Power distribution network area energy storage system control method considering active power distribution network bearing capacity

By building an optimal operating model for carrying capacity of the active distribution network and using the GA-PSO hybrid optimization algorithm, the grid problems caused by instability and intermittentity of the active distribution network are solved, and automatic optimization and power supply of the distribution network are realized. Improved load capacity.

CN118983835APending Publication Date: 2024-11-19YIZHENG POWER SUPPLY OF JIANGSU ELECTRIC POWER +2
View PDF 0 Cites 2 Cited by

Patent Information

Application Number
CN202411137140.3
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-19
Publication Date
2024-11-19

AI Technical Summary

Technical Problem

The instability and intermittent nature of the active distribution network leads to voltage fluctuations in the distribution network, frequency offsets and system collapse. It is difficult for traditional methods to quickly improve the load-bearing capacity of the distribution network to meet the rapidly developing active distribution network connections. Enter the demand.

Method used

The data acquisition module is used to monitor the power grid status in real time, build an optimal operating model for the bearing capacity of the active distribution network, and optimize the energy storage operation plan using Lagrangian relaxation algorithm and GA-PSO hybrid optimization algorithm, and automatically adjust the charge and discharge of the energy storage equipment. Strategy to optimize the bearing capacity of the active distribution network.

Benefits of technology

Automatic optimization of the active distribution network has been achieved, the power carrying capacity of the distribution network has been improved, the impact on the power grid has been reduced, and the stability and performance of the power grid has been improved.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118983835A_ABST
    Figure CN118983835A_ABST
Patent Text Reader

Abstract

The invention discloses a power distribution network area energy storage system control method considering active power distribution network bearing capacity, and relates to the technical field of intelligent control. Comprising the following steps: S1, acquiring data information of a power distribution network area energy storage system by using a data acquisition module, and monitoring the operation state of a power grid in real time; s2, constructing an active power distribution network bearing capacity optimal operation model based on the operation state of the power grid to analyze the bearing capacity; and S3, optimization solution is carried out, the central server carries out real-time control on the energy storage unit according to an optimization solution result, and the bearing capacity of the active power distribution network is improved. According to the invention, the energy storage capability of the distribution network area is greatly improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of distribution network, and in particular to a control method for a distribution network substation energy storage system taking into account the carrying capacity of an active distribution network. Background Art

[0002] With the rapid development of renewable energy, the number of active distribution networks connected to distribution network stations has gradually increased. However, the instability and intermittency of active distribution networks have brought challenges to the stable operation of distribution networks. When the output of active distribution networks exceeds the carrying capacity of distribution networks, it may cause problems such as grid voltage fluctuations, frequency deviations, and even system crashes; the intermittent characteristics of active distribution networks also make it difficult to meet continuous and stable power demand, especially during peak load periods.

[0003] The traditional method is to improve the carrying capacity of the distribution network by enhancing the infrastructure construction of the distribution network, such as increasing the capacity of substations, improving the line structure, etc.; however, this method not only requires a large amount of capital investment, but also has a long construction period, and it is difficult to meet the rapidly growing active distribution network access needs.

[0004] Therefore, the present invention proposes a distribution network substation energy storage system control method taking into account the carrying capacity of the active distribution network, so as to achieve effective management and utilization of the active distribution network and improve the power carrying capacity of the distribution network. Summary of the invention

[0005] In view of the shortcomings of the above-mentioned technology, the carrying capacity of the traditional active distribution network cannot be automatically optimized and is expensive. The present invention provides a distribution network substation energy storage system control method taking into account the carrying capacity of the active distribution network, which can realize the automatic optimization of the carrying capacity of the active distribution network and meet user needs.

[0006] The technical solution of the present invention is: a control method for a distribution network area energy storage system considering the carrying capacity of an active distribution network, comprising the following steps:

[0007] S1. Use the data acquisition module to obtain data information of the energy storage system in the distribution network area and monitor the operation status of the power grid in real time;

[0008] S2. With the goal of minimizing the number of reverse overloads in the distribution network lines, an optimal operation model of the active distribution network carrying capacity considering the grid carrying constraints is constructed based on the operation status of the grid;

[0009] S3, solving the optimal operation model of the active distribution network carrying capacity to obtain the energy storage operation plan;

[0010] Firstly, the Lagrangian relaxation algorithm is used to transform the constraint set of the optimal operation model of the active distribution network carrying capacity into an easily tractable form, and then the GA-PSO hybrid optimization algorithm is used to solve the optimal energy storage operation mode according to the simplified constraint set.

[0011] Collect relevant data on the carrying capacity of the active distribution network, including at least voltage, current, power and capacity.

[0012] The objective function is established with the goal of minimizing the number of lines with reverse power overload. The mathematical expression is:

[0013]

[0014] In the formula, Π j is the set of branch headend nodes with j as the terminal node; y ij A 0-1 variable that reflects whether branch ij is reversely overloaded, 1 indicates overload, and 0 indicates no overload.

[0015] The constraint conditions of the optimal operation model of the active distribution network carrying capacity include distribution network flow constraints, distribution network node voltage constraints, equipment operation constraints and energy storage charging and discharging constraints.

[0016] The mathematical expression of the distribution network power flow constraint is:

[0017]

[0018] In formula (1), is the active power at the mth node of the distribution network, V m is the node voltage at the mth node of the distribution network, V n The node voltage at the nth node of the distribution network, G mn is the conductance between the mth node and the nth node of the distribution network, B mn is the susceptance between the mth node and the nth node of the distribution network, θ mn is the power angle between the mth node and the nth node of the distribution network, α is the power flow constraint coefficient;

[0019]

[0020] In formula (2), is the reactive power at the mth node of the distribution network;

[0021] The distribution network node voltage constraint is:

[0022] V mmin ≤V m ≤V mmax (3)

[0023] In formula (3), V mmin is the voltage lower limit of the mth node in the distribution network, V mmax is the voltage upper limit of the mth node in the distribution network;

[0024] The equipment operation constraints, including line and transformer capacity constraints, require that the operating voltage and current cannot exceed the operating limit, the positive load cannot exceed the rated power, and the negative load cannot exceed 80% of the rated power, that is, the power should meet the following requirements:

[0025] -0.8P max ≤P≤P max (4)

[0026] In formula (4), P is the active power on the distribution network line, P max is the maximum active power on the distribution network line;

[0027] The energy storage charge and discharge constraints are:

[0028] S min ≤S t ≤S max (5)

[0029] In formula (5), S t is the state of charge at time t, S min is the lower limit of the charge state, S max is the upper limit of the state of charge;

[0030] The line transmission constraints are:

[0031] -P 1max ≤P 1 ≤P 1max (6)

[0032] In formula (6), P 1 is the transmission power corresponding to the first line of the target distribution network, P 1max The upper limit of the transmission power of the lines in the target distribution network;

[0033] As a further embodiment of the present invention, the workflow of the Lagrangian relaxation algorithm is:

[0034] 1) identifying complex constraints on distributed carrying capacity, wherein the complex constraints include at least the distribution network flow constraints, the distribution network node voltage constraints, the equipment operation constraints, and the energy storage charging and discharging constraints;

[0035] 2) Introduce Lagrange multipliers, multiply the constraints by the corresponding Lagrange multipliers and add them to the objective function; the calculation function after the Lagrange relaxation algorithm is:

[0036]

[0037] In formula (7), Z is the objective function of the relaxation problem, is the coefficient vector of the objective function, f is the decision variable vector, λ is the Lagrange multiplier, g is the random objective coefficient vector, A and B are matrices, and r is the objective coefficient vector;

[0038] 3) Solving the relaxation problem by using the GA-PSO hybrid optimization algorithm;

[0039] 4) Update the Lagrange multiplier, bring the current solution into the constraint condition and calculate the difference to obtain the constraint violation degree. If the constraint violation degree is exceeded, increase the corresponding Lagrange multiplier value. If the constraint violation degree is not exceeded, keep the corresponding Lagrange multiplier value.

[0040] 5) Iterate the solution and repeat 3) and 4) until the conditions are met; during the iteration process, the convergence is accelerated by adjusting the update strategy of the Lagrange multiplier.

[0041] As a further embodiment of the present invention,

[0042] The GA-PSO hybrid optimization algorithm includes a fitness evaluation module, a genetic algorithm GA module, a particle swarm optimization algorithm PSO module and a hybrid strategy module;

[0043] The workflow of the GA-PSO hybrid optimization algorithm is:

[0044] 1) Initializing the particle swarm, randomly generating a group of particles as the initial particle swarm according to the constraints of the active distribution network and the requirements of the carrying capacity optimization problem; each particle represents a possible energy storage system operation scheme, and the operation scheme at least includes the charging and discharging power and time of the energy storage device; setting parameters, and the parameters at least include the particle swarm size, the number of iterations, the crossover probability and the mutation probability;

[0045] 2) Performing fitness evaluation according to the fitness evaluation module, and calculating the fitness value of each particle according to the objective function of the optimization problem. The fitness function reflects the pros and cons of each energy storage system operation scheme, and the pros and cons indicators at least include network loss and voltage stability;

[0046] 3) performing operations according to the genetic algorithm GA module, including selection, crossover and mutation;

[0047] 4) operating according to the particle swarm optimization algorithm PSO module, updating the particle speed and position, comparing the fitness values ​​of the particles in each iteration, and updating the individual optimal solution and the global optimal solution;

[0048] 5) Perform iterative optimization according to the hybrid strategy module, repeatedly execute the genetic algorithm GA module operation and the particle swarm optimization algorithm PSO module operation, optimize the position and speed of particles through continuous iteration, and adjust the algorithm parameters, inertia weight and learning factor according to needs; combining the global search capability of the genetic algorithm GA module and the rapid convergence of the particle swarm optimization algorithm PSO module, the GA-PSO hybrid optimization algorithm can find a better energy storage system operation plan in a complex active distribution network environment, thereby improving the carrying capacity and performance of the distribution network;

[0049] 6) Determine the termination condition and when the maximum number of iterations is reached, determine whether to terminate the algorithm execution.

[0050] As a further embodiment of the present invention, the working method of the genetic algorithm GA module is:

[0051] 1) Selection operation, select excellent individuals to enter the next generation according to the fitness value;

[0052] 2) Crossover operation, generating a new individual by exchanging some genes of two parent individuals; the partial genes are some parameters of the energy storage system operation strategy; by exchanging the charge and discharge time and power parameters of the two parent individuals, a new strategy with different charge and discharge time and power distribution is generated;

[0053] 3) Mutation operation: randomly change some genes of individuals with a certain probability to increase the diversity of the population; mutate the charging and discharging power and time of some individuals, deviating from the original values. If the mutated strategy has a better fitness value, it will be retained in the population, otherwise it will be discarded or mutated again.

[0054] As a further embodiment of the present invention, the mathematical expression of the initialization particle position of the particle swarm optimization algorithm PSO module is:

[0055]

[0056] The mathematical expression for initializing particle velocity is:

[0057]

[0058] In formulas (8) and (9), i represents the i-th particle, j represents the dimension of the problem, Represents the initial position of the ith particle, Represented as the initialization particle speed of the i-th particle; each particle randomly generates an initial position, representing an initial energy storage operation plan, and the energy storage control parameters corresponding to each particle at least include the charging and discharging power time;

[0059] The mathematical operation expression for updating particle velocity is:

[0060]

[0061] In formula (10), M is the inertia weight, t is the number of iterations, and r 1 and r 2 is a random number, a 1 and a 2 is the acceleration factor, is the optimal position of particle i at the tth iteration, is the global optimal position of particle i at the tth iteration, M max is the maximum value of inertia weight, M min is the minimum value of the inertia weight, N is the current number of iterations, N max is the maximum number of iterations of the model;

[0062] The mathematical operation expression for updating the particle position is:

[0063]

[0064] Use the updated speed to update the particle's position to obtain a new energy storage operation plan;

[0065] The mathematical operation expression for updating the optimal position of the particle is:

[0066]

[0067] The mathematical operation expression for updating the global optimal position of the particle swarm is:

[0068]

[0069] In formula (13), r is a random number and a is an acceleration factor. In each iteration, the performance of the newly generated energy storage operation scheme, i.e., the carrying capacity of the active distribution network, is evaluated, and the carrying capacity of the active distribution network is compared with the individual historical optimal and global optimal solutions. If the performance of the new scheme is better, the individual optimal solution or the global optimal solution is updated.

[0070] As a further embodiment of the present invention, the hybrid strategy module includes parallel hybrid, serial hybrid and embedded hybrid; in parallel hybrid, the genetic algorithm GA module and the particle swarm optimization algorithm PSO module run independently, exchange information every ten iterations to share the optimal solution; serial hybrid first uses the genetic algorithm GA module and then switches to the particle swarm optimization algorithm PSO module; in embedded hybrid, the particle swarm optimization algorithm PSO module is embedded in the iterative process of the genetic algorithm GA module; the fitness evaluation module defines a fitness function to measure the advantage of each solution.

[0071] During operation, the present invention can effectively improve the carrying capacity of the active distribution network by constructing an optimal operation model of the carrying capacity of the active distribution network and a GA-PSO hybrid optimization algorithm. The energy storage unit can absorb excess electric energy when the active distribution network reaches a peak and release electric energy when the network reaches a trough, thereby balancing the fluctuations in the active distribution network, reducing the impact on the distribution network substations, and realizing automatic optimization of the carrying capacity of the active distribution network. BRIEF DESCRIPTION OF THE DRAWINGS

[0072] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative work, among which:

[0073] Figure 1 is a flow chart of the method of the present invention;

[0074] Figure 2 It is the working flow chart of the Lagrangian relaxation algorithm in the present invention;

[0075] Figure 3 It is the working principle diagram of the GA-PSO hybrid optimization algorithm in the present invention;

[0076] Figure 4 It is a working flow chart of the particle swarm optimization algorithm PSO module in the present invention;

[0077] Figure 5 This is a working principle diagram of the genetic algorithm GA module in the present invention. DETAILED DESCRIPTION

[0078] The preferred embodiments of the present invention are described below in conjunction with the accompanying drawings. It should be understood that the embodiments described herein are only used to illustrate and explain the present invention, and are not used to limit the present invention.

[0079] The present invention Figure 1 As shown, the control method of the distribution network area energy storage system considering the carrying capacity of the active distribution network includes the following steps:

[0080] S1. Use the data acquisition module to obtain data information of the energy storage system in the distribution network area and monitor the operation status of the power grid in real time;

[0081] S2. With the goal of minimizing the number of reverse overloads in the distribution network lines, an optimal operation model of the active distribution network carrying capacity considering the grid carrying constraints is constructed based on the operation status of the grid;

[0082] The constraints of the optimal operation model of the active distribution network carrying capacity include distribution network flow constraints, distribution network node voltage constraints, equipment operation constraints and energy storage charging and discharging constraints;

[0083] S3. Solve the optimal operation model of the active distribution network's carrying capacity to obtain an energy storage operation plan (i.e., based on the solution results, control the energy storage unit in real time through the central server to improve the carrying capacity of the active distribution network).

[0084] Firstly, the Lagrangian relaxation algorithm is used to transform the constraint set of the optimal operation model of the active distribution network carrying capacity into an easily tractable form, and then the GA-PSO hybrid optimization algorithm is used to solve the optimal energy storage operation mode according to the simplified constraint set.

[0085] In the present invention, the steps of constructing the optimal operation model of the active distribution network carrying capacity are:

[0086] 1) Collect relevant data on the carrying capacity of the active distribution network, including at least voltage, current, power and capacity;

[0087] 2) Establishing the optimal operation model of the active distribution network carrying capacity, determining the model inputs as current, voltage, power, charge and discharge state and energy storage capacity, and the model outputs as the maximum carrying capacity of the active distribution network, the distribution network stability index and the best optimization suggestion;

[0088] 3) Setting the bearing capacity constraint conditions, including the power flow constraint, the distribution network node voltage constraint, the line transmission constraint, the equipment operation constraint and the energy storage charging and discharging constraint; the mathematical expression of the distribution network power flow constraint is:

[0089]

[0090] In formula (1), is the active power at the mth node of the distribution network, V m is the node voltage at the mth node of the distribution network, V n The node voltage at the nth node of the distribution network, G mn is the conductance between the mth node and the nth node of the distribution network, B mn is the susceptance between the mth node and the nth node of the distribution network, θ mn is the power angle between the mth node and the nth node of the distribution network, α is the power flow constraint coefficient;

[0091]

[0092] In formula (2), is the reactive power at the mth node of the distribution network;

[0093] The distribution network node voltage constraint is:

[0094] V mmin ≤V m ≤V mmax (3)

[0095] In formula (3), V mmin is the voltage lower limit of the mth node in the distribution network, V mmax is the voltage upper limit of the mth node in the distribution network;

[0096] The equipment operation constraints, including line and transformer capacity constraints, require that the operating voltage and current cannot exceed the operating limit, the positive load cannot exceed the rated power, and the negative load cannot exceed 80% of the rated power, that is, the power should meet the following requirements:

[0097] -0.8P max ≤P≤P max (4)

[0098] In formula (4), P is the active power on the distribution network line, P max is the maximum active power on the distribution network line;

[0099] The energy storage charge and discharge constraints are:

[0100] S min ≤S t ≤S max (5)

[0101] In formula (5), S t is the state of charge at time t, S min is the lower limit of the charge state, S max is the upper limit of the state of charge;

[0102] The line transmission constraints are:

[0103] -P 1max ≤P 1 ≤P 1max (6)

[0104] In formula (6), P 1 is the transmission power corresponding to the first line of the target distribution network, P 1max The upper limit of the transmission power of the lines in the target distribution network;

[0105] 1) Establish an objective function with the goal of minimizing the number of lines with reverse power overload. The mathematical expression is:

[0106]

[0107] In the formula, Π jis the set of branch headend nodes with j as the terminal node; y ij A 0-1 variable that reflects whether branch ij is reversely overloaded, 1 indicates overload, and 0 indicates no overload.

[0108] 2) Result output: Considering the above bearing capacity constraint set, the constraint set is simplified by Lagrangian relaxation algorithm, and the optimal operation mode of energy storage under the substation is solved by the GA-PSO hybrid optimization algorithm.

[0109] In the present invention, the workflow of the Lagrangian relaxation algorithm is:

[0110] 1) identifying complex constraints on distributed carrying capacity, wherein the complex constraints include at least the distribution network flow constraints, the distribution network node voltage constraints, the equipment operation constraints, and the energy storage charging and discharging constraints;

[0111] 2) Introduce Lagrange multipliers, multiply the constraints by the corresponding Lagrange multipliers and add them to the objective function; the calculation function after the Lagrange relaxation algorithm is:

[0112]

[0113] In formula (7), Z is the objective function of the relaxation problem, is the coefficient vector of the objective function, f is the decision variable vector, λ is the Lagrange multiplier, g is the random objective coefficient vector, A and B are matrices, and r is the objective coefficient vector;

[0114] 3) Solving the relaxation problem by using the GA-PSO hybrid optimization algorithm, wherein the genetic algorithm GA module generates new solutions by selection, crossover and mutation, and evaluates the quality of the solution according to the fitness function; each particle in the particle swarm optimization algorithm PSO module represents a solution, and the particle updates its position and speed according to its own historical optimal position and the historical optimal position of the entire group, thereby gradually approaching the optimal solution, thereby finding the optimal solution;

[0115] 4) Update the Lagrange multiplier, bring the current solution into the constraint condition and calculate the difference to obtain the constraint violation degree. If the constraint violation degree is exceeded, increase the corresponding Lagrange multiplier value. If the constraint violation degree is not exceeded, keep the corresponding Lagrange multiplier value.

[0116] 5) Iterate the solution and repeat 3) and 4) until the conditions are met; during the iteration process, the convergence is accelerated by adjusting the update strategy of the Lagrange multiplier.

[0117] In a specific embodiment, the original calculation function after the Lagrangian relaxation algorithm is:

[0118]

[0119] In the formula, b is the coefficient vector, but the vector coefficient b is unique and cannot be automatically adjusted according to different constraints, which increases the amount of calculation and complexity. This problem is solved by introducing a random target coefficient vector g, and the target vector can be adaptively adjusted.

[0120] In a specific embodiment, the Lagrangian relaxation algorithm converts the constraints into Lagrangian multiplier form by introducing Lagrangian multipliers, and introduces the Lagrangian multipliers into the original function as a new variable, the original constraints are eliminated, and the problem is converted into an unconstrained optimization problem; in the Lagrangian relaxation algorithm, the Lagrangian multipliers connect the original objective function and the constraints, and by adjusting the values ​​of the Lagrangian multipliers, the transformed objective function approaches the optimal solution of the original problem while maintaining linearity; the Lagrangian relaxation algorithm provides a method for calculating the next term for solving integer programming and combinatorial optimization, and approaches the optimal solution of the original problem by solving the Lagrangian dual problem.

[0121] In the present invention, the GA-PSO hybrid optimization algorithm includes a fitness evaluation module, a genetic algorithm GA module, a particle swarm optimization algorithm PSO module and a hybrid strategy module;

[0122] The workflow of the GA-PSO hybrid optimization algorithm is:

[0123] 1) Initializing the particle swarm, randomly generating a group of particles as the initial particle swarm according to the constraints of the active distribution network and the requirements of the carrying capacity optimization problem; each particle represents a possible energy storage system operation scheme, and the operation scheme at least includes the charging and discharging power and time of the energy storage device; setting parameters, and the parameters at least include the particle swarm size, the number of iterations, the crossover probability and the mutation probability;

[0124] 2) Performing fitness evaluation according to the fitness evaluation module, and calculating the fitness value of each particle according to the objective function of the optimization problem. The fitness function reflects the pros and cons of each energy storage system operation scheme, and the pros and cons indicators at least include network loss and voltage stability;

[0125] 3) performing operations according to the genetic algorithm GA module, including selection, crossover and mutation;

[0126] 4) operating according to the particle swarm optimization algorithm PSO module, updating the particle speed and position, comparing the fitness values ​​of the particles in each iteration, and updating the individual optimal solution and the global optimal solution;

[0127] 5) Perform iterative optimization according to the hybrid strategy module, repeatedly execute the genetic algorithm GA module operation and the particle swarm optimization algorithm PSO module operation, optimize the position and speed of particles through continuous iteration, and adjust the algorithm parameters, inertia weight and learning factor according to needs; combining the global search capability of the genetic algorithm GA module and the rapid convergence of the particle swarm optimization algorithm PSO module, the GA-PSO hybrid optimization algorithm can find a better energy storage system operation plan in a complex active distribution network environment, thereby improving the carrying capacity and performance of the distribution network;

[0128] 6) Determine the termination condition and when the maximum number of iterations is reached, determine whether to terminate the algorithm execution.

[0129] In a specific embodiment, the GA-PSO hybrid optimization algorithm may adopt the following optimization strategies:

[0130] 1) Dynamically adjust the inertia weight. According to the running status of the algorithm and the search requirements, the size of the inertia weight is adjusted in real time. In the early stage of the search, in order to expand the search range, a larger inertia weight is usually set so that the particles can explore the entire solution space faster. As the search progresses, in order to speed up the convergence speed and avoid falling into the local optimum, the inertia weight will gradually decrease, so that the particles can perform a more detailed search near the current optimal solution.

[0131] 2) Elite strategy, retaining the best individuals in the population to avoid losing excellent genes during crossover and mutation;

[0132] 3) Domain search: Introducing a neighborhood search mechanism into particle swarm optimization to speed up local search;

[0133] 4) Multi-population strategy: Use multiple populations to search in parallel to increase the diversity of the algorithm and avoid falling into local optimality;

[0134] 5) Heuristic information: Use heuristic information to assist in generating the initial population and improve the quality of the initial solution.

[0135] In a specific embodiment, the elite strategy is implemented by the following steps:

[0136] 1) Identification of elite individuals. In each iteration, the algorithm calculates the fitness value of each individual according to the fitness function. Elite individuals usually refer to individuals with higher fitness values, which represent the optimal solution found in the current search process.

[0137] 2) Retention of elite individuals. After identifying elite individuals, the algorithm will retain these individuals to prevent them from being destroyed or lost in subsequent genetic algorithm GA module operations or particle swarm optimization algorithm PSO module operations;

[0138] 3) Utilization of elite individuals: elite individuals serve as guide particles in particle swarm optimization, guiding other particles to move to a better solution space, or serve as excellent genes in genetic algorithms, passing them to offspring individuals through crossover operations;

[0139] 4) Update of the elite population. The elite population needs to be updated regularly to ensure that it always contains the best individuals. The update operation can be performed based on the comparison of fitness values, and only individuals with higher fitness are retained.

[0140] In a specific embodiment, the basic principle of the multi-population strategy is:

[0141] 1) Population diversity: By dividing the entire population into multiple sub-populations, the multi-population strategy can increase the diversity of the population. Each sub-population performs independent searches in its own search space, which helps the algorithm explore more potential solution areas, thereby increasing the possibility of finding the global optimal solution;

[0142] 2) Parallel search: multiple subpopulations can perform search operations at the same time. This parallelism can significantly improve the execution efficiency of the algorithm. Each subpopulation can find a relatively good solution in its local search space, and then share information with other subpopulations through a certain information exchange mechanism to promote the discovery of the global optimal solution.

[0143] 3) Information interaction and collaboration: regularly migrate outstanding individuals from each subpopulation to other subpopulations to speed up the convergence of the global optimal solution. This information interaction helps prevent the algorithm from falling into the local optimal solution and improves the global search capability;

[0144] 4) Adaptive adjustment: dynamically adjust the number, size and migration strategy of sub-populations according to their performance to balance the algorithm’s search efficiency and global search capabilities.

[0145] In a specific embodiment, the main sources of the heuristic information are:

[0146] 1) The particle swarm optimization algorithm PSO module simulates the information sharing and collaboration behavior between individuals in a bird flock. Each particle updates its position and speed based on its own historical optimal solution and the global optimal solution of the entire population, which enables the particles to quickly converge to a better solution area in the search space;

[0147] 2) The genetic algorithm GA module generates new solutions by simulating genetic operations in the biological evolution process, avoiding premature convergence to the local optimal solution, and provides heuristic information to guide the search direction to a certain extent;

[0148] 3) By calculating the fitness value of the particles, the algorithm can determine the gap between the current solution and the optimal solution and adjust the search strategy.

[0149] In the present invention, the working method of the genetic algorithm GA module is:

[0150] 1) Selection operation, select excellent individuals to enter the next generation according to the fitness value;

[0151] 2) Crossover operation, generating a new individual by exchanging some genes of two parent individuals; the partial genes are some parameters of the energy storage system operation strategy; by exchanging the charge and discharge time and power parameters of the two parent individuals, a new strategy with different charge and discharge time and power distribution is generated;

[0152] 3) Mutation operation: randomly change some genes of individuals with a certain probability to increase the diversity of the population; mutate the charging and discharging power and time of some individuals, deviating from the original values. If the mutated strategy has a better fitness value, it will be retained in the population, otherwise it will be discarded or mutated again.

[0153] In a specific embodiment, the genetic algorithm GA module has the following operating steps:

[0154] 1) Before searching, the genetic algorithm GA module first represents the solution data of the solution space as genotype string structure data of the genetic space. Different combinations of these string structure data constitute different points.

[0155] 2) Generation of the initial population: randomly generate N initial string structure data, each string structure data is called an individual, and N individuals constitute a population; the genetic algorithm GA module starts to evolve with these N string structure data as the initial point;

[0156] 3) Fitness evaluation: Fitness indicates the quality of an individual or solution;

[0157] 4) Selection, the purpose is to select excellent individuals from the current group so that they have the opportunity to serve as parents to reproduce offspring for the next generation; the genetic algorithm GA module embodies Darwin's principle of survival of the fittest through the selection process. The principle of selection is that individuals with strong adaptability have a high probability of contributing one or more offspring to the next generation;

[0158] 5) Crossover: A new generation of individuals is obtained through crossover operation, and the new individuals combine the characteristics of the parent individuals;

[0159] 6) Mutation: First, randomly select an individual in the population, and randomly change the value of the string structure data with a certain probability for the selected individual. The probability of mutation in the genetic algorithm GA module is very low, with a value of about 0.5.

[0160] In the present invention, the mathematical expression of the initialization particle position of the particle swarm optimization algorithm PSO module is:

[0161]

[0162] The mathematical expression for initializing particle velocity is:

[0163]

[0164] In formulas (8) and (9), i represents the i-th particle, j represents the dimension of the problem, Represents the initial position of the ith particle, Represented as the initialization particle speed of the i-th particle; each particle randomly generates an initial position, representing an initial energy storage operation plan, and the energy storage control parameters corresponding to each particle at least include the charging and discharging power time;

[0165] The mathematical operation expression for updating particle velocity is:

[0166]

[0167] In formula (10), M is the inertia weight, t is the number of iterations, and r 1 and r 2 is a random number, a 1 and a 2 is the acceleration factor, is the optimal position of particle i at the tth iteration, is the global optimal position of particle i at the tth iteration, M max is the maximum value of inertia weight, M min is the minimum value of the inertia weight, N is the current number of iterations, N max is the maximum number of iterations of the model;

[0168] The mathematical operation expression for updating the particle position is:

[0169]

[0170] Use the updated speed to update the particle's position to obtain a new energy storage operation plan;

[0171] The mathematical operation expression for updating the optimal position of the particle is:

[0172]

[0173] The mathematical operation expression for updating the global optimal position of the particle swarm is:

[0174]

[0175] In formula (13), r is a random number and a is an acceleration factor. In each iteration, the performance of the newly generated energy storage operation scheme, i.e., the carrying capacity of the active distribution network, is evaluated, and the carrying capacity of the active distribution network is compared with the individual historical optimal and global optimal solutions. If the performance of the new scheme is better, the individual optimal solution or the global optimal solution is updated.

[0176] During the above work, the implementation of the particle swarm optimization algorithm PSO module formula is completed based on a high-performance processor CPU and a parallel calculator; the high-performance processor CPU accelerates the execution speed of the algorithm, processes data and iterations; the parallel calculator includes a multi-core processor and a dedicated accelerator, and the parallel calculator speeds up the update speed of particles; the implementation data results of the particle swarm optimization algorithm PSO module are shown in Table 1.

[0177] Table 1 Implementation data results of particle swarm optimization algorithm PSO module

[0178]

[0179] In a specific embodiment, the implementation process of the particle swarm optimization algorithm PSO module is:

[0180] 1) Initialize the population, including the upper and lower limits of the search space, two learning factors, the maximum number of iterations of the algorithm, and the upper and lower limits of each particle's velocity; randomly initialize the position and velocity of each particle in the population;

[0181] 2) Calculate the fitness value of each particle according to the fitness function, save the optimal position of each particle, save the individual best fitness value and the best position of the group so far;

[0182] 3) Update the speed and position according to the speed and position update formula;

[0183] 4) Calculate the fitness value of each particle after the update, and compare the best fitness value of each particle with the fitness value at its historical optimal position. If it is better, its current position is used as the optimal position of the particle;

[0184] 5) For each particle, compare the fitness value corresponding to its best position with the best fitness value of the population. If it is better, update the optimal position and the best fitness value of the population.

[0185] 6) Determine whether the searched structure meets the stopping condition. If so, output the optimal value. Otherwise, jump to (3) and continue running until the condition is met.

[0186] In the present invention, the hybrid strategy module includes parallel hybrid, serial hybrid and embedded hybrid; in the parallel hybrid, the genetic algorithm GA module and the particle swarm optimization algorithm PSO module run independently, and exchange information every ten iterations to share the optimal solution; in the serial hybrid, the genetic algorithm GA module is first used, and then switched to the particle swarm optimization algorithm PSO module; in the embedded hybrid, the particle swarm optimization algorithm PSO module is embedded in the iterative process of the genetic algorithm GA module; the fitness evaluation module defines a fitness function for measuring the advantage of each solution.

[0187] In a specific embodiment, a public transformer substation under a 10kV line in northern Jiangsu is selected for simulation, and a typical day with 4 reverse overload moments is selected for optimization and solution. When the reverse load rate of the line exceeds 80%, it is judged as reverse overload. After the optimal energy storage configuration is solved through optimization, the reverse load rate at the 4 moments is effectively reduced to below 80%. The optimization results of the active distribution network carrying capacity are shown in Table 2.

[0188] Table 2 Active distribution network carrying capacity optimization results

[0189]

[0190] Although the specific embodiments of the present invention are described above, it should be understood by those skilled in the art that these specific embodiments are only illustrative, and those skilled in the art may omit, replace, and change the details of the above methods and systems in various ways without departing from the principles and essence of the present invention. For example, merging the above method steps so as to perform substantially the same functions in substantially the same manner to achieve substantially the same results is within the scope of the present invention. Therefore, the scope of the present invention is limited only by the appended claims.

Claims

1. A control method for a distribution network area energy storage system considering the carrying capacity of an active distribution network, characterized in that: The steps include: S1. Use the data acquisition module to obtain data information of the energy storage system in the distribution network area and monitor the operation status of the power grid in real time; S2. With the goal of minimizing the number of reverse overloads in the distribution network lines, an optimal operation model of the active distribution network carrying capacity considering the grid carrying constraints is constructed based on the operation status of the grid; S3. Solve the optimal operation model of the active distribution network carrying capacity and obtain the energy storage operation plan.

2. The method for controlling the energy storage system in a distribution network area considering the carrying capacity of the active distribution network according to claim 1, characterized in that: The data information obtained from the energy storage system in the distribution network area includes voltage, current, power and capacity.

3. The method for controlling the energy storage system in a distribution network area considering the carrying capacity of the active distribution network according to claim 1, characterized in that: The objective function of the optimal operation model of the active distribution network carrying capacity is: In the formula, Π j is the set of branch headend nodes with j as the terminal node; y ij A 0-1 variable that reflects whether branch ij is reversely overloaded, 1 indicates overload, and 0 indicates no overload.

4. The method for controlling the energy storage system in a distribution network area considering the carrying capacity of the active distribution network according to claim 1, characterized in that: The optimal operation model of the active distribution network carrying capacity includes distribution network flow constraints, distribution network node voltage constraints, equipment operation constraints and energy storage charging and discharging constraints.

5. The method for controlling the energy storage system in a distribution network area considering the carrying capacity of the active distribution network according to claim 1, characterized in that: The power flow constraint of the distribution network is: In formula (1), is the active power at the mth node of the distribution network, V m is the node voltage at the mth node of the distribution network, V n The node voltage at the nth node of the distribution network, G mn is the conductance between the mth node and the nth node of the distribution network, B mn is the susceptance between the mth node and the nth node of the distribution network, θ mn is the power angle between the mth node and the nth node of the distribution network, α is the power flow constraint coefficient; In formula (2), is the reactive power at the mth node of the distribution network; The distribution network node voltage constraint is: In mmin ≤V m ≤V mmax (3) In formula (3), V mmin is the voltage lower limit of the mth node in the distribution network, V mmax is the voltage upper limit of the mth node in the distribution network; The equipment operation constraints are: -0.8P max ≤P≤P max (4) In formula (4), P is the active power on the distribution network line, P max is the maximum active power on the distribution network line; The energy storage charge and discharge constraints are: S min ≤S t ≤S max (5) In formula (5), S t is the state of charge at time t, S min is the lower limit of the charge state, S max is the upper limit of the state of charge; The line transmission constraints are: -P 1max ≤P1≤P 1max (6) In formula (6), P1 is the transmission power corresponding to the first line of the target distribution network, P 1max The upper limit of the line transmission power in the target distribution network.

6. The method for controlling the energy storage system in a distribution network area considering the carrying capacity of the active distribution network according to claim 1, characterized in that: Firstly, the Lagrangian relaxation algorithm is used to transform the constraint set of the optimal operation model of the active distribution network carrying capacity into an easily tractable form, and then the GA-PSO hybrid optimization algorithm is used to solve the optimal energy storage operation mode according to the simplified constraint set.

7. The method for controlling a distribution network area energy storage system considering the carrying capacity of an active distribution network according to claim 1, characterized in that: The workflow of the Lagrangian relaxation algorithm is: 1) Identify the constraints on distributed carrying capacity; 2) Introduce Lagrange multipliers, multiply the constraints by the corresponding Lagrange multipliers and add them to the objective function; the calculation function after the Lagrange relaxation algorithm is: In formula (7), Z is the objective function of the relaxation problem, is the coefficient vector of the objective function, f is the decision variable vector, λ is the Lagrange multiplier, g is the random objective coefficient vector, A and B are matrices, and r is the objective coefficient vector; 3) Solving the relaxation problem by using the GA-PSO hybrid optimization algorithm; 4) Update the Lagrange multiplier, bring the current solution into the constraint condition and calculate the difference to obtain the constraint violation degree. If the constraint violation degree is exceeded, increase the corresponding Lagrange multiplier value. If the constraint violation degree is not exceeded, keep the corresponding Lagrange multiplier value. 5) Iterate the solution and repeat 3) and 4) until the conditions are met; during the iteration process, the convergence is accelerated by adjusting the update strategy of the Lagrange multiplier.

8. The method for controlling a distribution network area energy storage system considering the carrying capacity of an active distribution network according to claim 1, characterized in that: The GA-PSO hybrid optimization algorithm includes a fitness evaluation module, a genetic algorithm GA module, a particle swarm optimization algorithm PSO module and a hybrid strategy module; The workflow of the GA-PSO hybrid optimization algorithm is: 1) Initializing the particle swarm, randomly generating a group of particles as the initial particle swarm according to the constraints of the active distribution network and the requirements of the carrying capacity optimization problem; each particle represents a possible energy storage system operation scheme, and the operation scheme at least includes the charging and discharging power and time of the energy storage device; setting parameters, and the parameters at least include the particle swarm size, the number of iterations, the crossover probability and the mutation probability; 2) Performing fitness evaluation according to the fitness evaluation module, and calculating the fitness value of each particle according to the objective function of the optimization problem. The fitness function reflects the pros and cons of each energy storage system operation scheme, and the pros and cons indicators at least include network loss and voltage stability; 3) performing operations according to the genetic algorithm GA module, including selection, crossover and mutation; 4) operating according to the particle swarm optimization algorithm PSO module, updating the particle speed and position, comparing the fitness values ​​of the particles in each iteration, and updating the individual optimal solution and the global optimal solution; 5) Perform iterative optimization according to the hybrid strategy module, repeatedly execute the genetic algorithm GA module operation and the particle swarm optimization algorithm PSO module operation, optimize the position and speed of particles through continuous iteration, and adjust the algorithm parameters, inertia weight and learning factor according to needs; combining the global search capability of the genetic algorithm GA module and the rapid convergence of the particle swarm optimization algorithm PSO module, the GA-PSO hybrid optimization algorithm can find a better energy storage system operation plan in a complex active distribution network environment, thereby improving the carrying capacity and performance of the distribution network; 6) Determine the termination condition and when the maximum number of iterations is reached, determine whether to terminate the algorithm execution.

9. The method for controlling the energy storage system in a distribution network area considering the carrying capacity of the active distribution network according to claim 1, characterized in that: The working method of the genetic algorithm GA module is: 1) Selection operation, select excellent individuals to enter the next generation according to the fitness value; 2) Crossover operation, generating a new individual by exchanging some genes of two parent individuals; the partial genes are some parameters of the energy storage system operation strategy; by exchanging the charge and discharge time and power parameters of the two parent individuals, a new strategy with different charge and discharge time and power distribution is generated; 3) Mutation operation: randomly change some genes of individuals with a certain probability to increase the diversity of the population; mutate the charging and discharging power and time of some individuals, deviating from the original values. If the mutated strategy has a better fitness value, it will be retained in the population, otherwise it will be discarded or mutated again.

10. The method for controlling a distribution network area energy storage system considering the carrying capacity of an active distribution network according to claim 1, characterized in that: The mathematical expression of the initialization particle position of the particle swarm optimization algorithm PSO module is: The mathematical expression for initializing particle velocity is: In formulas (8) and (9), i represents the i-th particle, j represents the dimension of the problem, Represents the initial position of the ith particle, Represented as the initialization particle speed of the i-th particle; each particle randomly generates an initial position, representing an initial energy storage operation plan, and the energy storage control parameters corresponding to each particle at least include the charging and discharging power time; The mathematical operation expression for updating particle velocity is: In formula (10), M is the inertia weight, t is the number of iterations, r1 and r2 are random numbers, a1 and a2 are acceleration factors, is the optimal position of particle i at the tth iteration, is the global optimal position of particle i at the tth iteration, M max is the maximum value of the inertia weight, M min is the minimum value of the inertia weight, N is the current number of iterations, N max is the maximum number of iterations of the model; the mathematical operation expression for updating the particle position is: Use the updated speed to update the particle's position to obtain a new energy storage operation plan; The mathematical operation expression for updating the optimal position of the particle is: The mathematical operation expression for updating the global optimal position of the particle swarm is: In formula (13), r is a random number and a is an acceleration factor; In each iteration process, the performance of the newly generated energy storage operation scheme, that is, the carrying capacity of the active distribution network, is evaluated, and the carrying capacity of the active distribution network is compared with the individual historical optimal and global optimal solutions. If the performance of the new scheme is better, the individual optimal solution or the global optimal solution is updated.

11. The method for controlling a distribution network area energy storage system considering the carrying capacity of an active distribution network according to claim 1, characterized in that: The hybrid strategy module includes parallel hybrid, serial hybrid and embedded hybrid; When mixed in parallel, the genetic algorithm GA module and the particle swarm optimization algorithm PSO module run independently, exchanging information and sharing the optimal solution every ten iterations; The serial hybrid first uses the genetic algorithm GA module and then switches to the particle swarm optimization algorithm PSO module; In embedded hybrid, the particle swarm optimization algorithm PSO module is embedded in the iterative process of the genetic algorithm GA module; The fitness evaluation module defines a fitness function for measuring the advantage of each solution.

Citation Information

Cited By

  • DDos attack identification method based on machine learning

    CN119232484A

  • Superconductive power equipment partial discharge positioning method and system based on temperature field distribution

    CN121231963A