Electric power system real-time simulation optimization scheduling method based on edge calculation

Abstract

The invention discloses an electric power system real-time simulation optimization scheduling method based on edge computing, and particularly relates to the technical field of edge computing management, and the method comprises the following steps: dividing an electric power system into a plurality of regional power grids according to a preset strategy, collecting the load information of each region, and carrying out the load analysis, activating a distributed computing potential evaluation mechanism after identifying abnormal load fluctuation; according to the method, resource utilization difference and task allocation imbalance analysis is carried out by obtaining regional power grid simulation cooperative task allocation information, and efficient and low-efficiency cooperative operation power grids are identified; for a low-efficiency collaborative power grid, task allocation of edge nodes and a center node is optimized, optimal scheduling of analogue simulation tasks is achieved, finally, a regional power grid is adjusted to be a high-efficiency collaborative operation power grid, and the power grid operation efficiency is improved; according to the method, tasks can be reasonably allocated according to the load and computing power of each node, overload or idle of some nodes is avoided, computing resources of edge and center nodes are fully utilized, and the simulation real-time performance and efficiency are improved.

Description

Technical Field

[0001] The present invention relates to the field of edge computing management technology, and more specifically, to a real-time simulation optimization scheduling method for a power system based on edge computing. Background Art

[0002] As the scale of power systems continues to expand, the operation and dispatch of power systems face more and more challenges. Traditional power system dispatch methods usually rely on centralized data processing modes, in which a large amount of data transmission and processing work is concentrated in the central server. This approach not only has processing delays and data transmission bottlenecks, but also easily leads to excessive concentration of computing resources in large-scale systems, increasing the computing burden and energy consumption, and limiting the real-time and flexibility of dispatch optimization.

[0003] In recent years, with the rapid development of edge computing technology, it has become a new type of distributed computing mode that can process and calculate data closer to the data source, greatly reducing data transmission delays and network bandwidth pressure. By deploying edge devices in various links of the power system, edge computing can collect real-time information such as the power system's operating status data, load demand, and equipment status, and perform local data processing and local simulation, thereby improving the dispatch response speed and optimization efficiency of the power system.

[0004] Therefore, the present invention proposes a real-time simulation optimization scheduling method for power systems based on edge computing. The real-time status of each component in the power system is simulated through edge nodes and central nodes, and the simulation tasks are optimally scheduled to achieve efficient, stable and economical operation of the simulation tasks, thereby providing a scientific basis for subsequent power scheduling in the power system. Summary of the invention

[0005] To achieve the above object, the present invention provides the following technical solutions:

[0006] The real-time simulation optimization dispatching method of the power system based on edge computing includes the following steps:

[0007] The power system is divided into regions according to a preset strategy to obtain multiple regional power grids, and then load information is collected from each regional power grid to obtain a load information set;

[0008] Perform load analysis based on the load information set to identify whether there is abnormal load fluctuation. If so, activate the distributed computing potential assessment mechanism;

[0009] After the distributed computing potential assessment mechanism is activated, the information of the regional power grid simulation collaborative task allocation is obtained to obtain the distributed collaborative information set, and then the resource utilization difference analysis of the edge nodes and the central nodes is carried out, as well as the task allocation imbalance analysis of the edge nodes and the central nodes is carried out;

[0010] Based on the results of resource utilization difference analysis and task allocation imbalance analysis, the collaborative task allocation of the regional power grid simulation is divided into efficient collaborative operation power grid and inefficient collaborative operation power grid.

[0011] For inefficient collaborative power grids, the allocation of simulated collaborative tasks to edge nodes and central nodes is optimized and scheduled, so that the regional power grid is re-divided into an efficient collaborative power grid.

[0012] In a preferred embodiment, load analysis refers to:

[0013] In a fixed time window, the load value at each time point is obtained from the load information set to obtain the load data time series, and then the absolute value of the corresponding data difference between two adjacent time points is obtained, and then the average and standard deviation of all absolute values ​​are calculated.

[0014] In a preferred embodiment, identifying whether there is abnormal load fluctuation refers to:

[0015] The average value of all absolute values ​​is compared with a preset threshold value one, and the standard deviation of all absolute values ​​is compared with a preset threshold value two. If the average value of all absolute values ​​is less than or equal to the preset threshold value one and the standard deviation of all absolute values ​​is less than or equal to the preset threshold value two, a normal signal is generated. If the average value of all absolute values ​​is less than or equal to the preset threshold value one and the standard deviation of all absolute values ​​is less than or equal to the preset threshold value two, an abnormal signal is generated. When an abnormal signal is generated, it indicates that signs of abnormal load fluctuations have been identified from the load information set of the regional power grid.

[0016] In a preferred embodiment, the result of performing resource utilization difference analysis on edge nodes and central nodes is a resource utilization difference index, and the result of performing task allocation imbalance analysis on edge nodes and central nodes is a task allocation imbalance index.

[0017] In a preferred embodiment, the logic for obtaining the resource utilization difference index is:

[0018] Get the resource utilization of each edge node and mark it as , Indicates the index number of the edge node, obtains the resource utilization of the central node and marks it as , and then by the difference value Represents the difference in resource utilization between edge node i and central node: ;

[0019] The difference value is calculated by the resource usage ratio of the edge node Perform dynamic adjustments to define the weight corresponding to edge node i for: ; N is the total number of edge nodes, represents the used resources of edge node i, where is the ratio of the total resources of edge node i to the total resources of the central node;

[0020] The adjusted difference is: ; represents the adjusted difference value;

[0021] In order to ensure that the difference value is not affected by the node scale, a nonlinear normalization method is used to obtain the processed difference value: ; represents the difference value after processing, k is a preset non-zero adjustment coefficient used to control the intensity of nonlinear normalization;

[0022] The difference in collaborative utilization between edge nodes is introduced to measure the problem of uneven resource allocation of edge nodes as a whole. The formula is: ; represents the absolute value of the difference in resource utilization between edge node i and edge node j, is the sum of the differences in resource utilization between edge nodes;

[0023] The final resource utilization difference index combines the central node difference and the edge node coordination difference and is defined as: ; is a preset non-zero adjustment factor used to weigh the impact of central node differences and edge node collaboration differences. Represents the resource utilization difference index.

[0024] In a preferred embodiment, the logic for obtaining the task allocation imbalance index is:

[0025] Get the number of tasks for edge node i , the number of real-time tasks of the central node , and the total number of tasks ,The total number of tasks is the sum of all tasks of edge and central nodes;

[0026] Task allocation ratio of edge node i for: ;

[0027] Task allocation ratio of central nodes for: ;

[0028] The ideal task allocation ratio is allocated based on the node task processing capacity:

[0029] ; ; represents the ideal task allocation ratio of edge node i, represents the ideal task allocation ratio of the central node, and They correspond to the task processing capabilities of edge node i and central node respectively, It represents the sum of the task processing capabilities of all edge nodes i and central nodes;

[0030] Calculate the fairness coefficient of task allocation for edge node i :

[0031] ;

[0032] Calculate the fairness coefficient of task allocation of central nodes :

[0033] ;

[0034] By assigning discrete Measures the degree of fluctuation of task ratios and distribution discreteness Defined as the variance of the task ratio of each node: ; represents the average value of the task allocation ratio of all edge nodes i and central nodes;

[0035] The task allocation imbalance index is generated by combining the fairness and discreteness calculation results:

[0036] ; Represents the task allocation imbalance index.

[0037] In a preferred embodiment, dividing the simulation collaborative task allocation of the regional power grid into a highly efficient collaborative operation power grid and a low-efficiency collaborative operation power grid means:

[0038] The task allocation imbalance index and resource utilization difference index of the regional power grid are obtained respectively and used as input variables together, the type corresponding to the simulation collaborative task allocation of the regional power grid is used as the output variable, the input variables are fuzzified, and the values ​​of the input variables are converted into fuzzy sets, the output variables are fuzzified, and the output variables are converted into fuzzy sets, fuzzy rules are formulated to describe the quality of collaborative operation under different data type combinations, the fuzzified input variables are inferred through fuzzy rules, and the type corresponding to the simulation collaborative task allocation of the regional power grid is determined.

[0039] In a preferred embodiment, a genetic algorithm is used to optimize the allocation of simulation collaborative tasks to edge nodes and central nodes for an inefficient collaborative power grid, so that the regional power grid is re-divided into an efficient collaborative power grid and the total execution time of the simulation collaborative tasks is minimized.

[0040] Technical effects and advantages of the present invention:

[0041] Load analysis is performed based on the load information set to identify whether there are abnormal load fluctuations. If so, the distributed computing potential assessment mechanism is activated. Activating the assessment mechanism helps to alleviate the pressure of power system simulation calculations by reallocating tasks or computing resources, and reduce the risks caused by single point failures or computing overloads. This improves the fault tolerance of the power system and enhances the reliability of the power system. By accurately identifying load fluctuations and evaluating computing potential, computing resources can be reasonably allocated when the load is too high to avoid unnecessary computing waste, thereby improving computing efficiency and achieving energy saving. At the same time, by analyzing load information and evaluating distributed computing potential, data support can be provided for the real-time response of the power grid.

[0042] Through precise task allocation strategies, the present invention can reasonably allocate tasks according to the load and computing power of each node, avoid overloading or idleness of certain nodes, and make full use of the computing resources of edge and central nodes. The optimized scheduling algorithm, namely the genetic algorithm, can ensure that the tasks are executed on the most suitable nodes, thereby reducing data transmission time and computing time, reducing the response time of the overall simulation task, and improving real-time performance and efficiency.

[0043] Edge computing distributes computing tasks to edge nodes close to data sources, which can process data generated by a large number of distributed devices in real time, reduce dependence on central servers, and improve the real-time performance of scheduling and simulation. Real-time simulation provides dynamic prediction capabilities for the operating status of the power system, helps identify possible operating risks or scheduling problems, and helps ensure the safety and economy of power grid operation.

[0044] Task allocation optimization can prevent the risk of power grid decoupling caused by overloading of some nodes, ensure balanced distribution of tasks, improve the reliability and stability of the coordinated use of edge and central nodes in the present invention, and avoid unnecessary repeated calculations and communications caused by the collapse of the simulation process, thereby saving energy consumption and optimizing the energy efficiency of the overall simulation calculation. Based on the optimized task scheduling strategy, the present invention can obtain accurate simulation results through rapid simulation, helping decision makers to make decisions and adjustments on subsequent power scheduling more quickly. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] In order to facilitate understanding by those skilled in the art, the present invention is further described below in conjunction with the accompanying drawings;

[0046] Figure 1 This is a schematic diagram of the edge computing-based real-time simulation optimization scheduling method for the power system in the present invention. DETAILED DESCRIPTION

[0047] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0048] Reference Figure 1 The following examples are obtained:

[0049] Example 1: Edge computing is a distributed computing method that sinks data storage and processing capabilities to the edge nodes of the network (such as smart devices, gateways, etc.) to reduce data transmission delays and improve real-time and local processing capabilities. Real-time simulation of power systems refers to the use of real-time data and computing technology to dynamically simulate the operating status of power systems to evaluate system performance, predict possible anomalies, and support scheduling decisions. Power dispatch in power systems is a mathematical model and algorithm that is pre-trained to reasonably allocate power generation, transmission, and distribution resources on the premise of meeting power supply needs, safety, and economy, so as to maximize the overall efficiency of the system.

[0050] As the complexity of power systems increases, real-time and accuracy become important requirements for dispatching. Edge computing distributes computing tasks to edge nodes close to data sources (such as substations, smart terminal devices, etc.), which can process data generated by a large number of distributed devices in real time, reduce dependence on central servers, and improve the real-time performance of dispatching and simulation.

[0051] Real-time simulation provides dynamic prediction capabilities for the operating status of the power system, helping to identify possible operating risks or scheduling problems. By combining preset power scheduling technologies, power generation plans, load distribution or grid operation strategies can be quickly adjusted based on real-time simulation results to ensure the safety and economy of grid operation.

[0052] Edge computing plays a key role as a bridge between real-time simulation and optimized scheduling. By implementing local real-time simulation at the edge node and cooperating with the central node, combined with centralized scheduling optimization, communication delay can be reduced, computing efficiency can be improved, and a closed-loop system from device perception to global optimization can be formed. This mode is particularly suitable for dealing with distributed energy access and frequent dynamic changes in the power grid. It should be noted that the optimization scheduling mentioned later in the present invention refers to the optimization scheduling of task allocation for the task allocation system of edge nodes and central nodes, rather than scheduling electricity. The scheduling of electricity can be based on the results after the implementation of the present invention, that is, the simulation results, and then through the pre-trained mathematical model and algorithm, under the premise of meeting the power supply demand, safety and economy, the power generation, transmission and distribution resources are reasonably allocated to maximize the overall efficiency of the system. This is the content after the execution of the present invention, which belongs to the prior art and will not be further elaborated here.

[0053] The present invention proposes a real-time simulation optimization scheduling method for a power system based on edge computing, which aims to optimize the task allocation of the task allocation system of edge nodes and central nodes, and includes the following steps:

[0054] The power system is divided into regions according to the preset strategy to obtain multiple regional power grids, and then the load information of each regional power grid is collected to obtain a load information set; the purpose of regional division is to divide the entire power system into multiple sub-regions according to characteristics such as geographical location or load demand for local optimization and scheduling. By dividing the regions, the management and scheduling of the power system can be more efficient, and the scheduling of each region can be more flexible and targeted. This division can also facilitate the allocation of edge computing and central computing tasks, because each region may have different resource requirements and computing loads.

[0055] Examples of regional division methods: Geographical division: Division is based on the geographical layout of the power system (such as cities, regions, etc.) so that each region contains a certain number of power equipment (such as substations, transmission lines, etc.).

[0056] Grid topology division: By analyzing the topological structure of the power system, the grid is divided into relatively independent sub-network areas that do not interfere with each other.

[0057] Collecting load information for each area is the basis for subsequent load analysis. Load information includes but is not limited to real-time power demand, load change trends, etc. By collecting data, we can have a comprehensive understanding of the load status of the power grid and identify abnormal fluctuations in power grid operation.

[0058] Load analysis is performed based on the load information set to identify whether there are abnormal load fluctuations. If so, the distributed computing potential assessment mechanism is activated; through the analysis of the load information set, abnormal load fluctuations in the power system are discovered in a timely manner and responded to. If abnormal fluctuations are found, the distributed computing potential assessment mechanism will be activated to evaluate the load change range and determine whether it exceeds the normal fluctuation range.

[0059] After the distributed computing potential assessment mechanism is activated, information is obtained on the allocation of collaborative tasks in regional power grid simulations to obtain a distributed collaborative information set, and then the resource utilization difference between edge nodes and central nodes is analyzed, as well as the task allocation imbalance analysis between edge nodes and central nodes. When abnormal load fluctuations are found, the corresponding abnormal distributed computing potential assessment mechanism of the regional power grid is automatically activated to evaluate the computing resource allocation in each region to decide how to efficiently allocate tasks and avoid excessive load or idle resources in some regions.

[0060] Obtain the simulation task coordination of each regional power grid in order to analyze the balance of task allocation. This step mainly collects the resource utilization of edge nodes and central nodes to evaluate the effect of existing task allocation. Task coordination analysis: edge node and central node comparison: Analyze the task execution of different computing nodes (such as edge nodes and central nodes) to identify areas with uneven resource utilization. Task scheduling: Analyze the task scheduling of each area to understand whether the tasks are reasonably allocated to each node to avoid overloading or idleness of some nodes.

[0061] Analyze the difference in resource utilization between edge nodes and central nodes: Analyze the difference in resource utilization between edge nodes and central nodes to find out the imbalance in computing resource usage. Resource differences may cause some nodes to be overloaded, while other nodes waste resources. Resource consumption comparison: Evaluate the difference in computing power resource consumption between edge computing nodes and central computing nodes. Load balancing optimization: Adjust the resource allocation between nodes through algorithms to ensure more balanced resource utilization.

[0062] Perform task allocation imbalance analysis for edge nodes and central nodes: Analyze whether there is an imbalance in task allocation, that is, whether some nodes are assigned too many or too few tasks, resulting in inefficient grid operation. Task allocation monitoring: Compare the task load of different nodes to check whether some nodes are underutilized. Dispatch strategy optimization: Optimize the task allocation strategy based on the analysis results to avoid over-reliance on a single node or region.

[0063] Based on the results of resource utilization difference analysis and task allocation imbalance analysis, the simulation collaborative task allocation of regional power grids is divided into efficient collaborative operation power grids and inefficient collaborative operation power grids; based on the previous analysis, the regional power grids are divided into two categories: efficient collaborative operation power grids and inefficient collaborative operation power grids. Efficient collaborative power grids have reasonable task allocation and maximized resource utilization, while inefficient collaborative power grids have problems such as unreasonable task allocation or idle resources. Meaning of classification type: Efficient collaborative operation power grid: reasonable task allocation, efficient use of computing resources, smooth task collaboration, and good load balance. Inefficient collaborative operation power grid: unreasonable task allocation, insufficient resource utilization, overloaded or idle load, affecting the overall scheduling efficiency.

[0064] For inefficient collaborative power grids, the allocation of simulation collaborative tasks to edge nodes and central nodes is optimized and scheduled, so that the regional power grid is reclassified as an efficient collaborative power grid. Optimize the scheduling of inefficient collaborative power grids and adjust their task allocation to turn them into efficient collaborative power grids, thereby improving the overall operation efficiency of the power grid. Task reallocation: Based on the previous analysis results, the tasks of the inefficient power grid are rescheduled and reasonably allocated to edge nodes and central nodes to ensure load balancing. Dynamic adjustment: Adjust the task allocation strategy in real time to ensure that the simulation can be quickly responded to when the power grid load fluctuates to avoid overload or waste of resources.

[0065] Load analysis means: in a fixed time window, the load value at each time point is obtained from the load information set to obtain the load data time series, and then the absolute value of the corresponding data difference between two adjacent time points is obtained, and then the average and standard deviation of all absolute values ​​are calculated.

[0066] Identifying whether there is abnormal load fluctuation means: comparing the average value of all absolute values ​​with a preset threshold value one, and comparing the standard deviation of all absolute values ​​with a preset threshold value two. If the average value of all absolute values ​​is less than or equal to the preset threshold value one and the standard deviation of all absolute values ​​is less than or equal to the preset threshold value two, a normal signal is generated. If the average value of all absolute values ​​is less than or equal to the preset threshold value one and the standard deviation of all absolute values ​​is less than or equal to the preset threshold value two, an abnormal signal is generated. When the abnormal signal is generated, it indicates that signs of abnormal load fluctuations have been identified from the load information set of the regional power grid.

[0067] Load analysis can monitor the load changes in the power grid system in real time and obtain and process load data in a timely manner. This is crucial to ensure the stable operation of the power grid. When the load fluctuates greatly or abnormally, it may affect the stability of the power supply and may even cause power grid failure. By calculating the difference, mean and standard deviation of the load values, a certain predictive ability can be provided for the power grid system. If the load fluctuation exceeds the normal range, the system can issue an early warning in time to remind the power grid operator to pay attention to the possibility of load abnormality, so as to take appropriate measures, such as load scheduling and load restriction. Regular and real-time load fluctuation analysis can help the system detect potential load anomalies and take timely measures to avoid abnormal fluctuations from causing excessive impact on the power grid. For example, if the load fluctuation is large and continuous, it may be necessary to adjust the power supply plan or alleviate the problem through load distribution and scheduling to prevent the collapse of the power grid.

[0068] Identifying load anomalies not only helps ensure the safe operation of the power grid, but also optimizes the allocation and scheduling of power grid resources, reduces energy waste, and improves the efficiency of power grid operation by analyzing the regularity of load changes. For example, the system can identify periods of time with smaller load fluctuations to adjust the load, thereby reducing operating costs. In the power system, abnormal load fluctuations may be a precursor to power equipment failure. By regularly detecting abnormal load fluctuations, potential equipment failures can be discovered in advance, equipment overload or damage can be prevented, maintenance costs can be reduced, and the impact of sudden failures can be reduced. Load analysis and fluctuation detection provide data support for power grid operation and load management. Decision makers can make more accurate judgments and scheduling decisions based on these data to ensure that the power grid can operate stably and meet electricity demand under various circumstances.

[0069] The result of analyzing the difference in resource utilization between edge nodes and central nodes is the resource utilization difference index, and the result of analyzing the imbalance in task allocation between edge nodes and central nodes is the task allocation imbalance index. The resource utilization difference index between edge nodes and central nodes can reflect the imbalance in resource allocation between different nodes in the system, and reveal the gap in resource utilization between edge nodes and central nodes. The task allocation imbalance index measures the imbalance in the distribution of tasks between edge nodes and central nodes in the system, reflecting possible over-centralization or decentralization problems. The combination of the two can provide a basis for optimizing resource allocation and task scheduling, and help improve the overall efficiency and load balancing of the system.

[0070] The logic for obtaining the resource utilization difference index is:

[0071] Get the resource utilization of each edge node and mark it as , Indicates the index number of the edge node, obtains the resource utilization of the central node and marks it as , and then by the difference value Represents the difference in resource utilization between edge node i and central node: ;

[0072] The difference value is calculated by the resource usage ratio of the edge node Perform dynamic adjustments to define the weight corresponding to edge node i for: ; N is the total number of edge nodes, represents the used resources of edge node i, where is the ratio of the total resources of edge node i to the total resources of the central node;

[0073] The adjusted difference is: ; represents the adjusted difference value;

[0074] In order to ensure that the difference value is not affected by the node scale, a nonlinear normalization method is used to obtain the processed difference value: ; represents the difference value after processing, k is a preset non-zero adjustment coefficient used to control the intensity of nonlinear normalization;

[0075] The difference in collaborative utilization between edge nodes is introduced to measure the problem of uneven resource allocation of edge nodes as a whole. The formula is: ; represents the absolute value of the difference in resource utilization between edge node i and edge node j, is the sum of the differences in resource utilization between edge nodes;

[0076] The final resource utilization difference index combines the central node difference and the edge node coordination difference and is defined as: ; is a preset non-zero adjustment factor used to weigh the impact of central node differences and edge node collaboration differences. Represents the resource utilization difference index.

[0077] The resource occupancy ratio of edge nodes and central nodes is combined to dynamically reflect the weight of actual differences. Nonlinear normalization enhances the sensitivity to large difference values ​​while maintaining reasonable compression of small difference values. Global collaborative analysis considers the collaborative differences between edge nodes and enhances the index's perception of overall resource unevenness. By adjusting the coefficients k and , can adapt to different power system scales and resource allocation scenarios. When the resource utilization difference index is high, it may be necessary to reallocate tasks and optimize resource allocation strategies. Specifically, the task allocation imbalance index needs to be considered, especially when the distributed computing potential is not fully utilized. A high RUDI value indicates that there is a large difference in resource utilization between edge nodes and central nodes, and there may be problems of uneven resource allocation or node load imbalance. A low RUDI value indicates that the resource utilization difference is small and the system resource allocation is relatively balanced.

[0078] The logic for obtaining the task allocation imbalance index is:

[0079] Get the number of tasks for edge node i , the number of real-time tasks of the central node , and the total number of tasks ,The total number of tasks is the sum of all tasks of edge and central nodes;

[0080] Task allocation ratio of edge node i for: ;

[0081] Task allocation ratio of central nodes for: ;

[0082] The ideal task allocation ratio is allocated based on the node task processing capacity:

[0083] ; ; represents the ideal task allocation ratio of edge node i, represents the ideal task allocation ratio of the central node, and They correspond to the task processing capabilities of edge node i and central node respectively, It represents the sum of the task processing capabilities of all edge nodes i and central nodes;

[0084] Calculate the fairness coefficient of task allocation for edge node i :

[0085] ;

[0086] Calculate the fairness coefficient of task allocation of central nodes :

[0087] ;

[0088] By assigning discrete Measures the degree of fluctuation of task ratios and distribution discreteness Defined as the variance of the task ratio of each node: ; represents the average value of the task allocation ratio of all edge nodes i and central nodes;

[0089] The task allocation imbalance index is generated by combining the fairness and discreteness calculation results:

[0090] ; Represents the task allocation imbalance index. A high TAII value indicates that the task allocation is seriously unbalanced, some nodes may be overloaded, and other node resources are idle. A low TAII value indicates that the task allocation is balanced, and the resources of each node are fully utilized, meeting the target allocation ratio. When the TAII value is high, it may be necessary to reallocate tasks to ensure balanced node loads. The TAII value is used to evaluate whether the existing allocation strategy is reasonable and make adjustments.

[0091] The division of the simulation collaborative task allocation of regional power grids into efficient collaborative operation grids and inefficient collaborative operation grids refers to:

[0092] The task allocation imbalance index and resource utilization difference index of the regional power grid are obtained separately and used as input variables together. The type corresponding to the simulation collaborative task allocation of the regional power grid is used as the output variable. The input variables are fuzzified and the values ​​of the input variables are converted into fuzzy sets. The output variables are fuzzified and the output variables are converted into fuzzy sets. Fuzzy rules are formulated to describe the quality of collaborative operation under different data type combinations. The fuzzified input variables are inferred through fuzzy rules to determine the type corresponding to the simulation collaborative task allocation of the regional power grid. The following are the detailed steps:

[0093] Input variables: task allocation imbalance index and resource utilization difference index. These two indexes are used as input variables to measure the balance of task allocation and resource utilization of regional power grids:

[0094] Task allocation imbalance index: measures whether tasks are evenly distributed among nodes in the power grid (such as edge nodes and central nodes). If the task allocation is unbalanced, some nodes may carry too many tasks while other nodes are idle or underloaded, which will lead to inefficient power grid operation.

[0095] Resource Utilization Difference Index: measures the difference in resource utilization of different nodes (edge ​​nodes and central nodes) in the power grid. If the resource utilization of some nodes is much higher than that of other nodes, it may cause resource waste or overload, affecting the overall performance of the system.

[0096] Fuzzification: Fuzzification of input and output variables. Fuzzification is the process of converting traditional quantitative data into fuzzy sets. This is because the task allocation imbalance index and resource utilization difference index are often not clear "high" or "low" standards, but are divided according to specific numerical ranges. Through fuzzification, the input data can be made more consistent with the uncertainty and complexity of actual power grid operation.

[0097] Fuzzification of input variables: The task allocation imbalance index may have several fuzzy levels, such as "low imbalance", "medium imbalance" and "high imbalance". The resource utilization variance index may have several fuzzy levels, such as "low variance", "medium variance" and "high variance".

[0098] Fuzzification of output variables: The type of cooperatively operated power grid, that is, the operating efficiency of the power grid, is divided into: "high-efficiency cooperative operation" and "low-efficiency cooperative operation".

[0099] Formulate fuzzy rules: Based on the fuzzification results of the task allocation imbalance index and the resource utilization difference index, formulate fuzzy rules to describe the impact of different input combinations on the type of coordinated operation of the power grid. Fuzzy rules determine how the input variables work together to affect the output variables (i.e., the efficiency of the coordinated operation of the power grid) through logical relationships. Some possible examples of fuzzy rules are as follows: If the task allocation imbalance index is "high imbalance" and the resource utilization difference index is "high difference", the output is "inefficient coordinated operation". If the task allocation imbalance index is "low imbalance" and the resource utilization difference index is "low difference", the output is "efficient coordinated operation". If the task allocation imbalance index is "medium imbalance" and the resource utilization difference index is "medium difference", the output may be "inefficient coordinated operation" or "efficient coordinated operation", depending on the rule settings in the fuzzy reasoning system.

[0100] Fuzzy reasoning: Fuzzy reasoning is to convert input variables (task allocation imbalance index and resource utilization difference index) into output variables (grid collaborative operation type) according to the established fuzzy rules. The reasoning process usually includes the following steps: Forward reasoning: Based on the fuzzy set of input variables (such as low imbalance, high difference), the corresponding output set (such as inefficient collaborative operation) is obtained through fuzzy rules. Backward reasoning: Backward reasoning is sometimes used to adjust input parameters and optimize grid collaborative task allocation to achieve the best operation effect.

[0101] Defuzzification of output variables: The output obtained by fuzzy reasoning is usually a fuzzy set, which represents a range or probability of the type of grid collaborative operation. For example, the system may output a probability of "efficient collaborative operation" of 70% and a probability of "inefficient collaborative operation" of 30%. In order to make a clear decision, it is usually necessary to defuzzify the output results and convert the fuzzy set into a clear decision result. In this case, the final result may be to choose "efficient collaborative operation" or "inefficient collaborative operation".

[0102] This collaborative task allocation and resource utilization difference analysis method based on fuzzy logic has the following significance in the optimal dispatching of power grids: Dynamic adaptability: Fuzzy logic can handle complex, dynamic and uncertain systems, and adapt to the changes in real-time load, task allocation and resource utilization of the power grid. Optimize the operation of the power grid: By evaluating the efficiency of the collaborative task allocation of the power grid, the scheduling of tasks in the power grid can be optimized, node overload or resource waste can be avoided, thereby improving the overall operation efficiency of the power grid. Risk warning and decision support: By judging the type of collaborative operation of the power grid, it can provide decision support for optimized dispatching and load distribution for power grid managers, reduce the risks in system operation, and ensure stable and safe operation of the power grid.

[0103] Genetic algorithms are used to optimize the allocation of simulation collaborative tasks to edge nodes and central nodes for inefficient collaborative power grids, so that the regional power grid is re-divided into an efficient collaborative power grid, and the total execution time of simulation collaborative tasks is minimized.

[0104] In the process of power grid simulation and task allocation, tasks usually need to be allocated between different nodes (including edge nodes and central nodes) to achieve optimal execution efficiency. In an inefficient collaborative operation power grid, task allocation may be unbalanced, resulting in excessive load on some nodes and long task execution time, which in turn affects the collaborative efficiency of the entire power grid. In order to improve the operating efficiency of the power grid, it is necessary to optimize task allocation so that the distribution of tasks between edge nodes and central nodes is more reasonable, and ultimately achieve efficient collaborative operation of the power grid.

[0105] The core of the genetic algorithm is the fitness function, which measures the quality of each solution. For this problem, the goal of fitness evaluation is to simulate the total execution time of collaborative tasks, that is, to minimize the total time. Assuming there are multiple simulation tasks, each task needs to be executed on a certain node, and the computing power of each node is different, the execution time of the task will be affected by the node load and computing power. The total execution time is the sum of the execution time of all tasks on their corresponding nodes.

[0106] The steps of the genetic algorithm are as follows: Initial population generation: First, a set of task allocation schemes are randomly generated as the initial population. Each individual represents a task allocation scheme, indicating which node each task is assigned to. The genome of each scheme can be an array, in which each element represents the corresponding node to which a task is assigned.

[0107] Fitness evaluation: For each individual, the inverse of the total execution time is obtained according to its task allocation plan as the fitness of the individual. The higher the fitness, the shorter the total execution time and the better the plan.

[0108] Selection: Selecting better individuals for reproduction based on fitness, usually by roulette wheel selection or tournament selection. During the selection process, individuals with higher fitness have a greater chance of being selected.

[0109] Crossover and mutation: Selected individuals undergo crossover operations to generate new individuals. The crossover operation simulates gene recombination and generates new individuals by exchanging some of the genetic information of the parent individuals. Crossover can enhance the algorithm's exploration ability. Mutating certain genes of individuals simulates the mutation phenomenon in nature and increases the diversity of solutions. Mutation operations are usually random and explore new solutions by adjusting task allocation.

[0110] Termination conditions: The genetic algorithm will iterate according to the set termination conditions. Common termination conditions include: reaching the maximum number of iterations; the fitness reaches a certain threshold; the fitness change does not increase significantly within a certain number of times.

[0111] Optimization goal: Through the continuous iteration of the genetic algorithm, the ultimate goal is to obtain a task allocation scheme that minimizes the total execution time of the simulation task of the power grid, thereby achieving efficient coordinated operation of the power grid. By simulating natural selection and genetic processes, it can effectively optimize the scheduling of inefficient coordinated operation of the power grid. The fitness evaluation formula is used to quantify the pros and cons of the task allocation scheme, ultimately minimizing the total execution time of the simulation task, thereby improving the overall operation efficiency of the power grid.

[0112] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters in the formula are set by technicians in this field according to actual conditions.

[0113] It should be understood that in the various embodiments of the present application, the size of the serial numbers of the above-mentioned processes does not 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 embodiments of the present application.

[0114] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0115] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0116] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A real-time simulation optimization dispatching method for power system based on edge computing, characterized in that: The following steps are involved: The power system is divided into regions according to a preset strategy to obtain multiple regional power grids, and then load information is collected from each regional power grid to obtain a load information set; Perform load analysis based on the load information set to identify whether there is abnormal load fluctuation. If so, activate the distributed computing potential assessment mechanism; After the distributed computing potential assessment mechanism is activated, the information of the regional power grid simulation collaborative task allocation is obtained to obtain the distributed collaborative information set, and then the resource utilization difference analysis of the edge nodes and the central nodes is carried out, as well as the task allocation imbalance analysis of the edge nodes and the central nodes is carried out; Based on the results of resource utilization difference analysis and task allocation imbalance analysis, the collaborative task allocation of the regional power grid simulation is divided into efficient collaborative operation power grid and inefficient collaborative operation power grid. For inefficient collaborative power grids, the allocation of simulated collaborative tasks to edge nodes and central nodes is optimized and scheduled, so that the regional power grid is re-divided into an efficient collaborative power grid.

2. The method for real-time simulation optimization and dispatching of a power system based on edge computing according to claim 1 is characterized in that: Load analysis refers to: In a fixed time window, the load value at each time point is obtained from the load information set to obtain the load data time series, and then the absolute value of the corresponding data difference between two adjacent time points is obtained, and then the average and standard deviation of all absolute values ​​are calculated.

3. The method for real-time simulation optimization and dispatching of power systems based on edge computing according to claim 2 is characterized in that: Identifying whether there is abnormal load fluctuation means: The average value of all absolute values ​​is compared with a preset threshold value one, and the standard deviation of all absolute values ​​is compared with a preset threshold value two. If the average value of all absolute values ​​is less than or equal to the preset threshold value one and the standard deviation of all absolute values ​​is less than or equal to the preset threshold value two, a normal signal is generated. If the average value of all absolute values ​​is less than or equal to the preset threshold value one and the standard deviation of all absolute values ​​is less than or equal to the preset threshold value two, an abnormal signal is generated. When an abnormal signal is generated, it indicates that signs of abnormal load fluctuations have been identified from the load information set of the regional power grid.

4. The real-time simulation optimization scheduling method for power system based on edge computing according to claim 3 is characterized in that: The result of analyzing the difference in resource utilization between edge nodes and central nodes is the resource utilization difference index, and the result of analyzing the imbalance in task allocation between edge nodes and central nodes is the task allocation imbalance index.

5. The method for real-time simulation optimization and dispatching of power systems based on edge computing according to claim 4 is characterized in that: The logic for obtaining the resource utilization difference index is: Get the resource utilization of each edge node and mark it as , Indicates the index number of the edge node, obtains the resource utilization of the central node and marks it as , and then by the difference value Represents the difference in resource utilization between edge node i and central node: ; The difference value is calculated by the resource usage ratio of the edge node Perform dynamic adjustments to define the weight corresponding to edge node i for: ; N is the total number of edge nodes, represents the used resources of edge node i, where is the ratio of the total resources of edge node i to the total resources of the central node; The adjusted difference is: ; represents the adjusted difference value; In order to ensure that the difference value is not affected by the node scale, a nonlinear normalization method is used to obtain the processed difference value: ; represents the difference value after processing, k is a preset non-zero adjustment coefficient used to control the intensity of nonlinear normalization; The difference in collaborative utilization between edge nodes is introduced to measure the problem of uneven resource allocation of edge nodes as a whole. The formula is: ; represents the absolute value of the difference in resource utilization between edge node i and edge node j, is the sum of the differences in resource utilization between edge nodes; The final resource utilization difference index combines the central node difference and the edge node coordination difference and is defined as: ; is a preset non-zero adjustment factor used to weigh the impact of central node differences and edge node collaboration differences. Represents the resource utilization difference index.

6. The method for real-time simulation optimization and dispatching of a power system based on edge computing according to claim 5 is characterized in that: The logic for obtaining the task allocation imbalance index is: Get the number of tasks for edge node i , the number of real-time tasks of the central node , and the total number of tasks ,The total number of tasks is the sum of all tasks of edge and central nodes; Task allocation ratio of edge node i for: ; Task allocation ratio of central nodes for: ; The ideal task allocation ratio is allocated based on the node task processing capacity: ; ; represents the ideal task allocation ratio of edge node i, represents the ideal task allocation ratio of the central node, and They correspond to the task processing capabilities of edge node i and central node respectively, It represents the sum of the task processing capabilities of all edge nodes i and central nodes; Calculate the fairness coefficient of task allocation for edge node i : ； Calculate the fairness coefficient of task allocation of central nodes : ； By assigning discrete Measures the degree of fluctuation of task ratios and distribution discreteness Defined as the variance of the task ratio of each node: ; represents the average value of the task allocation ratio of all edge nodes i and central nodes; Generate task allocation imbalance index: ; Represents the task allocation imbalance index.

7. The method for real-time simulation optimization and dispatching of a power system based on edge computing according to claim 6 is characterized in that: The division of the simulation collaborative task allocation of regional power grids into efficient collaborative operation grids and inefficient collaborative operation grids refers to: The task allocation imbalance index and resource utilization difference index of the regional power grid are obtained respectively and used as input variables together, the type corresponding to the simulation collaborative task allocation of the regional power grid is used as the output variable, the input variables are fuzzified, and the values ​​of the input variables are converted into fuzzy sets, the output variables are fuzzified, and the output variables are converted into fuzzy sets, fuzzy rules are formulated to describe the quality of collaborative operation under different data type combinations, the fuzzified input variables are inferred through fuzzy rules, and the type corresponding to the simulation collaborative task allocation of the regional power grid is determined.

8. The method for real-time simulation optimization and dispatching of a power system based on edge computing according to claim 7 is characterized in that: Genetic algorithms are used to optimize the allocation of simulation collaborative tasks to edge nodes and central nodes for inefficient collaborative power grids, so that the regional power grid is re-divided into an efficient collaborative power grid, and the total execution time of simulation collaborative tasks is minimized.

Images