Power grid adaptive scheduling method and system based on fault prediction and resource coordination

Abstract

The invention discloses a power grid adaptive scheduling method and system based on fault prediction and resource coordination, and the method comprises the steps: collecting power grid multi-source state data, carrying out the preprocessing of the collected power grid multi-source state data, carrying out the linear feature fusion of the preprocessed data, inputting the fused data into a multi-dimensional fault prediction model, outputting a fault prediction result; dynamically adjusting a power grid protection strategy according to a fault prediction result output by the multi-dimensional fault prediction model, and adaptively allocating communication resources according to predicted communication requirements and communication priorities; and integrating a power grid protection strategy and communication resource allocation, establishing a global scheduling target optimization function, and realizing cross-domain scheduling of power protection and communication resource management based on a solving result of the global scheduling target optimization function. According to the invention, the fault processing accuracy and the communication resource allocation balance can be improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of power systems and communication networks, and in particular relates to a power grid adaptive dispatching method and system based on fault prediction and resource coordination. Background Art

[0002] As the scale of the power grid continues to expand and the amount of new energy access increases, the structure and load characteristics of the power grid have become increasingly complex, resulting in a significant increase in the difficulty of implementing power grid fault prediction and protection strategies. At the same time, traditional power grid protection strategies usually rely on static settings and cannot adapt to complex scenarios with frequent changes in power grid structure and dynamic load fluctuations. Especially in the case of sudden faults, the response time of traditional protection strategies is long and cannot effectively prevent the spread of faults.

[0003] In addition, as an important part of ensuring the stable operation of the power grid, the power grid communication network has gradually increased the amount of data transmission and dynamic scheduling between communication devices with the development of emerging communication technologies such as 5G. Existing communication resource scheduling methods usually adopt fixed bandwidth allocation strategies, which makes it difficult to dynamically adjust bandwidth, frequency and latency in complex environments, resulting in communication bottlenecks and data transmission delays.

[0004] Prior art document 1 (publication number CN113781004A, invention name "A method and system for intelligent dispatching of power grids") provides a method for intelligent dispatching of power grids, including obtaining field data of power grids; using a power grid dispatching knowledge graph pre-stored with multiple dispatching instructions to detect the field data, and when it is detected that the field data satisfies the condition that at least one dispatching instruction in the power grid dispatching knowledge graph is triggered, the triggered dispatching instructions are output to the corresponding preset power dispatching execution mechanism in the power grid for automatic execution, so as to realize automatic dispatching of power grids. The present invention also provides a system for intelligent dispatching of power grids. The implementation of the present invention can realize automatic dispatching of power grids based on knowledge graphs, thereby overcoming the defects of traditional manual dispatching methods and improving efficiency.

[0005] Prior art document 2 (publication number CN116826961A, invention name "Power Grid Intelligent Dispatching and Operation and Maintenance System, Method and Storage Medium") provides a power grid intelligent dispatching and operation and maintenance system, method and storage medium, belonging to the field of power grid dispatching and operation and maintenance. The method includes obtaining data information of multiple subsystems corresponding to the main system; the present invention obtains data information of multiple subsystems, and constructs a knowledge graph based on the data information, constructs entity connections with all emergency plans of intelligent information dispatching based on the knowledge graph, sets alarm thresholds for all loads in multiple subsystems, monitors alarm information of all loads, and obtains the optimal fault solution based on the alarm information; the main system is used to collect and manage data information of multiple subsystems, which can effectively improve the efficiency and accuracy of load operation and maintenance, and the optimal fault solution for alarm information is matched in the knowledge graph to improve the timeliness and accuracy of system emergency processing, and effectively ensure the reliability and fault processing efficiency of the power grid system.

[0006] However, the high reliance of prior art document 1 on preset dispatching instructions and knowledge graphs makes the system lack flexibility and adaptability when facing dynamic changes or sudden failures. Once the field data exceeds the preset range or anomalies are not covered by the knowledge graph, the system may not be able to make effective decisions, resulting in dispatch failure. In addition, the construction and maintenance of the knowledge graph are complex. With the changes in the grid structure and dispatching rules, the update lag of the knowledge graph may cause the dispatching instructions to not match the actual situation on site, thereby affecting the accuracy and real-time performance of the dispatch.

[0007] The shortcomings of the prior art document 2 include: first, the system's high dependence on the data of each subsystem may affect the overall decision-making accuracy when the data is incomplete or not updated in time; second, the construction and maintenance of the knowledge graph are relatively complex. When the power grid structure changes dynamically, the graph update may lag behind, resulting in poor system adaptability; at the same time, the system only relies on the preset knowledge graph solution to handle faults, and lacks a flexible response strategy for complex linkage faults; in addition, the use of fixed alarm thresholds for load monitoring does not take into account the dynamic changes of the load, which may lead to false alarms or missed alarms; finally, since the knowledge graph matching process is relatively time-consuming, the system's real-time response speed and processing efficiency in emergency processing may be insufficient, reducing its application effect in high-timeliness power grid faults. Summary of the invention

[0008] The present invention aims to solve the problems such as delayed response of protection strategies, inaccurate fault handling and unbalanced allocation of communication resources caused by dynamic changes in the topology structure in the existing power grid system, and proposes a power grid adaptive scheduling method and system based on fault prediction and resource coordination. Through adaptive protection and communication scheduling management when the power grid topology structure changes dynamically and multi-node communication needs are met, it is possible to realize dynamic adjustment of power system protection strategies and real-time optimal allocation of communication resources in a complex power grid environment, thereby improving the operation efficiency and safety of the power grid.

[0009] The present invention realizes multi-dimensional real-time prediction of the power grid state through multi-dimensional feature fusion and time series analysis, so as to accurately evaluate the potential fault risks and load fluctuation trends. On this basis, the protection strategy parameters of each node of the power grid are dynamically adjusted before the fault occurs, which significantly improves the accuracy and response speed of fault handling. At the same time, according to the real-time communication requirements of each node of the power grid, the allocation and scheduling of communication resources are optimized to ensure the reliability and effectiveness of key data transmission. In addition, the present invention realizes the coordination between power protection and communication resource management through a global optimization algorithm, so as to find the best balance point between power system protection and communication scheduling, thereby realizing the optimal configuration and efficient utilization of the overall system resources. This intelligent adaptive scheduling method effectively improves the reliability and stability of power grid operation, and provides an innovative solution for the safe and efficient management of modern power systems.

[0010] The present invention adopts the following technical solution.

[0011] A first aspect of the present invention provides a power grid adaptive scheduling method based on fault prediction and resource coordination, comprising:

[0012] Collecting multi-source state data of the power grid, the multi-source state data including power parameters, environmental parameters and power grid communication parameters, preprocessing the collected multi-source state data of the power grid, performing linear feature fusion on the preprocessed data, inputting the fused data into a multi-dimensional fault prediction model, and outputting a fault prediction result;

[0013] Dynamically adjust the power grid protection strategy based on the fault prediction results output by the multi-dimensional fault prediction model, including adjustment of protection parameters, reallocation of protection priorities and optimization of fault response time;

[0014] Adaptively allocate communication resources based on predicted communication needs and communication priorities;

[0015] By integrating power grid protection strategy and communication resource allocation, a global dispatching objective optimization function is established, and cross-domain dispatching of power protection and communication resource management is achieved based on the solution of the global dispatching objective optimization function.

[0016] Optionally, a multi-dimensional fault prediction model is obtained by combining long short-term memory network and conditional random field for training, and the multi-dimensional fault prediction model includes:

[0017] LSTM layer, used to extract time series features from power grid status data:

[0018] h t =σ(W x ·x t +W h ·h t-1 +b h )

[0019] Among them, h t represents the hidden layer state at time t, x t is the input grid status data, W x is the input weight matrix, which is used to convert x t Mapped to the feature space of the hidden layer, W h is the hidden state weight matrix, which is used to transform the hidden state h at time t-1 t-1 The hidden state when mapped to t, b h is the hidden layer bias term, which is used to provide an offset for the linear change of the hidden layer, and σ represents the sigmoid activation function;

[0020] The CRF layer is used to model the correlation between the time series features extracted by the LSTM layer according to the following formula and output the prediction results:

[0021]

[0022]

[0023] ψ(y t ,y t-1 ,h t )=tanh(w1·y t +w2·y t-1 +w3·h t +b)

[0024] Among them, P(y|h) represents the probability of the CRF layer output sequence y when the input sequence is h, that is, the hidden layer output is h, and T is the time series feature h t The length of , w1, w2 and w3 are weight parameters, y t ,y t-1 are the outputs of the CRF layer at time t and t-1 respectively, and b is the output bias term.

[0025] Optionally, the fault prediction result includes at least one of the following: a fault type, a node where the fault occurs, and a fault risk level.

[0026] Optionally, the protection strategy adjustment algorithm based on prediction drive dynamically adjusts the power grid protection strategy according to the fault prediction result output by the multi-dimensional fault prediction model, including:

[0027] According to the fault risk level in the prediction results and the node where the fault occurs, the importance index of each node is calculated according to the following formula:

[0028] NII i =λ1·R i +λ2·L i +λ3·T i

[0029] Among them, R i is the failure risk level of the node, L i is the position weight of the node in the power grid topology, T i is the load type weight of the node, λ1, λ2, λ3 are weight factors;

[0030] Based on the prediction-driven protection strategy adjustment algorithm, the power grid protection strategy is adjusted according to the fault risk level and the importance index of each node.

[0031] Optionally, a protection strategy adjustment algorithm driven by prediction is used to adjust the power grid protection strategy according to the fault risk level and the importance index of each node, including:

[0032] The protection parameters include the triggering threshold of the protection action. The triggering threshold of the protection action is dynamically adjusted according to the node importance index according to the following formula:

[0033]

[0034] Among them, V b is the basic protection threshold, NII i is the importance index of the i-th node, N is the number of nodes in the power grid, NII j is the importance index of the jth node; and / or,

[0035] The protection priority of the nodes is reallocated according to the node importance and fault risk level according to the following formula:

[0036] φ i =α×NII i +β×R i

[0037] Among them, φ i is the protection priority of the ith node, NII i is the importance index of the i-th node, R iis the fault risk level of the i-th node, the higher the protection priority, the higher the response priority; α and β are the weights of the importance index and risk level, respectively, α+β=1; and / or,

[0038] The node failure response time is optimized according to the failure risk level using the following formula:

[0039]

[0040] Among them, Time i is the failure response time of the i-th node, Time base is the basic response delay, R i is the risk level of the ith node, γ is a positive constant, and is the fault response time adjustment coefficient.

[0041] Optionally, adaptively allocating communication resources according to predicted communication demands and communication priorities based on a context-aware adaptive scheduling algorithm includes:

[0042] The following communication traffic time series prediction model based on time series analysis is constructed to make real-time predictions on the communication traffic of different nodes and obtain the communication traffic prediction value:

[0043]

[0044] in, is the predicted value of communication traffic in the future period, T(tk) is the historical communication traffic data, ω k is the regression weight coefficient of the kth time step, K is the length of the sliding window, which indicates the number of time steps of historical data used for prediction;

[0045] According to the predicted value of communication traffic, the following context-aware adaptive scheduling algorithm is used to dynamically adjust the communication resource allocation strategy:

[0046]

[0047] Among them, B i is the allocated bandwidth of the node, B t is the total bandwidth resource, Q i is the communication priority of node i, θ i is the context requirement coefficient of node i, N is the number of nodes, The communication traffic forecast value for the future period.

[0048] Optionally, a global dispatch target optimization function is established by integrating power grid protection strategy and communication resource allocation, and cross-domain dispatch of power protection and communication resource management is realized based on the solution result of the global dispatch target optimization function, including:

[0049] Monitor the status, load distribution and communication resource utilization of each node in the power grid to form a global status graph;

[0050] Based on the global state diagram, the power grid protection strategy and communication resource allocation are scheduled based on a multi-objective optimization algorithm. The global scheduling objective optimization function of the multi-objective optimization algorithm is:

[0051]

[0052] Among them, T r is the fault response time of the system, η c is the communication resource utilization, L i is the load of node i, is the global average load;

[0053] According to the dispatch results, the power grid protection strategy adjustment and communication resource switching instructions are issued to the corresponding power grid equipment.

[0054] A second aspect of the present invention provides a power grid adaptive dispatching system, the system comprising a multi-dimensional fault prediction module, which is used to pre-process the collected power grid multi-source state data, perform linear feature fusion on the pre-processed data, input the fused data into a multi-dimensional fault prediction model, and output a fault prediction result, wherein the multi-source state data comprises power parameters, environmental parameters, and power grid communication parameters;

[0055] A dynamic protection strategy management module is used to dynamically adjust the power grid protection strategy according to the fault prediction results obtained by the multi-dimensional fault prediction module, including the adjustment of protection parameters, the reallocation of protection priorities and the optimization of fault response time;

[0056] An adaptive communication resource management module, used for adaptively allocating communication resources according to predicted communication needs and communication priorities;

[0057] The global resource collaborative scheduling module is used to establish a global scheduling target optimization function based on the comprehensive power grid protection strategy and communication resource allocation, and realize cross-domain scheduling of power protection and communication resource management based on the solution of the global scheduling target optimization function.

[0058] Optionally, in the multidimensional fault prediction module, a multidimensional fault prediction model is obtained by combining long short-term memory network and conditional random field for training, and the multidimensional fault prediction module includes:

[0059] LSTM layer, used to extract time series features from power grid status data:

[0060] h t =σ(W x ·x t +W h ·h t-1+b h )

[0061] Among them, h t represents the hidden layer state at time t, x t is the input grid status data, W x is the input weight matrix, which is used to convert x t Mapped to the feature space of the hidden layer, W h is the hidden state weight matrix, which is used to transform the hidden state h at time t-1 t-1 The hidden state when mapped to t, b h is the hidden layer bias term, which is used to provide an offset for the linear change of the hidden layer, and σ represents the sigmoid activation function;

[0062] The CRF layer is used to model the correlation between the time series features extracted by the LSTM layer according to the following formula and output the prediction results:

[0063]

[0064]

[0065] ψ(y t ,y t-1 ,h t )=tanh(w1·y t +w2·y t-1 +w3·h t +b)

[0066] Among them, P(y|h) represents the probability of the CRF layer output sequence y when the input sequence is h, that is, the hidden layer output is h, and T is the time series feature h t The length of , w1, w2 and w3 are weight parameters, y t ,y t-1 are the outputs of the CRF layer at time t and t-1 respectively, and b is the output bias term.

[0067] Optionally, the dynamic protection strategy management module dynamically adjusts the power grid protection strategy based on the prediction-driven protection strategy adjustment algorithm according to the fault prediction result obtained by the multi-dimensional fault prediction module, including:

[0068] The dynamic protection strategy management module calculates the importance index of each node according to the fault risk level and the node where the fault occurs in the prediction result according to the following formula:

[0069] NII i =λ1·R i +λ2·L i +λ3·T i

[0070] Among them, Ri is the failure risk level of the node, L i is the position weight of the node in the power grid topology, T i is the load type weight of the node, λ1, λ2, λ3 are weight factors;

[0071] The dynamic protection strategy management module is based on a prediction-driven protection strategy adjustment algorithm to adjust the power grid protection strategy according to the fault risk level and the importance index of each node.

[0072] The dynamic protection strategy management module is based on a prediction-driven protection strategy adjustment algorithm to adjust the grid protection strategy according to the fault risk level and the importance index of each node, including:

[0073] The protection parameters include the triggering threshold of the protection action. The triggering threshold of the protection action is dynamically adjusted according to the node importance index according to the following formula:

[0074]

[0075] Among them, V b is the basic protection threshold, NII i is the importance index of the i-th node, N is the number of nodes in the power grid, NII j is the importance index of the jth node; and / or,

[0076] The protection priority of the nodes is reallocated according to the node importance and fault risk level according to the following formula:

[0077] φ i =α×NII i +β×R i

[0078] Among them, φ i is the protection priority of the ith node, NII i is the importance index of the i-th node, R i is the fault risk level of the i-th node, the higher the protection priority, the higher the response priority; α and β are the weights of the importance index and risk level, respectively, α+β=1; and / or,

[0079] The node failure response time is optimized according to the failure risk level using the following formula:

[0080]

[0081] Among them, Time i is the failure response time of the i-th node, Time base is the basic response delay, R iis the risk level of the ith node, γ is a positive constant, and is the fault response time adjustment coefficient.

[0082] Optionally, the adaptive communication resource management module adaptively allocates communication resources according to predicted communication demands and communication priorities based on a context-aware adaptive scheduling algorithm, including:

[0083] The following communication traffic time series prediction model based on time series analysis is constructed to make real-time predictions on the communication traffic of different nodes and obtain the communication traffic prediction value:

[0084]

[0085] in, is the predicted value of communication traffic in the future period, T(tk) is the historical communication traffic data, ω k is the regression weight coefficient of the kth time step, K is the length of the sliding window, which indicates the number of time steps of historical data used for prediction;

[0086] According to the predicted value of communication traffic, the following context-aware adaptive scheduling algorithm is used to dynamically adjust the communication resource allocation strategy:

[0087]

[0088] Among them, B i is the allocated bandwidth of the node, B t is the total bandwidth resource, Q i is the communication priority of node i, θ i is the context requirement coefficient of node i, N is the number of nodes, The communication traffic forecast value for the future period.

[0089] Optionally, a global resource collaborative scheduling module is used to integrate power grid protection strategy and communication resource allocation, establish a global scheduling target optimization function, and realize cross-domain scheduling of power protection and communication resource management based on the solution result of the global scheduling target optimization function, including:

[0090] The global resource collaborative scheduling module monitors the status, load distribution and communication resource utilization of each node in the power grid to form a global status graph;

[0091] The global resource collaborative scheduling module schedules the power grid protection strategy and communication resource allocation based on the global state diagram and the multi-objective optimization algorithm. The global scheduling objective optimization function of the multi-objective optimization algorithm is:

[0092]

[0093] Among them, T r is the fault response time of the system, η cis the communication resource utilization, L i is the load of node i, is the global average load;

[0094] The global resource collaborative scheduling module issues power grid protection strategy adjustment and communication resource switching instructions to the corresponding power grid equipment based on the scheduling results.

[0095] The third aspect of the present invention provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the computer program is loaded into the processor, the method for implementing the above-mentioned power grid adaptive scheduling method based on fault prediction and resource coordination is implemented.

[0096] A fourth aspect of the present invention provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the above-mentioned method for adaptive power scheduling based on fault prediction and resource coordination.

[0097] Compared with the prior art, the beneficial effects of the present invention include at least:

[0098] By introducing a multidimensional fault prediction model based on the combination of long short-term memory network (LSTM) and conditional random field (CRF), accurate modeling of power grid status and early prediction of fault risks are achieved, effectively improving the accuracy of fault prediction; on this basis, the proposed prediction-driven dynamic protection strategy adaptively adjusts the protection threshold and response delay according to the fault risk level and node importance, thereby significantly shortening the fault response time and preventing the fault from spreading; for communication resource management, a context-aware adaptive scheduling algorithm is introduced, which dynamically allocates bandwidth and delay resources by real-time monitoring of the communication status and task requirements of each node, improves the utilization of communication resources, and ensures the priority transmission of critical mission data; in addition, the global resource collaborative scheduling mechanism based on the multi-objective optimization algorithm can uniformly dispatch power protection strategies and communication resource allocation across the entire network, forming a cross-domain collaborative optimization strategy, thereby improving the overall operation efficiency and safety of the system, and enhancing the adaptability and global stability in complex environments. BRIEF DESCRIPTION OF THE DRAWINGS

[0099] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for describing the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative labor. Among them:

[0100] Figure 1A schematic flow chart of a power grid adaptive scheduling method based on fault prediction and resource coordination provided by an embodiment of the present invention;

[0101] Figure 2 A schematic diagram of the structure of a power grid adaptive dispatching system provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0102] In order to make the purpose, technical scheme and advantages of the present invention clearer, the technical scheme of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. The embodiments described in this application are only embodiments of a part of the present invention, rather than all embodiments. Based on the spirit of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work belong to the protection scope of the present invention.

[0103] Combination Figure 1 As shown, embodiment 1 of the present invention provides a method for adaptive scheduling of a power grid based on fault prediction and resource coordination, comprising the following steps:

[0104] S1. Collect multi-source status data of the power grid, wherein the multi-source status data includes power parameters, environmental parameters and power grid communication parameters, pre-process the collected multi-source status data of the power grid, perform linear feature fusion on the pre-processed data, input the fused data into a multi-dimensional fault prediction model, and output a fault prediction result.

[0105] S1 specifically includes:

[0106] S1.1. Collect multi-source state data of the power grid, pre-process the collected multi-source state data of the power grid, and perform linear feature fusion on the pre-processed data.

[0107] Specifically, multi-source grid status data is collected from multiple nodes in the grid (such as backbone nodes, key nodes, and edge nodes), and the multi-source grid status data includes: power parameters, environmental parameters, and grid communication parameters. The power parameters include current, voltage, and load power; environmental parameters include meteorological conditions such as temperature, humidity, and wind speed; and grid communication parameters include data transmission rate, network delay, and packet loss rate between nodes.

[0108] Optionally, preprocessing of the collected various power grid data includes: denoising, normalization and anomaly detection of multi-source power grid status data, so as to ensure the consistency and accuracy of the data, provide data support for the construction and training of subsequent fault prediction models, and make its training results more accurate.

[0109] Optionally, the preprocessed data is subjected to linear feature fusion, including:

[0110] The multi-source power grid state data is reduced in dimension using principal component analysis, and the feature vector is expressed using a linear feature fusion formula:

[0111] F=α·P+β·E+γ·C (1)

[0112] Among them, P, E, and C are the characteristic vectors of power parameters, environmental parameters, and power grid communication parameters, respectively; α, β, and γ represent the weight factors of the characteristic vectors of power parameters, environmental parameters, and power grid communication parameters, respectively.

[0113] It should be noted that the embodiments of the invention do not limit the specific values ​​of α, β, and γ. Those skilled in the art may adjust the specific values ​​of α, β, and γ according to the importance of power parameters, environmental parameters, and grid communication parameters in actual applications.

[0114] In this embodiment, the multi-source power grid state data is subjected to dimensionality reduction processing through principal component analysis, which can reduce the dimension of the data, reduce the computational complexity and improve the computational speed, and retain the variability of the original data. It can also improve the performance of the subsequent fault prediction model, thereby further improving the accuracy and response speed of the entire dispatching system fault processing.

[0115] S1.2. Construct a multi-dimensional fault prediction model.

[0116] Specifically, a multi-dimensional fault prediction model is obtained by combining LSTM (Long Short-Term Memory) and CRF (Conditional Random Field) for training. The multi-dimensional fault prediction model includes:

[0117] LSTM layer, used to extract time series features from power grid status data:

[0118] h t =σ(W x ·x t +W h ·h t-1 +b h ) (2)

[0119] Among them, h t represents the hidden layer state at time t, x t is the input grid status data, W x is the input weight matrix, which is used to convert x t Mapped to the feature space of the hidden layer, W h is the hidden state weight matrix, which is used to transform the hidden state h at time t-1 t-1 The hidden state when mapped to t, b his the hidden layer bias term, which is used to provide an offset for the linear change of the hidden layer, and σ represents the sigmoid activation function.

[0120] The CRF layer is used to model the correlation between the time series features extracted by the LSTM layer according to the following formula and output the prediction results:

[0121]

[0122] ψ(y t ,y t-1 ,h t )=tanh(w1·y t +w2·y t-1 +w3·h t +b)

[0123] Among them, P(y|h) represents the probability of the CRF layer output sequence y when the input sequence is h, that is, the hidden layer output is h, and T is the time series feature h t The length of , w1, w2 and w3 are weight parameters, y t ,y t-1 are the outputs of the CRF layer at time t and t-1 respectively, and b is the output bias term.

[0124] Optionally, the prediction result includes at least one of the following: a fault type, a node where the fault occurs, a time period in which the fault occurs, and a fault risk level.

[0125] Fault types include overcurrent, overvoltage, short circuit, etc., and fault risk levels include high, medium and low.

[0126] This multi-dimensional fault prediction model predicts the state of the power grid at different time scales (short-term, medium-term, and long-term), and can identify the type, location, and severity of future faults in advance. Through the multi-dimensional fault prediction model that combines long short-term memory networks and conditional random fields, accurate modeling of the power grid state and early prediction of fault risks are achieved, effectively improving the accuracy of fault prediction.

[0127] Furthermore, a real-time warning signal is generated according to the risk level of the fault to warn of the fault.

[0128] S2. Dynamically adjust the power grid protection strategy according to the fault prediction results output by the multi-dimensional fault prediction model, including adjustment of protection parameters, reallocation of protection priorities and optimization of fault response time.

[0129] Optionally, the prediction-driven protection strategy adjustment algorithm in S2 dynamically adjusts the power grid protection strategy according to the fault prediction result output by the multi-dimensional fault prediction model, specifically including:

[0130] S2.1. According to the fault risk level in the prediction results and the node where the fault occurs, the importance index of each node is calculated according to the following formula:

[0131] NII i =λ1·R i +λ2·L i +λ3·T i (4)

[0132] Among them, R i is the failure risk level of the node, L i is the position weight of the node in the power grid topology, T i is the load type weight of the node, and λ1, λ2, and λ3 are weight factors.

[0133] It should be noted that the principle of setting the location weight is: whether the topological position of the node in the power grid has a key impact on the reliability and stability of the entire power grid; for example, the core node connects multiple important lines or other key nodes, and the weight is higher, while the edge node is at the end of the power grid and has less impact on other nodes, so the weight is lower. The principle of setting the load type weight is: different load types have different importance to the system. For example, industrial loads may have higher requirements for power quality, and residential loads are more sensitive to continuous power supply.

[0134] It should be noted that the embodiments of the present invention do not limit the specific values ​​of λ1, λ2, and λ3, and those skilled in the art may set the specific values ​​of λ1, λ2, and λ3 according to actual requirements.

[0135] Specifically, the node where the fault occurs includes the location of the node in the power grid topology (such as a trunk node, a branch node, a key load node, etc.).

[0136] S2.2. The prediction-driven protection strategy adjustment algorithm adjusts the grid protection strategy according to the risk level of the fault and the importance index of each node.

[0137] S2.2 specifically includes:

[0138] S2.21. Adjust the protection parameters, which include the triggering threshold of the protection action. The triggering threshold of the protection action is dynamically adjusted according to the node importance index:

[0139]

[0140] Among them, V b is the basic protection threshold, NII i is the importance index of the i-th node, N is the number of nodes in the power grid, NII j is the importance index of the jth node.

[0141] S2.22. Adjust the protection priority, dynamically assign response priority according to node importance and risk level, increase the protection priority of high-risk nodes, and reduce the priority of low-risk nodes.

[0142] Specifically, the protection priority is adjusted according to the following formula:

[0143] φ i =α×NII i +β×R i (6)

[0144] Among them, φ i is the protection priority of the ith node, NII i is the importance index of the i-th node, R i is the risk level of the i-th node. The higher the protection priority, the higher the response priority. α and β are the weights of the importance index and risk level respectively. α+β=1.

[0145] The protection priority is used to represent the urgency of the node task. The higher the protection priority, the more urgent the node task.

[0146] It should be noted that the embodiments of the present invention do not limit the specific values ​​of α and β, and those skilled in the art can set the specific values ​​of α and β according to actual needs.

[0147] S2.23. Adjust the fault response time and adopt different protection delay strategies. For high-risk nodes, a shorter protection response delay is adopted, and for low-risk nodes, a delayed processing strategy is adopted to ensure the optimal allocation of global resources.

[0148] Specifically, the fault response time is adjusted according to the following formula:

[0149]

[0150] Among them, Time i is the failure response time of the i-th node, Time base is the basic response delay, R i is the risk level of the ith node, γ is a positive constant, and is the fault response time adjustment coefficient.

[0151] In this embodiment, by introducing a multidimensional feature fusion model based on the combination of long short-term memory networks and conditional random fields, accurate modeling of the power grid state and early prediction of fault risks are achieved, effectively improving the accuracy of fault prediction; on this basis, a prediction-driven protection strategy adjustment algorithm is proposed to adaptively adjust the protection threshold and response delay according to the fault risk level and node importance, thereby significantly shortening the fault response time and preventing the fault from spreading.

[0152] S2 also includes: sending the adjusted protection strategy to the power grid equipment through the edge computing node or the central control node, so that the power grid equipment performs corresponding tasks according to the real-time control commands in the protection strategy and the adjusted protection parameters.

[0153] Specifically, the real-time control commands include load switching and backup power startup.

[0154] S3. Adaptively allocate communication resources based on predicted communication needs and communication priorities.

[0155] Optionally, in S3, adaptively allocating communication resources according to predicted communication demand and communication priority based on a context-aware adaptive scheduling algorithm specifically includes:

[0156] S3.1. Construct the following communication traffic time series prediction model based on time series analysis to make real-time predictions on the communication traffic of different nodes and obtain the communication traffic prediction value, which is the communication demand of the node:

[0157]

[0158] in, is the predicted value of communication traffic in the future period, T(tk) is the historical communication traffic data, ω k is the regression weight coefficient of the kth time step, K is the length of the sliding window, and represents the number of time steps of historical data used for prediction.

[0159] S3.2, according to the communication traffic prediction value and communication priority, use the context-aware adaptive scheduling algorithm to dynamically adjust the communication resource allocation strategy:

[0160]

[0161] Among them, B i is the allocated bandwidth of the node, B t is the total bandwidth resource, Q i is the communication priority of node i, θ i is the context requirement coefficient of node i, θ i The value is 0.18, N is the number of nodes, The communication traffic forecast value for the future period.

[0162] Communication priority is used to characterize the priority of node resource allocation. Nodes with high communication priority are given priority in allocating communication resources.

[0163] In this embodiment, by real-time monitoring of the communication status and task requirements of each node, bandwidth and latency resources are dynamically allocated to improve communication resource utilization and ensure priority transmission of mission-critical data.

[0164] S3 also includes the ability to flexibly allocate communication resources between different nodes through 5G network slicing or software-defined networking (SDN) technology, ensuring that each node can obtain resources that match its mission requirements.

[0165] S4. Integrate power grid protection strategy and communication resource allocation, establish a global dispatch target optimization function, and realize cross-domain dispatch of power protection and communication resource management based on the solution of the global dispatch target optimization function.

[0166] S4 specifically includes:

[0167] S4.1. Real-time monitoring of the status, load distribution, and communication resource utilization data of each node in the power grid to form a global status diagram.

[0168] S4.2. Based on the global state diagram, a multi-objective optimization algorithm based on constraints is used to comprehensively dispatch the power protection strategy and communication resource allocation. The multi-objective optimization function of the comprehensive dispatch is:

[0169]

[0170] Among them, T r is the fault response time of the system, η c is the communication resource utilization, L i is the load of node i, is the global average load. The system fault response time is the average of all node fault response times.

[0171] Specifically, the communication resource utilization is:

[0172]

[0173] Among them, B i is the allocated bandwidth of node i, B t is the total bandwidth resource, and N is the number of nodes.

[0174] Multi-objective optimization algorithms include the MODA algorithm.

[0175] S4.3. According to the dispatching results, protection strategy adjustment and communication resource switching instructions are sent to the corresponding power grid equipment in real time.

[0176] In this embodiment, cross-domain collaborative optimization of power protection and communication management is achieved through global resource collaborative scheduling based on a multi-objective optimization algorithm, thereby improving the overall operating efficiency and security of the entire network and enhancing the adaptability and global stability in complex environments.

[0177] Combination Figure 2As shown, embodiment 2 of the present invention provides a power grid intelligent adaptive scheduling system, including: a multi-dimensional fault prediction module, a dynamic protection strategy management module, an adaptive communication resource management module and a global resource collaborative scheduling module.

[0178] A multi-dimensional fault prediction module is used to pre-process the collected multi-source state data of the power grid, perform linear feature fusion on the pre-processed data, input the fused data into a multi-dimensional fault prediction model, and output a fault prediction result. The multi-source state data includes power parameters, environmental parameters, and power grid communication parameters;

[0179] A dynamic protection strategy management module is used to dynamically adjust the power grid protection strategy according to the fault prediction results obtained by the multi-dimensional fault prediction module, including the adjustment of protection parameters, the reallocation of protection priorities and the optimization of fault response time;

[0180] An adaptive communication resource management module, used for adaptively allocating communication resources according to predicted communication needs and communication priorities;

[0181] The global resource collaborative scheduling module is used to establish a global scheduling target optimization function based on the comprehensive power grid protection strategy and communication resource allocation, and realize cross-domain scheduling of power protection and communication resource management based on the solution of the global scheduling target optimization function.

[0182] Regarding the system in the above embodiment, the specific manner in which each unit performs operations has been described in detail in the embodiment of the method, and will not be elaborated here.

[0183] Embodiment 3 of the present invention provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the computer program is loaded into the processor, the method for adaptively scheduling a power grid based on fault prediction and resource coordination as described in Embodiment 1 is implemented.

[0184] Embodiment 4 of the present invention provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the method for adaptive scheduling of a power grid based on fault prediction and resource coordination according to embodiment 1 is implemented.

[0185] Compared with the prior art, the beneficial effects of the present invention include at least:

[0186] By introducing a multidimensional fault prediction model based on the combination of long short-term memory network (LSTM) and conditional random field (CRF), accurate modeling of power grid status and early prediction of fault risks are achieved, effectively improving the accuracy of fault prediction; on this basis, the proposed prediction-driven dynamic protection strategy adaptively adjusts the protection threshold and response delay according to the fault risk level and node importance, thereby significantly shortening the fault response time and preventing the fault from spreading; for communication resource management, a context-aware adaptive scheduling algorithm is introduced, which dynamically allocates bandwidth and delay resources by real-time monitoring of the communication status and task requirements of each node, improves the utilization of communication resources, and ensures the priority transmission of critical mission data; in addition, the global resource collaborative scheduling mechanism based on the multi-objective optimization algorithm can uniformly dispatch power protection strategies and communication resource allocation across the entire network, forming a cross-domain collaborative optimization strategy, thereby improving the overall operation efficiency and safety of the system, and enhancing the adaptability and global stability in complex environments.

[0187] It should be understood that the order of execution of the steps in the above embodiment does not necessarily mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiment of the present invention.

[0188] The present disclosure may be a system, a method and / or a computer program product. The computer program product may include a computer-readable storage medium carrying computer-readable program instructions for causing a processor to implement various aspects of the present disclosure.

[0189] A computer-readable storage medium may be a tangible device that can hold and store instructions used by an instruction execution device. A computer-readable storage medium may be, for example, but not limited to, an electrical storage device, a magnetic storage device, an optical storage device, an electromagnetic storage device, a semiconductor storage device, or any suitable combination of the foregoing. More specific examples of computer-readable storage media (a non-exhaustive list) include: a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), a static random access memory (SRAM), a portable compact disk read-only memory (CD-ROM), a digital versatile disk (DVD), a memory stick, a floppy disk, a mechanical encoding device, such as a punch card or a raised structure in a groove on which instructions are stored, and any suitable combination of the foregoing. As used herein, a computer-readable storage medium is not to be interpreted as a transient signal per se, such as a radio wave or other freely propagating electromagnetic wave, an electromagnetic wave propagating through a waveguide or other transmission medium (e.g., a light pulse through a fiber optic cable), or an electrical signal transmitted through a wire.

[0190] The computer-readable program instructions described herein can be downloaded from a computer-readable storage medium to each computing / processing device, or downloaded to an external computer or external storage device via a network, such as the Internet, a local area network, a wide area network, and / or a wireless network. The network can include copper transmission cables, optical fiber transmissions, wireless transmissions, routers, firewalls, switches, gateway computers, and / or edge servers. The network adapter card or network interface in each computing / processing device receives the computer-readable program instructions from the network and forwards the computer-readable program instructions for storage in the computer-readable storage medium in each computing / processing device.

[0191] The computer program instructions for performing the operation of the present disclosure may be assembly instructions, instruction set architecture (ISA) instructions, machine instructions, machine-related instructions, microcode, firmware instructions, state setting data, or source code or object code written in any combination of one or more programming languages, including object-oriented programming languages, such as Smalltalk, C++, etc., and conventional procedural programming languages, such as "C" language or similar programming languages. Computer-readable program instructions may be executed completely on a user's computer, partially on a user's computer, as an independent software package, partially on a user's computer, partially on a remote computer, or completely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer via any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., using an Internet service provider to connect via the Internet). In some embodiments, an electronic circuit, such as a programmable logic circuit, a field programmable gate array (FPGA), or a programmable logic array (PLA), may be customized by utilizing the state information of the computer-readable program instructions, and the electronic circuit may execute the computer-readable program instructions, thereby realizing various aspects of the present disclosure.

[0192] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, ordinary technicians in the relevant field should understand that the specific implementation methods of the present invention can still be modified or replaced by equivalents, and any modifications or equivalent replacements that do not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A power grid adaptive dispatching method based on fault prediction and resource coordination, characterized in that: include: Collecting multi-source state data of the power grid, the multi-source state data including power parameters, environmental parameters and power grid communication parameters, preprocessing the collected multi-source state data of the power grid, performing linear feature fusion on the preprocessed data, inputting the fused data into a multi-dimensional fault prediction model, and outputting a fault prediction result; Dynamically adjust the power grid protection strategy based on the fault prediction results output by the multi-dimensional fault prediction model, including adjustment of protection parameters, reallocation of protection priorities and optimization of fault response time; Adaptively allocate communication resources based on predicted communication needs and communication priorities; By integrating power grid protection strategy and communication resource allocation, a global dispatching objective optimization function is established, and cross-domain dispatching of power protection and communication resource management is achieved based on the solution of the global dispatching objective optimization function.

2. The method for adaptive dispatching of power grid based on fault prediction and resource coordination according to claim 1, characterized in that: A multi-dimensional fault prediction model is obtained by combining long short-term memory network and conditional random field training. The multi-dimensional fault prediction model includes: LSTM layer, used to extract time series features from power grid status data: h t =σ(W x ·x t +W h ·h t-1 +b h ) Among them, h t represents the hidden layer state at time t, x t is the input grid status data, W x is the input weight matrix, which is used to convert x t Mapped to the feature space of the hidden layer, W h is the hidden state weight matrix, which is used to transform the hidden state h at time t-1 t-1 The hidden state when mapped to t, b h is the hidden layer bias term, which is used to provide an offset for the linear change of the hidden layer, and σ represents the sigmoid activation function; The CRF layer is used to model the correlation between the time series features extracted by the LSTM layer according to the following formula and output the prediction results: ψ(y t ,y t-1 ,h t )=tanh(w1·y t +w2·y t-1 +w3·h t +b) Among them, P(y|h) represents the probability of the CRF layer output sequence y when the input sequence, that is, the hidden layer output is h, and T is the time series feature h t The length of , w1, w2 and w3 are weight parameters, y t ,y t-1 are the outputs of the CRF layer at time t and t-1 respectively, and b is the output bias term.

3. The power grid adaptive dispatching method based on fault prediction and resource coordination according to claim 1, characterized in that: The fault prediction result includes at least one of the following: the fault type, the node where the fault occurs, and the fault risk level.

4. The method for adaptive dispatching of power grid based on fault prediction and resource coordination according to claim 3, characterized in that: The prediction-driven protection strategy adjustment algorithm dynamically adjusts the power grid protection strategy according to the fault prediction results output by the multi-dimensional fault prediction model, including: According to the fault risk level in the prediction results and the node where the fault occurs, the importance index of each node is calculated according to the following formula: NII i =λ1·R i +λ2·L i +λ3·T i Among them, R i is the failure risk level of the node, L i is the position weight of the node in the power grid topology, T i is the load type weight of the node, λ1, λ2, λ3 are weight factors; Based on the prediction-driven protection strategy adjustment algorithm, the power grid protection strategy is adjusted according to the fault risk level and the importance index of each node.

5. The method for adaptive dispatching of power grid based on fault prediction and resource coordination according to claim 1, characterized in that: Based on the prediction-driven protection strategy adjustment algorithm, the power grid protection strategy is adjusted according to the fault risk level and the importance index of each node, including: The protection parameters include the triggering threshold of the protection action. The triggering threshold of the protection action is dynamically adjusted according to the node importance index according to the following formula: Among them, V b is the basic protection threshold, NII i is the importance index of the i-th node, N is the number of nodes in the power grid, NII j is the importance index of the jth node; and / or, The protection priority of the nodes is reallocated according to the node importance and fault risk level according to the following formula: f i =α×NII i +β×R i Among them, φ i is the protection priority of the ith node, NII i is the importance index of the i-th node, R i is the fault risk level of the i-th node, the higher the protection priority, the higher the response priority; α and β are the weights of the importance index and risk level, respectively, α+β=1; and / or, The node failure response time is optimized according to the failure risk level using the following formula: Among them, Time i is the failure response time of the i-th node, Time base is the basic response delay, R i is the risk level of the ith node, γ is a positive constant, and is the fault response time adjustment coefficient.

6. The method for adaptive dispatching of power grid based on fault prediction and resource coordination according to claim 1, characterized in that: The context-aware adaptive scheduling algorithm adaptively allocates communication resources according to the predicted communication demand and communication priority, including: The following communication traffic time series prediction model based on time series analysis is constructed to make real-time predictions on the communication traffic of different nodes and obtain the communication traffic prediction value: in, is the predicted value of communication traffic in the future period, T(tk) is the historical communication traffic data, ω k is the regression weight coefficient of the kth time step, K is the length of the sliding window, which indicates the number of time steps of historical data used for prediction; According to the predicted value of communication traffic, the following context-aware adaptive scheduling algorithm is used to dynamically adjust the communication resource allocation strategy: Among them, B i is the allocated bandwidth of the node, B t is the total bandwidth resource, Q i is the communication priority of node i, θ i is the context requirement coefficient of node i, N is the number of nodes, The communication traffic forecast value for the future period.

7. The method for adaptive dispatching of power grid based on fault prediction and resource coordination according to claim 1, characterized in that: Integrate power grid protection strategy and communication resource allocation, establish a global dispatch target optimization function, and realize cross-domain dispatch of power protection and communication resource management based on the solution of the global dispatch target optimization function, including: Monitor the status, load distribution and communication resource utilization of each node in the power grid to form a global status graph; Based on the global state diagram, the power grid protection strategy and communication resource allocation are scheduled based on a multi-objective optimization algorithm. The global scheduling objective optimization function of the multi-objective optimization algorithm is: Among them, T r is the fault response time of the system, η c is the communication resource utilization, L i is the load of node i, is the global average load; According to the dispatch results, the power grid protection strategy adjustment and communication resource switching instructions are issued to the corresponding power grid equipment.

8. A power grid adaptive dispatching system using the power grid adaptive dispatching method based on fault prediction and resource coordination as claimed in any one of claims 1 to 7, characterized in that: The system comprises: A multi-dimensional fault prediction module is used to pre-process the collected multi-source state data of the power grid, perform linear feature fusion on the pre-processed data, input the fused data into a multi-dimensional fault prediction model, and output a fault prediction result. The multi-source state data includes power parameters, environmental parameters, and power grid communication parameters; A dynamic protection strategy management module is used to dynamically adjust the power grid protection strategy according to the fault prediction results obtained by the multi-dimensional fault prediction module, including the adjustment of protection parameters, the reallocation of protection priorities and the optimization of fault response time; An adaptive communication resource management module, used to adaptively allocate communication resources according to predicted communication needs and communication priorities; The global resource collaborative scheduling module is used to establish a global scheduling target optimization function based on the comprehensive power grid protection strategy and communication resource allocation, and realize cross-domain scheduling of power protection and communication resource management based on the solution of the global scheduling target optimization function.

9. The power grid adaptive dispatching system according to claim 8, characterized in that: In the multidimensional fault prediction module, a multidimensional fault prediction model is obtained by combining long short-term memory network and conditional random field for training. The multidimensional fault prediction module includes: LSTM layer, used to extract time series features from power grid status data: h t =σ(W x ·x t +W h ·h t-1 +b h ) Among them, h t represents the hidden layer state at time t, x t is the input grid status data, W x is the input weight matrix, which is used to convert x t Mapped to the feature space of the hidden layer, W h is the hidden state weight matrix, which is used to transform the hidden state h at time t-1 t-1 The hidden state when mapped to t, b h is the hidden layer bias term, which is used to provide an offset for the linear change of the hidden layer, and σ represents the sigmoid activation function; The CRF layer is used to model the correlation between the time series features extracted by the LSTM layer according to the following formula and output the prediction results: ψ(y t ,y t-1 ,h t )=tanh(w1·y t +w2·y t-1 +w3·h t +b) Among them, P(y|h) represents the probability of the CRF layer output sequence y when the input sequence is the hidden layer output h, and T is the time series feature h t The length of , w1, w2 and w3 are weight parameters, y t ,y t-1 are the outputs of the CRF layer at time t and t-1 respectively, and b is the output bias term.

10. The power grid adaptive dispatching system according to claim 8, characterized in that: The dynamic protection strategy management module dynamically adjusts the power grid protection strategy based on the prediction-driven protection strategy adjustment algorithm according to the fault prediction results obtained by the multi-dimensional fault prediction module, including: The dynamic protection strategy management module calculates the importance index of each node according to the fault risk level and the node where the fault occurs in the prediction result according to the following formula: NII i =λ1·R i +λ2·L i +λ3·T i Among them, R i is the failure risk level of the node, L i is the position weight of the node in the power grid topology, T i is the load type weight of the node, λ1, λ2, λ3 are weight factors; The dynamic protection strategy management module is based on a prediction-driven protection strategy adjustment algorithm to adjust the power grid protection strategy according to the fault risk level and the importance index of each node.

11. The power grid adaptive dispatching system according to claim 8, characterized in that: The dynamic protection strategy management module is based on a prediction-driven protection strategy adjustment algorithm to adjust the grid protection strategy according to the fault risk level and the importance index of each node, including: The protection parameters include the triggering threshold of the protection action. The triggering threshold of the protection action is dynamically adjusted according to the node importance index according to the following formula: Among them, V b is the basic protection threshold, NII i is the importance index of the i-th node, N is the number of nodes in the power grid, NII j is the importance index of the jth node; and / or, The protection priority of the nodes is reallocated according to the node importance and fault risk level according to the following formula: f i =α×NII i +β×R i Among them, φ i is the protection priority of the ith node, NII i is the importance index of the i-th node, R i is the fault risk level of the i-th node, the higher the protection priority, the higher the response priority; α and β are the weights of the importance index and risk level, respectively, α+β=1; and / or, The node failure response time is optimized according to the failure risk level using the following formula: Among them, Time i is the failure response time of the i-th node, Time base is the basic response delay, R i is the risk level of the ith node, γ is a positive constant, and is the fault response time adjustment coefficient.

12. The power grid adaptive dispatching system according to claim 8, characterized in that: The adaptive communication resource management module adaptively allocates communication resources based on the predicted communication demand and communication priority based on the context-aware adaptive scheduling algorithm, including: The following communication traffic time series prediction model based on time series analysis is constructed to make real-time predictions on the communication traffic of different nodes and obtain the communication traffic prediction value: in, is the predicted value of communication traffic in the future period, T(tk) is the historical communication traffic data, ω k is the regression weight coefficient of the kth time step, K is the length of the sliding window, which indicates the number of time steps of historical data used for prediction; According to the predicted value of communication traffic, the following context-aware adaptive scheduling algorithm is used to dynamically adjust the communication resource allocation strategy: Among them, B i is the allocated bandwidth of the node, B t is the total bandwidth resource, Q i is the communication priority of node i, θ i is the context requirement coefficient of node i, N is the number of nodes, The communication traffic forecast value for the future period.

13. The power grid adaptive dispatching system according to claim 8, characterized in that: The global resource collaborative dispatch module is used to integrate power grid protection strategy and communication resource allocation, establish a global dispatch target optimization function, and realize cross-domain dispatch of power protection and communication resource management based on the solution of the global dispatch target optimization function, including: The global resource collaborative scheduling module monitors the status, load distribution and communication resource utilization of each node in the power grid to form a global status graph; The global resource collaborative scheduling module schedules the power grid protection strategy and communication resource allocation based on the global state diagram and the multi-objective optimization algorithm. The global scheduling objective optimization function of the multi-objective optimization algorithm is: Among them, T r is the fault response time of the system, η c is the communication resource utilization, L i is the load of node i, is the global average load; The global resource collaborative scheduling module issues power grid protection strategy adjustment and communication resource switching instructions to the corresponding power grid equipment based on the scheduling results.

14. An electronic device, comprising a processor and a storage medium, characterized in that: The storage medium is used to store instructions; The processor is used to operate according to the instructions to execute the steps of the power grid adaptive scheduling method based on fault prediction and resource coordination according to any one of claims 1 to 7.

15. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the steps of the power grid adaptive scheduling method based on fault prediction and resource coordination described in any one of claims 1 to 7 are implemented.

Images