A station end inspection path autonomous planning and positioning generation method and system

Abstract

The application discloses a kind of station end inspection path autonomous planning and positioning generation method and system, its method is through to substation operation fault multi-source data acquisition, preprocessing and prediction;Realize the change trend of judging substation operation fault related data in advance;While through difference calculation and expert customized defect level and data change difference relationship table matching inquiry fault type, maintenance deadline upper limit time, fault reason and maintenance time;Established with the shortest inspection maintenance path or the shortest inspection maintenance time establishes objective function;And solve optimal inspection maintenance path, realize the position data of the inspection station end collection point needed for maintenance is sent to maintenance personnel receiving end, provides optimal inspection maintenance path for maintenance personnel, improves the efficiency of maintenance personnel inspection maintenance.It solves the problem that existing station end monitoring unit inspection personnel is difficult to obtain substation fault severity through single data simple analysis, and lacks reasonable planning inspection path.

Description

Technical Field

[0001] This invention belongs to the field of substation-related data monitoring technology, specifically, it relates to a method and system for autonomous planning and positioning generation of substation inspection paths. Background Technology

[0002] The station-end monitoring unit is a device that takes the substation as its target and is responsible for the analysis of all monitoring data within the station and the management of monitoring devices and integrated monitoring units. It has control functions such as data comprehensive analysis, early warning, parameter setting, data retrieval, time synchronization, and forced restart, and can communicate in a formatted manner with other systems in the station control layer and the upper-level platform.

[0003] Existing autonomous planning of substation inspection routes uses substation robots or substation information data to analyze substation faults. However, simple analysis of single data to obtain substation faults cannot further determine the cause of substation operation failure, the upper limit of maintenance period, and maintenance time. This also makes it impossible for subsequent maintenance personnel to determine the optimal inspection and maintenance route for multiple inspection and maintenance points, resulting in low maintenance planning efficiency. Summary of the Invention

[0004] To address the problem that existing substation monitoring unit inspectors find it difficult to obtain the severity of substation faults through simple analysis of single data, and lack reasonable inspection route planning, this invention provides a method and system for autonomous planning and location generation of substation inspection routes.

[0005] To achieve the above-mentioned technical objectives, the technical solution adopted by the present invention is as follows: A method for autonomous planning and positioning generation of station inspection paths includes the following steps: S1. Acquire multi-source data on substation operation faults through the acquisition equipment at preset station-end acquisition points; S2. Preprocess the multi-source data of substation operation faults to obtain the preprocessed data related to substation operation faults and the location data of station-end collection points. S3. Input the preprocessed substation operation fault related data into the substation operation fault prediction model, and output the substation operation fault prediction data. S4. Subtract the substation operation fault prediction data from the predicted input substation operation fault related data to obtain the data change difference value. S5. Based on the data change difference, query the fault type, upper limit of the repair period, fault cause and repair time in the pre-stored defect level and data change difference relationship table; S6. Take the station-end data collection point location data, fault cause, upper limit of maintenance period and maintenance time into the objective function, and establish the objective function with the shortest inspection and maintenance path or the shortest inspection and maintenance time. S7. Calculate the optimal inspection and maintenance route, and send the location data of the inspection station end collection points required for maintenance to the maintenance personnel receiving end.

[0006] Furthermore, the data acquisition equipment at the substation end includes a temperature sensor, a voltage sensor, and a magnetic field sensor; the temperature sensor is used to acquire temperature data of the equipment at the substation end during operation, the voltage sensor is used to acquire voltage data of the equipment at the substation end during operation, and the magnetic field sensor is used to acquire magnetic field data of the equipment at the substation end during operation.

[0007] Multi-source data on substation operational faults includes temperature data, voltage data, magnetic field data, equipment ID, acquisition time, and the location of the acquisition point at the substation. Temperature sensors, voltage sensors, and magnetic field sensors are installed at different locations and for different equipment types at the substation. The changing trends of substation operational faults are predicted by segmenting the multi-source data from different equipment types and locations.

[0008] Furthermore, the detailed steps for preprocessing multi-source data on substation operational faults in step S2 include: S201. Deletion and replacement of outlier values ​​(deviation from expected values) in multi-source data of substation operation faults in the same channel; S202. When multi-source data on substation operation faults collected at the same time are missing, the expected value of historical data shall be used to supplement them. S203. For different types of substation operation faults, multi-source data is separated. The data of the same category is imported into a data table and stored in chronological order by matching the unit or field ID. S204. Filter out the location data of the station-end acquisition points. This facilitates subsequent prediction of multi-source data on substation operation faults and the transmission of corresponding station-end acquisition point location data.

[0009] Furthermore, the substation operation fault prediction model adopts an adaptive BP neural network model, which includes an input layer, a hidden layer, and an output layer. All layers propagate in one direction to each other. Each node is a single neuron, and the connection weights between layers reflect the connection strength between the neurons.

[0010] Furthermore, the training and learning phase of the adaptive BP neural network model includes forward propagation of the working signal and backward propagation of the error signal; Working signal forward propagation: Input information x i After processing by the hidden layer, the output value y of each neuron is obtained. j Hidden layer output y j Output layer output z l The formulas are as follows: in, The threshold value for hidden layer nodes. is the threshold of the output layer nodes; f(x) represents the activation function of the hidden layer, p(x) represents the activation function of the output layer; n is the number of inputs in the input layer, and m is the number of neurons in the hidden layer; Indicates the input layer node x i and hidden layer node y i The connection weights between them; v ji Represents the hidden layer node y i and output layer node z l The connection weights between them; Backpropagation of error signal: The network is adjusted based on the error function, modifying the connection weights and threshold parameters. The calculation formula is as follows: Where H is the number of samples, and the expected value of the l-th output node of the h-th sample is t. hl The actual output of the forward propagation through the neural network is z. hl .

[0011] Furthermore, the pre-stored table relating defect levels to data change differences should include at least the threshold change range, fault type, fault cause, maximum repair period, repair time, and device ID. After obtaining the data change difference in step S4, it will first determine whether the data change difference is greater than a preset threshold. Then, the device IDs with data change differences greater than the preset threshold will be imported into the pre-stored defect level and data change difference relationship table to determine the fault type, fault cause, upper limit of repair period, and repair time.

[0012] Furthermore, the detailed steps in step S6 for establishing the objective function based on the shortest inspection and maintenance path are as follows: Create a graph object with the currently required maintenance nodes and the paths between them to store the node and maintenance path data; Enter the maximum repair period and repair time as limiting conditions; The objective function formula for finding the shortest inspection and maintenance path is established using the Floyd algorithm.

[0013] Furthermore, the Floyd algorithm is used to establish the objective function of finding the shortest inspection and maintenance path. Through dynamic programming, intermediate nodes are gradually introduced to optimize the path. The core recursive formula is as follows: in, This represents the shortest path length from i to j when only the set of nodes {1,2,…,k} is allowed as intermediate nodes. k iterates from 1 to n, indicating that node k is gradually used as an intermediate node; Finally, when k=n, This is the length of the global shortest path from i to j.

[0014] Furthermore, in step S6, the objective function is established with the goal of minimizing the inspection and maintenance time: The path travel time between inspection nodes and the dwell time for maintenance at each node are converted into path lengths between nodes to create a graph object for storing data on path travel time between nodes and maintenance inspection nodes, as well as dwell time for maintenance at each node. The path travel time between inspection nodes and the dwell time for maintenance at each node are considered as the length of the inspection and maintenance path. Based on this, a method for calculating the shortest path is adopted. The optimal method for calculating inspection and maintenance paths is then developed.

[0015] At the same time, the maximum repair period for each node is entered as a constraint. The repair time has already been embedded in the node path in the previous steps, so it does not need to be considered as a constraint again. It is only necessary to consider whether the maximum repair period for each node is met when it is traversed. This is to prevent the inspection and repair time of nodes with serious problems from being delayed, which would exceed the repair launch time for the cause of the fault.

[0016] The objective function formula for finding the shortest inspection and maintenance path is established using the Floyd algorithm.

[0017] A substation inspection path autonomous planning and positioning generation system includes a substation operation fault multi-source data acquisition unit, a data preprocessing unit, a substation operation fault data change prediction unit, a change data difference calculation unit, a fault query unit, an inspection and maintenance path shortest objective function establishment unit, an inspection and maintenance time shortest objective function establishment unit, an optimal inspection and maintenance path calculation unit, an inspection station collection point location data transmission unit, and a data storage unit. The substation operation fault multi-source data acquisition unit acquires multi-source data on substation operation faults through acquisition equipment at preset station-end acquisition points. The data preprocessing unit is connected to the substation operation fault multi-source data acquisition unit to preprocess the substation operation fault multi-source data and obtain the preprocessed substation operation fault related data and station end acquisition point location data. The substation operation fault data change prediction unit communicates with the data preprocessing unit, inputs the preprocessed substation operation fault related data into the substation operation fault prediction model, and outputs substation operation fault prediction data. The data change difference calculation unit is connected to the substation operation fault data change prediction unit. It calculates the difference between the substation operation fault prediction data and the substation operation fault related data input for prediction to obtain the data change difference. The fault query unit is connected to the change data difference calculation unit. Based on the data change difference, it queries the fault type, the upper limit of the maintenance period, the cause of the fault, and the maintenance time in the pre-stored defect level and data change difference relationship table. The unit for establishing the shortest inspection and maintenance path objective function is connected to the fault query unit. The station-end data collection point location data, fault cause, upper limit of maintenance period and maintenance time are taken into account in the objective function to establish the objective function with the shortest inspection and maintenance path. The objective function for minimizing inspection and maintenance time is established by establishing a communication connection between the inspection and maintenance unit and the fault query unit. The station-end data collection point location data, fault causes, maintenance deadline, and maintenance time are taken into account in the objective function to establish the objective function with the shortest inspection and maintenance time. The optimal inspection and maintenance path calculation unit, the shortest inspection and maintenance path objective function establishment unit, and the shortest inspection and maintenance time objective function establishment unit are all equipped with communication connections for calculating the optimal inspection and maintenance path; The inspection station end data collection point location data sending unit is connected to the optimal inspection and maintenance path calculation unit, and is used to send the inspection station end data collection point location data required for maintenance to the maintenance personnel receiving end; The data storage unit is used to store relevant data in the table showing the relationship between defect levels and data change differences, as well as multi-source data on historical substation operation faults.

[0018] Compared with the prior art, the present invention has the following advantages: By collecting, preprocessing, and predicting multi-source data on substation operational faults, the system enables early assessment of trends in relevant data. Simultaneously, it uses difference calculations and expert-customized tables of defect levels and data change differences to query fault types, maximum repair time limits, fault causes, and repair times. An objective function is established based on minimizing the inspection and maintenance path or time. The optimal inspection and maintenance path is then calculated, enabling the transmission of data from the inspection station's data collection points to maintenance personnel, providing them with the optimal path and improving their efficiency. This addresses the problem of existing station-end monitoring units where inspectors struggle to determine the severity of substation faults through simple analysis of single data points and lack the ability to rationally plan inspection paths. Attached Figure Description

[0019] Figure 1 This is an overall flowchart of a station-side inspection path autonomous planning and positioning generation method in an embodiment of the present invention; Figure 2This is a structural block diagram of a station-side inspection path autonomous planning and positioning generation system according to an embodiment of the present invention; Figure 3 This is a partial schematic diagram of the multilayer feedforward BP network structure in an embodiment of the present invention. Detailed Implementation

[0020] To facilitate understanding by those skilled in the art, the present invention will be further described below with reference to embodiments and accompanying drawings. The content mentioned in the embodiments is not intended to limit the present invention.

[0021] like Figure 1 As shown, this embodiment provides a method for autonomous planning and location generation of station inspection paths, including the following steps: S1. Acquire multi-source data on substation operation faults through the acquisition equipment at preset station-end acquisition points; S2. Preprocess the multi-source data of substation operation faults to obtain the preprocessed data related to substation operation faults and the location data of station-end collection points. S3. Input the preprocessed substation operation fault related data into the substation operation fault prediction model, and output the substation operation fault prediction data. S4. Subtract the substation operation fault prediction data from the predicted input substation operation fault related data to obtain the data change difference value. S5. Based on the data change difference, query the fault type, the upper limit of the maintenance period (provide the upper limit of the maintenance time for the severity of node faults to prevent nodes with high severity from exceeding the maintenance time), the cause of the fault, and the maintenance time in the pre-stored defect level and data change difference relationship table. S6. Take the station-end data collection point location data, fault cause, upper limit of maintenance period and maintenance time into the objective function, and establish the objective function with the shortest inspection and maintenance path or the shortest inspection and maintenance time. S7. Calculate the optimal inspection and maintenance route, and send the location data of the inspection station points required for maintenance to the maintenance personnel's receiving end. Send the node path sequence and station point location data to the maintenance personnel to improve maintenance efficiency.

[0022] The data acquisition equipment at the substation end includes temperature sensors, voltage sensors, and magnetic field sensors. The temperature sensors are used to collect temperature data of the equipment at the substation end during operation, the voltage sensors are used to collect voltage data of the equipment at the substation end during operation, and the magnetic field sensors are used to collect magnetic field data of the equipment at the substation end during operation.

[0023] Multi-source data on substation operational faults includes temperature data, voltage data, magnetic field data, equipment ID, acquisition time, and the location of the acquisition point at the substation. Temperature sensors, voltage sensors, and magnetic field sensors are installed at different locations and for different equipment types at the substation. The changing trends of substation operational faults are predicted by segmenting the multi-source data from different equipment types and locations.

[0024] The detailed steps for preprocessing multi-source data on substation operational faults in step S2 include: S201. Deletion and replacement of outlier values ​​(deviation from expected values) in multi-source data of substation operation faults in the same channel; S202. When multi-source data on substation operation faults collected at the same time are missing, the expected value of historical data shall be used to supplement them. S203. For different types of substation operation faults, multi-source data is separated. The data of the same category is imported into a data table and stored in chronological order by matching the unit or field ID. S204. Filter out the location data of the station-end acquisition points. This facilitates subsequent prediction of multi-source data on substation operation faults and the transmission of corresponding station-end acquisition point location data.

[0025] like Figure 2 As shown, the substation operation fault prediction model adopts an adaptive BP neural network model, which includes an input layer, a hidden layer, and an output layer. All layers propagate in one direction to each other. Each node is a single neuron, and the connection weights between layers reflect the connection strength between the neurons.

[0026] The training and learning phase of the adaptive BP neural network model includes forward propagation of the working signal and backward propagation of the error signal; Working signal forward propagation: Input information x i After processing by the hidden layer, the output value y of each neuron is obtained. j Hidden layer output y j Output layer output z l The formulas are as follows: in, The threshold value for hidden layer nodes. is the threshold of the output layer nodes; f(x) represents the activation function of the hidden layer, p(x) represents the activation function of the output layer; n is the number of inputs in the input layer, and m is the number of neurons in the hidden layer; Indicates the input layer node x i and hidden layer node y i The connection weights between them; vji Represents the hidden layer node y i and output layer node z l The connection weights between them; Backpropagation of error signal: The network is adjusted based on the error function, modifying the connection weights and threshold parameters. The calculation formula is as follows: Where H is the number of samples, and the expected value of the l-th output node of the h-th sample is t. hl The actual output of the forward propagation through the neural network is z. hl .

[0027] The pre-stored table of the relationship between defect level and data change difference should include at least the threshold change range, fault type, fault cause, upper limit of repair period, repair time and equipment ID; After obtaining the data change difference in step S4, it will first determine whether the data change difference is greater than a preset threshold. The preset threshold is the minimum value in the threshold change range of the pre-stored defect level and data change difference relationship table. Then, the device IDs with data change differences greater than the preset threshold will be imported into the pre-stored defect level and data change difference relationship table to determine the fault type, fault cause, upper limit of repair period, and repair time.

[0028] The defect levels and causes of failures corresponding to changes in temperature difference data of electrical equipment at inspection nodes are shown in Table 1 below: Table 1 Relationship between common electrical equipment defect levels and temperature Therefore, abnormal changes in voltage and electromagnetic field at electrical equipment testing points correspond to different degrees and causes of faults; this can be used as the upper limit of the maintenance period and the standard for maintenance time during inspections.

[0029] The detailed steps for establishing the objective function based on the shortest inspection and maintenance path in step S6 are as follows: Create a graph object with the currently required maintenance nodes and the paths between them to store the node and maintenance path data; Enter the maximum repair period and repair time as limiting conditions; The objective function formula for finding the shortest inspection and maintenance path is established using the Floyd algorithm.

[0030] The Floyd-Warshall algorithm is used to establish the objective function of finding the shortest inspection and maintenance path. Through dynamic programming, intermediate nodes are gradually introduced to optimize the path. The core recursive formula is as follows: in, This represents the shortest path length from i to j when only the set of nodes {1,2,…,k} is allowed as intermediate nodes. This indicates that the path traversal is accelerated by simultaneously traversing the shortest path from both the initial node i and the final node j.

[0031] k iterates from 1 to n, indicating that node k is gradually used as an intermediate node; Finally, when k=n, This represents the global shortest path length from node i to node j. The recursive formula above, during the traversal calculation, first iterates through all path schemes, then sorts them by shortest path, and finally selects the shortest path that meets the constraints of the node's maintenance deadline and maintenance time as the maintenance path calculation result.

[0032] In step S6, the objective function is established to minimize the inspection and maintenance time: The path travel time between inspection nodes and the dwell time for maintenance at each node are converted into path lengths between nodes to create a graph object for storing data on path travel time between nodes and maintenance inspection nodes, as well as dwell time for maintenance at each node. The path travel time between inspection nodes and the dwell time for maintenance at each node are considered as the length of the inspection and maintenance path. Based on this, a method for calculating the shortest path is adopted. The optimal method for calculating inspection and maintenance paths is then developed.

[0033] At the same time, the maximum repair period for each node is entered as a constraint. The repair time has already been embedded in the node path in the previous steps, so it does not need to be considered as a constraint again. It is only necessary to consider whether the maximum repair period for each node is met when it is traversed. This is to prevent the inspection and repair time of nodes with serious problems from being delayed, which would exceed the repair launch time for the cause of the fault.

[0034] The objective function formula for finding the shortest inspection and maintenance path is established using the Floyd algorithm.

[0035] like Figure 3 As shown, a substation inspection path autonomous planning and positioning generation system includes a substation operation fault multi-source data acquisition unit, a data preprocessing unit, a substation operation fault data change prediction unit, a change data difference calculation unit, a fault query unit, an inspection and maintenance path shortest objective function establishment unit, an inspection and maintenance time shortest objective function establishment unit, an optimal inspection and maintenance path calculation unit, an inspection station end acquisition point location data transmission unit, and a data storage unit. The substation operation fault multi-source data acquisition unit acquires multi-source data on substation operation faults through acquisition equipment at preset station-end acquisition points. The data preprocessing unit is connected to the substation operation fault multi-source data acquisition unit to preprocess the substation operation fault multi-source data and obtain the preprocessed substation operation fault related data and station end acquisition point location data. The substation operation fault data change prediction unit communicates with the data preprocessing unit, inputs the preprocessed substation operation fault related data into the substation operation fault prediction model, and outputs substation operation fault prediction data. The data change difference calculation unit is connected to the substation operation fault data change prediction unit. It calculates the difference between the substation operation fault prediction data and the substation operation fault related data input for prediction to obtain the data change difference. The fault query unit is connected to the change data difference calculation unit. Based on the data change difference, it queries the fault type, the upper limit of the maintenance period, the cause of the fault, and the maintenance time in the pre-stored defect level and data change difference relationship table. The unit for establishing the shortest inspection and maintenance path objective function is connected to the fault query unit. The station-end data collection point location data, fault cause, upper limit of maintenance period and maintenance time are taken into account in the objective function to establish the objective function with the shortest inspection and maintenance path. The objective function for minimizing inspection and maintenance time is established by establishing a communication connection between the inspection and maintenance unit and the fault query unit. The station-end data collection point location data, fault causes, maintenance deadline, and maintenance time are taken into account in the objective function to establish the objective function with the shortest inspection and maintenance time. The optimal inspection and maintenance path calculation unit, the shortest inspection and maintenance path objective function establishment unit, and the shortest inspection and maintenance time objective function establishment unit are all equipped with communication connections for calculating the optimal inspection and maintenance path; The inspection station end data collection point location data sending unit is connected to the optimal inspection and maintenance path calculation unit, and is used to send the inspection station end data collection point location data required for maintenance to the maintenance personnel receiving end; The data storage unit is used to store relevant data in the table showing the relationship between defect levels and data change differences, as well as multi-source data on historical substation operation faults.

[0036] Compared with the prior art, the present invention has the following advantages: By collecting, preprocessing, and predicting multi-source data on substation operational faults, the system enables early assessment of trends in relevant data. Simultaneously, it uses difference calculations and expert-customized tables of defect levels and data change differences to query fault types, maximum repair time limits, fault causes, and repair times. An objective function is established based on minimizing the inspection and maintenance path or time. The optimal inspection and maintenance path is then calculated, enabling the transmission of data from the inspection station's data collection points to maintenance personnel, providing them with the optimal path and improving their efficiency. This addresses the problem of existing station-end monitoring units where inspectors struggle to determine the severity of substation faults through simple analysis of single data points and lack the ability to rationally plan inspection paths.

[0037] The above provides a detailed description of a method and system for autonomous planning and positioning of station inspection paths provided in this application. The specific embodiments described are only intended to aid in understanding the method and its core ideas. It should be noted that those skilled in the art can make various improvements and modifications to this application without departing from its principles, and these improvements and modifications also fall within the scope of protection of the claims of this application.

Claims

1. A method for autonomous planning and positioning generation of station inspection paths, characterized in that, Including the following steps: S1. Acquire multi-source data on substation operation faults through the acquisition equipment at preset station-end acquisition points; S2. Preprocess the multi-source data of substation operation faults to obtain the preprocessed data related to substation operation faults and the location data of station-end collection points. S3. Input the preprocessed substation operation fault related data into the substation operation fault prediction model, and output the substation operation fault prediction data. S4. Subtract the substation operation fault prediction data from the predicted input substation operation fault related data to obtain the data change difference value. S5. Based on the data change difference, query the fault type, upper limit of the repair period, fault cause and repair time in the pre-stored defect level and data change difference relationship table; S6. Take the station-end data collection point location data, fault cause, upper limit of maintenance period and maintenance time into the objective function, and establish the objective function with the shortest inspection and maintenance path or the shortest inspection and maintenance time. S7. Calculate the optimal inspection and maintenance route, and send the location data of the inspection station end collection points required for maintenance to the maintenance personnel receiving end.

2. The method for autonomous planning and positioning of station inspection paths according to claim 1, characterized in that, The data acquisition equipment at the substation end includes temperature sensors, voltage sensors, and magnetic field sensors; the temperature sensors are used to collect temperature data of the equipment at the substation end during operation, the voltage sensors are used to collect voltage data of the equipment at the substation end during operation, and the magnetic field sensors are used to collect magnetic field data of the equipment at the substation end during operation. Multi-source data for substation operation faults includes temperature data, voltage data, magnetic field data, equipment ID, acquisition time, and location data of the acquisition point at the substation.

3. The method for autonomous planning and positioning of station inspection paths according to claim 2, characterized in that, The detailed steps for preprocessing multi-source data on substation operational faults in step S2 include: S201. Deletion and replacement of abnormal values ​​in multi-source data of substation operation faults in the same channel; S202. When multi-source data on substation operation faults collected at the same time are missing, the expected value of historical data shall be used to supplement them. S203. For different types of substation operation faults, multi-source data is separated. The data of the same category is imported into a data table and stored in chronological order by matching the unit or field ID. S204. Filter out the location data of the station-end acquisition points. This facilitates subsequent prediction of multi-source data on substation operation faults and the transmission of corresponding station-end acquisition point location data.

4. The method for autonomous planning and positioning of station inspection paths according to claim 3, characterized in that, The substation operation fault prediction model adopts an adaptive BP neural network model, which includes an input layer, a hidden layer, and an output layer. All layers propagate in one direction to each other. Each node is a single neuron, and the connection weights between layers reflect the connection strength between the neurons.

5. The method for autonomous planning and positioning of station inspection paths according to claim 4, characterized in that, The training and learning phase of the adaptive BP neural network model includes forward propagation of the working signal and backward propagation of the error signal; Working signal forward propagation: Input information x i After processing by the hidden layer, the output value y of each neuron is obtained. j Hidden layer output y j Output layer output z l The formulas are as follows: ； in, The threshold value for hidden layer nodes. is the threshold of the output layer nodes; f(x) represents the activation function of the hidden layer, p(x) represents the activation function of the output layer; n is the number of inputs in the input layer, and m is the number of neurons in the hidden layer; Indicates the input layer node x i and hidden layer node y i The connection weights between them; v ji Represents the hidden layer node y i and output layer node z l The connection weights between them; Backpropagation of error signal: The network is adjusted based on the error function, modifying the connection weights and threshold parameters. The calculation formula is as follows: Where H is the number of samples, and the expected value of the l-th output node of the h-th sample is t. hl The actual output of the forward propagation through the neural network is z. hl .

6. The method for autonomous planning and positioning of station inspection paths according to claim 5, characterized in that, The pre-stored table of the relationship between defect level and data change difference should include at least the threshold change range, fault type, fault cause, upper limit of repair period, repair time and equipment ID; After obtaining the data change difference in step S4, it will first determine whether the data change difference is greater than a preset threshold. Then, the device IDs with data change differences greater than the preset threshold will be imported into the pre-stored defect level and data change difference relationship table to determine the fault type, fault cause, upper limit of repair period, and repair time.

7. The method for autonomous planning and positioning of station inspection paths according to claim 6, characterized in that, The detailed steps for establishing the objective function based on the shortest inspection and maintenance path in step S6 are as follows: Create a graph object with the currently required maintenance nodes and the paths between them to store the node and maintenance path data; Enter the maximum repair period and repair time as limiting conditions; The objective function formula for finding the shortest inspection and maintenance path is established using the Floyd algorithm.

8. The method for autonomous planning and positioning of station inspection paths according to claim 7, characterized in that, The Floyd-Warshall algorithm is used to establish the objective function of finding the shortest inspection and maintenance path. Through dynamic programming, intermediate nodes are gradually introduced to optimize the path. The core recursive formula is as follows: in, This represents the shortest path length from i to j when only the set of nodes {1,2,…,k} is allowed as intermediate nodes. k iterates from 1 to n, indicating that node k is gradually used as an intermediate node; When k=n This is the global shortest path length from i to j.

9. The method for autonomous planning and positioning of station inspection paths according to claim 6, characterized in that, In step S6, the objective function is established to minimize the inspection and maintenance time: Convert the path travel time between inspection nodes and the dwell time for maintenance at each node into the path length between nodes to create a graph object, which is used to store the path travel time between nodes, the dwell time for maintenance at each node, and the data of nodes, maintenance inspection nodes, and the dwell time for maintenance at each node. At the same time, the maximum repair period for the node should be entered as a limiting condition; The objective function formula for finding the shortest inspection and maintenance path is established using the Floyd algorithm.

10. A station-side inspection path autonomous planning and positioning generation system, characterized in that, It includes a multi-source data acquisition unit for substation operation faults, a data preprocessing unit, a substation operation fault data change prediction unit, a change data difference calculation unit, a fault query unit, a shortest objective function establishment unit for inspection and maintenance paths, a shortest objective function establishment unit for inspection and maintenance time, an optimal inspection and maintenance path calculation unit, an inspection station end acquisition point location data transmission unit, and a data storage unit. The substation operation fault multi-source data acquisition unit acquires multi-source data on substation operation faults through acquisition equipment at preset station-end acquisition points. The data preprocessing unit is connected to the substation operation fault multi-source data acquisition unit to preprocess the substation operation fault multi-source data and obtain the preprocessed substation operation fault related data and station end acquisition point location data. The substation operation fault data change prediction unit communicates with the data preprocessing unit, inputs the preprocessed substation operation fault related data into the substation operation fault prediction model, and outputs substation operation fault prediction data. The data change difference calculation unit is connected to the substation operation fault data change prediction unit. It calculates the difference between the substation operation fault prediction data and the substation operation fault related data input for prediction to obtain the data change difference. The fault query unit is connected to the change data difference calculation unit. Based on the data change difference, it queries the fault type, the upper limit of the maintenance period, the cause of the fault, and the maintenance time in the pre-stored defect level and data change difference relationship table. The unit for establishing the shortest inspection and maintenance path objective function is connected to the fault query unit. The station-end data collection point location data, fault cause, upper limit of maintenance period and maintenance time are taken into account in the objective function to establish the objective function with the shortest inspection and maintenance path. The objective function for minimizing inspection and maintenance time is established by establishing a communication connection between the inspection and maintenance unit and the fault query unit. The station-end data collection point location data, fault causes, maintenance deadline, and maintenance time are taken into account in the objective function to establish the objective function with the shortest inspection and maintenance time. The optimal inspection and maintenance path calculation unit is used to calculate the optimal inspection and maintenance path. The inspection station end data collection point location data sending unit is connected to the optimal inspection and maintenance path calculation unit, and is used to send the inspection station end data collection point location data required for maintenance to the maintenance personnel receiving end; The data storage unit is used to store relevant data in the table showing the relationship between defect levels and data change differences, as well as multi-source data on historical substation operation faults.

Images