Method of system for intelligent linkage of substation equipment

Abstract

The invention provides a method of a substation equipment intelligent linkage system. The system is provided with a data acquisition module, a communication module, a processing and analysis module, a linkage module and a man-machine interaction module. The implementation method of the system comprises the following steps: step 1, installing a sensor in a substation site and accessing an acquisition module; 2, the communication module transmits the collected data to a processing and analyzing module, and the collected original data is processed; step 3, based on the topological structure and historical operation data of the substation equipment, performing data analysis through a processing and analysis module; 4, linkage is carried out on a result obtained through analysis and a linkage module; 5, a manager carries out man-machine interaction control and checking in a man-machine interaction module; according to the system and method, a deep learning algorithm is introduced to realize dynamic intelligent decision; equipment implicit association is excavated, a closed-loop control system is constructed, a real-time feedback optimization strategy is combined, and the adaptability and reliability of the system are enhanced.

Description

Technical Field

[0001] The present invention relates to the technical field of power system automation, and specifically to a system method for intelligent linkage of substation equipment. Background Art

[0002] In the existing substation operation process, there are many deficiencies in the data interaction and linkage control between equipment. On the one hand, traditional linkage methods are mostly based on pre-set fixed rules, which are difficult to cope with complex and changeable operating conditions. When the equipment operating environment changes, such as sudden fluctuations in grid load and abnormal external climatic conditions, the linkage strategy under the fixed rules cannot make adaptive adjustments in time, resulting in low equipment operation efficiency and may even cause failures. On the other hand, the data collected by each equipment point often lacks in-depth analysis and effective integration, and cannot fully explore the potential correlation between the equipment behind the data, resulting in poor coordination between equipment and inability to achieve optimal resource allocation. For this reason, the applicant proposes a system method for intelligent linkage of substation equipment according to demand. Through intelligent analysis and processing of real-time data of equipment points, accurate and efficient linkage between equipment is achieved, the safety and reliability of substation operation are improved, and the allocation of power resources is optimized. Summary of the Invention

[0003] In order to solve the above technical problems, the present invention proposes a system method for intelligent linkage of substation equipment. The system is composed of a data acquisition module, a communication module, a processing and analysis module, a linkage module and a human-computer interaction module. The system method installs sensors in the substation and connects to the acquisition module. The communication module transmits the collected data to the processing and analysis module, processes the collected raw data, and the processing and analysis module performs data analysis based on the topological structure and historical operation data of the substation equipment, and then links the analysis results with the linkage module. Finally, the management personnel perform human-computer interaction control and viewing in the human-computer interaction module. The system introduces deep learning algorithms to realize dynamic intelligent decision-making; explores the implicit associations of equipment, builds a closed-loop control system, and combines real-time feedback optimization strategies to enhance system adaptability and reliability.

[0004] To achieve the above object, the technical solution adopted by the present invention is:

[0005] A method for a system for intelligent linkage of substation equipment, characterized in that: the system for intelligent linkage of substation equipment is provided with a data acquisition module, a communication module, a processing and analysis module, a linkage module and a human-computer interaction module; the steps for implementing the system are as follows:

[0006] Step 1: Install sensors in the substation and connect them to the data collection module. The data collection module packages the collected data at a certain time interval and transmits it to the substation data processing center through the communication module connected to the industrial Ethernet to ensure the real-time and accuracy of the data.

[0007] Step 2: The communication module transmits the collected data to the processing and analysis module. The processing and analysis module uses a data program written in C++ to process the collected raw data. Within a set reasonable threshold range, it removes noise data and outliers that are obviously inconsistent with the actual situation. It also uses a convolutional neural network to extract features from the normalized data. The data is input into a network structure consisting of multiple convolutional layers and pooling layers. The network automatically learns the characteristic patterns in the data and extracts key feature vectors that can reflect the operating status of the equipment.

[0008] Step 3: Based on the topological structure and historical operating data of the substation equipment, the processing and analysis module performs data analysis. The processing and analysis module uses a graph neural network to analyze the correlation between the equipment point data and determine the influence relationship and linkage priority between each device.

[0009] Step 4: Link the analysis results with the linkage module; pre-configure the linkage module configuration information in the telemetry point configuration of the intelligent linkage service, and associate the results obtained by the processing and analysis modules with the device measurement points that need to be controlled through the linkage module; the C++ data service sends the collected real-time data to emq, and the intelligent auxiliary platform application service listens to the emq real-time data topic message, performs business processing on the received message, retains only the real-time data of the telemetry type measurement point, and calls the intelligent auxiliary measurement point remote control method to control the equipment and perform related storage operations;

[0010] Step 5. The manager performs human-machine interaction control and viewing in the human-machine interaction module. The manager performs manual control of relevant equipment in the human-machine interaction module according to needs and views relevant operating data.

[0011] Furthermore, the convolutional neural network algorithm is used in step 2 of the method for implementing the intelligent linkage system of substation equipment. The convolutional neural network algorithm uses deep learning and a unique structural design to perform system image recognition, target detection, semantic segmentation, and speech recognition. It simulates the image processing mechanism of the human visual system and gradually converts raw data into abstract high-level features through layer-by-layer feature extraction to complete data analysis and processing tasks.

[0012] Furthermore, in step 3 of the method for implementing the intelligent linkage system of substation equipment, a graph neural network algorithm is used, including a graph convolutional network and a graph attention network, specifically:

[0013] The calculation formula of the graph convolutional network is:

[0014]

[0015] in: For the The node feature matrix of the layer;

[0016] is the normalized adjacency matrix

[0017] A is the original adjacency matrix;

[0018] D is a diagonal matrix, and the diagonal elements are the degrees of the nodes;

[0019] For the The learnable parameter matrix of the layer;

[0020] Σ is the activation function;

[0021] Graph attention network, the formula for calculating the attention coefficient αij between node i and neighbor node j is:

[0022]

[0023] Where: hi and hj are the features of node i and neighbor node j respectively;

[0024] W is the learnable weight matrix;

[0025] a is the attention vector;

[0026] Ni is the set of neighbor nodes of node i;

[0027] || indicates a splicing operation;

[0028] LeakyReLU is the activation function.

[0029] 1. A method for implementing a system for intelligent linkage of substation equipment according to claim 1, characterized in that: the specific process of step 4 of the method for implementing the system for intelligent linkage of substation equipment is as follows: in the intelligent linkage service, the real-time data measurement point type is filtered to retain only the real-time data of the telemetry points; the real-time data threshold is filtered through the linkage configuration information configured by the platform; and the frequent reporting characteristics of the telemetry points are fully considered to perform secondary processing on the data, which specifically includes the following three points:

[0030] 1) For the value reported by the measuring point, if it has not changed, filter out this data;

[0031] 2) Filter the numerical range reported by the measuring points, and only perform linkage processing when the value exceeds the range;

[0032] 3) Increase the time interval of linkage control. If the control is repeated within a certain period of time, no operation will be performed.

[0033] The benefits of this application are:

[0034] 1. The system and method of intelligent linkage of substation equipment can improve the level and effectiveness of engineering safety supervision, reduce the incidence of safety accidents, and ensure the safety of equipment in the substation;

[0035] 2. The system and method for intelligent linkage of substation equipment can make corresponding control strategies according to the on-site environment, reducing costs within the substation while ensuring the safety of some places within the substation;

[0036] 3. The system and method for intelligent linkage of substation equipment are easy to access, can make full use of the original installed equipment and basic systems, and are low in cost. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 Schematic diagram of the method flow of the system of the present invention;

[0038] Figure 2 Schematic diagram of the intelligent linkage module of the intelligent assistance platform of the system of the present invention. DETAILED DESCRIPTION

[0039] The present invention is further described in detail below with reference to the accompanying drawings and specific embodiments:

[0040] like Figure 1 Figure 2 As shown, a method for a system of intelligent linkage of substation equipment is shown, characterized in that: the system of intelligent linkage of substation equipment is equipped with a data acquisition module, a communication module, a processing and analysis module, a linkage module and a human-computer interaction module; the implementation method of the system comprises the following steps:

[0041] Step 1: Install sensors in the substation and connect them to the data collection module. The data collection module packages the collected data at a certain time interval and transmits it to the substation data processing center through the communication module connected to the industrial Ethernet to ensure the real-time and accuracy of the data.

[0042] Step 2: The communication module transmits the collected data to the processing and analysis module. The processing and analysis module uses a data program written in C++ to process the collected raw data. Within a set reasonable threshold range, it removes noise data and outliers that are obviously inconsistent with the actual situation. It also uses a convolutional neural network to extract features from the normalized data. The data is input into a network structure consisting of multiple convolutional layers and pooling layers. The network automatically learns the characteristic patterns in the data and extracts key feature vectors that can reflect the operating status of the equipment.

[0043] The convolutional neural network algorithm shown uses deep learning and a unique structural design to perform systematic image recognition, target detection, semantic segmentation, and speech recognition. It simulates the image processing mechanism of the human visual system and gradually converts raw data into abstract high-level features through layer-by-layer feature extraction to complete data analysis and processing tasks. Specifically:

[0044] 1) Convolutional layer;

[0045] The convolution layer is the core component of the convolutional neural network. Its design is inspired by the perception characteristics of neurons in the human visual cortex of local areas. It extracts local features by sliding the convolution kernel on the input data. Its principle is to slide the convolution kernel on the input data to perform convolution operations and extract local features of the data. Its calculation formula is: let the input feature map be X, the convolution kernel be K, and the output feature map be Y. For two-dimensional convolution, the output Yi,j at position (i, j) is calculated as:

[0046] Yi,j=∑m,nXi+m,j+nKm,n

[0047] Where: m and n are the indices of the convolution kernel;

[0048] 2) Pooling layer;

[0049] The pooling layer is mainly used to downsample the feature map, while retaining important features, reducing the amount of data and calculation, and reducing the risk of overfitting of the model. Its principle is to downsample the input feature map to reduce the amount of data while retaining the main features. Common pooling methods include maximum pooling and average pooling. Maximum pooling takes the maximum value in a region as the output; for example, in a 2×2 pooling window, the maximum value in the window is taken as the output value; average pooling calculates the average value in a region as the output.

[0050] 3) Fully connected layer;

[0051] The fully connected layer is usually located at the end of the convolutional neural network. It integrates the high-level features extracted by the previous convolutional layer and pooling layer and maps them to the output space to complete specific tasks such as classification and regression. Its principle is to integrate the features extracted by the previous convolutional layer and pooling layer and map them to the output space. It is usually used for classification or regression tasks. Its calculation formula is: let the input vector be x, the weight matrix be W, and the bias be b, then the output is

[0052] y=f(Wx+b)

[0053] Where: f is the activation function;

[0054] 4) Activation function;

[0055] The activation function introduces nonlinear factors into the neural network, enabling the model to learn and represent complex nonlinear relationships, thus having the ability to handle various practical problems. The principle is to introduce nonlinear factors into the neural network so that the model can learn more complex functional relationships. Commonly used activation functions include ReLU, Sigmoid, Tanh, etc. Take the ReLU function as an example:

[0056] ReLU function: The calculation formula is

[0057] f(x)=max(0,x)

[0058] That is, when x>0, f(x)=x; when x≤0, f(x)=0; it has the advantages of simple calculation and fast convergence speed;

[0059] 5) Training of convolutional neural networks;

[0060] The principle of convolutional neural network training is to use the backpropagation algorithm to update the parameters in the network, such as the weights of the convolution kernel, the weights and biases of the fully connected layers, so as to minimize the loss function. The process is as follows: first, the training data is input into the network, and the forward propagation is used to calculate the output result; then, the loss function value is calculated based on the output result and the true label; then, the gradient of the loss function with respect to each parameter is calculated through the backpropagation algorithm; finally, the parameters are updated according to the gradient. This process is repeated until the model converges to the convolution layer;

[0061] Step 3: Based on the topological structure and historical operation data of the substation equipment, the processing and analysis module performs data analysis. The processing and analysis module uses a graph neural network to analyze the correlation between the equipment point data and determine the influence relationship and linkage priority between each device. The graph neural network algorithm includes a graph convolutional network and a graph attention network. Specifically:

[0062] The calculation formula of the graph convolutional network is:

[0063]

[0064] in: For the The node feature matrix of the layer;

[0065] is the normalized adjacency matrix

[0066] A is the original adjacency matrix;

[0067] D is a diagonal matrix, and the diagonal elements are the degrees of the nodes;

[0068] For the The learnable parameter matrix of the layer,

[0069] Σ is the activation function;

[0070] As a cutting-edge technology in the field of deep learning, graph neural networks are neural network models specifically designed to process graph-structured data. They play an increasingly important role in the development of artificial intelligence and have become a key technology for breakthroughs in many fields.

[0071] Graph attention network, the formula for calculating the attention coefficient αij between node i and neighbor node j is:

[0072]

[0073] Where: hi and hj are the features of node i and neighbor node j respectively;

[0074] W is the learnable weight matrix;

[0075] a is the attention vector;

[0076] Ni is the set of neighbor nodes of node i || represents the splicing operation;

[0077] LeakyReLU is the activation function.

[0078] In 2017, the GraphSAGE (Graph Sample and Aggregate) model was proposed. By sampling and aggregating neighbor node features, it addressed the low computational efficiency of graph convolutional networks on large graphs and improved the model's scalability. In the same year, the graph attention network was introduced, introducing an attention mechanism that enables the model to adaptively learn the importance weights of different neighbor nodes to the current node, further enhancing the ability of graph neural networks to capture graph structural information.

[0079] Step 4: Link the analysis results with the linkage module; pre-configure the linkage module configuration information in the telemetry point configuration of the intelligent linkage service, and associate the results of the processing and analysis modules with the device measurement points that need to be controlled through the linkage module; the C++ data service sends the collected real-time data to emq, and the intelligent auxiliary platform application service listens to the emq real-time data topic message, performs business processing on the received message, retains only the real-time data of the telemetry type measurement point, and calls the intelligent auxiliary measurement point remote control method to control the equipment and perform related storage operations; in the intelligent linkage service, the real-time data measurement point type is filtered, and only the real-time data of the telemetry point is retained. After the linkage configuration information configured by the platform, the threshold of the real-time data is filtered, and the frequent reporting characteristics of the telemetry point are fully considered, and the data is processed twice, which specifically includes the following three points:

[0080] 1) For the value reported by the measuring point, if it has not changed, filter out this data;

[0081] 2) Filter the numerical range reported by the measuring points, and only perform linkage processing when the value exceeds the range;

[0082] 3) Increase the time interval of linkage control. If the control is repeated within a certain period of time, no operation will be performed;

[0083] Step 5. The manager performs human-machine interaction control and viewing in the human-machine interaction module. The manager performs manual control of relevant equipment in the human-machine interaction module according to needs and views relevant operating data.

[0084] The system and method for intelligent linkage of substation equipment shown in the figure first collects real-time data of equipment points through sensors and transmits it to the processing center via 5G or industrial Ethernet; then performs preprocessing such as data cleaning and normalization, and uses convolutional neural networks to extract features; based on the topological structure and historical data, a device association model is constructed with the help of graph neural networks to clarify the influence relationship and linkage priority between devices; then, combined with operating rules, thresholds and association analysis, intelligent linkage decisions are generated; finally, instructions are issued to the equipment for execution, and the decision is optimized based on feedback; the system sets up data acquisition modules, communication modules, processing and analysis modules, linkage control modules and human-computer interaction modules, which are respectively responsible for data acquisition, transmission, analysis and decision-making, instruction execution and human-computer interaction. Breaking through the traditional fixed-rule linkage model, deep learning algorithms are introduced to achieve dynamic intelligent decision-making; implicit equipment associations are mined through graph neural networks, which is more accurate and efficient than traditional methods; a closed-loop control system is constructed, combined with real-time feedback optimization strategies to enhance system adaptability and reliability; in the field of power systems, this patented technology can significantly improve the fault handling capabilities of substations and reduce the risk of fault expansion; dynamically adjust the linkage strategy according to the grid operating conditions to improve operational flexibility; optimize power resource allocation and reduce operating costs; ensure the safety of equipment and personnel, and improve the overall reliability of the power system, with high economic and social value.

[0085] The above description is merely a preferred embodiment of the present invention and does not constitute any other form of limitation to the present invention. Any modification or equivalent variation based on the technical essence of the present invention shall still fall within the scope of protection claimed by the present invention.

Claims

1. A method for intelligent linkage of substation equipment, characterized by: The system for intelligent linkage of substation equipment is equipped with a data acquisition module, a communication module, a processing and analysis module, a linkage module and a human-computer interaction module; the implementation method of the system is as follows: Step 1: Install sensors in the substation and connect them to the data collection module. The data collection module packages the collected data at a certain time interval and transmits it to the substation data processing center through the communication module connected to the industrial Ethernet to ensure the real-time and accuracy of the data. Step 2: The communication module transmits the collected data to the processing and analysis module. The processing and analysis module uses a data program written in C++ to process the collected raw data. Within a set reasonable threshold range, it removes noise data and outliers that are obviously inconsistent with the actual situation. It also uses a convolutional neural network to extract features from the normalized data. The data is input into a network structure consisting of multiple convolutional layers and pooling layers. The network automatically learns the characteristic patterns in the data and extracts key feature vectors that can reflect the operating status of the equipment. Step 3: Based on the topological structure and historical operating data of the substation equipment, the processing and analysis module performs data analysis. The processing and analysis module uses a graph neural network to analyze the correlation between the equipment point data and determine the influence relationship and linkage priority between each device. Step 4: Link the analysis results with the linkage module; pre-configure the linkage module configuration information in the telemetry point configuration of the intelligent linkage service, and associate the results obtained by the processing and analysis modules with the device measurement points that need to be controlled through the linkage module; the C++ data service sends the collected real-time data to emq, and the intelligent auxiliary platform application service listens to the emq real-time data topic message, performs business processing on the received message, retains only the real-time data of the telemetry type measurement point, and calls the intelligent auxiliary measurement point remote control method to control the equipment and perform related storage operations; Step 5. The manager performs human-machine interaction control and viewing in the human-machine interaction module. The manager performs manual control of relevant equipment in the human-machine interaction module according to needs and views relevant operating data.

2. The method of a system for intelligent linkage of substation equipment according to claim 1, characterized in that: Step 2 of the method for implementing the system of intelligent linkage of substation equipment adopts a convolutional neural network algorithm; the convolutional neural network algorithm adopts deep learning and a unique structural design to perform system image recognition, target detection, semantic segmentation, and speech recognition. It simulates the image processing mechanism of the human visual system, and through layer-by-layer feature extraction, it gradually converts the original data into abstract high-level features to complete the data analysis and processing tasks.

3. The method of a system for intelligent linkage of substation equipment according to claim 1, characterized in that: In step 3 of the method for implementing the intelligent linkage system of substation equipment, a graph neural network algorithm is used, including a graph convolutional network and a graph attention network, specifically: The calculation formula of the graph convolutional network is: ； in: For the The node feature matrix of the layer; is the normalized adjacency matrix A is the original adjacency matrix; D is a diagonal matrix, and the diagonal elements are the degrees of the nodes; For the The learnable parameter matrix of the layer; Σ is the activation function; Graph attention network, the formula for calculating the attention coefficient αij between node i and neighbor node j is: ； Where: hi and hj are the features of node i and neighbor node j respectively; W is the learnable weight matrix; a is the attention vector; Ni is the set of neighbor nodes of node i; || indicates a concatenation operation; LeakyReLU is the activation function.

4. The method of a system for intelligent linkage of substation equipment according to claim 1, characterized in that: The specific process of step 4 of the method for implementing the intelligent linkage system of substation equipment is as follows: in the intelligent linkage service, the real-time data measurement point type is filtered to retain only the real-time data of the telemetry points. The real-time data threshold is filtered according to the linkage configuration information configured by the platform. In addition, the data is processed twice considering the frequent reporting characteristics of the telemetry points. Specifically, the following three points are included: 1) For the value reported by the measuring point, if it has not changed, filter out this data; 2) Filter the numerical range reported by the measuring point, and only perform linkage processing when it exceeds the range; 3) Increase the time interval of linkage control. If the control is repeated within a certain period of time, no operation will be performed.

Images