Intelligent planning method for power distribution network

Abstract

The invention discloses an intelligent planning method for a power distribution network, and the method comprises the steps: dividing the power distribution network into a plurality of functional blocks, collecting and classifying the power data of power consumers, generating a change curve based on the data, predicting the power demands, and obtaining a prediction curve of the power data size. Performing digital twinborn modeling and computing power simulation on each functional block, and recording simulation computing power indexes; establishing a fitness function according to the power data prediction curve and the calculation power index, and evaluating the twin communication matching degree of the function block and the power distribution network; according to the fitness index, evaluating the priority of the time sequence nodes and layering the time sequence nodes, finally determining the time sequence nodes needing to be adjusted, and dynamically planning the power distribution network. According to the method, regional power utilization characteristics can be accurately mastered, the system state and the computing power demand are evaluated in advance, the communication matching degree is accurately evaluated, potential problems are found in time, resource allocation is optimized, the operation stability and efficiency of the power distribution network are improved, and reasonable utilization and efficient management of resources are achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and particularly to an intelligent planning method for a distribution network. Background Art

[0002] Intelligent distribution network planning is an optimization planning method that fully considers a series of new boundary condition constraints such as distributed power sources, distributed energy storage, electric vehicle charging and swapping facilities, and demand-side response on the basis of traditional distribution network planning. The intelligent distribution network planning method should fully consider various uncertainty factors under the new boundary conditions, model and analyze the uncertainty information, establish an objective function for optimal planning, and on this basis, use modern heuristic algorithms to solve and obtain the global optimal solution.

[0003] The invention patent with the patent application number 202311219045.3 discloses an intelligent planning method for a distribution network. By collecting power users in the target area, multiple types of power users are determined. The power data source change curves of multiple types of power users are collected and predicted, and a data volume prediction curve is output. A distribution network twin is obtained and computing power simulation is performed, and a simulated computing power index under a preset failure rate condition is output. Fitness calculation is performed according to the data volume prediction curve and the simulated computing power index to obtain a fitness recognition result, where the fitness recognition result is multiple identified timing nodes for sub-network expansion. The distribution network to be planned is dynamically planned according to the multiple timing nodes for sub-network expansion. It solves the technical problems in the prior art that the intelligence of distribution network planning is relatively low, the processing device cannot be updated in a timely manner based on the data processing volume, resulting in anomalies in power grid data processing and affecting the stability of the distribution network.

[0004] However, in actual distribution network management, the scale of the distribution network is constantly expanding, the number of power users is increasing day by day, and the power consumption demands are becoming more and more complex and diverse. The existing distribution network planning methods have a low level of intelligence and are difficult to accurately adapt to the constantly changing power grid operation conditions. Since the processing device cannot be updated in a timely manner according to the actual data processing volume, anomalies frequently occur during the power grid data processing process, seriously affecting the stable operation of the distribution network. Summary of the Invention

[0005] In this application, by providing an intelligent planning method for a distribution network, the distribution network is finely divided into functional blocks, the power consumption data of power users is accurately collected and classified, a change curve is generated to predict the power consumption demand, digital twin modeling and computing power simulation are performed, a fitness function is established to evaluate the matching degree, and hierarchical timing nodes are used to dynamically plan the distribution network to cope with the complex and changeable power grid operation conditions.

[0006] This application provides an intelligent planning method for a distribution network, including: S1. Divide the distribution network into multiple functional blocks, collect the power consumption data of power users in the functional blocks and classify them; S2. Generate power data change curves based on the power consumption data of each functional block, predict the power consumption demand of each functional block, and generate the predicted curve of the power data volume corresponding to each functional block; S3. Conduct digital twin modeling and computing power simulation for each functional block, and record the simulated computing power indicators; S4. According to the predicted curve of the power data volume and the simulated computing power indicators, establish a fitness function for each functional block to communicate with the distribution network twin body, and evaluate its matching degree with the digital twin body communication; S5. Evaluate the priority of each time series node according to the fitness index, and stratify the time series nodes according to the evaluation results; S6. Determine the time series nodes that need to be adjusted in the distribution network, and conduct dynamic programming for the distribution network.

[0007] Preferably, in S3, conducting digital twin modeling and computing power simulation for each functional block specifically includes: S3a. Collect the physical parameters and historical operation data of the distribution network, and establish a distribution network topology model and an electrical characteristic model; S3b. Integrate the topology model and the electrical characteristic model to form a complete distribution network digital twin model; S3c. Establish a data interface with the distribution network monitoring system to obtain the operation data of the power grid in real time; S3d. Dynamically adjust the digital twin model according to the real-time data and the change of the operation state of the power grid; S3e. Analyze the possible fault types in the distribution network, determine the location and influence range of the faults for different fault types, and set the failure rates of different devices according to the historical fault data and equipment reliability indicators; S3f. Use MATLAB to conduct simulation calculations of the distribution network to determine the computing resources for computing power simulation; S3g. Inject faults into the digital twin model according to the set failure rates and fault scenarios, and simulate the change of the operation state of the distribution network when the faults occur; S3h. Run the digital twin model to conduct power flow calculation and fault analysis simulation tasks of the distribution network, and record the time of the simulation calculation and the consumption of computing resources; S3i. For different failure rates and fault scenarios, repeat the above fault injection and simulation calculation processes to obtain multiple sets of computing power performance data; S3j. Define the indicators for evaluating the computing power performance of the distribution network, and calculate the computing power indicators under different failure rates according to the simulated recorded data; S3k, analyze the changing trend of computing power indicators under different failure rates, and evaluate the computing power bearing capacity of the distribution network under different fault conditions.

[0008] Preferably, in S4, according to the power data volume prediction curve and the simulated computing power indicators, it includes: adopting the weighted comprehensive scoring method to combine power demand and computing power performance:

[0009] is the fitness score of the i-th functional block, is the power demand prediction index, is the load volatility, is the computing power performance, 、 、 are the weight coefficients, adjusted according to actual needs, and Norm() is the normalization process.

[0010] Preferably, S6 specifically includes: S6a, traverse all time series nodes of each functional block; S6b, compare the fitness index of each time series node with the lower limit adjustment threshold and the upper limit adjustment threshold; S6c, if the fitness index of the time series node is lower than the lower limit adjustment threshold, mark the time series node as a weak time series node; S6d, if the fitness index of the time series node is higher than the upper limit adjustment threshold, mark the time series node as a strong time series node.

[0011] Preferably, for the step of dividing the distribution network into multiple functional blocks, step S1 further includes: S11, set an initial upper limit of the load capacity for each functional block, and obtain the load change situation and the operation status information of the equipment within the functional block in real time; S12, compare the actual load of the functional block with the set upper limit of the load capacity. When the load within the functional block reaches the capacity upper limit, immediately trigger the modification process; S13, merge the overloaded functional block with the adjacent functional block with a low load capacity and re-divide the functional blocks; S14, evaluate the computing reliability of the re-divided functional blocks.

[0012] Preferably, S13 includes: according to the topological structure of the distribution network, determine the set of adjacent functional blocks to the overloaded functional block i , for each functional block j in the set of adjacent functional blocks , calculate its remaining load capacity, and select the functional block with the largest remaining load capacity as the merging object; merge functional block i and functional block to form a new functional block, and the upper limit of the load capacity of the new functional block is the sum of the original upper limits of the load capacities of the two functional blocks minus the functional block The actual load, and at the same time update the topological structure information of the distribution network.

[0013] Preferably, in S13, when re-dividing the functional blocks, it further includes: S131, deploy high-precision sensors inside each functional block, and transmit the collected data to the central intelligent control center of the distribution network in real time; S132, perform cross-regional collaborative demand determination on the functional blocks to identify the functional blocks that can work collaboratively and share areas; wherein, define as the load growth rate threshold, and calculate the load growth rate of the functional block of the cross-regional collaborative demand determination index: calculate the load growth rate r of the functional block within a period of time Δt, and the formula is:

[0014] P(t) is the real-time power. When r exceeds the preset threshold , it is determined that there is a high-load collaborative demand; S133, evaluate each functional block in the shared area to determine the combination of functional blocks participating in the cooperation; S134, calculate the electricity consumption energy of the functional blocks with collaborative demands, and allocate resources according to the required electricity consumption energy in the merged shared area; S135, the central intelligent control center adjusts the topological structure of the distribution network in real time through the remote control system; S136, after the collaborative work ends, restore the independent operation of each functional block.

[0015] Preferably, the identification of the functional blocks that can work collaboratively and share areas includes: based on the geographic information system and electrical topological structure of the distribution network, use the Dijkstra algorithm to calculate the electrical distance between the functional blocks ; comprehensively consider the line capacity , electrical distance factors, construct a shared area identification model, and calculate a shared area score for each functional block , the formula is:

[0016] , is the weight coefficient, is the maximum line capacity, sort the scores in descending order, and select the functional blocks starting from the highest score to form a shared area.

[0017] Preferably, in S134, calculate the power consumption of functional blocks with collaborative requirements, and allocate resources for the merged shared area according to the required power consumption. It further includes: S1341, analyze the power transmission path, capacity limit, and time characteristics of functional blocks between different regions, and construct a mathematical model describing the power transmission law between regions; S1342, according to the priority of time sequence nodes, count the supply data of the shared area, and according to the time sequence priority rule, preferentially allocate power from the area with sufficient power supply; S1343, regard the transmission capacity and time sequence priority as the key factors affecting power scheduling, and set them as dynamic variables; S1344, with the time sequence priority and power supply reliability as the goals, construct a multi-objective optimization function; S1345, continuously collect real-time monitoring data of the power grid operation, analyze the power grid state and change trend, and dynamically adjust the weights and value ranges of each dynamic variable according to the analysis results; S1346, when the power cooperation requirement requested by the functional block exceeds the supply capacity of the shared area, automatically trigger multi-objective optimization calculation.

[0018] Preferably, S1344 specifically includes: the time sequence priority goal is to determine the power supply stability of high-priority tasks. Let the proportion of time when the power supply of high-priority tasks is insufficient be , then the time sequence priority objective function is ; the reliability goal is to set the power grid outage probability as , and the reliability objective function is ; the economic objective is to set the power scheduling cost as Cost, and the objective function is ; the comprehensive multi-objective optimization function is , where , , are the weights of each goal.

[0019] One or more technical solutions provided in this application have at least the following technical effects or advantages: By dividing the distribution network into functional blocks, classifying data collection, and predicting demand, the electricity consumption characteristics of each region can be accurately grasped. Digital twin modeling and computing power simulation can evaluate the system operation state and computing power requirements in advance, ensuring real-time monitoring and decision-making of the distribution network. Based on the power data prediction curve and computing power index, a fitness function is established to accurately evaluate the communication matching degree between the functional block and the digital twin body. Stratify the time sequence nodes according to the fitness index and dynamically plan, which can timely discover potential problems, optimize resource allocation, improve the operation stability and efficiency of the distribution network, and realize the reasonable utilization and efficient management of resources.

[0020] By setting the upper limit of the load capacity of the function block and monitoring the load changes in real time, the overload situation of the load can be detected in time, and the modification process can be quickly triggered to avoid equipment overload and power supply risks. Merging the overloaded function block with adjacent low-load function blocks and re-dividing them can optimize the load distribution and improve the resource utilization rate. At the same time, evaluating the reliability of the function block after re-division can ensure the stable operation of the entire distribution network, reduce the probability of power outage faults, enhance the power supply quality, provide users with higher-quality and reliable power services, and promote the development of the distribution network towards intelligence and high efficiency.

[0021] By deploying high-precision and multi-type sensors inside the function block, key data can be accurately collected and transmitted to the control center in real time, providing a reliable basis for subsequent analysis and decision-making. The cross-region collaborative demand determination and shared area identification method comprehensively considers various factors, can accurately identify collaborative demands and reasonably divide shared areas. Comprehensively evaluating and determining the combination of function blocks in the shared area, and adopting a multi-objective optimization model, taking into account both collaborative effects and costs. Calculating the electricity energy consumption and reasonably allocating resources, and adjusting the topological structure in real time, improving the operation efficiency and reliability of the distribution network. After the collaboration ends, the function block operation is restored according to the actual situation, ensuring the flexible and stable operation of the power system.

[0022] By analyzing and modeling the power transmission path, capacity limit and time characteristics between function blocks in the distribution network, the power transmission law can be accurately grasped. Allocating electric energy based on time sequence priority to ensure stable power supply for high-priority tasks and taking into account the needs of sub-priority tasks. Setting the transmission capacity and time sequence priority as dynamic variables and constructing a multi-objective optimization function can dynamically adapt to the changes in the power grid. Continuously collecting real-time data to adjust the variable weights and value ranges makes the scheduling more flexible. When the cooperative power demand exceeds the supply capacity, the optimization calculation is automatically triggered to obtain the optimal scheduling plan, effectively improving the overall scheduling efficiency and operation reliability of the distribution network. Brief Description of the Drawings

[0023] Figure 1 It is a flow schematic diagram of a distribution network intelligent planning method according to an embodiment of the present invention. Detailed Embodiments

[0024] To facilitate the understanding of the present invention, the present application will be described more comprehensively with reference to the relevant drawings; the drawings show preferred embodiments of the present invention, however, the present invention can be implemented in many different forms and is not limited to the embodiments described herein; on the contrary, these embodiments are provided to make the disclosure of the present invention more thorough and comprehensive.

[0025] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the technical field to which this invention belongs; the terms used in the description of this invention herein are only for the purpose of describing specific embodiments and are not intended to limit this invention; the term "and / or" used herein includes any and all combinations of one or more of the related listed items.

[0026] Embodiment 1: Figure 1 It is a schematic flow diagram of an intelligent distribution network planning method according to an embodiment of this invention.

[0027] As Figure 1 shown, an intelligent distribution network planning method includes the following steps: S1. Divide the distribution network into multiple functional blocks, collect the electricity consumption data of the power users in the functional blocks and classify them.

[0028] Among them, based on the geographical area, load distribution and grid structure, divide the distribution network into multiple functional blocks. Each functional block should contain a relatively independent power supply area and load to ensure a low coupling degree between the functional blocks. Clearly identify the load nodes, power source nodes and switch nodes within each functional block.

[0029] For each functional block, collect its electricity consumption data, including historical electricity consumption and electricity consumption patterns. According to the electricity consumption data, divide the users within each functional block into high electricity consumption users, medium electricity consumption users and low electricity consumption users.

[0030] S2. Generate a power data change curve based on the electricity consumption data of each functional block, predict the electricity consumption demand of each functional block, and generate a predicted curve of the electricity consumption data volume corresponding to each functional block.

[0031] Specifically, sort out the electricity consumption data of each functional block according to the time series and draw a power data change curve. Extract key features from the change curve, such as peak value (the highest point in the curve, indicating the maximum electricity consumption during this time period), valley value (the lowest point in the curve, indicating the minimum electricity consumption during this time period), average load (calculate the average value of the electricity consumption over a period of time, reflecting the average electricity consumption level during this time period), load volatility (calculate the standard deviation or coefficient of variation of the electricity consumption change, reflecting the degree of electricity consumption fluctuation), etc.

[0032] Use the ARIMA autoregressive integrated moving average model to divide the collected electricity consumption data into a training set (the training set accounts for a large proportion) and a test set, and specify a future time point or time period as the prediction target. Use the trained model to generate the predicted value of the electricity consumption demand at this time point or time period, and sort out the predicted electricity consumption demand values according to the time series to ensure that each predicted value has a corresponding time stamp.

[0033] The same tool is used to draw the prediction curve as that for the power data change curve. The horizontal axis represents time, and the vertical axis represents the power load value. According to the sorted prediction results, the power consumption prediction curves of each functional block are drawn.

[0034] S3. Perform digital twin modeling and computing power simulation for each functional block, and record the simulated computing power indicators.

[0035] Among them, digital twin modeling is to create an accurate mapping of the physical distribution network in the virtual space for functional block simulation. Before digital twin modeling, the boundaries of each functional block need to be clarified, including the power supply area, load nodes, power source nodes, and switch nodes. Determine that the modeling objectives are load prediction and fault simulation, simulate the operating state of the actual distribution network, and obtain a virtual model that can reflect the operating state of the distribution network in real time. Perform computing power simulation in the virtual model, assuming that some devices may fail, and calculate the computing power indicators under different failure rates.

[0036] Computing power simulation is used to evaluate the operating efficiency of the digital twin model under given computing resources. Use OpenDSS to evaluate the simulation time of the digital twin model for typical scenarios of the distribution network (such as peak load and fault recovery) under different computing resources, and determine the minimum computing power configuration that meets the requirements of real-time monitoring and decision-making of the distribution network. Evaluate the time complexity of the main computational tasks of the model, and establish the relationship between computing resources (CPU, memory) and simulation time.

[0037] Specifically, perform digital twin modeling and computing power simulation for each functional block, including: S3a. Collect the physical parameters and historical operation data of the distribution network, and establish a topological model and electrical characteristic model of the distribution network.

[0038] Among them, the physical parameters include line length, conductor type, transformer capacity, switchgear specifications, etc.; historical operation data, such as load curves, voltage measurement values, current measurement values, switch operation records, etc.

[0039] Specifically, clean the collected data, remove noise and outliers, normalize the data to make data with different dimensions comparable, and sort the data according to the time series position. Abstract substations, load points, switchgear, etc. in the distribution network into nodes, assign a unique identifier to each node, and record its physical location, electrical parameters (such as rated voltage, rated current), etc. Abstract the parts connecting nodes such as transmission lines and cables into edges, and record parameters such as the length, conductor type, resistance, and reactance of the edges. Use the method of graph theory to represent the relationship between nodes and edges, construct the topological structure of the distribution network, and use the adjacency matrix data structure to store the topological information (when using the adjacency matrix to store the topological information, if the number of distribution network nodes is large, a sparse matrix storage method can be considered to save storage space.

[0040] Based on parameters such as the resistance and reactance of the line, establish an equivalent circuit model of the line. Considering the distributed parameter characteristics of the line, a distributed parameter model can be used for long lines. Establish an equivalent circuit model of the transformer, considering parameters such as the turns ratio and short-circuit impedance of the transformer. According to the operating characteristics of the transformer, simulate the voltage and current changes under different loads. Based on the historical data of the load, establish a dynamic model of the load. For important loads, use a more detailed load characteristic curve to describe them.

[0041] S3b. Integrate the topological model and the electrical characteristic model to form a complete digital twin model of the distribution network.

[0042] Among them, when using historical operation data to verify the digital twin model, indicators such as root mean square error (RMSE) and mean absolute error (MAE) can be used to evaluate the accuracy of the model.

[0043] S3c. Establish a data interface with the distribution network monitoring system to obtain the operation data of the power grid in real time.

[0044] Among them, input the real-time data into the digital twin model to update the operation state of the model.

[0045] S3d. Dynamically adjust the digital twin model according to the real-time data and the changes in the operation state of the power grid.

[0046] For example, when a topological change (such as a switch operation) occurs in the power grid, update the topological model in a timely manner.

[0047] S3e. Analyze the possible fault types in the distribution network. For different fault types, determine the location and scope of influence of the fault. According to the historical fault data and equipment reliability indicators, set the failure rates of different equipment.

[0048] S3f. Use MATLAB to perform simulation calculations on the distribution network to determine the computing resources for computing power simulation.

[0049] S3g. Inject faults into the digital twin model according to the set failure rates and fault scenarios to simulate the changes in the operation state of the distribution network when a fault occurs.

[0050] S3h. Run the digital twin model to perform power flow calculation and fault analysis simulation tasks on the distribution network, and record the time of the simulation calculation and the consumption of computing resources.

[0051] S3i. For different failure rates and fault scenarios, repeat the above process of fault injection and simulation calculation to obtain multiple sets of computing power performance data.

[0052] S3j. Define the metrics for evaluating the computing power performance of the distribution network, and calculate the computing power metrics under different failure rates based on the simulated recorded data.

[0053] Among them, the defined computing power metrics include simulation calculation time, average CPU utilization rate, peak memory occupancy, etc. Calculate the computing power metrics under different failure rates according to the simulated recorded data. For example, the average CPU utilization rate can be calculated as , where is the CPU utilization rate at time t, and T is the total simulation time. In actual calculation, the discrete summation method can be used for approximate calculation.

[0054] S3k. Analyze the changing trends of the computing power metrics under different failure rates, and evaluate the computing power bearing capacity of the distribution network under different fault conditions.

[0055] Specifically, the linear regression method can be used to analyze the relationship between the computing power metrics and the failure rate, establish a mathematical model, and predict the computing power performance under different failure rates. For example, for the simulation calculation time T and the failure rate λ, assume their relationship is T = aλ + b, and determine the coefficients a and b by least squares fitting.

[0056] S4. According to the power data volume prediction curve and the simulated computing power metrics, establish a fitness function for the communication connection between each functional block and the distribution network twin body, and evaluate its matching degree with the digital twin body communication.

[0057] Specifically, extract the key features (peak value, valley value, average load, load volatility) from the power data volume prediction curve, and extract the key parameters (simulation calculation time, average CPU utilization rate, peak memory occupancy, etc.) according to the simulated computing power metrics; evaluate the fault tolerance of the functional block according to the relationship between the failure rate and the computing power metrics (linear regression model T = aλ + b).

[0058] Adopt the weighted comprehensive scoring method to combine the power demand and the computing power performance:

[0059] is the fitness score of the i-th functional block, is the power demand prediction index (such as peak load), is the load volatility (reflecting stability, which can be obtained by calculating the standard deviation or coefficient of variation of the load change), is the computing power performance (such as simulation time or CPU utilization rate), , , are the weight coefficients, adjusted according to actual needs (for example, in high load areas, more attention is paid to the computing power stability, then the weight should be higher than and , and ), Norm() is the normalization process to ensure consistent dimensions. Use historical data to test the rationality of the function and adjust the weights to match the actual priorities. For example, when the failure rate is high, increase .

[0060] S5. Evaluate the priorities of each time series node according to the fitness index, and stratify the time series nodes according to the evaluation results.

[0061] Specifically, divide the predicted time point or time period into discrete time series nodes (the specific time is defined by yourself according to the actual situation), and for each time series node of the function block , use the fitness function established in step S4 to calculate its fitness index , sort the fitness indexes of the time series nodes according to the size of the fitness index, and evaluate the priorities of each time series node. According to the priorities, divide the time series nodes into different levels. For example, the time series nodes with high priorities can be placed in the first layer, those with medium priorities in the second layer, and those with low priorities in the third layer.

[0062] Among them, , if the fitness index is lower than the lower adjustment threshold, it means that the computing power of this time series node is insufficient or the load exceeds the limit, which may cause a failure, and it has a lower priority. Immediately expand the capacity, adjust the topology or allocate additional computing power resources to ensure the safe and stable operation of the distribution network; , if the fitness index is between the lower adjustment threshold and the upper adjustment threshold, it means that this time series node runs stably but has not exceeded the limit, and it has a medium priority. Continuously monitor it to timely discover potential problems; , if the fitness index is higher than the upper adjustment threshold, it means that this time series node has sufficient resources and runs stably, and it has a higher priority. The topology can be adjusted and computing power resources can be allocated to optimize the overall operation efficiency of the distribution network. The lower adjustment threshold α and the upper adjustment threshold β are determined through historical data statistics.

[0063] S6. Determine the time series nodes that need to be adjusted in the distribution network and perform dynamic programming on the distribution network.

[0064] Specifically, to determine the time series nodes that need to be adjusted in the distribution network, including: S6a. Traverse all the time series nodes of each function block.

[0065] S6b. Compare the fitness index of each time series node with the lower adjustment threshold and the upper adjustment threshold.

[0066] S6c. If the fitness index of the time series node is lower than the lower adjustment threshold, mark this time series node as a weak time series node.

[0067] Among them, for the nodes marked as weak timing nodes, corresponding processing measures should be taken immediately, such as capacity expansion, topology adjustment or additional computing power resources allocation.

[0068] S6d, if the fitness index of the timing node is higher than the upper limit adjustment threshold, then mark this timing node as a strong timing node.

[0069] Among them, for the nodes marked as strong timing nodes, topology adjustment and allocation of computing power resources can be considered to optimize the resource allocation and operation efficiency of the distribution network. For example, allocate some computing power resources of the strong timing node to other nodes in need to achieve reasonable utilization of resources.

[0070] It should be noted that the determination method of the adjusted timing nodes is instantaneously dynamic. Therefore, when traversing the timing nodes, the operation characteristics of the distribution network and the data sampling frequency should be considered to ensure that the adjusted timing nodes take into account both real-time performance and stability.

[0071] The technical solutions in the above embodiments of the present application at least have the following technical effects or advantages: By dividing the distribution network into functional blocks, classifying data collection, and predicting demands, the electricity consumption characteristics of each region can be accurately grasped. Digital twin modeling and computing power simulation can evaluate the system operation status and computing power requirements in advance to ensure real-time monitoring and decision-making of the distribution network. Based on the power data prediction curve and computing power index, a fitness function is established to accurately evaluate the communication matching degree between the functional block and the digital twin. By stratifying the timing nodes according to the fitness index and dynamically planning, potential problems can be discovered in time, resource allocation can be optimized, the operation stability and efficiency of the distribution network can be improved, and reasonable utilization and efficient management of resources can be achieved.

[0072] Embodiment 2: In Embodiment 1, when the actual load of a functional block reaches the capacity upper limit and triggers the modification process, after merging the overloaded functional block with adjacent low-load functional blocks and re-partitioning, although the load pressure can be alleviated to a certain extent, there are significant differences in the load characteristics, equipment aging degree, operating environment, etc. of different functional blocks before merging. Moreover, the reliability assessment of the functional blocks after re-partitioning is only based on the conventional reliability index calculation method, without fully considering the comprehensive impact of these differences on the operating stability and reliability of the functional blocks. Under different degrees of differences, the adaptability of the functional blocks to the new load and the collaborative working ability between devices are different. Using a unified reliability assessment method will inevitably lead to inaccurate assessment results and it is difficult to comprehensively reflect the actual operating conditions of the functional blocks. Due to the diversity of the load changes, equipment failures, etc. of different functional blocks before and after merging and during operation in history, there must be differences in the required indicators for judging the reliability of the functional blocks after re-partitioning. In order to more accurately assess the reliability of the functional blocks after re-partitioning and obtain a more scientific and reasonable distribution network operation management plan, it is necessary to further optimize and improve by considering factors such as the degree of difference before merging the functional blocks and the load distribution after re-partitioning at the same time.

[0073] In some embodiments, when dividing the distribution network into multiple functional blocks, step S1 further includes: S11, set an initial load capacity upper limit for each functional block, and obtain the load change situation and the operating state information of the equipment in the functional block in real time.

[0074] Specifically, for each functional block i, according to the rated capacity of the power supply equipment in the functional block and a certain safety margin coefficient k (k ranges from 0.8 to 0.9), calculate the initial load capacity upper limit :

[0075] For example, if the rated capacity of the transformer in a certain functional block is 1000 kVA and the safety margin coefficient k = 0.8$, then the initial load capacity upper limit of this functional block is = 0.8 1000 = 800$kVA.

[0076] Through devices such as smart meters and sensors installed in the functional block, collect load data in real time , where t represents time. At the same time, collect the operating state information of the equipment, such as the switch state (on / off), equipment temperature, fault alarm signal, etc. These information can be collected and transmitted through the SCADA system (Supervisory Control and Data Acquisition system).

[0077] S12. Compare the actual load of the function block with the set upper limit of the load capacity. When the load in the function block reaches the capacity upper limit, immediately trigger the modification process.

[0078] Specifically, within each time interval (such as 1 minute), calculate the actual load of function block i and the ratio to the upper limit of the load capacity :

[0079] When , it indicates that the load of function block i has reached or exceeded the capacity upper limit, and at this time, trigger the modification process.

[0080] S13. Merge the function block with overloaded load with the adjacent function block with a lower load capacity and re - divide the function blocks.

[0081] Specifically, according to the topological structure of the distribution network, determine the set of adjacent function blocks to the overloaded function block i. For each function block j in the set of adjacent function blocks , calculate its remaining load capacity , and select the function block with the largest remaining load capacity as the merge object, that is: .

[0082] Merge function block i and function block to form a new function block. The upper limit of the load capacity of the new function block is the sum of the original upper limits of the load capacities of the two function blocks minus the actual load of function block : . At the same time, update the topological structure information of the distribution network, and manage and monitor the merged function block as a whole.

[0083] S14. Evaluate the calculation reliability of the re - divided function blocks.

[0084] Among them, the reliability index is calculated as follows: Common distribution network reliability indexes include the power supply reliability rate (ASAI), the system average interruption frequency (SAIFI), and the system average interruption duration (SAIDI), etc. Taking the power supply reliability rate as an example for calculation, the calculation formula of the power supply reliability rate ASAI is:

[0085] n is the number of users in the function block, is the power outage time of the k - th user within the statistical period is the duration of the statistical period of the k - th user (usually one year).

[0086] Specifically, for the reliability assessment based on Monte Carlo simulation, for each component (such as transformers, lines, etc.) within the functional block, a probability distribution model of the failure rate λ and repair rate μ is established. According to the reliability model of the component, random numbers are generated to simulate the failure and repair processes of the component. Conduct N simulations, and record the power outage time and power supply status of the functional block for each simulation. According to the simulation results, reliability indicators such as the power supply reliability rate, the system average outage frequency, and the system average outage duration are calculated.

[0087] For example, for a certain simulation, if there are 100 users in the functional block and the statistical period is one year (8760 hours), and the total power outage time obtained from the simulation is 87.6 hours, then the power supply reliability rate for this simulation is:

[0088] By taking the average of the results of N simulations, the final reliability assessment result of the functional block is obtained.

[0089] The technical solutions in the embodiments of the present application described above have at least the following technical effects or advantages: By setting the upper limit of the load capacity of the functional block and monitoring the load changes in real time, the situation of overloaded load can be detected in time, the modification process can be quickly triggered, and equipment overload and power supply risks can be avoided. Merging and re-dividing the overloaded functional block with adjacent low-load functional blocks can optimize the load distribution and improve the resource utilization rate. At the same time, evaluating the reliability of the functional block after re-division can ensure the stable operation of the entire distribution network, reduce the probability of power outage faults, enhance the power supply quality, provide better-quality and more reliable power services for users, and promote the development of the distribution network towards intelligence and high efficiency.

[0090] Embodiment 3: Although the dynamic adjustment of the functional block is achieved in Embodiment 2, each functional block still operates relatively independently and lacks a collaborative working mechanism. In actual operation, when facing emergencies such as high load and equipment failures, or when multiple functional blocks need to cooperate, this independent operation mode exposes its disadvantages. The independently operating functional blocks cannot respond in time, are difficult to quickly provide sufficient support or achieve effective resource allocation, and may lead to problems such as unstable power supply and untimely fault handling. To make up for this deficiency, break the isolation between functional blocks, establish an efficient collaborative working mechanism, and achieve resource sharing and complementary advantages, Embodiment 3 is specifically introduced.

[0091] In some embodiments, in step S13, when re-dividing the functional block, it further includes: S131, deploy high-precision sensors inside each functional block, and transmit the collected data to the central intelligent control center of the distribution network in real time.

[0092] Among them, high-precision sensors include load sensors (accurately measuring real-time power, current, voltage, etc.), equipment status sensors (monitoring transformer temperature, number of switch operations, degree of line aging, etc.), and environmental sensors (recording factors such as temperature, humidity, and wind speed that may affect power operation).

[0093] S132. Determine the cross-region collaboration requirements for function blocks, and identify function blocks that can share regions for collaborative work.

[0094] Among them, define as the load growth rate threshold, and calculate the load growth rate of the calculation function block of the cross-region collaboration requirement determination index: Calculate the load growth rate r of the function block within a period of time Δt. The formula is:

[0095] P(t) is the real-time power. When r exceeds the preset threshold , it is determined that there is a high-load collaboration requirement. According to the historical fault data and real-time status data of the equipment, use the Bayesian algorithm to calculate the fault probability of the equipment . For example, when calculating the fault probability of a transformer, factors such as temperature and load rate can be considered.

[0096] Specifically, identify function blocks that can share regions for collaborative work, including: Based on the geographical information system (GIS) and electrical topological structure of the distribution network, use the Dijkstra algorithm to calculate the electrical distance between function blocks . Considering the line capacity , electrical distance factors, construct a shared region identification model, and use the weighted scoring method to calculate a shared region score for each function block . The formula is:

[0097] , is the weight coefficient, is the maximum line capacity. Sort the scores in descending order, and select function blocks starting from the highest score to form a shared region.

[0098] S133. Evaluate each function block in the shared region to determine the combination of function blocks participating in the cooperation.

[0099] Specifically, evaluate each function block in the shared region, including: Calculate the load capacity, and calculate the remaining load capacity of the function block , is the upper limit of the load capacity of the function block.

[0100] Calculate the reliability of the computing device, and use the FTA (Fault Tree Analysis) method to evaluate the reliability of the device (The specific calculation method is determined according to the device structure and failure mode).

[0101] Calculate the power generation capacity. For the distributed power generation function block, its power generation capacity refers to the electrical energy that the function block can generate within a specific time, and calculate the output power of the distributed power generation .

[0102] Among them, determine the combination of function blocks participating in cooperation. Specifically, use the genetic algorithm to establish a multi-objective optimization model to simultaneously achieve the maximization of the collaborative effect (such as minimizing the total load gap) and the minimization of the resource allocation cost. The objective function is: , is the load gap of the i-th function block, is the cost of the j-th resource allocation, w is the weight coefficient. By minimizing the objective function, while meeting the load demand, the resource allocation cost can be reduced as much as possible, and the economy and efficiency of collaborative work can be improved. The specific method for determining the combination of function blocks participating in cooperation based on the genetic algorithm refers to the existing technology and will not be elaborated in this article.

[0103] S134, calculate the electrical energy consumption of the function blocks with collaborative requirements, and allocate resources according to the required electrical energy in the merged shared area.

[0104] Among them, the calculation formula for the electrical energy consumption of the function blocks with collaborative requirements is: W = P × t, where the electrical energy W is equal to the electric power P multiplied by the power consumption time t, S135, the central intelligent control center adjusts the topology of the distribution network in real time through the remote control system.

[0105] S136, after the collaborative work ends, resume the independent operation of each function block.

[0106] Specifically, after the collaborative work is completed, according to the evaluation of the collaborative work effect and the actual demand situation, each functional block is restored to independent operation. A comprehensive evaluation is carried out on various indicators during the collaborative work, such as the improvement of power consumption efficiency, the rationality of resource allocation, and the improvement degree of system stability. By comparing the relevant data before and after the collaborative work, it is analyzed whether the collaborative work meets the expected goals. Consider the actual demand changes after the collaborative work ends, such as whether the load demand has returned to the normal level, whether there are new production tasks or energy supply requirements, etc. If the collaborative work effect is good and there are no major changes in the actual demand, the central intelligent control center issues an instruction to restore each functional block to the independent operation state before the collaborative work, and adjust the operation parameters and control strategies of each functional block. If the collaborative work effect is not ideal or there are major changes in the actual demand, the collaborative strategy is optimized and adjusted, or some collaborative operation modes are continued to be maintained to meet the actual demand. During the restoration process, closely monitor the operation status of each functional block to ensure the smooth progress of the restoration work.

[0107] The technical solutions in the embodiments of the present application at least have the following technical effects or advantages: By deploying high-precision multi-type sensors inside the functional block, key data can be accurately collected and transmitted to the control center in real time, providing a reliable basis for subsequent analysis and decision-making. The cross-region collaborative demand determination and shared area identification method comprehensively considers various factors, can accurately identify collaborative demands and reasonably divide shared areas. A comprehensive evaluation and determination of combinations of functional blocks in the shared area are carried out, and a multi-objective optimization model is adopted to take into account both collaborative effects and costs. Calculate the electricity energy and rationally allocate resources, and adjust the topological structure in real time, improving the operation efficiency and reliability of the distribution network. After the collaboration ends, the operation of the functional block is restored according to the actual situation, ensuring the flexible and stable operation of the power system.

[0108] Embodiment 4: In Embodiment 3, when the distribution network faces cross-region electric energy demand allocation, a conventional method based on fixed rules and simple priorities is used for electric energy scheduling. However, in actual operation, the electric energy transmission paths in the regions where different functional blocks are located are complex and changeable, the capacity limits of each path are affected by various factors and change dynamically over time, and the electricity demand and power generation capacity vary significantly in different time periods. Moreover, only relying on simple priorities for electric energy allocation cannot fully consider the dynamic adjustment of transmission capacity and the real-time changes in the grid operation state, resulting in an inflexible and inaccurate scheduling plan when facing complex and changeable grid conditions, making it difficult to meet the actual electric energy allocation requirements and prone to problems such as unstable electric energy supply and low transmission efficiency. In order to perform cross-region electric energy scheduling of the distribution network more scientifically and efficiently, fully considering the electric energy transmission characteristics, dynamic change factors, and multi-objective optimization, it is necessary to further optimize and improve the scheduling plan, so Embodiment 4 is proposed.

[0109] In some embodiments, in step S134, the electrical energy consumption of the functional blocks with collaborative requirements is calculated, and the merged shared area is allocated resources according to the required electrical energy, further including: S1341. Analyze the electrical energy transmission path, capacity limit, and time characteristics between different regions of the functional blocks, and construct a mathematical model describing the law of electrical energy transmission between regions.

[0110] Specifically, conduct a detailed survey of the electrical energy transmission paths between the regions where different functional blocks are located in the distribution network, record information such as the line orientation and connection nodes, measure the capacity limits of each transmission path through actual measurement, and set the capacity limit of the i-th path as , analyze the time characteristics, including the electricity demand and power generation capacity in different time periods. For example, count the electricity consumption in different time periods of each day in the past year to obtain the electricity demand function D(t) (unit: kW) in each time period, where t represents time.

[0111] Based on circuit theory, establish the node current equation and loop voltage equation for electrical energy transmission. Taking the node current equation as an example, for node j, there is , represents the current flowing from node i to node j, and n is the number of nodes connected to node j. Combining graph theory knowledge, abstract the distribution network into a graph structure, where nodes represent the regions where functional blocks are located, edges represent transmission paths, and the weights of the edges are transmission capacities. Calculate the shortest transmission path and the maximum transmissible electrical energy between different regions through graph theory algorithms (such as Dijkstra algorithm).

[0112] S1342. According to the priority of the time sequence nodes, count the supply data of the shared area, and according to the time sequence priority rule, preferentially allocate electrical energy from the regions with sufficient electrical energy supply.

[0113] Specifically, collect the electrical energy supply data of the shared area and its surrounding areas, including the output power of the power generation equipment (unit: kW), the remaining power of the energy storage device (unit: kWh), etc., where k represents the region number. According to the priority of the time sequence nodes, divide the tasks into high-priority and low-priority tasks. Let the high-priority task set be H and the low-priority task set be L.

[0114] Preferentially allocate electrical energy from the regions with sufficient electrical energy supply. For the high-priority task h∈H, calculate the electrical energy required at time t, sort the regions from high to low according to the degree of sufficient supply, and sequentially allocate electrical energy from the regions with sufficient supply until or the electrical energy in the supply region is insufficient. For the low-priority task , after the high-priority tasks are allocated, allocate according to the remaining electrical energy situation.

[0115] S1343. Take the transmission capacity and timing priority as the key factors affecting power scheduling, and set them as dynamic variables.

[0116] Among them, the key factors are determined as follows: take the transmission capacity C(t) and timing priority as the key factors affecting power scheduling. The transmission capacity C(t) can be dynamically adjusted according to the actual load capacity of the line and the operation requirements of the power grid. Let the timing priority factor be , for high-priority tasks, = 1; for sub-priority tasks, = 0.5 (which can be adjusted according to the actual situation).

[0117] S1344. Take the timing priority and power supply reliability as the goals, and construct a multi-objective optimization function.

[0118] Specifically, the timing priority goal is to ensure the stability of power supply for high-priority tasks. Let the proportion of time when the power supply for high-priority tasks is insufficient be , then the timing priority objective function is . The reliability goal is to set the power grid outage probability as , and the reliability objective function is . The economic goal is to set the power scheduling cost as Cost, including transmission loss cost, generation cost, etc. The economic objective function is .

[0119] The comprehensive multi-objective optimization function is , where , , are the weights of each objective, which can be adjusted according to the actual situation.

[0120] S1345. Continuously collect real-time monitoring data of the power grid operation, analyze the power grid status and change trend, and dynamically adjust the weights and value ranges of each dynamic variable according to the analysis results.

[0121] Specifically, continuously collect real-time monitoring data of the power grid operation, including line current, voltage, power, power consumption demand in each region, etc. Use data analysis algorithms to analyze the real-time data and predict the operation status and change trend of the power grid. For example, use a time series model to predict the power consumption demand in the next period of time , according to the analysis results, dynamically adjust the weights and value ranges of each dynamic variable. For example, when it is predicted that the load of a certain line will increase, reduce the value range of the transmission capacity variable C(t) of this line, and correspondingly increase the weight of the timing priority factor in the objective function .

[0122] S1346, when the power cooperation demand requested by the functional block exceeds the supply capacity of the shared area, the multi-objective optimization calculation is automatically triggered.

[0123] Specifically, when the functional block requests the cooperative power demand exceeds the supply capacity S(t) of the shared area, that is , the multi-objective optimization calculation is automatically triggered. The genetic algorithm is selected to initialize the population, and the values of each dynamic variable are used as the chromosome genes. The objective function F is continuously optimized through operations such as crossover and mutation. After multiple iterative optimizations, the optimal scheduling scheme is obtained, including the power distribution of each transmission path, the power supply time of each area, etc.

[0124] The technical solutions in the above embodiments of the present application have at least the following technical effects or advantages: By analyzing and modeling the power transmission path, capacity limit and time characteristics among functional blocks in the distribution network, the power transmission law is accurately grasped. The power is allocated based on the time sequence priority to ensure stable power supply for high-priority tasks and take into account the requirements of sub-priority tasks. The transmission capacity and time sequence priority are set as dynamic variables to construct a multi-objective optimization function, which can dynamically adapt to the changes in the power grid. Continuously collect real-time data to adjust the variable weights and value ranges to make the scheduling more flexible. When the cooperative power demand exceeds the supply capacity, the optimization calculation is automatically triggered to obtain the optimal scheduling scheme, effectively improving the overall scheduling efficiency and operation reliability of the distribution network.

[0125] The above is only the preferred embodiment of the present invention and is not used to limit the present invention. For those skilled in the art, the present invention can have various changes and modifications. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A distribution network intelligent planning method, characterized in that: include: S1, divide the distribution network into multiple functional blocks, collect and classify the power consumption data of power users in the functional blocks; S2, generating a power data change curve based on the power consumption data of each functional block, predicting the power demand of each functional block, and generating a power consumption data amount prediction curve corresponding to each functional block; S3, conduct digital twin modeling and computing power simulation for each functional block, and record the simulation computing power indicators; S4, based on the power data volume prediction curve and the simulation computing power index, establish a fitness function for each functional block to communicate with the distribution network twin and evaluate its matching degree with the digital twin communication; S5, evaluate the priority of each time series node according to the fitness index, and stratify the time series nodes according to the evaluation results; S6, determining the timing nodes that need to be adjusted in the distribution network, and dynamically planning the distribution network.

2. The distribution network intelligent planning method according to claim 1, characterized in that: S3 performs digital twin modeling and computing power simulation on each functional block, specifically including: S3a, collects physical parameters and historical operation data of the distribution network, and establishes a distribution network topology model and electrical characteristic model; S3b, integrates the topology model and the electrical characteristic model to form a complete digital twin model of the distribution network; S3c, establish a data interface with the distribution network monitoring system to obtain the operation data of the power grid in real time; S3d, dynamically adjusts the digital twin model based on real-time data and changes in the grid’s operating status; S3e, analyzes the possible fault types in the distribution network, determines the location and impact range of the fault for different fault types, and sets the failure rate of different equipment based on historical fault data and equipment reliability indicators; S3f, use MATLAB to perform simulation calculations on the distribution network and determine the computing resources used for computing power simulation; S3g, injects faults into the digital twin model according to the set fault rate and fault scenario to simulate the changes in the operating status of the distribution network when the fault occurs; S3h, run the digital twin model to perform power flow calculation and fault analysis simulation tasks of the distribution network, and record the simulation calculation time and computing resource consumption; S3i, for different failure rates and failure scenarios, repeatedly executes the above fault injection and simulation calculation process to obtain multiple sets of computing power performance data; S3j, defines the indicators used to evaluate the computing power performance of the distribution network, and calculates the computing power indicators under different failure rates based on the simulated recorded data; S3k, analyzes the changing trend of computing power indicators under different failure rates, and evaluates the computing power tolerance of the distribution network under different fault conditions.

3. The distribution network intelligent planning method according to claim 1, characterized in that: In S4, based on the power data volume prediction curve and the simulated computing power index, it includes: using a weighted comprehensive scoring method to combine power demand with computing power performance: is the fitness score of the ith functional block, is the electricity demand forecast indicator, is the load fluctuation rate, For computing power performance, , , is the weight coefficient, which can be adjusted according to actual needs. Norm() is the normalization process.

4. The distribution network intelligent planning method according to claim 1, characterized in that: The S6 specifically includes: S6a, traversing all timing nodes of each functional block; S6b, comparing the fitness index of each timing node with the lower limit adjustment threshold and the upper limit adjustment threshold; S6c, if the fitness index of the timing node is lower than the lower limit adjustment threshold, marking the timing node as a weak timing node; S6d, if the fitness index of the timing node is higher than the upper limit adjustment threshold, marking the timing node as a strong timing node.

5. The distribution network intelligent planning method according to claim 1, characterized in that: The step S1 of dividing the distribution network into a plurality of functional blocks further comprises: S11, setting an initial load capacity upper limit for each functional block, and obtaining load changes within the functional block and equipment operation status information in real time; S12, comparing the actual load of the function block with the set load capacity upper limit, when the load in the function block reaches the capacity upper limit, immediately triggering the modification process; S13, merging the overload function with the adjacent function block with low load capacity to re-divide the function block; S14, evaluating the computational reliability of the re-divided functional blocks.

6. The distribution network intelligent planning method according to claim 5, characterized in that: The step S13 includes: determining a set of function blocks adjacent to the load overload function block i according to the topological structure of the distribution network; , for the set of adjacent functional blocks For each functional block j in the , calculate its remaining load capacity and select the functional block with the largest remaining load capacity As the merge object; function block i and function block Merge to form a new functional block. The upper limit of the load capacity of the new functional block is the sum of the upper limits of the original load capacity of the two functional blocks minus the upper limit of the functional block The actual load of the distribution network is updated at the same time.

7. The distribution network intelligent planning method according to claim 5, characterized in that: In the S13, re-dividing the functional blocks further includes: S131, deploys high-precision sensors in each functional block to transmit the collected data to the central intelligent control center of the distribution network in real time; S132, determining the cross-regional collaboration requirements of the functional blocks, identifying the functional blocks that can perform collaborative work and share the area; wherein, definition is the load growth rate threshold, the load growth rate of the calculation function block of the cross-region collaborative demand determination index: the load growth rate r of the calculation function block within a period of time Δt is calculated as follows: P(t) is the real-time power. When r exceeds the preset threshold When , it is determined that there is a high load coordination demand; S133, evaluating each functional block in the shared area to determine a combination of functional blocks participating in the cooperation; S134, calculating the power consumption of the functional blocks with coordination requirements, and allocating resources in the merged shared area according to the required power consumption; S135, the central intelligent control center adjusts the topological structure of the distribution network in real time through the remote control system; S136, after the collaborative work is completed, each functional block is restored to independent operation.

8. The method for intelligent planning of a distribution network according to claim 7, characterized in that: The identification of functional blocks that can perform collaborative work and share areas includes: calculating the electrical distance between functional blocks using the Dijkstra algorithm based on the geographic information system and electrical topology of the distribution network ; Comprehensively consider the line capacity , Electrical distance Factors, build a shared area recognition model, and calculate a shared area score for each functional block , the formula is: , is the weight coefficient, For the maximum capacity of the line, the scores are sorted in descending order, and functional blocks are selected to form the shared area starting from the highest score.

9. The distribution network intelligent planning method according to claim 7, characterized in that: In the S134, the power consumption of the functional blocks with coordination requirements is calculated, and the resources of the merged shared area are allocated according to the required power consumption, and the following further includes: S1341, analyzing the power transmission path, capacity limitation and time characteristics of the functional blocks between different areas, and constructing a mathematical model describing the law of power transmission between areas; S1342, according to the priority of the timing node, counting the supply data of the shared area, and according to the timing priority rule, preferentially allocating electric energy from the area with sufficient electric energy supply; S1343, taking transmission capacity and timing priority as key factors affecting power scheduling, and setting them as dynamic variables; S1344, construct a multi-objective optimization function with timing priority and power supply reliability as the goals; S1345, continuously collect real-time monitoring data of power grid operation, analyze power grid status and change trends, and dynamically adjust the weight and value range of each dynamic variable according to the analysis results; S1346, when the electric energy cooperation demand requested by the functional block exceeds the supply capacity of the shared area, the multi-objective optimization calculation is automatically triggered.

10. The distribution network intelligent planning method according to claim 9, characterized in that: The S1344 specifically includes: the timing priority target is to determine the power supply stability of the high-priority task, and the proportion of the time when the power supply of the high-priority task is insufficient is set to , then the timing priority objective function is ; The reliability target is to assume that the power outage probability of the power grid is , the reliability objective function is ; The economic goal is to set the power dispatch cost as Cost, and the objective function is ; The comprehensive multi-objective optimization function is ,in , , is the weight of each target.

Images