Power distribution network self-healing control method based on forward detection and backtracking topology search algorithm

Abstract

The invention discloses a power distribution network self-healing control method based on a forward detection and backtracking topology search algorithm, and belongs to the field of power distribution network self-healing control. The method comprises the following steps: constructing a power distribution network simplified model; analyzing the network topology structure of the power distribution network; wherein the topology analysis method adopts a forward detection and backtracking topology search algorithm; constructing a distribution network self-healing reconstruction strategy optimization model based on a fuzzy weighting method; and solving the distribution network self-healing reconstruction strategy optimization model based on an improved sparrow search (ISSA) algorithm. The method can adapt to topology analysis of a large-scale network, effectively realizes formulation and optimization of a self-healing strategy of the power distribution network, and achieves rapid self-healing control of the power distribution network.

Description

Technical Field

[0001] The present invention relates to a self-healing control method for a distribution network based on a forward exploration and backtracking topology search algorithm, and belongs to the field of self-healing control of a distribution network. Background Art

[0002] As the last link in power production and supply, the distribution network directly faces users and is a bridge connecting power production and electricity users. It is related to the quality of power supply safety and reliability. Therefore, it is urgent to establish a perfect distribution network control strategy to form a self-healing intelligent distribution network with flexible operation mode, which can complete fault isolation and recovery in a short time, within a small range, and with less manual intervention after a fault occurs.

[0003] The self-healing of a distribution network refers to the ability of the distribution network to continuously evaluate its own operating state and complete self-prevention and self-recovery without or with less human intervention. An intelligent distribution network with self-healing function is stronger and more "elastic" than a traditional distribution network, and can effectively prevent and resist the impacts and risks brought by natural disasters and emergencies to the power system. When formulating a self-healing control strategy for a distribution network, it is necessary to analyze the topological structure of the existing distribution network, and the connection relationship and connection sequence between switches and feeders are the basis for a computer to calculate network power flow. Through topology search, a grid node-switch model containing grid descriptions can be established. The existing method for forming a topology node (TN), also known as a node fusion algorithm, merges equipotential points in the node-switch model to establish a node-branch model for power grid analysis and calculation.

[0004] In the process of constructing a network model, search algorithms are often used. Topology search is usually based on graph theory algorithms and describes a graph using an adjacency list or an adjacency matrix. Most of the search algorithms that use matrix search algorithms for the topology search of a distribution network and implement the topology analysis of the distribution network through an adjacency matrix and its improved algorithms need to perform matrix multiplication, with a large amount of calculation, and storing the adjacency matrix requires a large amount of memory. Therefore, matrix search algorithms are not applicable to the topology analysis of large-scale networks; the depth first search (DFS) algorithm and the breadth first search (BFS) algorithm are common search algorithms in graph theory. They are simple and easy to implement and are applicable to the topology analysis of most networks, and are widely used in the topology analysis of power grids. However, traditional search algorithms based on DFS / BFS belong to blind search algorithms, and are mostly used for global analysis of the entire network. The search range is large and the number of searches is relatively large, resulting in low efficiency of topology analysis.

[0005] In addition, the self-healing control of the distribution network directly affects whether the best rapid power restoration plan can be given for the power outage in non-fault areas and whether it can effectively resist the impact of natural disasters and emergencies. However, due to the diverse operation modes and numerous switches in the distribution network, factors such as the load restoration amount, network loss, and self-healing efficiency need to be comprehensively considered when formulating strategies to seek the optimal solution. It can be seen that the problem of fault self-healing in the distribution network is ultimately a discrete multi-objective, multi-constraint, multi-dimensional non-linear optimization problem, which relatively increases the difficulty of self-healing control of the distribution network. Summary of the Invention

[0006] The present invention provides a self-healing control method for a distribution network based on a forward exploration and backtracking topology search algorithm to achieve rapid self-healing control of the distribution network.

[0007] The technical solution of the present invention is as follows:

[0008] A self-healing control method for a distribution network based on a forward exploration and backtracking topology search algorithm includes: constructing a simplified model of the distribution network; analyzing the network topology structure of the distribution network, wherein the topology analysis method uses a forward exploration and backtracking topology search algorithm; constructing an optimization model for the self-healing reconstruction strategy of the distribution network based on the fuzzy weighted method; and solving the optimization model for the self-healing reconstruction strategy of the distribution network based on an improved sparrow search (ISSA) algorithm.

[0009] The construction of the simplified model of the distribution network includes: based on the definition of a dissipative network, simply describing a complex distribution network as an undirected graph representing the topological relationship and operating state of the distribution network.

[0010] The analysis of the network topology structure of the distribution network includes:

[0011] Specifying the rules for network layering and hierarchical search

[0012] Using the information of the substations and voltage levels associated with the equipment, stratifying the distribution network equipment according to different voltage levels; taking the tie switch shared by two feeders as the benchmark, starting from the outlet switch of each feeder as the starting point of the search, and setting the tie switch as the target node of the search.

[0013] Using the forward exploration and backtracking topology search algorithm to conduct searches within the same level

[0014] In the search process, select one side to first enter the node of the current voltage level as the starting point of the search, regard the node on the other side that first enters this level as the end target of the search, search for the path between the two points within this level; when the target point is found, trace back reversely from this node to the upper-level set, and establish the mutual connection layer by layer according to the information of the nodes along the way, and construct a complete topological structure. If the target node cannot be found, it is necessary to transfer to the next level to continue the search until the topological analysis of the entire distribution network is finally completed.

[0015] The construction of the optimization model for the self-healing reconstruction strategy of the distribution network based on the fuzzy weighting method includes: constructing a multi-objective function, assigning weight coefficients to different objective functions through fuzzy evaluation of indicators, and finally obtaining the optimization model for the self-healing reconstruction strategy of the distribution network by weighted summation. The indicators include: the number of switch transpositions, the amount of lost load, the load distribution balance degree, the network loss, the load transfer amount in the non-lost power area, the reserve degree for re-faults, the voltage quality, and the self-healing ability index; considering the network structure constraints, the feeder capacity constraints, the branch current constraints, the node voltage constraints, and the power flow equation constraints.

[0016] The solution of the optimization model for the self-healing reconstruction strategy of the distribution network based on the improved sparrow search algorithm lies in introducing the global optimal solution of the previous generation and the dynamic weight ω, which can avoid the algorithm from falling into local optimum and improve the convergence speed.

[0017] The beneficial effects of the present invention are:

[0018] Based on the forward-backtracking topology search algorithm for the topology analysis of the distribution network, the target node is searched from the source point through the forward process, and the path between two points is obtained through the backtracking process. Compared with the general depth-first search (DFS) and breadth-first search (BFS) algorithms, the search times can be significantly reduced, and it can adapt to the topology analysis of large-scale networks with small computational complexity, significantly improving the search efficiency; based on the improved ISSA algorithm, the solution of complex non-linear models with multiple objectives and multiple constraints can be effectively realized, and the formulation and optimization of the self-healing strategy of the distribution network can be achieved, reaching the fast self-healing control of the distribution network. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 is a block diagram of the method of the present invention;

[0020] Figure 2 is a flowchart of the forward-backtracking topology search algorithm;

[0021] Figure 3 is a flowchart of the solution of the model based on the improved sparrow search (ISSA) algorithm. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0022] The following will further illustrate the invention in conjunction with the drawings and embodiments, but the content of the present invention is not limited to the described scope.

[0023] Embodiment 1: As Figures 1-3As shown in the figure, the present invention provides a self-healing control method for a distribution network based on a forward exploration and backtracking topology search algorithm, including: constructing a simplified model of the distribution network; analyzing the network topology structure of the distribution network, wherein the topology analysis method uses a forward exploration and backtracking topology search algorithm; constructing an optimization model for the self-healing reconstruction strategy of the distribution network based on the fuzzy weighted method; and solving the optimization model for the self-healing reconstruction strategy of the distribution network based on the improved sparrow search (ISSA) algorithm.

[0024] Further, the construction of the simplified model of the distribution network includes: based on the definition of a dissipative network, simply describing a complex distribution network as an undirected graph representing the topological relationship and operating state of the distribution network.

[0025] Further, the analysis of the network topology structure of the distribution network includes: stipulating the rules for network layering and layer-by-layer search; and using the forward exploration and backtracking topology search algorithm to perform search within the same layer.

[0026] Further, the stipulation of the rules for network layering and layer-by-layer search includes: using the information of the substation and voltage level associated with the equipment to layer the distribution network equipment according to different voltage levels; based on the tie switches shared by two feeders, taking the outlet switches of each feeder as the starting points of the search and the tie switches as the target position points of the search.

[0027] Further, the search within the same layer using the forward exploration and backtracking topology search algorithm includes: in the search process, selecting one side to first enter the node of the current voltage layer as the starting point of the search, regarding the node on the other side that first enters this layer as the end target of the search, and searching for the path between the two points within this layer; if the target point is found, then trace back from this node to the upper layer set in reverse, and establish the interconnections layer by layer according to the information of the nodes along the way, and construct a complete topological structure. If the target node cannot be found, then it is necessary to transfer to the next layer for continued search until the topological analysis of the entire distribution network is finally completed.

[0028] Further, the construction of the optimization model for the self-healing reconstruction strategy of the distribution network based on the fuzzy weighted method includes: constructing a multi-objective function, assigning weight coefficients to different objective functions by performing a fuzzy evaluation of the indicators, and finally obtaining the optimization model for the self-healing reconstruction strategy of the distribution network by weighted summation; considering the constraints of network structure, feeder capacity, branch current, node voltage, and power flow equation.

[0029] To make the technical means, creative features, achieved purposes, and effects of the present invention easy to understand, the present invention will be further described below in conjunction with specific embodiments:

[0030] 1. Construction of a simplified model of the distribution network

[0031] Before analyzing the fault self-healing strategy of the distribution network, it is necessary to simplify it into a graph structure composed of edges and vertices. Based on the concept of dissipative network, the distribution network can be regarded as a dissipative system, where switches are the nodes of the graph, and the lines between switches and distribution transformers are uniformly regarded as dissipative elements.

[0032] II. Distribution Network Topology Analysis Method Using the Forward Exploration and Backtracking Topology Search Algorithm

[0033] In the distribution network, when determining the tie switches shared by two feeders, starting from the outlet switches of each feeder and taking the tie switches as the end points, search for the connected paths between two points inside the feeder.

[0034] The distribution network is divided into three levels: 110 kV, 35 kV, and 10 kV and below. The components at each level have different impacts on the reliability of load points. The failure of upper-level components is likely to cause the loss of power of lower-level buses and affect the power supply to users; while the failure of the lower-level network has little impact on the upper-level network due to the normal operation of automatic switches. Therefore, it is necessary to ensure the reliability of component equipment level by level according to the voltage level to guarantee the power supply quality for users.

[0035] After entering a certain level, select the points that initially enter this level on both sides as the source point and the target point respectively, and search for the connected path within this level. For the clear target topology search of closing the loop, the present invention proposes the forward exploration and backtracking algorithm, and the process is as Figure 2 shown, and the steps are as follows.

[0036] (1) Select the source point and the target point, and start the search from the source point. First, obtain the connection nodes of the source point equipment and the equipment connected to it, and form the first-level search set.

[0037] (2) Then, search for the connection points of the equipment in the first-level search set, and search for the equipment corresponding to these connection points within the remaining range, and classify them into the second-level search set.

[0038] (3) Continue like this until the target point equipment is found or no new equipment can be found anymore.

[0039] (4) If the target point is found, trace back from the target point to the upper-level set in reverse, and confirm the relationship layer by layer according to the node information to construct the topology structure; if not found, it means that the source point and the target point are not connected at this level, and it is necessary to enter the next level for search.

[0040] III. Optimization Model of Distribution Network Self-Healing Reconfiguration Strategy Based on Fuzzy Weighting Method

[0041] The fault self-healing of the distribution network is a multi-objective problem. The difference between its optimization function and the traditional fault recovery objective function is that in addition to considering some conventional indicators, the self-healing ability indicators of the distribution network also need to be taken into account, so that the selected fault self-healing scheme can better meet the requirements of the power grid for the "self-healing" ability, improve the ability to prevent re-faults, and achieve a balanced load at each power source point as much as possible. Since the fault strategies corresponding to different self-healing objectives have different optimization objective functions. Therefore, it is necessary to establish the fault self-healing objective function according to different self-healing objectives and focuses. The present invention mainly considers the following indicators:

[0042] (1) The number of switch transpositions Y op

[0043] Fault self-healing is achieved by manipulating switches, which will increase the wear of the switches and thus affect their service life. Given that the distribution network cannot fully realize the remote operation of switches, when calculating the switch operation cost, it is necessary to distinguish between remotely controllable switches and switches that need to be operated locally. Based on the self-healing objectives and core requirements, the present invention constructs a fault self-healing objective function and focuses on the following indicators:

[0044] Y op = Y0 + K op Y1

[0045] where Y0 is the number of remotely operated switches; Y1 is the number of locally operated switches; K op is the equivalent penalty coefficient for the impact of local operation on self-healing.

[0046] (2) The amount of lost load L loss

[0047] When constructing the power grid, feeder margins are reserved to cope with faults. However, in the case of severe faults, the margins may not be sufficient to supply power to all non-fault areas. At this time, it is necessary to reduce the power outage loss, consider the load priority, and give priority to restoring the power supply of important loads:

[0048] L loss = K loss ∑L 1i + ∑L 2j

[0049] where L 1i is the i-th important load that has not been restored, and L 2j is the j-th ordinary load that has not been restored, and K loss is the penalty coefficient for the loss of important loads.

[0050] (3) The load distribution balance degree J

[0051] The present invention proposes an innovative load distribution balance index to ensure the balanced distribution of loads at each power source point during the self-healing process of the distribution network. The objective function is to minimize the difference between the maximum and minimum feeder current-carrying capacities. The specific approach is as follows: Sort the feeder current-carrying capacities and calculate the ratio J of the minimum to the maximum value:

[0052] J = L1 / L n

[0053] where n is the number of lines, L1 is the minimum value of the current-carrying capacities of each feeder, and L n is the maximum value of the current-carrying capacities of each feeder.

[0054] (4) Network loss w

[0055] Network loss is a key indicator for the economic operation of the power system. When selecting a fault self-healing scheme for the distribution network, it is necessary to focus on reducing power losses and improving economic benefits. Given the characteristic that the load varies at all times, the present invention selects the load state at a specific moment to calculate the power loss, and uses the active power loss of the system as the measure of network loss.

[0056]

[0057] where m is the total number of branches, r i is the resistance of the branch, P i , Q i , |V i | are the active power, reactive power, and the amplitude of the node voltage of the branch.

[0058] (5) Load transfer amount L in the non-power-loss area t

[0059] If the non-fault power outage area is too large and exceeds the power supply capacity of the adjacent power source, it may be necessary to transfer the load in some normal power supply areas over a long distance to other power sources to enhance the backup power supply for the adjacent power-loss areas. This part of the transferred load is the load transfer amount in the non-power-loss area, which aims to balance the power supply economy and reliability requirements. Its functional expression is:

[0060]

[0061] where is the load transfer amount at the i-th location, and T is the number of load transfers.

[0062] (6) Reserve degree R for re-faults

[0063] When the system fails, normal lines are needed to supply power instead. Therefore, the power grid lines need to have spare capacity to achieve self-healing. To evaluate the reserve for re-faults of the lines after a fault, the maximum feeder load is considered as the measure of the recovery reserve:

[0064]

[0065] Among them, S j is the actual load power of the feeder, and S max is the limit load power of the feeder.

[0066] (7) Voltage quality V

[0067] After the fault self-healing, the key to the stable operation of the system lies in the voltage quality of each node. The degree of deviation of the node voltage from the system power supply voltage is used as the evaluation index, that is, the absolute value of the maximum voltage drop percentage of the node:

[0068]

[0069] Among them, V i represents the actual voltage of each node, and V0 represents the system power supply voltage.

[0070] (8) Self-healing ability index

[0071] The self-healing ability index measures the ability of the network to resist re-faults and risks after a network fault, and is shown by the invulnerability. As the invulnerability improves, the self-healing ability index also increases. Given that the invulnerability value of the distribution network is often low, it is amplified as the self-healing ability index value.

[0072] K = imv(G)

[0073] Different measurement standards and fault self-healing requirements for distribution networks are different. The influence and focus of the above indicators on the self-healing strategy are all different. Therefore, each indicator is represented by a membership degree in the interval of 0-1, and a fuzzy evaluation is carried out on each indicator to obtain the corresponding weight.

[0074] The distribution network standards and self-healing requirements are different, and the influence and focus of each indicator on the self-healing strategy are different. Therefore, it is represented by a membership degree in the interval of 0-1, and a fuzzy evaluation is carried out to determine the weight. Finally, the objective function of the comprehensive optimization of the fault self-healing strategy is obtained by fuzzy weighted summation:

[0075]

[0076] The present invention uses the analytic hierarchy process to determine the weight of each indicator. Among them, α(·) is the membership function of each indicator, and λ j represents the weight coefficient of each indicator, and satisfies:

[0077]

[0078] Constraints for distribution network fault self-healing:

[0079] (1) Network structure constraint: The distribution network must maintain a radial structure during operation to avoid the occurrence of a loop network.

[0080] When i≠j,

[0081] Among them, F i and F j are respectively the sets of loads powered by components i and j.

[0082] (2) Constraint of feeder capacity: The actual operating capacity of the feeder needs to meet the constraint of its maximum capacity, and the inequality needs to be satisfied:

[0083] S k ≤S kmax

[0084] Among them, S k represents the actual power of the kth line: S kmax is the maximum allowable passing power of the kth line.

[0085] (3) Branch current constraint: The branch current shall not exceed its maximum current-carrying capacity, and it needs to satisfy:

[0086] I j ≤I jmax

[0087] Among them, I j represents the actual current-carrying capacity of the jth branch in the system operation; I jmax represents the maximum allowable passing current-carrying capacity of the jth branch.

[0088] (4) Node voltage constraint:

[0089] U min ≤U i ≤U max

[0090] Among them, U i represents the actual voltage value of the ith node in the system operation; U min and U max are the minimum and maximum voltages allowed at the ith node.

[0091] (5) Constraint of power flow equation:

[0092]

[0093] Among them, P i , Q i are respectively the injected active and reactive powers of node i; are respectively the voltage vector forms of nodes i and j; is the admittance matrix element; N is the set of system nodes.

[0094] IV. Solving the model based on the improved sparrow search (ISSA) algorithm

[0095] In the present invention, the optimization model of the distribution network self-healing reconstruction strategy belongs to a discrete, multi-objective, multi-dimensional and non-linear optimization problem with multiple constraints, and an improved ISSA algorithm is used for solution. The algorithm flow is as Figure 3 shown.

[0096] The sparrow search algorithm is a swarm intelligence optimization algorithm that simulates the foraging behavior of sparrows, including three types of individuals: discoverers, followers, and vigilant ones, and updates their positions according to their respective rules.

[0097] (1) Initialize the parameters and initialize the sparrow population using the Bernoulli chaotic map.

[0098] The chaotic map used is the Bernoulli equation:

[0099]

[0100] where X n is the current value of the nth generation chaotic sequence generated, and λ is the control parameter with a value of 0.3.

[0101] (2) According to the initial distribution network parameters, calculate the fitness value of each sparrow, and sort to select the current optimal fitness value, the worst fitness value and their corresponding positions.

[0102] (3) Update the positions of the discoverers, followers, and vigilant ones respectively.

[0103] Introducing the global optimal solution of the previous generation and the dynamic weight ω can avoid the algorithm falling into local optimum and improve the convergence speed.

[0104] ω = (ω min + ω max ) / 2 + (ω max + ω min ) cos(tπ / MaxIter)

[0105]

[0106] where represents the position of individual i at the t-th iteration in the j-th dimension, ω max , ω min are the inertia weights before and after iteration; t is the current iteration number, MaxIter is the maximum number; ω max = 0.95, ω min = 0.4, the algorithm has the optimal performance under these conditions. rand is a random number in [0, 1], R2 is the warning value, R2 ∈ [0, 1], M s ∈ [0, 1] is the safety value; Q is a random number obeying the normal distribution in [0, 1]; is the optimal position of the sparrow in the j-th dimension at the t-th iteration.

[0107] Follower position update:

[0108]

[0109] wherein, The position of follower individual i in dimension j at the t-th iteration. Is the worst position of the sparrows at the t-th iteration, Is the optimal position at the (t + 1)-th iteration; A + Is a 1×d random vector matrix, and the elements are randomly 1 or -1; L is a step size control factor used to adjust the amplitude of position update. Indicates that the i-th follower has not obtained food and has a low fitness and needs to find another area; Then it means that the i-th follower forages randomly nearby.

[0110] Scout position update:

[0111]

[0112] wherein, Is the position of sparrow i in the j-th dimension at the t-th iteration; Is the optimal position of the sparrows at the t-th iteration; Is the worst position of the sparrows at the t-th iteration; β is the step size; f s Is the current fitness; f g Is the global best value; Is a random number in [-1, 1]; f w Is the global worst value; ε is a very small constant to avoid the denominator being zero. If a certain condition is greater than a certain value, f g > f s Indicates that the sparrow is at the edge of the population and is easily preyed upon; if f s > f g Indicates that the sparrow is at the center of the population and approaches its companions to avoid danger due to perceiving a threat.

[0113] (4) Update the fitness, select the best sparrow position as the optimal solution and record it, and update the global optimal solution.

[0114] (5) Determine whether to terminate. If it meets the condition, output the optimal solution and end; if not, return to step 2 to continue the iteration until the condition is met and output the optimal solution.

[0115] The specific implementation manners of the present invention have been described in detail above in conjunction with the accompanying drawings. However, the present invention is not limited to the above implementation manners, and various changes can be made without departing from the gist of the present invention within the scope of knowledge possessed by those of ordinary skill in the art.

Claims

1. A self-healing control method for a distribution network based on a look-ahead backtracking topological search algorithm, characterized in that Including: Constructing a simplified model of the distribution network; Analyzing the network topology of the distribution network; among them, the topology analysis method adopts the forward exploration and backtracking topology search algorithm; Constructing an optimization model for the self-healing reconstruction strategy of the distribution network based on the fuzzy weighted method; Solving the optimization model for the self-healing reconstruction strategy of the distribution network based on the improved sparrow search algorithm.

2. The distribution network self-healing control method based on the look-ahead backtracking topological search algorithm according to claim 1, characterized in that, The construction of the simplified model of the distribution network includes: Based on the definition of the dissipative network, a complex distribution network is simply described as an undirected graph representing the topological relationship and operating state of the distribution network.

3. The self-healing control method for a distribution network based on the look-ahead backtracking topology search algorithm according to claim 1, characterized in that The analysis of the network topology of the distribution network includes: Specifying the rules for network layering and hierarchical search Using the information of the substations and voltage levels associated with the equipment, the distribution network equipment is layered according to different voltage levels; taking the tie switch shared by two feeders as the benchmark, starting from the outlet switch of each feeder as the starting point of the search, and setting the tie switch as the target node of the search; Using the forward exploration and backtracking topology search algorithm to perform searches within the same layer In the search process, select one side to enter the node of the current voltage level first as the starting point of the search, regard the node on the other side that enters this level first as the end target of the search, and search for the path between the two points within this level; If the target point is found, trace back from this node to the upper-level set in reverse, and establish the interconnections layer by layer according to the information of the nodes along the way, and construct a complete topological structure; if the target node is not found, it is necessary to transfer to the next level to continue the search until the topological analysis of the entire distribution network is finally completed.

4. The self-healing control method for a distribution network based on the look-ahead backtracking topology search algorithm according to claim 1, wherein The construction of the optimization model for the self-healing reconstruction strategy of the distribution network based on the fuzzy weighted method includes: Constructing a multi-objective function, assigning weight coefficients to different objective functions by performing a fuzzy evaluation of the indicators, and finally obtaining the optimization model for the self-healing reconstruction strategy of the distribution network by weighted summation; Considering the constraints of network structure, feeder capacity, branch current, node voltage, and power flow equation.

5. The self-healing control method for a distribution network based on the look-ahead backtracking topology search algorithm according to claim 4, wherein The indicators in the multi-objective function include: the number of switch transpositions, the amount of lost load, the load distribution balance degree, network loss, the load transfer amount in the non-lost power area, the reserve degree for re-faults, voltage quality, and self-healing ability indicators.

6. The self-healing control method for a distribution network based on the look-ahead backtracking topology search algorithm according to claim 1, wherein The solution of the optimization model for the self-healing reconstruction strategy of the distribution network based on the improved sparrow search algorithm includes: (1) Initializing the parameters and initializing the sparrow population using the Bernoulli chaotic map; (2) According to the initial distribution network parameters, calculating the fitness value of each sparrow, sorting to select the current optimal fitness value, the worst fitness value and their corresponding positions; (3) Updating the positions of the discoverers, followers and vigilants respectively, and introducing the global optimal solution of the previous generation and the dynamic weight ω during the update process; ω=(ω min +ω max ) / 2+(ω max +ω min )cos(tπ / MaxIter) ω max , ω min are the inertia weights before and after iteration, t is the current iteration number, and MaxIter is the maximum number of iterations; (4) Updating the fitness, selecting the best sparrow position as the optimal solution and recording it, and updating the global optimal solution; (5) Judging whether to terminate. If so, output the optimal solution and end; if not, return to step (2) to continue the iteration until the conditions are met and the optimal solution is output.

Images