Dual-mode HPLC (High Performance Liquid Chromatography) fusion terminal area topology intelligent identification system

Abstract

The invention discloses a dual-mode HPLC (High Performance Liquid Chromatography) fusion terminal area topology intelligent identification system, relates to the technical field of intelligent power grids, and solves the technical problems that a single index is depended, the relationship between nodes is difficult to understand from multiple angles, and meanwhile, the node data is difficult to group and manage through dual-mode HPLC communication, and the information of subordinate nodes is difficult to effectively integrate. In combination with analysis of power utilization characteristics and environment characteristics, the relationship between nodes can be understood from multiple angles instead of depending on a single index. The comprehensive method is helpful to better identify potential problems. The Pearson's correlation coefficient can quantify the correlation between the power consumption characteristics, provide a clear correlation degree, analyze the environmental characteristics between the nodes in combination with a DTW algorithm, capture a complex dynamic mode, and receive and process node data from different levels in real time through dual-mode HPLC communication.

Description

Technical Field

[0001] The invention belongs to the field of smart grids, and in particular is a dual-mode HPLC fusion terminal area topology intelligent recognition system. Background Art

[0002] With the rapid development of smart grid and Internet of Things (IoT) technologies, traditional distribution networks are facing growing electricity demand and complex load management challenges. In order to improve the reliability, flexibility and efficiency of the power supply system, various new communication and data processing technologies have emerged. Among them, dual-mode HPLC (High-Performance Liquid Chromatography) communication technology, as an efficient data transmission method, can achieve high-speed information transmission on existing power lines, providing the possibility for intelligent identification of substation topology. In modern distribution systems, information interaction between nodes is crucial. By real-time monitoring of the power consumption characteristics and environmental characteristics of each node and effectively integrating them, accurate data support can be provided to decision makers, thereby optimizing resource allocation, improving fault response capabilities and enhancing overall power supply security. Therefore, it is very necessary to establish a terminal substation topology intelligent identification system based on dual-mode HPLC communication.

[0003] Most of the existing intelligent identification systems for substation topology analyze the correlation between nodes through power consumption characteristics, relying on a single indicator, making it difficult to understand the relationship between nodes from multiple perspectives. At the same time, it is difficult to group and manage node data through dual-mode HPLC communication, and it is difficult to effectively integrate the information of subordinate nodes. Summary of the invention

[0004] The present invention aims to solve at least one of the technical problems existing in the prior art; to this end, the present invention proposes a dual-mode HPLC fusion terminal area topology intelligent identification system, which is used to solve the technical problems of relying on a single indicator and being difficult to understand the relationship between nodes from multiple angles. At the same time, it is difficult to group and manage node data through dual-mode HPLC communication and it is difficult to effectively integrate the information of subordinate nodes.

[0005] To solve the above problems, the first aspect of the present invention provides a dual-mode HPLC fusion terminal station area topology intelligent identification system, comprising:

[0006] Data collection module: collects the power consumption data and environmental data of nodes through smart meters and environmental sensors arranged at the nodes in the substation area, and extracts the power consumption characteristics and environmental characteristics of the nodes in the time series according to the time series factors;

[0007] Substation topology detection module: Based on the power consumption characteristics and environmental characteristics of the node time series, the Pearson correlation coefficient correlation analysis is used to perform power consumption characteristics correlation analysis between nodes. The DTW dynamic time warping algorithm of the environmental characteristics between nodes is used to perform environmental characteristics correlation analysis between nodes. Based on the power consumption characteristics correlation analysis and environmental characteristics correlation analysis, the substation topology is determined.

[0008] Node classification module: According to the level of nodes in the area topology, nodes at different levels are grouped by analyzing the correlation between nodes at different levels, and the node data is stored in the storage mechanism of high-level nodes through the location data in the node environment characteristics;

[0009] Dual-mode HPLC communication module: The high-level nodes in the station area package and store the data received by the nodes of different levels. The high-level nodes communicate through dual-mode HPLC to transmit data between nodes.

[0010] As a further solution of the present invention: also include:

[0011] Node fault analysis module: Based on the grouping results of nodes at different levels, a fault diagnosis model is established to perform sub-diagnosis on the faults of nodes in the substation topology.

[0012] As a further solution of the present invention: the data acquisition module collects the power consumption data and environmental data of the node through the smart meter and environmental sensor arranged at the node of the substation, and extracts the power consumption characteristics and environmental characteristics of the node in the time series according to the time series factor, including the following steps:

[0013] Smart meters and environmental sensors deployed at the nodes in the substation area collect the node's power consumption data and environmental data. The power consumption data includes: power, current, voltage, and power consumption time. The environmental data includes: ambient temperature, ambient humidity, and node location data.

[0014] Add data collection timestamp to power consumption data and set it as the index of corresponding data to obtain power time series graph and voltage time series graph. According to the obtained fluctuation graph, obtain the daily average load, peak-to-valley difference and load factor of each node. According to the daily average load data of the node, count the power time series data of the node whose daily average load exceeds the threshold range for consecutive preset days, and use the daily average load, peak-to-valley difference, load factor and power time series data of the node as the power consumption characteristics of the node.

[0015] According to the average daily load of each node, the environmental data of the node whose average daily load exceeds the threshold range for a preset number of consecutive days is obtained, and according to the corresponding environmental data, the environmental temperature time series data and the environmental humidity time series data are obtained;

[0016] The area is divided into grids, and based on the node location data, the distances from the nodes in the area to the central nodes of each substation are counted, and the nodes in the area are divided into the substations corresponding to the nearest central nodes.

[0017] As a further solution of the present invention: the area topology structure detection module performs a correlation analysis of power consumption characteristics between various nodes through a Pearson correlation coefficient correlation analysis, including the following steps:

[0018] The daily average load, peak-to-valley difference, load factor and power time series data of the node are combined into power consumption characteristic array data;

[0019] The calculation formula of Pearson's correlation coefficient is: , where xi and yi are the values ​​of the i-th corresponding data in the electrical characteristic arrays of the two nodes, and x and are the average values ​​of the corresponding data in the electrical characteristic arrays of the two nodes in the samples;

[0020] The correlation coefficient of all electrical characteristic array data between any two nodes in each grid is calculated, and the mean value of the correlation coefficient of all electrical characteristic array data is used as the power consumption correlation coefficient between any two nodes in the grid.

[0021] As a further solution of the present invention: using the DTW dynamic time warping algorithm of the environmental features between the nodes to perform correlation analysis of the environmental features between the nodes, the following steps are included:

[0022] According to the ambient temperature time series data and the ambient humidity time series data, the time series is divided into several detection time intervals, and the corresponding time series graph is drawn;

[0023] Let d(a j , b j ) represents the distance between the time series data of node a in the detection time interval j and the time series data of node b in the detection time interval j. The value of d(a, b) is the difference between the mean values ​​of the time series data of node a and node b in the detection time interval j. The DTW dynamic time warping equation is set as:

[0024] dp(a j , b j ) = min{d(a j-1 , b j ), d(a j-1 , b j-1 ), d(a j , b j-1 )} + d(a j , b j )

[0025] Among them, dp(a j , b j ) is the DTW distance between node a and node b in the detection time interval j;

[0026] The sum of the DTW distances of the ambient temperature time series data and the ambient humidity time series data between the nodes in the detection time interval j is calculated, and the environmental feature correlation coefficient between the nodes is calculated by the following formula:

[0027]

[0028] Among them, u is the environmental feature correlation coefficient between nodes, Sdtw is the sum of the DTW distances of the ambient temperature time series data and the ambient humidity time series data between nodes in the detection time interval j, Tj is the sum of the mean values ​​of the ambient temperature time series data between nodes in the detection time interval j, and Rj is the sum of the mean values ​​of the ambient humidity time series data between nodes in the detection time interval j;

[0029] By calculating the environmental feature correlation coefficient between any two nodes in each grid, the environmental feature correlation analysis between each node is performed.

[0030] As a further solution of the present invention: determining the topological structure of the substation area according to the correlation analysis of power consumption characteristics and environmental characteristics includes the following steps:

[0031] According to the results of power consumption feature correlation analysis and environmental feature correlation analysis, the correlation between nodes is calculated using the following formula:

[0032]

[0033] Among them, α is the correlation analysis value between nodes, g is the power consumption correlation coefficient between any two nodes in the grid, and e is the Euler number;

[0034] Nodes whose correlation analysis values ​​between nodes in the grid are greater than a preset threshold are divided into the same group, nodes between different grid node groups in the station area are randomly selected for correlation analysis, and node groups whose correlation analysis values ​​are greater than a preset threshold are merged into the same group;

[0035] Calculate the average value of the daily average load within a continuous preset time for different groups, set the nodes in the group whose average daily average load is greater than the first preset threshold as high-level nodes, set the nodes in the group whose average daily average load is less than the second preset threshold as low-level nodes, set the nodes in the remaining groups in the substation as middle-level nodes, and determine the substation topology structure.

[0036] As a further solution of the present invention: the node classification module groups nodes at different levels according to the levels of the nodes in the station area topology by analyzing the correlation between nodes at different levels, and stores the node data in the storage mechanism of the high-level node through the location data in the node environment characteristics, including the following steps:

[0037] According to the level of the nodes in the area topology, the correlation analysis values ​​between the low-level nodes and the middle-level nodes are calculated respectively, and the nodes whose correlation analysis values ​​between the nodes are less than the threshold are divided into the same group, so as to group the low-level nodes and the middle-level nodes;

[0038] Obtain the node location data in the node environment characteristics, store the data of the low-level nodes and the middle-level nodes in the storage mechanism of the nearest high-level node, and when the node data stored in the storage mechanism of the high-level node is greater than the threshold, transfer the node data of 10% of the nodes stored therein to the storage mechanism of the nearest high-level node for storage.

[0039] As a further solution of the present invention: the dual-mode HPLC communication module packages and stores the data of different-level node groups received by the high-level node in the station area, and the high-level node transmits data between nodes through dual-mode HPLC communication, including the following steps:

[0040] High-level nodes use HPLC high-speed power line carrier communication and HRF high-speed wireless communication technology to perform data transmission between nodes through mutual backup of HPLC and HRF channels;

[0041] The high-level nodes in the station area package and store the data received from the node groups in different levels, number the nodes in the node group, use the node number and the node position as the label of the node data, and add a label to the data in each group data packet.

[0042] As a further solution of the present invention: the node fault analysis module establishes a fault diagnosis model for the grouping results of nodes at different levels and performs sub-diagnosis on the faults of nodes in the substation topology, including the following steps:

[0043] Obtain historical data of power consumption characteristic data of nodes in groups of nodes at different levels, and add normal and fault labels to the historical data according to whether the historical data is normal or faulty;

[0044] The power consumption characteristic data of nodes in node groups at different levels are used to form a data set, and support vector machines for node groups at different levels are trained respectively. After training, fault diagnosis models for each node group are established for support vector machines for node groups at different levels, and faults of nodes in node groups in the substation topology are diagnosed separately.

[0045] Compared with the prior art, the present invention has the following beneficial effects:

[0046] The present invention combines the analysis of power consumption characteristics and environmental characteristics to facilitate understanding the relationship between nodes from multiple perspectives, rather than relying solely on a single indicator. This comprehensive approach helps to better identify potential problems. The Pearson correlation coefficient can quantify the correlation between power consumption characteristics and provide a clear degree of correlation so that the similarities between different nodes can be clearly represented. Combined with the DTW algorithm, the nonlinear and time-varying data of the environmental characteristics between nodes can be effectively analyzed, thereby capturing complex dynamic patterns.

[0047] The present invention groups nodes of different levels to facilitate clearer management and maintenance of data in the substation area; high-level nodes serve as data aggregation centers to effectively integrate the information of subordinate nodes and simplify the data processing process; the data of multiple low-level nodes are concentrated in high-level nodes, which helps to reduce the network burden and improve data transmission efficiency. High-level nodes can obtain data from subordinate nodes regularly or on demand, thereby optimizing bandwidth usage. Through dual-mode HPLC communication, high-level nodes can receive and process node data from different levels in real time, thereby realizing real-time monitoring of the substation status and taking timely measures to deal with abnormal situations. The data information from different low levels is centrally stored in high-level nodes, which can increase the redundancy of the system. If a low-level node fails, its data can still be obtained from other similar types of devices, thereby improving the reliability of the overall system. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] 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 work.

[0049] Figure 1 Schematic diagram of the system framework of the present invention. DETAILED DESCRIPTION

[0050] The technical solution of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than 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.

[0051] See also Figure 1 The first embodiment of the present invention provides a dual-mode HPLC fusion terminal area topology intelligent identification system, comprising:

[0052] Data collection module: collects the power consumption data and environmental data of nodes through smart meters and environmental sensors arranged at the nodes in the substation area, and extracts the power consumption characteristics and environmental characteristics of the nodes in the time series according to the time series factors;

[0053] Substation topology detection module: Based on the power consumption characteristics and environmental characteristics of the node time series, the Pearson correlation coefficient correlation analysis is used to perform power consumption characteristics correlation analysis between nodes. The DTW dynamic time warping algorithm of the environmental characteristics between nodes is used to perform environmental characteristics correlation analysis between nodes. Based on the power consumption characteristics correlation analysis and environmental characteristics correlation analysis, the substation topology is determined.

[0054] Node classification module: According to the level of nodes in the area topology, nodes at different levels are grouped by analyzing the correlation between nodes at different levels, and the node data is stored in the storage mechanism of high-level nodes through the location data in the node environment characteristics;

[0055] Dual-mode HPLC communication module: The high-level nodes in the station area package and store the data received by the nodes of different levels. The high-level nodes communicate through dual-mode HPLC to transmit data between nodes.

[0056] Specifically, in this embodiment, the smart meters and environmental sensors arranged at the substation nodes collect the power consumption data and environmental data of the nodes. The data may include the following fields: power consumption data: power, current, voltage, and power consumption time (timestamp).

[0057] Environmental data: ambient temperature (temperature), ambient humidity (humidity) and node location (such as city name or coordinates);

[0058] Ensure that the timestamp columns of the power consumption data and environmental data are in date and time format and are set as indexes. Based on the time series factors, extract the power consumption characteristics and environmental characteristics of the node's time series.

[0059] Through the area topology detection module, according to the power consumption characteristics and environmental characteristics of the node time series, through the Pearson correlation coefficient correlation analysis, the Pearson correlation coefficient matrix between each node can be calculated using the .corr() function in Pandas. Here, we need to calculate each feature data in the power consumption characteristics separately, and perform the power consumption characteristics correlation analysis between each node. Through the DTW dynamic time warping algorithm of the environmental characteristics between each node, fastdtw or other libraries can be used to calculate the DTW distance. Perform the environmental characteristics correlation analysis between each node, and determine the area topology based on the power consumption characteristics correlation analysis and the environmental characteristics correlation analysis.

[0060] Combining the analysis of power consumption characteristics and environmental characteristics makes it easier to understand the relationship between nodes from multiple perspectives, rather than relying on just a single indicator. This comprehensive approach helps better identify potential problems.

[0061] The Pearson correlation coefficient can quantify the correlation between power consumption characteristics and provide a clear correlation degree, so that the similarity between different nodes can be clearly represented. Combined with the DTW algorithm, it can effectively analyze the nonlinear and time-varying data of the environmental characteristics between nodes, thereby capturing complex dynamic patterns.

[0062] The node classification module is used to group nodes at different levels according to their levels in the station area topology by analyzing the correlation between nodes at different levels. The node data is stored in the storage mechanism of the high-level node through the location data in the node environmental characteristics. The dual-mode HPLC communication module packages and stores the data of different levels of node groups received by the high-level nodes in the station area. The high-level nodes carry out data transmission between nodes through dual-mode HPLC communication.

[0063] Grouping nodes at different levels facilitates clearer management and maintenance of data within the area. High-level nodes serve as data aggregation centers, effectively integrating information from subordinate nodes and simplifying data processing procedures.

[0064] Centralizing the data of multiple low-level nodes into high-level nodes helps reduce network burden and improve data transmission efficiency. High-level nodes can obtain data from subordinate nodes regularly or on demand, thereby optimizing bandwidth usage.

[0065] Through dual-mode HPLC communication, high-level nodes can receive and process node data from different levels in real time, thereby achieving real-time monitoring of the status of the station area and taking timely measures to deal with abnormal situations. Centralized storage of data information from different low-level nodes in high-level nodes can increase the redundancy of the system. If a low-level node fails, its data can still be obtained from other similar types of devices, thereby improving the reliability of the overall system.

[0066] In one embodiment of the present invention, it also includes:

[0067] Node fault analysis module: Based on the grouping results of nodes at different levels, a fault diagnosis model is established to perform sub-diagnosis on the faults of nodes in the substation topology.

[0068] In one embodiment of the present invention, the data acquisition module collects the power consumption data and environmental data of the node through the smart meter and environmental sensor arranged at the node of the substation, and extracts the power consumption characteristics and environmental characteristics of the node in the time series according to the time series factor, including the following steps:

[0069] Smart meters and environmental sensors deployed at the nodes in the substation area collect the node's power consumption data and environmental data. The power consumption data includes: power, current, voltage, and power consumption time. The environmental data includes: ambient temperature, ambient humidity, and node location data.

[0070] Add data collection timestamp to power consumption data and set it as the index of corresponding data to obtain power time series graph and voltage time series graph. According to the obtained fluctuation graph, obtain the daily average load, peak-to-valley difference and load factor of each node. According to the daily average load data of the node, count the power time series data of the node whose daily average load exceeds the threshold range for consecutive preset days, and use the daily average load, peak-to-valley difference, load factor and power time series data of the node as the power consumption characteristics of the node.

[0071] According to the average daily load of each node, the environmental data of the node whose average daily load exceeds the threshold range for a preset number of consecutive days is obtained, and according to the corresponding environmental data, the environmental temperature time series data and the environmental humidity time series data are obtained;

[0072] The area is divided into grids, and based on the node location data, the distances from the nodes in the area to the central nodes of each substation are counted, and the nodes in the area are divided into the substations corresponding to the nearest central nodes.

[0073] Specifically, in this embodiment, the daily average power is calculated as the daily average load of the node;

[0074] The difference between the daily peak and valley values ​​is calculated as the peak-to-valley difference;

[0075] The ratio of the average daily load to the daily peak is calculated as the load factor.

[0076] According to the daily average load of each node, the environmental data of the nodes whose daily average load exceeds the threshold range for seven consecutive days are obtained, and according to the corresponding environmental data, the environmental temperature time series data and the environmental humidity time series data are obtained. The upper limit of the threshold range of the daily average load is set to 1.5 times the daily average load of the node in the past year, and the upper limit of the threshold range of the daily average load is set to 0.6 times the daily average load of the node in the past year;

[0077] The area is divided into grids, and based on the node location data, the distances from the nodes in the area to the central nodes of each substation are counted, and the nodes in the area are divided into the substations corresponding to the nearest central nodes.

[0078] In one embodiment of the present invention, the area topology detection module performs a correlation analysis of power consumption characteristics between nodes through a Pearson correlation coefficient correlation analysis, including the following steps:

[0079] The daily average load, peak-to-valley difference, load factor and power time series data of the node are combined into power consumption characteristic array data;

[0080] The calculation formula of Pearson's correlation coefficient is: , where xi and yi are the values ​​of the i-th corresponding data in the electrical characteristic arrays of the two nodes, and x and are the average values ​​of the corresponding data in the electrical characteristic arrays of the two nodes in the samples;

[0081] The correlation coefficient of all electrical characteristic array data between any two nodes in each grid is calculated, and the mean value of the correlation coefficient of all electrical characteristic array data is used as the power consumption correlation coefficient between any two nodes in the grid.

[0082] In one embodiment of the present invention, the correlation analysis of the environmental features between the nodes is performed by using the DTW dynamic time warping algorithm of the environmental features between the nodes, including the following steps:

[0083] According to the ambient temperature time series data and the ambient humidity time series data, the time series is divided into several detection time intervals, and the corresponding time series graph is drawn;

[0084] Let d(a j , b j ) represents the distance between the time series data of node a in the detection time interval j and the time series data of node b in the detection time interval j. The value of d(a, b) is the difference between the mean values ​​of the time series data of node a and node b in the detection time interval j. The DTW dynamic time warping equation is set as:

[0085] dp(a j , b j ) = min{d(a j-1 , b j ), d(a j-1 , b j-1 ), d(a j , b j-1 )} + d(a j , b j )

[0086] Among them, dp(a j , b j ) is the DTW distance between node a and node b in the detection time interval j;

[0087] The sum of the DTW distances of the ambient temperature time series data and the ambient humidity time series data between the nodes in the detection time interval j is calculated, and the environmental feature correlation coefficient between the nodes is calculated by the following formula:

[0088]

[0089] Among them, u is the environmental feature correlation coefficient between nodes, Sdtw is the sum of the DTW distances of the ambient temperature time series data and the ambient humidity time series data between nodes in the detection time interval j, Tj is the sum of the mean values ​​of the ambient temperature time series data between nodes in the detection time interval j, and Rj is the sum of the mean values ​​of the ambient humidity time series data between nodes in the detection time interval j;

[0090] By calculating the environmental feature correlation coefficient between any two nodes in each grid, the environmental feature correlation analysis between each node is performed.

[0091] In one embodiment of the present invention, determining the topology of the substation area according to the correlation analysis of power consumption characteristics and the correlation analysis of environmental characteristics includes the following steps:

[0092] According to the results of power consumption feature correlation analysis and environmental feature correlation analysis, the correlation between nodes is calculated using the following formula:

[0093]

[0094] Among them, α is the correlation analysis value between nodes, g is the power consumption correlation coefficient between any two nodes in the grid, and e is the Euler number;

[0095] Nodes whose correlation analysis values ​​between nodes in the grid are greater than a preset threshold are divided into the same group, nodes between different grid node groups in the station area are randomly selected for correlation analysis, and node groups whose correlation analysis values ​​are greater than a preset threshold are merged into the same group;

[0096] Calculate the average value of the daily average load within a continuous preset time for different groups, set the nodes in the group whose average daily average load is greater than the first preset threshold as high-level nodes, set the nodes in the group whose average daily average load is less than the second preset threshold as low-level nodes, set the nodes in the remaining groups in the substation as middle-level nodes, and determine the substation topology structure.

[0097] Specifically, in this embodiment, nodes whose correlation analysis values ​​between nodes in the grid are greater than a preset threshold are divided into the same group, nodes between different grid node groups in the station area are randomly selected for correlation analysis, and node groups whose correlation analysis values ​​are greater than a preset threshold are merged into the same group, and the preset threshold of the correlation analysis value is set to 0.6;

[0098] Calculate the average value of the daily average load within a continuous preset time of different groups, set the nodes in the group whose average daily average load is greater than the first preset threshold as high-level nodes, set the nodes in the group whose average daily average load is less than the second preset threshold as low-level nodes, set the nodes in the remaining groups in the substation as middle-level nodes, and determine the substation topology; the first preset threshold is set to 80 kWh, and the second preset threshold is set to 600kWh.

[0099] In one embodiment of the present invention, the node classification module groups nodes at different levels according to the levels of the nodes in the station area topology by analyzing the correlation between nodes at different levels, and stores the node data in the storage mechanism of the high-level node through the location data in the node environment characteristics, including the following steps:

[0100] According to the level of the node in the station area topology, the correlation analysis values ​​between each node in the low-level nodes and the middle-level nodes are calculated respectively, and the low-level nodes and the middle-level nodes are grouped by dividing the nodes whose correlation analysis values ​​between the nodes are less than the threshold into the same group; in this embodiment, the preset threshold of the correlation analysis value is set to 0.6.

[0101] Obtain the node location data in the node environment characteristics, store the data of the low-level nodes and the middle-level nodes in the storage mechanism of the nearest high-level node, and when the node data stored in the storage mechanism of the high-level node is greater than the threshold, transfer the node data of 10% of the nodes stored therein to the storage mechanism of the nearest high-level node for storage.

[0102] In one embodiment of the present invention, the dual-mode HPLC communication module packages and stores the data received by the high-level node in the station area from the different-level node groups, and the high-level node transmits data between nodes through the dual-mode HPLC communication, including the following steps:

[0103] High-level nodes use HPLC high-speed power line carrier communication and HRF high-speed wireless communication technology to perform data transmission between nodes through mutual backup of HPLC and HRF channels;

[0104] The high-level nodes in the station area package and store the data received from the node groups in different levels, number the nodes in the node group, use the node number and the node position as the label of the node data, and add a label to the data in each group data packet.

[0105] In one embodiment of the present invention, the node fault analysis module establishes a fault diagnosis model for the grouping results of nodes at different levels and performs sub-diagnosis on the faults of nodes in the substation topology, including the following steps:

[0106] Obtain historical data of power consumption characteristic data of nodes in groups of nodes at different levels, and add normal and fault labels to the historical data according to whether the historical data is normal or faulty;

[0107] The power consumption characteristic data of nodes in node groups at different levels are used to form a data set, and support vector machines for node groups at different levels are trained respectively. After training, fault diagnosis models for each node group are established for support vector machines for node groups at different levels, and faults of nodes in node groups in the substation topology are diagnosed separately.

[0108] The above embodiments are only used to illustrate the technical method of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical method of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical method of the present invention.

Claims

1. A dual-mode HPLC fusion terminal area topology intelligent identification system, characterized in that: include: Data collection module: collects the power consumption data and environmental data of nodes through smart meters and environmental sensors arranged at the nodes in the substation area, and extracts the power consumption characteristics and environmental characteristics of the nodes in the time series according to the time series factors; Substation topology detection module: Based on the power consumption characteristics and environmental characteristics of the node time series, the Pearson correlation coefficient correlation analysis is used to perform power consumption characteristics correlation analysis between nodes. The DTW dynamic time warping algorithm of the environmental characteristics between nodes is used to perform environmental characteristics correlation analysis between nodes. Based on the power consumption characteristics correlation analysis and environmental characteristics correlation analysis, the substation topology is determined. Node classification module: According to the level of nodes in the area topology, nodes at different levels are grouped by analyzing the correlation between nodes at different levels, and the node data is stored in the storage mechanism of high-level nodes through the location data in the node environment characteristics; Dual-mode HPLC communication module: The high-level nodes in the station area package and store the data received by the nodes of different levels. The high-level nodes communicate through dual-mode HPLC to transmit data between nodes.

2. A dual-mode HPLC fusion terminal area topology intelligent identification system according to claim 1, characterized in that: Also includes: Node fault analysis module: Based on the grouping results of nodes at different levels, a fault diagnosis model is established to perform sub-diagnosis of node faults in the substation topology.

3. A dual-mode HPLC fusion terminal area topology intelligent identification system according to claim 1, characterized in that: The data acquisition module collects the power consumption data and environmental data of the nodes through the smart meters and environmental sensors arranged at the nodes in the substation area, and extracts the power consumption characteristics and environmental characteristics of the nodes in the time series according to the time series factors, including the following steps: Smart meters and environmental sensors deployed at the nodes in the substation area collect the node's power consumption data and environmental data. The power consumption data includes: power, current, voltage, and power consumption time. The environmental data includes: ambient temperature, ambient humidity, and node location data. Add data collection timestamp to power consumption data and set it as the index of corresponding data to obtain power time series graph and voltage time series graph. According to the obtained fluctuation graph, obtain the daily average load, peak-to-valley difference and load factor of each node. According to the daily average load data of the node, count the power time series data of the node whose daily average load exceeds the threshold range for consecutive preset days, and use the daily average load, peak-to-valley difference, load factor and power time series data of the node as the power consumption characteristics of the node. According to the average daily load of each node, the environmental data of the node whose average daily load exceeds the threshold range for a preset number of consecutive days is obtained, and according to the corresponding environmental data, the environmental temperature time series data and the environmental humidity time series data are obtained; The area is divided into grids, and based on the node location data, the distances from the nodes in the area to the central nodes of each substation are counted, and the nodes in the area are divided into the substations corresponding to the nearest central nodes.

4. A dual-mode HPLC fusion terminal area topology intelligent identification system according to claim 3, characterized in that: The area topology detection module performs a correlation analysis of the power consumption characteristics between each node through the Pearson correlation coefficient correlation analysis, including the following steps: The daily average load, peak-to-valley difference, load factor and power time series data of the node are combined into power consumption characteristic array data; The calculation formula of Pearson's correlation coefficient is: , where xi and yi are the values ​​of the i-th corresponding data in the electrical characteristic arrays of the two nodes, and x and are the average values ​​of the corresponding data in the electrical characteristic arrays of the two nodes in the samples; The correlation coefficient of all electrical characteristic array data between any two nodes in each grid is calculated, and the mean value of the correlation coefficient of all electrical characteristic array data is used as the power consumption correlation coefficient between any two nodes in the grid.

5. A dual-mode HPLC fusion terminal area topology intelligent identification system according to claim 3, characterized in that: The DTW dynamic time warping algorithm of the environmental features between the nodes is used to perform correlation analysis of the environmental features between the nodes, including the following steps: According to the ambient temperature time series data and the ambient humidity time series data, the time series is divided into several detection time intervals, and the corresponding time series graph is drawn; Let d(a j , b j ) represents the distance between the time series data of node a in the detection time interval j and the time series data of node b in the detection time interval j. The value of d(a, b) is the difference between the mean values ​​of the time series data of node a and node b in the detection time interval j. The DTW dynamic time warping equation is set as: dp(a j , b j ) = min{d(a j-1 , b j ), d(a j-1 , b j-1 ), d(a j , b j-1 )} + d(a j , b j ); Among them, dp(a j , b j ) is the DTW distance between node a and node b in the detection time interval j; The sum of the DTW distances of the ambient temperature time series data and the ambient humidity time series data between the nodes in the detection time interval j is calculated, and the environmental feature correlation coefficient between the nodes is calculated by the following formula: ； Among them, u is the environmental feature correlation coefficient between nodes, Sdtw is the sum of the DTW distances of the ambient temperature time series data and the ambient humidity time series data between nodes in the detection time interval j, Tj is the sum of the mean values ​​of the ambient temperature time series data between nodes in the detection time interval j, and Rj is the sum of the mean values ​​of the ambient humidity time series data between nodes in the detection time interval j; By calculating the environmental feature correlation coefficient between any two nodes in each grid, the environmental feature correlation analysis between each node is performed.

6. A dual-mode HPLC fusion terminal area topology intelligent identification system according to claim 5, characterized in that: According to the correlation analysis of power consumption characteristics and environmental characteristics, the topology of the substation area is determined, including the following steps: According to the results of power consumption feature correlation analysis and environmental feature correlation analysis, the correlation between nodes is calculated using the following formula: ； Among them, α is the correlation analysis value between nodes, g is the power consumption correlation coefficient between any two nodes in the grid, and e is the Euler number; Nodes whose correlation analysis values ​​between nodes in the grid are greater than a preset threshold are divided into the same group, nodes between different grid node groups in the station area are randomly selected for correlation analysis, and node groups whose correlation analysis values ​​are greater than a preset threshold are merged into the same group; Calculate the average value of the daily average load within a continuous preset time for different groups, set the nodes in the group whose average daily average load is greater than the first preset threshold as high-level nodes, set the nodes in the group whose average daily average load is less than the second preset threshold as low-level nodes, set the nodes in the remaining groups in the substation as middle-level nodes, and determine the substation topology structure.

7. A dual-mode HPLC fusion terminal area topology intelligent identification system according to claim 1, characterized in that: The node classification module: according to the level of the node in the station area topology, by analyzing the correlation between nodes at different levels, the nodes at different levels are grouped, and the node data is stored in the storage mechanism of the high-level node through the location data in the node environment characteristics, including the following steps: According to the level of the nodes in the area topology, the correlation analysis values ​​between the low-level nodes and the middle-level nodes are calculated respectively, and the nodes whose correlation analysis values ​​between the nodes are less than the threshold are divided into the same group, so as to group the low-level nodes and the middle-level nodes; Obtain the node location data in the node environment characteristics, store the data of the low-level nodes and the middle-level nodes in the storage mechanism of the nearest high-level node, and when the node data stored in the storage mechanism of the high-level node is greater than the threshold, transfer the node data of 10% of the nodes stored therein to the storage mechanism of the nearest high-level node for storage.

8. A dual-mode HPLC fusion terminal area topology intelligent identification system according to claim 1, characterized in that: The dual-mode HPLC communication module packages and stores the data received by the high-level nodes in the station area from the different-level node groups, and the high-level nodes communicate through the dual-mode HPLC to transmit data between nodes, including the following steps: High-level nodes use HPLC high-speed power line carrier communication and HRF high-speed wireless communication technology to perform data transmission between nodes through mutual backup of HPLC and HRF channels; The high-level nodes in the station area package and store the data received from the node groups in different levels, number the nodes in the node group, use the node number and the node position as the label of the node data, and add a label to the data in each group data packet.

9. A dual-mode HPLC fusion terminal area topology intelligent identification system according to claim 2, characterized in that: The node fault analysis module: according to the grouping results of nodes at different levels, according to the grouping of nodes at different levels, a fault diagnosis model is established to perform sub-diagnosis on the faults of nodes in the substation topology, including the following steps: Obtain historical data of power consumption characteristic data of nodes in groups of nodes at different levels, and add normal and fault labels to the historical data according to whether the historical data is normal or faulty; The power consumption characteristic data of nodes in node groups at different levels are used to form a data set, and support vector machines for node groups at different levels are trained respectively. After training, fault diagnosis models for each node group are established for support vector machines for node groups at different levels, and faults of nodes in node groups in the substation topology are diagnosed separately.

Images