Transmission and distribution cooperative control method and system based on elastic bearing capacity of power grid

Abstract

The invention relates to the technical field of power grid control, in particular to a transmission and distribution cooperative control method and system based on the elastic bearing capacity of a power grid, and the method comprises the steps: collecting the node data of the power grid, calculating the line power, and carrying out the real-time data transmission through a cloud; elastic bearing capacity is calculated based on line power, a power transmission layer power flow optimization model is established by using mixed integer quadratic cone programming, and power transmission layer power optimization is carried out; real-time optimization is supported through elastic bearing capacity of nodes of a power distribution layer and calculation of elastic bearing capacity of the power distribution layer, power flow distribution can be dynamically adjusted, load distribution can be optimized, power supply capacity and load balancing capacity of a power distribution network can be improved through efficient utilization of distributed energy, large-scale constraint problems can be efficiently processed through a Gurobi optimization solver, and the power distribution layer can be optimized in real time. Real-time power grid optimization dispatching is achieved, the power transmission efficiency of the whole network is remarkably improved through transmission and distribution collaborative optimization, and the power supply reliability of nodes is optimized through dynamic adjustment of distributed energy and power flow.

Description

Technical Field

[0001] The present invention relates to the technical field of power grid control, and in particular to a transmission and distribution collaborative control method and system based on the elastic carrying capacity of the power grid. Background Art

[0002] With the continuous growth of global power demand and the increase in the proportion of new energy access, the complexity and dynamics of the power grid have been significantly enhanced. The traditional transmission and distribution network control methods are facing unprecedented challenges. As an important part of the power system, the efficient, safe and reliable operation of the transmission and distribution network has a crucial impact on the overall performance of the power grid.

[0003] However, traditional power grid control methods usually adopt the strategy of independently optimizing the transmission network and the distribution network, ignoring the coupling relationship between the two, and it is difficult to achieve transmission and distribution collaborative optimization. At the same time, with the large-scale access of distributed energy, the randomness and uncertainty of load fluctuations in the distribution network are aggravated, and it is difficult to meet the abnormal identification requirements of multi-dimensional dynamic data in modern power grids. Summary of the Invention

[0004] In view of the problems existing in the above-mentioned prior art, the present invention is proposed.

[0005] Therefore, the technical problem to be solved by the present invention is to provide a transmission and distribution collaborative control method based on the elastic carrying capacity of the power grid to solve the problems that traditional power grid control methods usually adopt the strategy of independently optimizing the transmission network and the distribution network, ignoring the coupling relationship between the two, and it is difficult to achieve transmission and distribution collaborative optimization. At the same time, with the large-scale access of distributed energy, the randomness and uncertainty of load fluctuations in the distribution network are aggravated, and it is difficult to meet the abnormal identification requirements of multi-dimensional dynamic data in modern power grids.

[0006] To solve the above technical problems, the present invention provides the following technical solutions. A transmission and distribution collaborative control method based on the elastic carrying capacity of the power grid includes: collecting power grid node data, calculating line power and performing real-time data transmission through the cloud; calculating the elastic carrying capacity based on the line power, and using mixed integer second-order cone programming to establish a power flow optimization model for the transmission layer to perform power optimization for the transmission layer; based on the power optimization of the transmission layer, establishing a local optimization model for the distribution layer to perform distribution layer optimization, and performing iterative optimization based on the optimization results and the power flow optimization model of the transmission layer; constructing an anomaly detection model based on a convolutional neural network to identify abnormal lines, using the Dijkstra shortest path algorithm based on nodes to perform line update, and marking and visually displaying abnormal nodes and lines.

[0007] As a preferred solution of a transmission and distribution collaborative control method based on the elastic carrying capacity of the power grid according to the present invention, wherein: calculating the line power includes extracting substation nodes through the power grid topology map, screening nodes of 110 kV including those greater than 110 kV as transmission layer nodes, and the nodes of the transmission layer include all substations, high-voltage buses connected to transmission lines, and access points of power plants;

[0008] Taking distribution transformers below 110 kV, distributed energy sources, industrial and residential load access points, and low-voltage distribution lines as distribution layer nodes;

[0009] Based on power flow analysis, calculating the power of the lines of the transmission layer nodes, expressed as:

[0010]

[0011] wherein, P ij represents the active power of the transmission line from node i to node j, V i represents the voltage of node i, V j represents the voltage of node j, Z ij represents the line impedance between node i and node j, θ i represents the voltage phase angle at both ends of the line of node i, θ j represents the voltage phase angle at both ends of the line of node j, Q ij represents the reactive power of the transmission line from node i to node j;

[0012] Based on phase power flow analysis, calculating the power of the lines of the distribution layer nodes, expressed as:

[0013]

[0014] wherein P' ij represents the power of the distribution layer line from node i to node j;

[0015] The real-time data transmission through the cloud includes deploying edge computing nodes at transmission substations and main distribution transformers, calculating the node power and performing real-time data transmission through the cloud using the low-latency MQTT protocol.

[0016] As a preferred solution of a transmission and distribution collaborative control method based on the elastic carrying capacity of the power grid according to the present invention, wherein: calculating the elastic carrying capacity includes calculating the elastic carrying capacity of the transmission layer nodes based on the transmission power of the node lines, expressed as:

[0017]

[0018] wherein, represents the elastic carrying capacity of the i-th node of the transmission layer, C i represents the equipment capacity of the i-th node, L irepresents the current load of the i-th node, and N(i) represents the set of neighbor nodes of the i-th node;

[0019] At the same time, calculate the elastic bearing capacity of the distribution layer nodes, which is expressed as:

[0020]

[0021] Among them, represents the elastic bearing capacity of the i-th node in the distribution layer, and P der represents the distributed energy output power of node i;

[0022] The establishment of the power flow optimization model for the transmission layer using mixed-integer quadratic cone programming includes minimizing the power transmission loss of the transmission line as the objective function, taking the actual transmission power of the line not exceeding the rated capacity as the constraint condition, and taking the difference between the input power and the output power of the node equal to the net power supply of the node as the power balance constraint condition, which is expressed as:

[0023]

[0024] Among them, represents the actual power supply of the transmission layer node i;

[0025] Perform quadratic cone constraint through the Euclidean norm of the active power and reactive power of the line, which is expressed as:

[0026]

[0027] Among them, represents the power capacity of the line between the transmission layer nodes i and j;

[0028] Solve the objective function and constraint conditions through the Gurobi optimization solver, and use the MISOCP solver module to calculate the optimal transmission layer power flow and the optimal transmission layer node power supply in reverse based on the power balance constraint condition formula;

[0029] The power optimization for the transmission layer includes sending the optimization results of the transmission layer calculated by the cloud to the corresponding edge nodes for distribution layer optimization.

[0030] As a preferred solution of the transmission and distribution collaborative control method based on the grid elastic bearing capacity described in the present invention, wherein: for the power optimization of the transmission layer, establishing a local optimization model for the distribution layer includes using the optimal transmission layer node power supply as the constraint condition for the distribution layer power supply capacity, establishing a local optimization model for the distribution layer, and using maximizing the node elastic bearing capacity and minimizing the load shedding amount as the objective function, and defining the constraint condition for node power balance, which is expressed as:

[0031]

[0032] Among them, represents the actual power supply of distribution layer node i;

[0033] It is defined that the voltage of the node is within the range of the power grid design specification, and the power transmitted by the line does not exceed the rated capacity as the constraint condition;

[0034] The optimization task is divided into stages according to the physical connection network topology of the distribution layer nodes. Among them, stage 1 represents from the main distribution transformer to the first-level load nodes, and stage 2 represents from the first-level load nodes to the second-level load nodes, and so on recursively;

[0035] The dynamic programming algorithm is used for recursive calculation, which is expressed as:

[0036]

[0037] Among them, represents the optimal node power supply value function of distribution layer node i, represents the optimal node power supply value function of node j;

[0038] The power supply of the distribution layer nodes, the balance of the power flow with adjacent nodes and the distributed energy output, the power flow of the line not exceeding the rated capacity, and the voltage of the nodes need to be kept within the safety standard range are used as optimization constraints;

[0039] The iterative optimization based on the optimization results and the power flow optimization model of the transmission layer includes starting from the bottom-layer nodes of the network topology, recursively pushing up layer by layer, and calculating the optimal power supply of the distribution layer nodes and the adjusted distributed energy output power according to the recurrence equation;

[0040] The distribution layer nodes package and transmit the optimization results to the cloud. The cloud re-adjusts the power flow of the transmission layer according to the optimal power supply of the distribution layer nodes, which is expressed as:

[0041]

[0042] Among them, M(i) represents the set of neighbor nodes of distribution layer node k, represents the node power supply of distribution layer node k;

[0043] Based on the reallocated P ij Re-perform the power flow optimization model of the transmission layer and perform iterative optimization until P der does not change, and stop the iteration;

[0044] Output the power flow of the transmission lines in the transmission layer and the power supply of the transmission nodes, the power supply of the distribution nodes in the distribution layer, and the output power of the distributed energy.

[0045] As a preferred solution of a transmission and distribution collaborative control method based on the elastic carrying capacity of the power grid according to the present invention, wherein: the abnormal detection model constructed based on the convolutional neural network includes an input layer, a convolutional layer, a pooling layer, and an output layer;

[0046] Among them, the input layer includes data of node voltage, current, power flow, and elastic carrying capacity value based on a time acquisition window; the convolutional layer extracts feature patterns in the local time series; the pooling layer reduces the computational complexity through dimensionality reduction; the output layer maps the convolutional features to an abnormal probability and outputs the abnormal probability using the Sigmoid activation function;

[0047] The identification of abnormal lines includes training the abnormal detection model using the training data calibrated for abnormalities, selecting the cross-entropy loss function to calculate the calculation loss between the class probabilities predicted by the CNN and the actual labels, using the Adam optimizer for gradient descent optimization, updating the weights of the CNN model, and stopping the iteration when the loss of the model no longer decreases significantly during continuous iteration, and outputting the model parameters to update the model;

[0048] Based on the input acquisition data of the distribution nodes and predict the abnormal probability.

[0049] As a preferred solution of a transmission and distribution collaborative control method based on the elastic carrying capacity of the power grid according to the present invention, wherein: the Dijkstra shortest path algorithm based on nodes includes determining an abnormal threshold based on the sum of the mean and standard deviation of the historical abnormal probabilities. If the predicted value of the abnormal probability is greater than or equal to the abnormal threshold, the corresponding node is determined as an abnormal node, and the associated line combination of the abnormal node is marked as an abnormal associated line set and isolated;

[0050] Based on each node, use the Dijkstra shortest path algorithm to calculate the current shortest path length from the source node to node i in the set of unvisited nodes, select the node u with the smallest current distance value and remove u from the set of unvisited nodes, traverse the neighbor nodes of node u, and select the minimum path node from node u to the neighbor node;

[0051] Re-select the node with the smallest current distance value from the set of unvisited nodes again, gradually calculate and determine the shortest paths from the source node to all nodes, determine the shortest path of the target node, and verify the abnormality judgments of all nodes and lines in the path.

[0052] As a preferred solution of a transmission and distribution collaborative control method based on the elastic carrying capacity of the power grid according to the present invention, wherein: the marking and visual display of abnormal nodes and lines includes adding the shortest paths excluding abnormal nodes and lines determined by the Dijkstra shortest path algorithm as alternative paths to the current network topology map;

[0053] Execute power flow analysis to reallocate power flow and node power supply, and mark the faulty nodes and lines identified by anomaly detection as offline;

[0054] Display the new power flow distribution and node power supply status on the visualization platform.

[0055] Another object of the present invention is to provide a transmission and distribution collaborative control system based on the elastic carrying capacity of the power grid. By integrating multiple modules such as data acquisition, power flow optimization at the transmission layer, local optimization at the distribution layer, transmission and distribution collaborative iterative optimization, anomaly detection, dynamic path update, and visualization display, real-time monitoring and management of each node and line in the power grid are achieved. The system can optimize the power flow distribution at the transmission layer based on minimizing power transmission losses and satisfying the constraints of power balance; at the same time, by maximizing the node elastic carrying capacity and minimizing the load shedding amount, the distribution network is optimized in stages, and the power supply distribution is optimized layer by layer recursively. In addition, the system can also use a convolutional neural network to construct an anomaly detection model to timely identify faulty nodes and lines, and use the Dijkstra shortest path algorithm to recalculate the power supply path to ensure the stability and security of the power grid operation. Finally, through the visualization display module, users can intuitively understand the status and changes of the entire power grid.

[0056] To solve the above technical problems, the present invention provides the following technical solutions: A transmission and distribution collaborative control system based on the elastic carrying capacity of the power grid, comprising: a data acquisition module, a power flow optimization module at the transmission layer, a local optimization module at the distribution layer, a transmission and distribution collaborative iterative optimization module, an anomaly detection module, a dynamic path update module, and a visualization display module;

[0057] The data acquisition module collects real-time data of each node and line in the power grid;

[0058] The power flow optimization module at the transmission layer uses mixed-integer second-order cone programming to construct a power flow optimization model at the transmission layer, and optimizes the power flow distribution at the transmission layer based on minimizing power transmission losses and satisfying the constraints of power balance;

[0059] The local optimization module at the distribution layer establishes a local optimization model at the distribution layer based on the power flow optimization results at the transmission layer, maximizes the node elastic carrying capacity and minimizes the load shedding amount, optimizes the distribution network in stages, and optimizes the power supply distribution layer by layer recursively;

[0060] The transmission and distribution collaborative iterative optimization module, based on the local optimization results at the distribution layer, reversely adjusts the power flow optimization model at the transmission layer for iterative optimization, and continuously iterates until the changes in power distribution and node power supply tend to be stable;

[0061] The abnormal detection module uses a convolutional neural network (CNN) to build an abnormal detection model. It inputs the time series data of nodes and lines, outputs the abnormal probabilities of each line and node, and determines whether there is an abnormality by combining with a threshold value.

[0062] The dynamic path update module uses the Dijkstra shortest path algorithm. After removing abnormal nodes and lines, it recalculates the power supply path and adds the updated path as an alternative path to the current network topology.

[0063] The visualization display module shows the power supply status of nodes and the distribution of line power flow through colors or icons.

[0064] A computer device includes a memory and a processor. The memory stores a computer program. The processor, when executing the computer program, implements the steps of a transmission and distribution collaborative control method based on the elastic bearing capacity of the power grid as described above.

[0065] A computer-readable storage medium stores a computer program. The computer program, when executed by a processor, implements the steps of a transmission and distribution collaborative control method based on the elastic bearing capacity of the power grid as described above.

[0066] The beneficial effects of the present invention are as follows: Through the elastic bearing capacity of distribution layer nodes, the calculation of the elastic bearing capacity of the distribution layer supports real-time optimization, can dynamically adjust the power flow distribution, optimize the load distribution, improve the power supply capacity and load balancing ability of the distribution network through the efficient utilization of distributed energy, can efficiently handle large-scale constraint problems through the Gurobi optimization solver, realize real-time power grid optimization scheduling, significantly improve the efficiency of power transmission across the network through transmission and distribution collaborative optimization, optimize the power supply reliability of nodes through the dynamic adjustment of distributed energy and power flow, reduce the risk of power supply interruption, and avoid the efficiency loss of independent optimization of the two layers through the two-way interaction of optimization between the transmission layer and the distribution layer, thus achieving the maximization of overall benefits. BRIEF DESCRIPTION OF THE DRAWINGS

[0067] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for description in the embodiments will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0068] Figure 1 It is a flowchart showing a transmission and distribution collaborative control method based on the elastic bearing capacity of the power grid provided by an embodiment of the present invention.

[0069] Figure 2Schematic diagram of the module structure of a transmission and distribution collaborative control method based on the elastic carrying capacity of the power grid provided by an embodiment of the present invention. Detailed implementation manners

[0070] To make the above objects, features, and advantages of the present invention more obvious and understandable, the following will describe the detailed implementation manners of the present invention in conjunction with the accompanying drawings of the specification. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0071] Example 1, referring to Figure 1 , which is an embodiment of the present invention. This embodiment provides a transmission and distribution collaborative control method based on the elastic carrying capacity of the power grid, including:

[0072] S1: Collect power grid node data, calculate line power, and perform real-time data transmission through the cloud.

[0073] It should be noted that as shown in S1 in Figure 1 , collecting power grid node data, calculating line power, and performing real-time data transmission through the cloud includes extracting substation nodes through the power grid topology diagram, and screening nodes with a voltage level of 110 kV or above as transmission layer nodes. The nodes in the transmission layer include all substations (voltage level ≥ 110 kV), high-voltage buses connected to transmission lines, and access points of power plants;

[0074] Connect distribution transformers below 110 kV, distributed energy sources (such as photovoltaic systems, energy storage devices), industrial and residential load access points, and low-voltage distribution lines as distribution layer nodes;

[0075] Based on power flow analysis, calculate the power of the lines of the transmission layer nodes, expressed as:

[0076]

[0077] Among them, P ij represents the active power of the transmission line from node i to node j, V i represents the voltage of node i, V j represents the voltage of node j, Z ij represents the line impedance between node i and node j, θ i represents the voltage phase angle at both ends of the line of node i, θ j represents the voltage phase angle at both ends of the line of node j, Q ij represents the reactive power of the transmission line from node i to node j.

[0078] Furthermore, based on phase power flow analysis, calculate the power of the lines of the distribution layer nodes, expressed as:

[0079]

[0080] where P' ij represents the power of the line from i to j in the distribution layer;

[0081] Edge computing nodes are deployed at each transmission substation and main distribution transformer, the power of the computing nodes is calculated, and real-time data transmission is carried out through the cloud using the low-latency MQTT protocol.

[0082] Furthermore, the nodes in the transmission layer and the distribution layer are respectively responsible for long-distance transmission of high power, ensuring the effective transmission and distribution of electric energy between regions in the municipal-level power grid to specific users and distributed energy access points, directly serving the effective transmission of individual users. The hierarchical design can effectively reduce the complexity of optimization calculation and achieve the rapid solution of local problems, providing technology for the calculation and optimization of elastic bearing capacity. In the cloud collaborative calculation, the power flow optimization of the transmission layer provides input for the distribution layer scheduling, and the elastic data fed back by the distribution layer provides assistance for the decision-making of the transmission layer;

[0083] The phase power flow analysis simplifies the power flow calculation formula for the low-voltage distribution network, reduces the calculation complexity, and facilitates real-time online calculation. Through the power flow analysis of the distribution layer, the power contribution of distributed energy can be accurately calculated, the load demand and energy supply can be coordinated, the power flow of the distribution layer can be monitored and optimized in real time, which helps to quickly locate abnormal nodes and make dynamic adjustments, enhancing the elasticity and reliability of the distribution network. Through edge computing combined with the low-latency MQTT protocol, the response time of data transmission and processing is shortened, adapting to the dynamic changes of the power grid. The coordinated optimization of transmission and distribution avoids the unbalanced distribution of power resources and improves the collaborative efficiency of transmission and distribution.

[0084] S2: Calculate the elastic bearing capacity based on the line power, and use mixed-integer second-order cone programming to establish a power flow optimization model for the transmission layer to optimize the power of the transmission layer.

[0085] It should be noted that, as Figure 1 shown in S2, calculating the elastic bearing capacity based on the line power and using mixed-integer second-order cone programming to establish a power flow optimization model for the transmission layer to optimize the power of the transmission layer includes calculating the elastic bearing capacity of the nodes in the transmission layer based on the transmission power of the node lines, expressed as:

[0086]

[0087] where, represents the elastic bearing capacity of the i-th node in the transmission layer, C i represents the equipment capacity of the i-th node, L i represents the current load of the i-th node, and N(i) represents the set of neighbor nodes of the i-th node;

[0088] Calculate the elastic bearing capacity of the distribution layer nodes simultaneously, expressed as:

[0089]

[0090] Among them, represents the elastic bearing capacity of the i-th node in the distribution layer, and P der represents the distributed energy output power of node i.

[0091] Furthermore, use mixed-integer second-order cone programming (MISOCP) to establish a power flow optimization model for the transmission layer. Based on minimizing the power transmission loss of the transmission line as the objective function, the actual transmitted power of the line does not exceed its rated capacity as a constraint condition, and the difference between the input power and the output power of each node must be equal to the net power supply of the node as a power balance constraint condition, expressed as:

[0092]

[0093] Among them, represents the actual power supply of the transmission layer node i;

[0094] Perform second-order cone constraints through the Euclidean norm of the active and reactive powers of the line, expressed as:

[0095]

[0096] Among them, represents the power capacity of the line between the transmission layer nodes i and j;

[0097] Solve the objective function and constraint conditions through the Gurobi optimization solver, and use the MISOCP solver module to calculate the optimal transmission layer power flow and the optimal transmission layer node power supply in reverse based on the power balance constraint condition formula;

[0098] Send the optimized results of the transmission layer calculated by the cloud to the corresponding edge nodes for distribution layer optimization.

[0099] Furthermore, by calculating the elastic bearing capacity of the transmission layer nodes, the remaining bearing capacity of each node in the transmission layer after meeting the current load demand can be determined, facilitating the reasonable distribution of the load. It indicates that the capacity of the node equipment is insufficient and there is an overload risk. Early planning for dispatching optimization or capacity increase can be carried out. Through the dynamic monitoring of the elastic bearing capacity, the power grid faults, load fluctuations, and emergency power supply demands can be better coped with. Through the elastic bearing capacity of the distribution layer nodes, the calculation of the elastic bearing capacity of the distribution layer supports real-time optimization, can dynamically adjust the power flow distribution, optimize the load distribution, improve the power supply capacity and load balancing capacity of the distribution network through the efficient utilization of distributed energy, reduce the energy loss on the line by optimizing the power flow path, improve the transmission efficiency, ensure that all transmission lines are within the safe operating range by considering the line rated capacity constraint, and prevent equipment failures caused by overload. The traditional linear programming model's inability to handle non-linear and second-order cone constraints is solved by MISOCP (Mixed Integer Second-Order Cone Programming), adapting to the complexity of power grid power flow optimization. The Gurobi optimization solver can efficiently handle large-scale constraint problems to achieve real-time power grid optimization dispatching. The overall network power transmission efficiency is significantly improved, the loss is reduced, and the power supply capacity is enhanced through the coordinated optimization of transmission and distribution. Through the collaborative optimization of the cloud and edge nodes, the integrated intelligent management of transmission and distribution is realized, promoting the digital and intelligent transformation of the power grid.

[0100] S3: Based on the power optimization of the transmission layer, establish a local optimization model for the distribution layer to optimize the distribution layer, and perform iterative optimization based on the optimization results and the power flow optimization model of the transmission layer.

[0101] It should be noted that, as Figure 1 shown in S3, based on the power optimization of the transmission layer, establish a local optimization model for the distribution layer to optimize the distribution layer, and perform iterative optimization based on the optimization results and the power flow optimization model of the transmission layer, including: Based on the optimal power supply of the transmission layer nodes as the constraint condition for the power supply capacity of the distribution layer, establish a local optimization model for the distribution layer. Based on maximizing the node elastic bearing capacity and minimizing the load shedding amount as the objective function, define the constraint condition of node power balance, expressed as:

[0102]

[0103] Among them, represents the actual power supply of distribution layer node i;

[0104] At the same time, define that the voltage of each node is within the upper and lower limit values of the power grid design specification, and the power transmitted by the line shall not exceed its rated capacity as the constraint condition;

[0105] The optimization tasks are divided into stages according to the physical connection network topology of the distribution layer nodes. Among them, stage 1 is from the main distribution transformer to the first-level load nodes, stage 2 is from the first-level load nodes to the second-level load nodes, and so on recursively.

[0106] The dynamic programming algorithm is used for recursive calculation, expressed as:

[0107]

[0108] Among them, represents the optimal node power supply value function of distribution layer node i, represents the optimal node power supply value function of node j;

[0109] The power supply of the distribution layer nodes, the balance between the power flow of adjacent nodes and the distributed energy output, and the power flow of the line shall not exceed its rated capacity, and the voltage of the nodes shall be maintained within the safety standard range are used as optimization constraints.

[0110] Furthermore, starting from the bottom layer nodes of the network topology, recursively upwards, calculate the optimal power supply of each distribution layer node and the adjusted distributed energy output power according to the recurrence equation;

[0111] The distribution layer nodes package and transmit the optimization results to the cloud. The cloud re-adjusts the power flow of the transmission layer according to the optimal power supply of the distribution layer nodes, expressed as:

[0112]

[0113] Among them, M(i) represents the set of neighbor nodes of distribution layer node k, represents the node power supply of distribution layer node k;

[0114] Based on the redistributed P ij Re-perform the power flow optimization model of the transmission layer and conduct iterative optimization until the change of P der is no longer obvious, and stop the iteration;

[0115] Output the power flow of each transmission line in the transmission layer and the power supply of each transmission node, the power supply of each distribution node in the distribution layer, and the output power of the distributed energy.

[0116] Furthermore, the optimized node elastic bearing capacity improves the ability of the distribution network to cope with load fluctuations and emergencies. After the node bearing capacity is clarified, the load can be distributed according to priorities, avoiding unnecessary load shedding. By preferentially supplying power to important nodes and loads, economic and social losses caused by the curtailment of critical loads are avoided. Starting from the bottom-layer nodes and performing recursive calculations, the power supply optimization of the upper-layer nodes will inherit the optimization results of the lower-layer nodes, improving the overall optimization quality. Based on the physical topology of the distribution network, the recursive strategy is flexibly adjusted to adapt to various distribution network structures. The safety of node operation is ensured through power balance and voltage limits, preventing problems such as overload, undervoltage, or overvoltage. The power supply reliability of nodes is optimized through the dynamic adjustment of distributed energy and power flow, reducing the risk of power supply interruption. The two-way interaction between the optimization of the transmission layer and the distribution layer avoids the efficiency loss of independent optimization of the two layers and maximizes the overall benefits. Through multiple iterative optimizations, the system dynamically responds to the uncertainties of load fluctuations and distributed energy output. By establishing a local optimization model for the distribution layer, dynamic programming recursion, and coordinated iteration optimization of transmission and distribution, the entire system achieves accurate load distribution, minimized losses, dynamic response to load fluctuations, and efficient utilization of distributed energy, comprehensively improving the operation efficiency, power supply reliability, and safety of the power grid, providing technical support for building a modern intelligent power grid.

[0117] S4: Build an anomaly detection model based on a convolutional neural network to identify abnormal lines. Based on nodes, use the Dijkstra shortest path algorithm to update the lines, and mark the abnormal nodes and lines for visual display.

[0118] It should be noted that, as Figure 1 shown in S4, building an anomaly detection model based on a convolutional neural network to identify abnormal lines includes: building an anomaly detection model based on a convolutional neural network, including an input layer, a convolutional layer, a pooling layer, and an output layer;

[0119] Among them, the input layer includes data on node voltage, current, power flow, and elastic bearing capacity values based on a time acquisition window. The convolutional layer extracts feature patterns in local time series. The pooling layer reduces the computational complexity through dimensionality reduction. The output layer maps the convolutional features to anomaly probabilities and uses the Sigmoid activation function to output anomaly probabilities;

[0120] Use the training data calibrated for anomalies to train the anomaly detection model. Select the cross-entropy loss function to calculate the computational loss between the class probabilities predicted by the CNN and the actual labels. Use the Adam optimizer for gradient descent optimization to update the weights of the CNN model. Stop the iteration when the loss of the model no longer decreases significantly during consecutive iterations, and output the model parameters to update the model;

[0121] Predict the anomaly probabilities based on the input acquisition data of distribution nodes;

[0122] By means of data input based on a time acquisition window, the model can capture the dynamic change characteristics of power grid operation, which is applicable to detecting short-term fluctuations and long-term trend anomalies. Through convolutional operations, the model can identify tiny anomaly patterns (such as voltage offsets and current surges), and can accurately detect them even in large-scale data. For the complex boundary between normal and abnormal conditions in the power grid, the output probability distribution can reflect the uncertainty of the model, supporting more accurate classification decisions. The intelligent monitoring system based on the CNN model can meet the requirements of large-scale distributed power grids and provide technical support for the intelligent operation of power grids.

[0123] Furthermore, based on each node, the Dijkstra shortest path algorithm is used for line update, including: determining the anomaly threshold based on the sum of the mean and standard deviation of historical anomaly probabilities. If the predicted value of the anomaly probability is greater than or equal to the anomaly threshold, the corresponding node is determined as an abnormal node, and the associated lines of the abnormal node are marked as an abnormal associated line set and isolated.

[0124] Based on each node, the Dijkstra shortest path algorithm is used to calculate the current shortest path length from the source node to node i in the set of unvisited nodes, select the node u with the smallest current distance value and remove u from the set of unvisited nodes, traverse the neighbor nodes of node u, and select the node with the minimum path from node u to the neighbor node.

[0125] Re-select the node with the smallest current distance value from the set of unvisited nodes again, gradually calculate and determine the shortest paths from the source node to all other nodes until the shortest path to the target node is determined, and verify that all nodes and lines in the path are not judged as abnormal.

[0126] By means of data input based on a time acquisition window, the model can capture the dynamic change characteristics of power grid operation, which is applicable to detecting short-term fluctuations and long-term trend anomalies. Through the greedy strategy of the Dijkstra algorithm, the shortest path can be quickly generated in a complex network with multiple nodes and multiple lines, meeting the real-time scheduling requirements of the power grid. By dynamically reconstructing the path to bypass abnormal nodes and lines, the influence range is controlled within the smallest area to ensure the power supply stability of most users. After calibrating the fault range through anomaly detection, the dynamic path planning quickly generates a reliable power supply path, realizing an efficient closed-loop from detection to optimization. Through the combination of anomaly detection and path optimization, the real-time monitoring, judgment, and optimized scheduling of the power grid operation state are realized, comprehensively improving the intelligent level of the power grid.

[0127] Even further, marking and visualizing abnormal nodes and lines includes adding the shortest path excluding abnormal nodes and lines determined by the Dijkstra shortest path algorithm as an alternative path to the current network topology diagram.

[0128] Execute power flow analysis to reallocate power flow and node power supply, and mark the faulty nodes and lines identified by anomaly detection as offline;

[0129] Display the new power flow distribution and node power supply status on the visualization platform.

[0130] By using the Dijkstra shortest path algorithm, the shortest path calculated after removing abnormal nodes and lines is used as an alternative path, which is dynamically added to the current network topology map. The optimal line is selected by calculating the shortest path, reducing power loss and lowering the economic and energy costs of power grid anomaly recovery. Through power flow analysis, the power flow is reallocated, the optimal transmission path is selected, reducing line load imbalance and power loss. Through the introduction of alternative paths and dynamic adjustment, the power supply reliability in abnormal situations is significantly improved, reducing the risk of large-scale power outages. Through power flow analysis and topology adjustment, dispatchers can quickly adjust operation strategies according to real-time demands and flexibly adapt to different load scenarios and sudden faults.

[0131] The above is a schematic solution of a transmission and distribution collaborative control method based on the elastic carrying capacity of the power grid in this embodiment. It should be noted that the technical solution of the system of a transmission and distribution collaborative control method based on the elastic carrying capacity of the power grid belongs to the same concept as the above-mentioned technical solution of a transmission and distribution collaborative control method. For the details not described in detail in the technical solution of the system of a transmission and distribution collaborative control method based on the elastic carrying capacity of the power grid in this embodiment, reference can be made to the description of the technical solution of the above-mentioned transmission and distribution collaborative control method.

[0132] Example 2, referring to Figure 2 , which is an embodiment of the present invention. This embodiment provides a transmission and distribution collaborative control system based on the elastic carrying capacity of the power grid, including: a data acquisition module, a power flow optimization module for the transmission layer, a local optimization module for the distribution layer, a transmission and distribution collaborative iterative optimization module, an anomaly detection module, a dynamic path update module, and a visualization display module;

[0133] The data acquisition module collects real-time data of each node and line in the power grid;

[0134] The power flow optimization module for the transmission layer uses mixed-integer second-order cone programming to construct a power flow optimization model for the transmission layer, and optimizes the power flow distribution of the transmission layer based on the constraints of minimizing power transmission loss and satisfying power balance;

[0135] The local optimization module for the distribution layer establishes a local optimization model for the distribution layer based on the power flow optimization results of the transmission layer, maximizes the node elastic carrying capacity and minimizes the load curtailment amount, and optimizes the distribution network in stages, and recursively optimizes the power supply distribution layer by layer;

[0136] The above-mentioned transmission and distribution collaborative iterative optimization module, based on the local optimization results of the distribution layer, reversely adjusts the power flow optimization model of the transmission layer, conducts iterative optimization, and continuously iterates until the changes in power distribution and node power supply tend to be stable;

[0137] The above-mentioned anomaly detection module uses a convolutional neural network (CNN) to construct an anomaly detection model, inputs the time series data of nodes and lines, outputs the anomaly probabilities of each line and node, and combines thresholds to determine whether there are anomalies;

[0138] The above-mentioned dynamic path update module uses the Dijkstra shortest path algorithm, recalculates the power supply path after removing abnormal nodes and lines, and adds the updated path as an alternative path to the current network topology;

[0139] The above-mentioned visualization display module displays the node power supply status and line power flow distribution through colors or icons.

[0140] This embodiment also provides a computing device applicable to a situation of a transmission and distribution collaborative control method based on the elastic carrying capacity of the power grid, including:

[0141] A memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to implement a transmission and distribution collaborative control method based on the elastic carrying capacity of the power grid as proposed in the above embodiment.

[0142] This embodiment also provides a storage medium, on which a computer program is stored, and when the program is executed by a processor, it implements a transmission and distribution collaborative control method based on the elastic carrying capacity of the power grid as proposed in the above embodiment.

[0143] The storage medium proposed in this embodiment and a transmission and distribution collaborative control method based on the elastic carrying capacity of the power grid proposed in the above embodiment belong to the same inventive concept. Technical details not described in detail in this embodiment can be referred to in the above embodiment, and this embodiment has the same beneficial effects as the above embodiment.

[0144] When the above-mentioned functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or a part of this technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in various embodiments of the present invention. The foregoing storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memories (ROM, Read-Only Memory), random access memories (RAM, Random Access Memory), magnetic disks, or optical discs that can store program codes.

[0145] The logic and / or steps described otherwise herein, for example, can be considered as a defined sequence list of executable instructions for implementing logical functions, and can be specifically implemented in any computer-readable medium for use by an instruction execution system, apparatus, or device (such as a computer-based system, a system including a processor, or other systems that can fetch and execute instructions from the instruction execution system, apparatus, or device), or used in combination with these instruction execution systems, apparatuses, or devices. For the purposes of this specification, a "computer-readable medium" can be any device that can contain, store, communicate, propagate, or transport a program for use by or in connection with an instruction execution system, apparatus, or device.

[0146] It should be understood that each part of the present invention can be implemented by hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, any one or a combination of the following well-known technologies in the art can be used: discrete logic circuits having logic gate circuits for implementing logical functions on data signals, application-specific integrated circuits having appropriate combinational logic gate circuits, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0147] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical solutions of the present invention can be modified or equivalently replaced without departing from the spirit and scope of the technical solutions of the present invention, and they should all be covered by the scope of the claims of the present invention.

Claims

1. A transmission and distribution coordinated control method based on the elastic bearing capacity of a power grid, characterized in that: include: Collect grid node data, calculate line power and transmit real-time data through the cloud; The elastic bearing capacity is calculated based on the line power, and the transmission layer power flow optimization model is established using mixed integer quadratic cone programming to optimize the transmission layer power. Based on the power optimization of the transmission layer, a local optimization model of the distribution layer is established to optimize the distribution layer, and iterative optimization is performed based on the optimization results and the power flow optimization model of the transmission layer; An anomaly detection model is built based on a convolutional neural network to identify abnormal lines. The Dijkstra shortest path algorithm is used based on nodes to update lines, and abnormal nodes and lines are marked for visual display.

2. A transmission and distribution coordinated control method based on the elastic bearing capacity of a power grid as claimed in claim 1, characterized in that: The calculation of line power includes extracting substation nodes through the power grid topology map, screening 110kV and nodes greater than 110kV as transmission layer nodes, wherein the nodes of the transmission layer include all substations, high-voltage buses connected to the transmission lines, and access points of power plants; Distribution transformers below 110kV, distributed energy resources, industrial and residential load access points, and low-voltage distribution lines are considered as distribution layer nodes; The power of the transmission layer node line is calculated based on the power flow analysis and is expressed as: Among them, P ij represents the active power of transmission line i to j, V i represents the voltage at node i, V j represents the voltage at node j, Z ij represents the line impedance between node i and node j, θ i represents the voltage phase angle across the line at node i, θ i represents the voltage phase angle across the line at node j, Q ij represents the reactive power of transmission line i to j; The power of the distribution layer node line is analyzed based on the phase flow, which is expressed as: Where P' ij Represents the power of the distribution layer line i to j; The real-time data transmission through the cloud includes deploying edge computing nodes at transmission substations and main distribution transformers, calculating node power and transmitting real-time data through the cloud using the low-latency MQTT protocol.

3. A transmission and distribution coordinated control method based on the elastic bearing capacity of a power grid as claimed in claim 2, characterized in that: The calculation of the elastic bearing capacity includes calculating the elastic bearing capacity of the transmission layer node based on the transmission power of the node line, which is expressed as: in, represents the elastic bearing capacity of the ith node of the transmission layer, C i represents the equipment capacity of the ith node, L i represents the current load of the i-th node, and N(i) represents the set of neighbor nodes of the i-th node; At the same time, the elastic bearing capacity of the distribution layer nodes is calculated and expressed as: in, represents the elastic bearing capacity of the ith node of the distribution layer, P der represents the distributed energy output power of node i; The use of mixed integer quadratic cone programming to establish a transmission layer power flow optimization model includes minimizing the power transmission loss of the transmission line as the objective function, taking the actual transmission power of the line not exceeding the rated capacity as a constraint condition, and taking the difference between the input power and the output power of the node equal to the net power supply of the node as a power balance constraint condition, which is expressed as: in, represents the actual power supply of node i in the transmission layer; The secondary cone constraint is performed by the Euclidean norm of the line active power and reactive power, which is expressed as: in, represents the power capacity of the line between node i and node j in the transmission layer; The objective function and constraints are optimized by Gurobi solver, and the MISOCP solver module is used to reversely calculate the optimal transmission layer power flow and the optimal transmission layer node power supply based on the power balance constraint formula; The transmission layer power optimization includes sending the transmission layer optimization results calculated in the cloud to the corresponding edge nodes to perform distribution layer optimization.

4. A transmission and distribution coordinated control method based on the elastic bearing capacity of a power grid as claimed in claim 3, characterized in that: The transmission layer power optimization and establishment of a local optimization model for the distribution layer include establishing a local optimization model for the distribution layer based on the optimal transmission layer node power supply as a constraint condition for the power supply capacity of the distribution layer, and defining the constraint condition for node power balance based on maximizing the node elastic bearing capacity and minimizing the load reduction as the objective function, which is expressed as: in, Indicates the actual power supply of node i in the distribution layer; The voltage of the defined node is within the range of the grid design specification, and the power transmitted by the line does not exceed the rated capacity as a constraint; The optimization task is divided into stages according to the physical connection network topology of the distribution layer nodes, where stage 1 represents from the main distribution transformer to the primary load node, stage 2 represents from the primary load node to the secondary load node, and so on recursively; Use dynamic programming algorithm for recursive calculation, expressed as: in, represents the optimal node power supply value function of node i in the distribution layer, represents the optimal node power supply value function of node j; The power supply of the distribution layer nodes is balanced with the power flow of the adjacent nodes and the output balance of distributed energy resources, and the power flow of the line does not exceed the rated capacity, and the voltage of the node must be kept within the safety standard range as optimization constraints; The iterative optimization based on the optimization results and the transmission layer power flow optimization model includes starting from the bottom node of the network topology, recursively upward layer by layer, and calculating the optimal power supply of the distribution layer node and the adjusted distributed energy output power according to the recursive equation; The distribution layer nodes package the optimization results and transmit them to the cloud. The cloud readjusts the transmission layer power flow according to the optimal power supply of the distribution layer nodes, which is expressed as: Where M(i) represents the set of neighbor nodes of node k at the distribution layer. Indicates the node power supply of node k in the distribution layer; Based on the reallocation of P ij Re-optimize the transmission layer power flow model and perform iterative optimization until P der When there is no change, stop iterating; The output transmission layer includes the power flow of the transmission lines and the power supply of the transmission nodes, the power supply of the distribution nodes and the output power of the distributed energy at the distribution layer.

5. A transmission and distribution coordinated control method based on the elastic bearing capacity of a power grid as claimed in claim 4, characterized in that: The anomaly detection model based on the convolutional neural network includes an input layer, a convolution layer, a pooling layer and an output layer; The input layer includes node voltage, current, power flow and elastic bearing capacity value data based on the time acquisition window; the convolution layer extracts the characteristic patterns in the local time series; the pooling layer reduces the computational complexity by reducing the dimension; the output layer maps the convolution features to abnormal probability and uses the Sigmoid activation function to output the abnormal probability; The abnormal line identification includes training the abnormal detection model using abnormally calibrated training data, selecting a cross entropy loss function to calculate the computational loss between the category probability predicted by the CNN and the actual label, using the Adam optimizer to perform gradient descent optimization, updating the weights of the CNN model, stopping the iteration if the loss of the model no longer decreases significantly during the continuous iteration process, and outputting the model parameters to update the model; Collect data based on the input of distribution nodes and predict the abnormal probability.

6. A transmission and distribution coordinated control method based on the elastic bearing capacity of a power grid as claimed in claim 5, characterized in that: The node-based Dijkstra shortest path algorithm includes determining an abnormal threshold based on the sum of the mean and standard deviation of the historical abnormal probability. If the predicted value of the abnormal probability is greater than or equal to the abnormal threshold, the corresponding node is judged as an abnormal node, and the associated line combination of the abnormal node is marked as an abnormal associated line set, and isolated; Based on each node, Dijkstra's shortest path algorithm is used to calculate the current shortest path length from the source node to the node i from the unvisited node set, select the node u with the smallest current distance value and remove u from the unvisited node set, traverse the neighbor nodes of node u, and select the node with the minimum path from node u to the neighbor node; Reselect the node with the smallest current distance value from the set of unvisited nodes, gradually calculate and determine the shortest path from the source node to all nodes, determine the shortest path to the target node, and verify the abnormal judgment of all nodes and lines in the path.

7. A transmission and distribution coordinated control method based on the elastic bearing capacity of a power grid as claimed in claim 6, characterized in that: The marking of abnormal nodes and lines for visual display includes adding the shortest path determined based on the Dijkstra shortest path algorithm and excluding abnormal nodes and lines as a backup path to the current network topology diagram; Perform power flow analysis to redistribute power flow and node power supply, and mark faulty nodes and lines identified by anomaly detection as offline; The new power flow distribution and node power supply status are displayed on the visualization platform.

8. A system for coordinated control of transmission and distribution based on the elastic bearing capacity of a power grid according to any one of claims 1 to 7, characterized in that: include: Data acquisition module, transmission layer power flow optimization module, distribution layer local optimization module, transmission and distribution collaborative iterative optimization module, anomaly detection module, dynamic path update module and visualization display module; The data acquisition module collects real-time data of each node and line in the power grid; The transmission layer power flow optimization module uses mixed integer quadratic cone programming to construct a transmission layer power flow optimization model, and optimizes the power flow distribution of the transmission layer based on minimizing power transmission loss and satisfying power balance constraints; The distribution layer local optimization module establishes a local optimization model for the distribution layer based on the power transmission layer flow optimization results, maximizes the node elastic bearing capacity and minimizes the load reduction, optimizes the distribution network in stages, and recursively optimizes the power supply distribution layer by layer; The transmission and distribution collaborative iterative optimization module reversely adjusts the transmission layer power flow optimization model based on the local optimization results of the distribution layer, performs iterative optimization, and continues to iterate until the changes in power distribution and node power supply tend to be stable; The anomaly detection module uses a convolutional neural network (CNN) to build an anomaly detection model, inputs the time series data of nodes and lines, outputs the anomaly probability of each line and node, and determines whether it is abnormal in combination with a threshold; The dynamic path updating module uses the Dijkstra shortest path algorithm to recalculate the power supply path after removing abnormal nodes and lines, and adds the updated path as a backup path to the current network topology; The visual display module displays the node power supply status and line power flow distribution through colors or icons.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of a transmission and distribution coordinated control method based on the elastic carrying capacity of the power grid as described in any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of a transmission and distribution coordinated control method based on the elastic carrying capacity of a power grid as described in any one of claims 1 to 7 are implemented.

Images