Fast power supply recovery method and device based on source-load-storage cooperative power grid partition optimization

Abstract

The invention belongs to the technical field of power supply recovery, and particularly relates to a rapid power supply recovery method and device based on source-load-storage cooperative power grid partition optimization, and the method comprises the steps: S1, respectively constructing an electrical coupling degree index, a source-load-storage power balance index, a grid structure index and a node scale index, and comprehensively forming a power grid partition comprehensive index; s2, taking the comprehensive index of the power grid partition as a fitness function, designing an energy entropy driven hybrid selection mechanism to improve a simulated annealing genetic algorithm, and solving the power grid partition by using the improved simulated annealing genetic algorithm; s3, on the basis of the power grid partitions, taking each partition as a unit, and establishing an optimal scheduling model taking the maximum recovery of the important load as a target; and S4, solving the optimal scheduling model by adopting a particle swarm algorithm to obtain a rapid power supply recovery scheme. Priority recovery of the key load can be realized, and rapid recovery capability and reliability of the power grid are enhanced.

Description

Technical Field

[0001] The present invention belongs to the technical field of power supply restoration, and particularly relates to a fast power supply restoration method and device based on coordinated grid partition optimization of source, load and energy storage. Background Art

[0002] When the power grid is operating normally, it may encounter unconventional events, which refer to sudden disasters (such as extreme weather, earthquakes, cyber attacks) or major equipment failures (such as main transformer explosions, transmission line cascading outages) that exceed the conventional disaster tolerance capacity of the power grid. Such events often lead to the destruction of the power grid topology and local or even global power outages. Therefore, it is necessary to quickly restore power supply. Traditional power supply restoration technologies mainly rely on three types of methods: relay protection devices, centralized control, and distributed control. Relay protection passively isolates faults by presetting thresholds and cannot dynamically adapt to complex and changeable unconventional event scenarios; centralized control relies on a central decision-making system and is prone to response delays and single-point failures in case of communication interruptions or large-scale failures; although distributed control improves local autonomy, it lacks a global coordination mechanism, and frequent policy conflicts occur between regions. Existing methods generally have defects such as slow restoration speed, poor strategy adaptability, and low important load guarantee rate in extreme events.

[0003] The coordinated restoration method based on power grid partition improves the restoration efficiency by dividing autonomous regions. However, existing research has significant limitations: partition indicators mostly focus on a single dimension (such as electrical distance, static source-load balance), ignoring key factors such as the dynamic regulation ability of energy storage and the voltage stability of nodes; partition algorithms (such as traditional genetic algorithms, tabu search) are prone to falling into local optima and are difficult to balance topological compactness and functional autonomy; the cross-region mutual assistance mechanism is lacking after partitioning, resulting in insufficient distributed power generation consumption capacity and limited flexibility of restoration strategies.

[0004] Research on the island operation of distributed power generation mostly focuses on local power balance and realizes island division through real-time monitoring and predictive scheduling. However, existing models overly simplify the objective function and do not incorporate power supply reliability, voltage deviation, energy storage state of charge attenuation, etc. into the multi-objective optimization framework. In addition, the coordinated mechanism between island division and power grid partition has not been established, and a three-dimensional restoration architecture of "partition autonomy - island complementarity - global mutual assistance" cannot be formed, restricting the spatio-temporal coordination potential of source-load-energy storage resources.

[0005] In summary, there are three core contradictions in the current technical system: the structural imbalance between single-dimensional partition indicators and multi-dimensional operation requirements; the game dilemma between algorithm convergence and global optimization ability; the lack of coordination between local restoration strategies and global resource allocation. These defects make it difficult for existing methods to achieve fast and accurate power supply restoration in unconventional events, and there is an urgent need to construct a new solution that integrates multi-dimensional evaluation indicators, intelligent optimization algorithms, and cross-region coordination mechanisms. Summary of the Invention

[0006] In view of the deficiencies in the above prior art, the object of the present invention is to provide a rapid power supply restoration method and device based on source-load-storage coordinated power grid zoning optimization, which can achieve the priority restoration of critical loads and enhance the rapid restoration ability and reliability of the power grid.

[0007] To achieve the above object, the present invention provides a rapid power supply restoration method based on source-load-storage coordinated power grid zoning optimization, including the following steps: S1. Respectively construct an electrical coupling degree index C, a source-load-storage power balance index , a grid structure index and a node scale index , and comprehensively form a comprehensive power grid zoning index , expressed as: (1); In the formula, , , , are the weight coefficients of C, , , respectively; S2. Take as the fitness function, and design a hybrid selection mechanism driven by energy entropy to improve the simulated annealing genetic algorithm, and use the improved simulated annealing genetic algorithm to solve the power grid zoning; S3. On the basis of the power grid zoning, taking each zone as a unit, establish an optimal scheduling model with the goal of maximizing the restoration of important loads; S4. Use the particle swarm algorithm to solve the optimal scheduling model to obtain a rapid power supply restoration plan.

[0008] As a preferred solution of the present invention, in S1, the construction process of the electrical coupling degree index C is as follows: The electrical coupling degree between node i and node j is expressed as: (2); In the formula, is the power interaction between node i and node j; , are the node voltages of node i and node j respectively; In the power grid zoning, when node i is in zone A and node j is in zone B, the inter-zone coupling degree between zone A and zone B is expressed as: (3); In the formula, is the number of connection nodes between zone A and zone B; When both node i and node j are located in partition A, the intra-partition coupling degree of partition A is expressed as: (4); In the formula, is the number of node pairs within partition A; The electrical coupling degree index C is comprehensively formed as: (5); In the formula, is 's weight coefficient; is 's weight coefficient.

[0009] As a preferred embodiment of the present invention, in the above S1, the source-load-storage power balance index is expressed as: (6); In the formula, , , are weight coefficients; represents the minimum net power difference of partition k; is the total load of partition k; is the available energy margin of energy storage in partition k; is the total capacity of energy storage in partition k; is the supply-demand matching degree of partition k; is the maximum allowable matching deviation; Among them, is expressed as: (7); In the formula, is the set of power generation nodes within partition k, and x represents one of the power generation nodes; is the set of energy storage nodes within partition k, and m represents one of the energy storage nodes; is the set of load nodes within partition k, and y represents one of the load nodes; is the output of power generation node x; is the discharge power of energy storage node m; is the demand of load node y; is expressed as: (8); In the formula, is the current state of charge of energy storage node m; is the minimum allowable state of charge of energy storage node m; is the capacity of energy storage node m; It is expressed as: (9); Wherein, T is the time window; represents the power generation power of partition k at time t; represents the energy storage discharge power of partition k at time t; represents the load demand of partition k at time t.

[0010] As a preferred solution of the present invention, in the above S1, the grid structure index is expressed as: (10); Wherein, Q is the modularity index; when node i and node j are in partition k, is the electrical distance between node i and node j, is the power transmission capacity between node i and node j; is the weight coefficient for balancing the importance of topological compactness and power transmission capacity.

[0011] As a preferred solution of the present invention, in the above S1, the node scale index is expressed as: (11); Wherein, represents the average number of nodes: K is the total number of partitions; is the number of nodes in partition k; (12); Wherein, N is the total number of nodes in the whole network.

[0012] As a preferred solution of the present invention, in the above S2, taking as the fitness function, and taking the partition result as the problem to be solved for optimization, the cooling function and Metropolis criterion in the simulated annealing genetic algorithm are expressed as: (13); (14); Wherein, , respectively represent the temperatures of the z-th generation and the (z + 1)-th generation; a is the annealing temperature coefficient, 0 < a < 1; , respectively represent the fitness values of the z-th generation and the (z + 1)-th generation; P represents the acceptance probability of transferring from the z-th generation to the (z + 1)-th generation in the simulated annealing algorithm; Design an energy entropy-driven hybrid selection mechanism, and the energy entropy calculation formula is as follows: (15); Wherein, represents the energy entropy of the current population; is the Boltzmann probability of individual u, that is, the probability that individual u is selected at the current temperature ; M is the population size, that is, the total number of individuals in the population; , are the individual energies of individual u and individual v respectively, that is, the reciprocals of the fitness values; Set the threshold of energy entropy , according to the selection-acceptance scheme: If , roulette wheel selection is adopted; if , tournament selection is adopted; Combine the Metropolis criterion to accept the new solution, but the acceptance probability is negatively correlated with the energy entropy, expressed as: (16); Wherein, is the energy difference between the new solution and the current solution; is the maximum possible value of the energy entropy.

[0013] As a preferred solution of the present invention, in the S3, the objective function F of the optimal scheduling model is: (17); (18); Wherein, is the weight of load h, load h is located in partition k, and K is the total number of partitions; is the load recovery amount of load h; is the time penalty coefficient; is the total time from the occurrence of the fault to the restoration of power supply; is the start-up time of the black start power source; is the key switch operation time, and the key switches include sectional switches, tie switches, black start power source access switches, bus tie switches, and automatic reclosing switches; is the power balance adjustment time within the partition; The calculation method of (19); Wherein, is the start-up time of the black start power source within partition k; is whether the black start power source is enabled in partition k, which is a 01 variable, indicates that the black start power source in partition k is enabled, Indicates that the black start power supply of partition k is not enabled; The calculation method of is: (20); In the formula, Is the operation time of switch s; Indicates whether switch s is operated, and it is a 01 variable. Indicates that switch s is operated. Indicates that switch s is not operated; S is the total number of switches. The calculation method of is: (21); In the formula, Is the power capacity of partition k; Is the energy storage output within partition k; Is the load demand within partition k; Is the regulation rate of partition k.

[0014] As a preferred solution of the present invention, the constraint conditions of the optimal scheduling model include: Power balance constraint: (22); In the formula, Is the load of partition k at time t; Is the wind power output of partition k at time t; Is the photovoltaic output of partition k at time t; Is the charge / discharge power of the energy storage within partition k at time t; Is the power transmitted between partition k and the superior power grid at time t; Is the power transmission between each partition of partition k at time t; Energy storage power and capacity constraints: (23); In the formula, And respectively Are the energy storage capacities at time t and t-1; Is the charge / discharge variable of the energy storage at time t. When The energy storage is in the charging state. When The energy storage is in the discharging state; Is the energy storage charging efficiency; Is the time interval; Is the energy storage discharging efficiency; Voltage and frequency constraints: (24); In the formula, U represents voltage and f represents frequency; Energy storage state of charge: (25); In the formula, and are the upper and lower limits of the state of charge; is the initial state of charge; is the state of charge of energy storage I; Partition autonomy constraint: (26); In the formula, O is the partition autonomy constraint index; is the available capacity within partition k; is the maximum load of partition k.

[0015] As a preferred solution of the present invention, in S4, the solution process is as follows: S4.1. Input the partition result, that is, input the pre-divided power grid partition information, including the list of nodes in each area, the source-load-storage configuration parameters, and the capacity limit of the tie lines between partitions; S4.2. Initialize the PSO algorithm parameters, that is, set the parameters of the particle swarm optimization algorithm, including the particle swarm size, particle dimension, maximum number of iterations, initial value of the inertia weight, learning factor, and variable range; S4.3. Randomly initialize the particle positions and velocities, and randomly generate the initial particle swarm according to the variable range, where the position of each particle represents a restoration strategy; S4.4. Calculate the particle fitness value, and evaluate the performance of each particle based on the objective function and constraint conditions of the optimal scheduling model; S4.5. Update the individual historical optimal solution, compare the current fitness value of each particle with its historical optimal value, and update the individual historical optimal position and individual historical optimal fitness value if it is better; S4.6. Update the global historical optimal solution, screen the global optimal solution from all particles, the global historical optimal position, and the global historical optimal fitness value; S4.7. Determine whether the condition is met, whether the maximum number of iterations is reached or the difference in fitness values between two iterations is the smallest. If the condition is met, output the optimal solution and terminate the iteration. If not, update the position vector and velocity vector of each particle and return to step S4.4 to continue the iteration.

[0016] The fast power supply restoration device based on source-load-storage coordinated power grid partition optimization includes a memory, a processor, and a computer program stored on the memory and capable of running on the processor. The above method is implemented by the processor executing the computer program.

[0017] The beneficial effects of the present invention are: The present invention constructs a set of power grid zoning indicators that comprehensively consider electrical coupling degree, source-load-storage power balance, grid structure, and node scale. Compared with the traditional zoning method based on a single indicator, it can more comprehensively and accurately reflect the actual operating state and power supply capacity of the power grid, effectively avoid problems such as power imbalance and voltage over-limit caused by a single indicator, significantly improve the scientificity and rationality of power grid zoning, and lay a solid foundation for subsequent optimal dispatching.

[0018] The present invention introduces the idea of simulated annealing to improve the genetic algorithm, forming the SAGA algorithm, which effectively overcomes the defects of premature convergence and local optimality of the traditional genetic algorithm, significantly improves the optimization ability and performance of the algorithm, makes the power grid zoning results better, can better meet the zoning requirements of the power grid under different operating conditions, and enhances the flexibility and adaptability of power grid operation.

[0019] The present invention establishes a multi-objective zoning rapid restoration optimization model with each zone as a unit, assigns higher weights to critical loads, and gives priority to restoring their power supply, so as to effectively prevent and reduce the harm caused by power outages, relieve the pressure on the power grid, minimize economic losses, significantly improve the rapid restoration ability and power supply reliability of the power grid in the face of unconventional events, and better meet the urgent need of modern power grids for high-reliability and rapid power restoration.

[0020] The present invention also fully considers the dynamic characteristics and uncertain factors of power grid operation, leaving room for further development of dynamic zoning methods, introduction of robust optimization or stochastic optimization methods, etc., making the entire scheme more forward-looking, practical and sustainable, and better able to adapt to the new trends and challenges of future power grid development. Brief Description of the Drawings

[0021] Figure 1 is the process schematic diagram of the present invention; Figure 2 is the flowchart of the SAGA algorithm in the present invention; Figure 3 is the flowchart of particle swarm optimization solution combined with power grid zoning in the present invention; Figure 4 is the structure diagram of each zone of the power grid with distributed energy storage in the verification process of the present invention; Figure 5 is the corresponding structure diagram of the power grid zoning of Scheme 1 in the verification process of the present invention; Figure 6 is the corresponding structure diagram of the power grid zoning of Scheme 2 in the verification process of the present invention; Figure 7 is the schematic diagram of the iterative curve of comprehensive indicators in the verification process of the present invention; Figure 8It is a schematic diagram comparing the load recovery times of two schemes during the verification process of the present invention; Figure 9 It is a schematic diagram comparing the voltage levels of two schemes during the verification process of the present invention. Detailed implementation manners

[0022] The embodiments of the present invention will be further described below with reference to the accompanying drawings: Embodiment 1: As Figure 1 shown, the fast power supply restoration method based on source-load-storage coordinated power grid zoning optimization includes the following steps: S1. Respectively construct the electrical coupling degree index C, the source-load-storage power balance index , the grid structure index and the node scale index , and comprehensively form the comprehensive power grid zoning index , expressed as: (1); In the formula, , , , are the weight coefficients of C, , , respectively; S2. Take as the fitness function, and design a hybrid selection mechanism driven by energy entropy to improve the simulated annealing genetic algorithm. Use the improved simulated annealing genetic algorithm (SAGA algorithm) to solve the power grid zoning; S3. On the basis of the power grid zoning, take each zone as a unit to establish an optimal scheduling model with the goal of maximizing the restoration of important loads; S4. Use the particle swarm algorithm to solve the optimal scheduling model to obtain the fast power supply restoration plan.

[0023] In S1, the construction process of the electrical coupling degree index C is as follows: One of the goals of zoning is to minimize the electrical coupling degree between different regions. By reasonably dividing the regions, the electrical coupling degree between nodes within each region is relatively high, while the electrical coupling degree between different regions is relatively low, so as to achieve the optimization of the system.

[0024] The electrical coupling degree between node i and node j is expressed as: (2); In the formula, is the power interaction between node i and node j; , are the node voltages of node i and node j respectively; If the electrical coupling degree between two nodes is low, it indicates that the electrical connection between them is weak. When partitioning, these nodes can be preferentially considered for different regions because the mutual influence between them is small, and it will not have a great impact on the overall performance of the system after partitioning. On the contrary, if the electrical coupling degree between two nodes is high, it indicates that the electrical connection between them is strong. When partitioning, these nodes should be avoided being assigned to different regions as much as possible because the mutual influence between them is large, and it may lead to a decline in system performance or an increase in the risk of fault propagation after partitioning.

[0025] In power grid partitioning, minimizing the inter-region coupling degree and maximizing the intra-region coupling degree are two complementary goals. Minimizing the inter-region coupling degree is to reduce the mutual influence between partitions, while maximizing the intra-region coupling degree is to enhance the autonomy and stability within the partition.

[0026] In power grid partitioning, when node i is in partition A (the initial partition here) and node j is in partition B, the inter-region coupling degree between partition A and partition B is expressed as: (3); In the formula, is the number of connection node pairs between partition A and partition B; The value of can measure the average electrical coupling strength between two partitions. The smaller the value, the weaker the inter-region interaction and the stronger the partition autonomy.

[0027] When both node i and node j are in partition A, the intra-region coupling degree of partition A is expressed as: (4); In the formula, is the number of node pairs within partition A; To optimize both minimizing the inter-region coupling degree and maximizing the intra-region coupling degree simultaneously, these two indicators can be combined to comprehensively form the electrical coupling degree indicator C: (5); In the formula, is 's weight coefficient; is 's weight coefficient, which is used to balance the importance of the inter-region and intra-region coupling degrees.

[0028] The smaller the value of C, the better. Adjust the weights according to actual needs. If the power grid has high requirements for fault isolation, then is larger. If the power grid has high requirements for the stability within the partition, then Larger.

[0029] In S1, to measure the source-load-storage power balance level, not only the static power matching but also the dynamic capacity margin should be considered, because it reflects the adjustable capacity of the energy storage within the partition and the energy balance potential on the time scale. Similarly, the supply-demand matching degree is very important for measuring the power balance level, which can quantify the spatio-temporal matching characteristics of power generation, energy storage, and load. Considering the partition interaction ability can evaluate the power support ability of the partition in an emergency. The source-load-storage power balance index is expressed as: (6); In the formula, , , are weight coefficients (which can be set according to requirements, the same below); represents the minimum net power difference of partition k; is the total load of partition k; is the available energy margin of the energy storage in partition k; is the total capacity of the energy storage in partition k; is the supply-demand matching degree of partition k; is the maximum allowable matching deviation; Among them, is expressed as: (7); In the formula, is the set of power generation nodes in partition k, and x represents one of the power generation nodes; is the set of energy storage nodes in partition k, and m represents one of the energy storage nodes; is the set of load nodes in partition k, and y represents one of the load nodes; is the output of power generation node x; is the discharge power of energy storage node m (a positive value indicates discharge, and a negative value indicates charge); is the demand of load node y; is expressed as: (8); In the formula, is the current state of charge of energy storage node m (the proportion of the current stored capacity of the energy storage to its total capacity); is the minimum allowable state of charge of energy storage node m; is the capacity of energy storage node m; is expressed as: (9); In the formula, T is the time window (usually taking the value of 24, representing 1 day). represents the power generation of partition k at time t; represents the energy storage discharge power of partition k at time t; represents the load demand of partition k at time t.

[0030] In S1, based on the traditional modularity function used to measure the internal connection tightness of partitions and the isolation degree between partitions, electrical characteristics are introduced, and the electrical distance and power transmission capacity are incorporated into the calculation of the modularity function. According to the load distribution and power generation output, the topology and electrical characteristic weights are dynamically adjusted to improve the adaptability of the index, and the grid structure index is expressed as: (10); In the formula, Q is the modularity index, which is used to measure the degree of partition modularity; when node i and node j are in partition k, is the electrical distance between node i and node j, is the power transmission capacity between node i and node j; is the weight coefficient for balancing the importance of topological tightness and power transmission capacity. The modularity index is an existing index and can be obtained according to the existing technology.

[0031] In S1, the node scale index is expressed as: (11); In the formula, represents the average number of nodes: K is the total number of partitions; is the number of nodes in partition k; (12); In the formula, N is the total number of nodes in the whole network.

[0032] The genetic algorithm is a global optimization algorithm, and its global search ability can ensure that it gradually approaches the optimal solution as the number of iterations increases. When the number of iterations increases, the evolution ability shows a decline phenomenon, resulting in a single population type and obvious lack of diversification. The algorithm is likely to converge locally, thus resulting in the premature phenomenon. The simulated annealing algorithm (SA) has a core idea of finding an approximate global solution to the optimization problem by simulating the high-temperature annealing process, jumping out of the local optimum and continuing to search until the global optimum is obtained.

[0033] In S2, taking as the fitness function, optimizing with the partition result as the problem to be solved, the cooling function and Metropolis criterion in the simulated annealing genetic algorithm are expressed as: (13); (14); Wherein, , respectively represent the temperatures of the z-th generation and the (z + 1)-th generation; a is the annealing temperature coefficient, 0 < a < 1; , respectively represent the fitness values of the z-th generation and the (z + 1)-th generation; P represents the acceptance probability of transferring from the z-th generation to the (z + 1)-th generation in the simulated annealing algorithm; Design an energy entropy-driven hybrid selection mechanism, and the calculation formula of energy entropy is as follows: (15); Wherein, represents the energy entropy of the current population, which measures the population diversity, and the larger the value, the higher the diversity; is the Boltzmann probability of individual u, that is, the probability that individual u is selected at the current temperature ; M is the population size, that is, the total number of individuals in the population; , are the individual energies of individual u and individual v respectively, that is, the reciprocals of the fitness values; ln represents the natural logarithm; Set the threshold of energy entropy , according to the selection-acceptance rule: If , roulette wheel selection is adopted, focusing on diversity; if , tournament selection is adopted, focusing on convergence; Roulette wheel selection is a selection method based on fitness proportion, and the probability of each individual being selected is proportional to its fitness value. The higher the fitness value of an individual, the greater the probability of being selected. Tournament selection is a selection method based on competition. Each time, several individuals are randomly selected from the population for comparison, and the individual with the highest fitness value is selected to enter the next generation population.

[0034] Combine the Metropolis criterion to accept the new solution, but the acceptance probability is negatively correlated with the energy entropy, expressed as: (16); Wherein, is the energy difference between the new solution and the current solution; is the maximum possible value of the energy entropy, used for normalization .

[0035] , it can be set that , is an empirical coefficient, which is dynamically adjusted. A higher threshold is set in the initial stage to maintain diversity, and the threshold is lowered in the later stage to accelerate convergence.

[0036] Combined with the existing genetic algorithm process, the complete flowchart of the SAGA algorithm is as Figure 2 shown. The SA parameters refer to the simulated annealing parameters, and the GA parameters refer to the genetic algorithm parameters. Each individual (solution) consists of multiple genes, and each gene represents the partition label of a node. Then, through the combined mutation and split mutation of genetic operations, the optimal partition result is finally output. The basis for accepting a new solution is , which is the difference in fitness values between the new solution and the current solution, and

[0037] In S3, the objective function F of the optimal scheduling model is: (17); (18); In the formula, is the weight of load h, and load h is located in partition k; is the load recovery amount of load h; is the time penalty coefficient, which is used to balance the recovery amount and the recovery speed; is the total time from the occurrence of the fault to the restoration of power supply; is the start-up time of the black start power source; is the key switch operation time. The key switches include sectional switches, tie switches, black start power source access switches, bus tie switches, and automatic reclosing switches; is the power balance adjustment time within the partition; A black start power source refers to a power source that can start generating electricity relying on its own power resources without relying on external power sources after the entire power system has been powered off due to a fault, and gradually drive other generator sets without self-starting capabilities, ultimately realizing the restoration of the entire power system.

[0038] A sectional switch is a device used to divide a power line into several sections and is usually installed on a distribution line.

[0039] A tie switch is a switching device that connects two or more distribution lines and allows different lines to be connected when needed.

[0040] A black start power source access switch is a key device used to connect a black start power source (such as a hydropower or diesel generator) to the power system.

[0041] A bus tie switch is a switching device that connects different buses within a substation and is used to achieve electrical connection between the buses.

[0042] An automatic reclosing switch is a switching device that can automatically close a disconnected line after detecting an instantaneous fault on the line.

[0043] The obtained fast power supply restoration plan includes switch operation decisions, which ensure the shortest restoration time with the least load shedding under the set constraints; it also includes the power restoration sequence, considering the warm-up time and ramp-up time required for the power sources to determine the wake-up sequence of the black-start power sources; and the transmission of tie lines during inter-regional mutual assistance, where the area containing critical loads is restored first to ensure that the critical loads are not powered off.

[0044] The calculation method of is: (19); In the formula, is the start-up time of the black-start power source in zone k; indicates whether the black-start power source in zone k is enabled, which is a binary variable, means the black-start power source in zone k is enabled, means the black-start power source in zone k is not enabled; The calculation method of is: (20); In the formula, is the operation time of switch s; indicates whether switch s is operated, which is a binary variable, means switch s is operated, means switch s is not operated; S is the total number of switches; The calculation method of is: (21); In the formula, is the power capacity of zone k; is the energy storage output in zone k; is the load demand in zone k; is the regulation rate of zone k.

[0045] The constraint conditions of the optimal scheduling model include: Power balance constraint: (22); In the formula, is the load in zone k at time t; is the wind power output in zone k at time t; is the photovoltaic power output in zone k at time t; is the charge / discharge power of the energy storage in zone k at time t; is the power transmitted between zone k and the superior power grid at time t; is the power transmission between each zone in zone k at time t; Energy storage power and capacity constraints: (23); Wherein, , respectively are the energy storage capacities at time t and t-1; is the charge and discharge variable of the energy storage at time t. When , the energy storage is in the charging state. When , the energy storage is in the discharging state; is the energy storage charging efficiency; is the time interval; is the energy storage discharging efficiency; Voltage and frequency constraints: (24); Wherein, U represents voltage and f represents frequency; p.u. (p.u) is the per-unit value, which is a unit used to represent the relative value of electrical quantities.

[0046] Energy storage state of charge: (25); Wherein, , are the upper and lower limits of the state of charge; is the initial state of charge; is the state of charge of energy storage I; Partition autonomy constraint: (26); Wherein, O is the partition autonomy constraint index; is the available capacity within partition k; is the maximum load of partition k.

[0047] In S4, the solution process is as follows: S4.1. Input the partition result, that is, input the pre-divided power grid partition information, including the list of nodes in each area, the source-load-storage configuration parameters, and the capacity limit of the inter-zone tie lines; S4.2. Initialize the PSO algorithm parameters, that is, set the particle swarm optimization algorithm parameters, including the particle swarm size, particle dimension, maximum number of iterations, initial value of the inertia weight, learning factor, and variable range; S4.3. Randomly initialize the particle positions and velocities, and randomly generate the initial particle swarm according to the variable range, where the position of each particle represents a restoration strategy; S4.4. Calculate the particle fitness value, and evaluate the performance of each particle based on the objective function and constraint conditions of the optimal scheduling model; S4.5. Update the individual historical optimal solution. Compare the current fitness value of each particle with its historical optimal value. If it is better, update the individual historical optimal position and the individual historical optimal fitness value. S4.6. Update the global historical optimal solution. Select the global optimal solution, the global historical optimal position, and the global historical optimal fitness value from all the particles. S4.7. Determine whether the conditions are met, that is, whether the maximum number of iterations is reached or the difference in fitness values between two iterations is the smallest. If the conditions are met, output the optimal solution and terminate the iteration. If not, update the position vector and velocity vector of each particle and return to step S4.4 to continue the iteration.

[0048] The solution flow chart is as Figure 3 shown.

[0049] The verification process is as follows: The structure diagrams of each partition of the power grid with distributed energy storage are constructed, as Figure 4 shown. It mainly includes distributed wind power, photovoltaic, energy storage, and conventional loads. According to the characteristics of the internal nodes of the main grid, the loads, power plants, and distributed energy storage are planned into different regions. Different colored frame lines represent different partitions. According to the loads corresponding to different partitions, fast power restoration strategies for each partition are designed to ensure that the important loads in each partition are quickly restored to power.

[0050] The power grid is partitioned by combining the given typical day data with the IEEE 39 - node system. Four elements, namely electrical coupling degree, node scale, grid structure, and power balance index, are considered. Different weight combinations will produce different power grid partition results. In the case of no emphasis, the weights can be set the same to ensure performance. Scheme 1: Partition the power grid based on the source - load comprehensive index without considering energy storage. Scheme 2: Partition the power grid based on the comprehensive evaluation index considering energy storage. The partition results of different schemes are shown in Table 1. The structure diagrams corresponding to the power grid partitions of Scheme 1 and Scheme 2 are as Figure 5 and Figure 6 shown.

[0051] Table 1 Results of power grid partitioning for different schemes

[0052] As can be seen from Table 1: The power grids of both schemes are divided into 3 regions, and there are no isolated nodes. However, the node distribution in Scheme 2 is more balanced and reasonable compared to Scheme 1. Under the same result of power grid zoning, Scheme 2 considers the power balance of source-load-storage. Its modularity index decreases by 0.02% compared to Scheme 1, the strength of the regional node structure is slightly weaker, but the power balance index increases by 4.2%, the functional index is significantly improved, and the node scale index also increases by 8%, making the node distribution more reasonable. The comprehensive index of Scheme 1 is 0.7089, and the comprehensive index of Scheme 2 is 0.7306, which is 3.1% higher than the former. Therefore, considering the comprehensive index of the modularity, electrical coupling degree index, grid structure index, and node scale index of source-load-storage can ensure good structural properties of the zoning and fully utilize the power complementary characteristics between nodes, meeting the comprehensive requirements of both structure and function.

[0053] Taking Scheme 2 as an example to compare the GA algorithm and the improved SAGA algorithm, the iterative curve of the comprehensive index is as Figure 7 shown. The GA algorithm reaches the maximum value of 0.6952 at about the 20th time. The SAGA algorithm reaches the maximum value of 0.7192 at about the 10th time. Obviously, the improved algorithm is better than the GA in terms of optimization speed and results.

[0054] The multi-objective zoning rapid restoration optimization model (optimization scheduling model) proposed in this embodiment is solved by the particle swarm optimization algorithm. Taking the power grid zoning situation corresponding to Scheme 2 as an example, on this basis, in order to more intuitively show the effectiveness and superiority of the power grid zoning, two different schemes are used for comparison. Scheme 1: Without considering the power grid zoning, power supply is restored. Scheme 2: Considering each power grid zone, rapid power supply restoration is carried out with each zone as a unit.

[0055] As Figure 8 shown, when power supply restoration is completed without zoning, it takes 7 minutes, while when power supply restoration is carried out with each zone as a unit, the load restoration time is about 5.5 minutes, which is 23% shorter than Scheme 1, greatly saving the time for power supply restoration.

[0056] The comparison of the voltage levels of the two schemes is as Figure 9 shown. When not considering the power grid zoning, the average voltage level of the system nodes is relatively low, the minimum value is 0.9134 p.u, and the average voltage volatility of the system is 5.06%. After the large power grid is zoned, the minimum value of the system nodes is 0.9291 p.u, and the average voltage volatility of the system is 4.33%. Therefore, power grid zoning has a certain lifting effect on the node voltage and can also suppress voltage fluctuations to a certain extent.

[0057] Example 2: Based on Example 1, the following improvements are made in this example: Based on Equation (8), further calculate the corrected value of the available energy margin of energy storage in partition k , and in the calculation of Equation (6), use to replace , where: (27); In the formula, is the cyclic attenuation coefficient of energy storage node m, which is related to the battery chemical characteristics; represents the absolute value of the change in the state of charge in the current charge-discharge cycle; is the threshold of the maximum allowable depth of discharge; is the total number of energy storage devices in partition k. Considering the energy storage life attenuation during the partition optimization stage, the dual optimization of power supply restoration economy and equipment service life is realized.

[0058] Embodiment 3: A fast power supply restoration device based on source-load-storage collaborative power grid partition optimization, including a memory, a processor, and a computer program stored on the memory and capable of running on the processor. The method in Embodiment 1 or Embodiment 2 is implemented by the processor executing the computer program.

Claims

1. A fast power supply restoration method based on source-load-storage coordinated grid partition optimization is characterized by The following steps are involved: S1. Construct the electrical coupling index C and the source-load-storage power balance index respectively. , grid structure indicators and node size indicators , and comprehensively form comprehensive indicators of power grid division , expressed as: （1）； In the formula, , , , They are C, , , The weight coefficient of S2. As the fitness function, a hybrid selection mechanism driven by energy entropy is designed to improve the simulated annealing genetic algorithm, and the improved simulated annealing genetic algorithm is used to solve the power grid partitioning. S3. Based on the grid division, each division is taken as a unit to establish an optimization dispatching model with the goal of maximizing the restoration of important loads; S4. Use particle swarm algorithm to solve the optimization scheduling model and obtain a fast power supply restoration plan.

2. The rapid power supply restoration method based on source-load-storage coordinated power grid partition optimization according to claim 1 is characterized in that: In the above S1, the construction process of the electrical coupling index C is as follows: The electrical coupling between nodes i and j It is expressed as: （2）； In the formula, is the power interaction between node i and node j; , are the node voltages of node i and node j respectively; In the power grid partition, when node i is located in partition A and node j is located in partition B, the inter-regional coupling degree between partition A and partition B is It is expressed as: （3）； In the formula, is the number of connected node pairs between partition A and partition B; When both nodes i and j are located in partition A, the coupling degree within the region of partition A is It is expressed as: （4）； In the formula, is the number of node pairs in partition A; The electrical coupling index C is formed comprehensively: （5）； In the formula, yes The weight coefficient of yes The weight coefficient of .

3. The rapid power supply restoration method based on source-load-storage coordinated grid partition optimization according to claim 1 is characterized in that: In S1, the source-load-storage power balance index It is expressed as: （6）； In the formula, , , is the weight coefficient; represents the difference in the minimum net power of partition k; is the total load of partition k; is the available energy margin of energy storage in partition k; is the total energy storage capacity of partition k; is the supply-demand matching degree of partition k; is the maximum allowable matching deviation; in, It is expressed as: （7）； In the formula, is the set of power generation nodes in partition k, and x represents one of the power generation nodes; is the set of energy storage nodes in partition k, and m represents one of the energy storage nodes; is the set of load nodes in partition k, and y represents one of the load nodes; is the output of power generation node x; is the discharge power of energy storage node m; is the demand of load node y; It is expressed as: （8）； In the formula, is the current state of charge of the energy storage node m; is the minimum allowable state of charge of the energy storage node m; is the capacity of energy storage node m; It is expressed as: （9）； Where T is the time window; represents the power generation of partition k at time t; represents the energy storage discharge power of partition k at time t; represents the load demand of partition k at time t.

4. The rapid power supply restoration method based on source-load-storage coordinated power grid partition optimization according to claim 1 is characterized in that: In S1, the grid structure index It is expressed as: （10）； Where Q is the modularity index; when nodes i and j are in partition k, is the electrical distance between node i and node j, is the power transfer capability between node i and node j; is a weighting factor that balances the importance of topology compactness and power transfer capability.

5. The rapid power supply restoration method based on source-load-storage coordinated grid partition optimization according to claim 1 is characterized in that: In S1, the node scale index It is expressed as: （11）； In the formula, Indicates the average number of nodes: K is the total number of partitions; is the number of nodes in partition k; （12）； In the formula, N is the total number of nodes in the entire network.

6. The rapid power supply restoration method based on source-load-storage coordinated grid partition optimization according to claim 1 is characterized in that: In the S2, is the fitness function, and the partition result is used as the problem to be solved for optimization. The cooling function and Metropolis criterion in the simulated annealing genetic algorithm are expressed as: （13）； （14）； In the formula, , denote the temperatures of the zth generation and z+1th generation respectively; a is the annealing temperature coefficient, 0<a<1; , Represent the fitness values ​​of the zth generation and the z+1th generation respectively; P represents the acceptance probability of transferring from the zth generation to the z+1th generation in the simulated annealing algorithm; Design an energy entropy driven hybrid selection mechanism. The energy entropy calculation formula is as follows: （15）； In the formula, Represents the energy entropy of the current population; is the Boltzmann probability of individual u, that is, individual u at the current temperature The probability of being selected; M is the population size, that is, the total number of individuals in the population; , are the individual energies of individual u and individual v, which are the reciprocals of the fitness values; Set the threshold of energy entropy , according to the selection - accept the plan: like , using roulette wheel selection; if , using tournament selection; Combined with the Metropolis criterion, accept the new solution, but accept the probability Negatively correlated with energy entropy, expressed as: （16）； In the formula, is the energy difference between the new solution and the current solution; is the maximum possible value of energy entropy.

7. The rapid power supply restoration method based on source-load-storage coordinated grid partition optimization according to claim 1 is characterized in that: In S3, the objective function F of the optimization scheduling model is: （17）； （18）； In the formula, is the weight of load h, load h is located in partition k, K is the total number of partitions; is the load recovery amount of load h; is the time penalty coefficient; It is the total time from the occurrence of fault to the restoration of power supply; It is the black start power supply start time; It is the operation time of key switches, including section switches, tie switches, black start power supply access switches, bus tie switches, and automatic reclosing switches; Adjust time for power balance within the partition; The calculation method is: （19）； In the formula, is the start-up time of the black start power supply in partition k; Whether to enable black start power supply for partition k, which is 01 variable. Indicates that partition k enables black start power supply. Indicates that black start power is not enabled for partition k; The calculation method is: （20）； In the formula, is the operating time of switch s; Indicates whether switch s is operated, which is a 01 variable. Indicates that switch s is operated. Indicates that switch s is not operated; S is the total number of switches; The calculation method is: （21）； In the formula, is the power capacity of partition k; The energy storage output in partition k; is the load demand in partition k; is the adjustment rate of partition k.

8. The rapid power supply restoration method based on source-load-storage coordinated grid partition optimization according to claim 7 is characterized in that: The constraints of the optimization scheduling model include: Power balance constraints: （22）； In the formula, is the load of partition k at time t; is the wind power output of zone k at time t; is the photovoltaic output of partition k at time t; is the charging / discharging power of the energy storage in partition k at time t; The power transmitted between partition k and the upper grid at time t; is the power transfer between partitions in partition k at time t; Energy storage power and capacity constraints: （23）； In the formula, ,respectively is the energy storage capacity at time t and time t-1; is the charge and discharge variable of the energy storage at time t, when When the energy storage is in charging state, When the energy storage is in discharge state; Charging efficiency for energy storage; is the time interval; is the energy storage discharge efficiency; Voltage and frequency constraints: （24）； In the formula, U represents voltage and f represents frequency; Energy storage charge state: （25）； In the formula, , The upper and lower limits of the state of charge; is the initial state of charge; is the state of charge of energy storage I; Partition autonomy constraints: （26）； In the formula, O is the partition autonomy constraint index; is the available capacity in partition k; is the maximum load of partition k.

9. The rapid power supply restoration method based on source-load-storage coordinated grid partition optimization according to claim 8 is characterized in that: In the above S4, the solution process is: S4.

1. Input the partitioning result, i.e. input the pre-divided grid partitioning information, including the node list of each area, source-load-storage configuration parameters and the capacity limit of the interconnection line between partitions; S4.2, initialize PSO algorithm parameters, that is, set the particle swarm optimization algorithm parameters, including particle swarm size, particle dimension, maximum number of iterations, initial value of inertia weight, learning factor and variable range; S4.3, randomly initialize the particle position and velocity, and randomly generate an initial particle group according to the variable range, where the position of each particle represents a recovery strategy; S4.4, calculate the fitness value of the particle, and evaluate the performance of each particle based on the objective function and constraints of the optimization scheduling model; S4.5, update the individual historical optimal solution, compare the current fitness value of each particle with its historical optimal value, and if it is better, update the individual historical optimal position and the individual historical optimal fitness value; S4.6, update the historical optimal solution of the group, select the global optimal solution, the historical optimal position of the group and the historical optimal fitness value of the group from all particles; S4.7, determine whether the conditions are met, whether the maximum number of iterations is reached or the fitness value difference between two iterations is the smallest. If the conditions are met, output the optimal solution and terminate the iteration. If not, update the position vector and velocity vector of each particle and return to step S4.4 to continue iteration.

10. A rapid power supply restoration device based on source-load-storage coordinated power grid partition optimization, characterized in that: The method comprises a memory, a processor and a computer program stored in the memory and capable of running on the processor, and the method according to any one of claims 1 to 9 is implemented by executing the computer program by the processor.

Images