sequence; Performing feature extraction on the feature data through the first detection network to obtain a first feature map; Performing feature extraction on the first feature map by using the second detection network to obtain a second feature map; A Leaky Relu activation layer is used to perform state evaluation on the second feature map to obtain a fault state evaluation result corresponding to the high-voltage parallel cable line.

2. The method for online monitoring of fault status of a high-voltage parallel cable line according to claim 1, characterized in that: The electrical quantity parameters include a first grounding current, a second grounding current, a load current, a sheath induced voltage, a fault time, a start time, and an end time. The step of performing data preprocessing on each of the electrical quantity parameters based on the sampling interval to obtain corresponding characteristic data includes: performing difference processing on each of the second grounding currents and the associated first grounding current to obtain a first difference; Performing ratio processing on the first difference and the sampling interval respectively to obtain the ground current change rate corresponding to each monitoring point; Inputting two load currents connected in parallel at the same node in each of the load currents into a preset load current imbalance function to obtain a plurality of load current imbalances; Inputting each of the load currents into a preset phase-to-phase current imbalance function respectively to obtain the phase-to-phase current imbalance corresponding to each of the monitoring points; Performing ratio processing on each of the second grounding currents and the associated load current respectively to obtain the grounding current load ratio corresponding to each of the monitoring points; Each of the grounding current change rates, each of the load current imbalances, each of the phase current imbalances, each of the grounding current load ratios, each of the sheath induced voltages, each of the second grounding currents, each of the load currents, each of the fault times, each of the start times and each of the end times are used as characteristic data.

3. The method for online monitoring of fault status of a high-voltage parallel cable line according to claim 1, characterized in that: The first detection network includes a 10×10 standard convolution layer and a data normalization layer connected in sequence, and the step of extracting features from the feature data through the first detection network to obtain a first feature map includes: A 10×10 standard convolutional layer is used to extract features from the feature data to obtain a first data feature map; The first data feature map is normalized by a data normalization layer to obtain a first feature map.

4. The method for online monitoring of fault status of a high-voltage parallel cable line according to claim 1, characterized in that: The second detection network includes a first detection module, a 1×1 convolutional layer, a second detection module, a 1×1 convolutional layer and a feature fusion layer. The step of extracting features from the first feature map through the second detection network to obtain a second feature map includes: Using the first detection module to perform feature extraction on the first feature map to obtain a first detection feature map, wherein the first detection module includes a first extraction branch, a feature fusion layer, a batch normalization layer, and a Tanh activation layer; Downsampling the first detection feature map through a 1×1 convolutional layer to obtain a first sampling feature map; Performing feature extraction on the first detection feature map through the second detection module to obtain a second detection feature map, wherein the second detection module includes a second extraction branch, a 1×1 convolution layer, a feature fusion layer, a batch normalization layer and a Relu activation layer; Performing feature extraction on the second detection feature map through a 1×1 convolutional layer to obtain a second sampling feature map; A feature fusion layer is used to perform feature fusion on the first sampling feature map and the second sampling feature map to obtain a second feature map.

5. The method for online monitoring of fault status of high-voltage parallel cable lines according to claim 4 is characterized in that: The step of using the first detection module to extract features from the first feature map to obtain a first detection feature map includes: Performing feature extraction on the first feature map through the first extraction branch to obtain a first residual feature map, wherein the first extraction branch includes a 1×1 convolutional layer, a 3×3 convolutional layer, a batch normalization layer, a Tanh activation layer, and a 1×1 standard convolutional layer connected in sequence; Using a feature fusion layer to perform feature fusion on the first residual feature map and the first feature map to obtain a first fused feature map; The first fusion feature map is nonlinearly transformed through a batch normalization layer and a Relu activation layer in sequence to obtain a first detection feature map.

6. The method for online monitoring of fault status of high-voltage parallel cable lines according to claim 4, characterized in that: The step of extracting features from the first detection feature map by the second detection module to obtain a second detection feature map comprises: Performing feature extraction on the first detection feature map through the second extraction branch to obtain a second residual feature map, wherein the second extraction branch includes a 1×1 standard convolution layer, a batch normalization layer, a ReLU activation layer, a 3×3 convolution layer, a ReLU activation layer, and a 1×1 standard convolution layer connected in sequence; Performing feature extraction on the first detection feature map through a 1×1 convolutional layer to obtain a third residual feature map; Using a feature fusion layer to perform feature fusion on the second residual feature map and the third residual feature map to obtain a second fused feature map; The second fusion feature map is nonlinearly mapped through a batch normalization layer and a Relu activation layer in sequence to obtain a second detection feature map.

7. A high-voltage parallel cable line fault status online monitoring system, characterized in that: include: An acquisition module is used to obtain electrical quantity parameters of each monitoring point of the high-voltage parallel cable line within a preset sampling interval, and perform data preprocessing on each of the electrical quantity parameters based on the sampling interval to obtain corresponding characteristic data; An input module, used for inputting the feature data into a pre-trained line fault state assessment model, wherein the line fault state assessment model comprises a first detection network, a second detection network and a Leaky Relu activation layer connected in sequence; A first extraction module, configured to extract features from the feature data through the first detection network to obtain a first feature map; A second extraction module, configured to extract features from the first feature map through the second detection network to obtain a second feature map; An evaluation module is used to use a Leaky Relu activation layer to perform a state evaluation on the second feature map to obtain a fault state evaluation result corresponding to the high-voltage parallel cable line.

8. An electronic device, characterized in that: It comprises 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 steps of the method for online monitoring of the fault status of a high-voltage parallel cable line as described in any one of claims 1 to 6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed, the method for online monitoring of the fault status of a high-voltage parallel cable line as described in any one of claims 1 to 6 is implemented.

10. A computer program product, characterized in that The computer program product includes a computer program stored on a non-transitory computer-readable storage medium, and the computer program includes program instructions, wherein when the program instructions are executed by a computer, the computer executes the online monitoring method for the fault status of a high-voltage parallel cable line as described in any one of claims 1-6.

Images