A power distribution network reconstruction method based on SA and CS hybrid algorithm

Abstract

This invention discloses a distribution network reconfiguration method based on a hybrid SA and CS algorithm. The method acquires basic reconfiguration data, operational status data, operational constraint data, and switch status data of the target distribution network, and establishes basic loop breakpoint codes; generates candidate switch combinations, and updates the breakpoint codes corresponding to the candidate switch combinations using a cuckoo search algorithm; performs topology feasibility repair, adjusting the breakpoint codes based on closed-loop status, islanding status, node voltage over-limit status, and line overload status; calculates a comprehensive evaluation value and operational constraint improvement for the repaired candidate switch combinations; accepts the candidate switch combinations when the comprehensive evaluation value improves, and if the comprehensive evaluation value does not improve but the operational constraint improvement is greater than zero, determines whether to accept them according to the acceptance criteria of the simulated annealing algorithm. This method can reduce unfeasible candidate switch combinations, improve the adaptability of the distribution network reconfiguration scheme to node voltage and line capacity constraints, and output a target reconfigured topology that satisfies radial topology and operational constraints.

Description

Technical Field

[0001] This invention relates to the field of distribution network optimization operation technology, and specifically to a distribution network reconfiguration method based on a hybrid SA and CS algorithm. Background Technology

[0002] With the continuous increase in the scale of photovoltaic, wind power, and energy storage connected to the distribution network, the operation of the distribution network is jointly affected by the output fluctuations of distributed generation sources, load changes, line power flow transfer, and node voltage variations. Distribution network reconfiguration, by changing the on / off states of sectionalizing switches and tie switches, adjusts the distribution network topology and is an important operation control means to reduce network losses, improve voltage, and optimize power flow distribution.

[0003] In distribution networks containing distributed generation, reconfiguration calculations need to simultaneously consider the radial operation structure of the distribution network, node connectivity, line capacity, node voltage, number of switching operations, and the location of distributed generation connections. Distribution network reconfiguration problems typically manifest as optimization problems coupling discrete and continuous power flow variables, with a large number of switching state combinations, and different switching combinations having varying impacts on power flow distribution and node voltage.

[0004] Simulated Annealing (SA) has the ability to accept suboptimal candidate solutions and escape local search regions, while Cuckoo Search (CS) has strong global search capabilities and fast candidate solution update capabilities. Combining the two for distribution network reconfiguration can provide an algorithmic foundation for optimizing switch combinations. However, in distribution network reconfiguration scenarios, candidate switch combinations must satisfy radial topology, islanded operation, voltage limits, and line capacity limits. Simply superimposing the simulated annealing and cuckoo search algorithms may still generate a large number of unexecutable switch combinations during the candidate solution update process, and the simulated annealing acceptance process may also retain candidate solutions that do not improve the constraints on distribution network operation. Summary of the Invention

[0005] This invention addresses the shortcomings of existing technologies by providing a distribution network reconfiguration method based on a hybrid SA and CS algorithm. This method solves the problems in existing technologies, such as the Cuckoo Search algorithm easily generating unexecutable candidate solutions corresponding to closed loops, islanding, low-voltage nodes, and line overloads during the candidate switch combination update process of distribution network reconfiguration, as well as the problem that the ordinary simulated annealing acceptance criterion only accepts non-optimal candidate solutions based on the target evaluation value and annealing temperature, resulting in some accepted candidate solutions not improving node voltage overruns, line overloads, and project executability.

[0006] To achieve the above objectives, this invention provides a distribution network reconfiguration method based on a hybrid SA and CS algorithm, comprising the following steps: S1. Obtain the basic data for reconfiguration of the target distribution network, operating status data, operating constraint data, and switch status data; S2. Identify the basic loops formed in the target distribution network after the tie switch is closed based on the reconstructed basic data, and establish basic loop breakpoint codes; S3. Generate an initial candidate switch combination set based on the basic loop breakpoint encoding, and perform topological feasibility verification on the candidate switch combinations in the initial candidate switch combination set based on the reconstructed basic data and switch state data to obtain a feasible candidate switch combination set. S4. The candidate switch combinations in the feasible candidate switch combination set are used as the nest positions of the cuckoo search algorithm, and the basic loop breakpoint codes corresponding to the candidate switch combinations are updated by the cuckoo search algorithm to obtain the updated candidate switch combinations. S5. Based on the reconstructed basic data, operating status data and operating constraint data, perform topology feasibility repair on the updated candidate switch combination, adjust the basic loop breakpoint coding, and obtain the repaired candidate switch combination. S6. Perform power flow calculation on the candidate switch combination to obtain the comprehensive evaluation value and operational constraint improvement amount corresponding to the candidate switch combination; S7. When the comprehensive evaluation value of the candidate switch combination to be repaired is better than the comprehensive evaluation value of the current switch combination, the candidate switch combination to be repaired is accepted, and the current candidate switch combination set is updated with the candidate switch combination to be repaired. S8. When the comprehensive evaluation value of the candidate switch combination to be repaired is not better than the comprehensive evaluation value of the current switch combination, and the improvement amount of the operating constraints is greater than zero, the candidate switch combination to be repaired is judged according to the acceptance criteria of the simulated annealing algorithm based on the current annealing temperature, the difference in comprehensive evaluation values ​​and the improvement amount of operating constraints; when the judgment result is to accept, the current candidate switch combination set is updated with the candidate switch combination to be repaired; otherwise, the candidate switch combination to be repaired is discarded. S9. When the comprehensive evaluation value of the candidate switch combination to be repaired is not better than the comprehensive evaluation value of the current switch combination, and the improvement amount of the operating constraints is not greater than zero, the candidate switch combination to be repaired is discarded. S10. When the preset iteration stop condition is reached, select a candidate switch combination from the current candidate switch combination set that satisfies the following conditions: node connectivity, radial structure, no islanding operation, node voltage allowable range, line allowable transmission capacity, and switch action limit, as the target switch combination, and output the reconstructed distribution network topology based on the target switch combination.

[0007] To optimize the above technical solution, the specific measures also include: Further, in step S1, the reconstructed basic data is used to characterize the connection relationship between nodes, lines, sectionalizing switches and tie switches in the target distribution network, including node number, line start node, line end node, line impedance, sectionalizing switch position and tie switch position; The operational status data is used to characterize the load access status and distributed generation access status of the target distribution network during the reconfiguration calculation period, including load active power, load reactive power, distributed generation access nodes, and distributed generation output data. The operational constraint data is used to characterize the allowable range of node voltage, allowable transmission capacity of lines, and switch action limits of the target distribution network, including the upper limit of node voltage, the lower limit of node voltage, the line capacity limit, and the limit of the number of switch actions; The switch status data is used to characterize the open / closed state of the sectionalizing switch and the tie switch before reconstruction, including the open / closed state of the sectionalizing switch and the open / closed state of the tie switch before reconstruction.

[0008] Further, in step S2, the basic loop breakpoint code is used to represent the disconnected switches in each basic loop, and the establishment process includes: Close the tie switches in the target distribution network and identify the basic loops formed by the closure of the tie switches; The segment switches and tie switches that can serve as breakpoints in each basic loop are combined into a candidate breakpoint set for the corresponding basic loop. The switch number in the candidate set of breakpoints for each basic loop is used as the breakpoint code value for that basic loop; Arrange the breakpoint codes according to the numbering order of the basic loop to form the basic loop breakpoint codes corresponding to the candidate switch combinations.

[0009] Furthermore, in step S3, the topology feasibility verification of the candidate switch combinations includes: Determine the set of closed circuits and the set of open circuits based on the candidate switch combinations; Based on the set of closed circuits, determine whether all load nodes in the target distribution network have a power supply path to the power source nodes; Determine whether the candidate switch combination satisfies the radial structure based on the number of closed circuits, the number of nodes, and the connection relationship of power supply nodes; Determine whether the candidate switch combination forms an island node based on the power supply path of each load node; Candidate switch combinations that simultaneously satisfy the conditions of having a power supply path to the load node, having a radial structure, and not having formed an island node are written into the set of feasible candidate switch combinations.

[0010] Further, in step S4, updating the basic loop breakpoint code corresponding to the candidate switch combination using the cuckoo search algorithm includes: The number of basic loops that need to be adjusted in this update is determined based on the current iteration number and the step size factor. Select the breakpoint code bits that need to be adjusted from the basic loop breakpoint codes corresponding to the candidate switch combinations; Select a new switch number from the breakpoint candidate set corresponding to the selected breakpoint code bit; Write the newly selected switch number into the corresponding breakpoint code bit to form the updated basic loop breakpoint code; An updated candidate switch combination is generated based on the updated basic loop breakpoint code.

[0011] Further, in step S5, the topology feasibility repair adjusts the basic loop breakpoint coding based on the closed-loop state, islanding state, node voltage over-limit state, and line overload state of the updated candidate switch combinations. Adjusting the basic loop breakpoint coding based on the closed-loop state and islanding state includes: When updating the candidate switch combination to form a closed loop, identify the set of branches that form the closed loop, and select the switch that meets the node voltage allowable range and the line allowable transmission capacity after being disconnected from the set of branches as the disconnecting switch; When updating candidate switch combinations to form an island, identify the island node set, close the switch that can reconnect the island node set with the power node, and select a switch as the disconnect switch in the basic loop formed after closing the switch. The basic loop breakpoint code is updated based on the disconnected switch, so that the updated candidate switch combination is restored to a radial topology and the islanded nodes are eliminated.

[0012] Furthermore, in step S5, adjusting the basic loop breakpoint coding based on the node voltage over-limit status and line overload status includes: When the updated candidate switch combination satisfies the radial structure and there is a node voltage over-limit, the power supply path from the voltage over-limit node to the power supply node is determined, and the breakpoint code value is adjusted in the basic loop related to the power supply path. When the updated candidate switch combination satisfies the radial structure and there is a line overload, the power supply path of the overloaded line is determined, and the breakpoint coding value is adjusted in the basic loop related to the upstream and downstream nodes of the overloaded line. Power flow calculations were performed on the adjusted candidate switch combinations, and the candidate switch combinations that met the allowable range of node voltage and the allowable transmission capacity of the line were selected as the repair candidate switch combinations.

[0013] Furthermore, in step S6, the comprehensive evaluation value is determined based on the active power loss evaluation quantity, the node voltage deviation evaluation quantity, the line operation constraint deviation evaluation quantity, and the switch action evaluation quantity; The active power loss evaluation quantity is determined based on the active power loss of the line obtained from the power flow calculation. The node voltage deviation evaluation quantity is determined based on the deviation between each node voltage and the corresponding node voltage allowable range; The deviation evaluation value of the line operation constraint is determined based on the degree of deviation between the load rate of each line and the allowable transmission capacity of the corresponding line. The switching action evaluation quantity is determined based on the amount of change of the candidate switch combination relative to the initial state of the switch.

[0014] Further, in step S6, the improvement amount of the operating constraints is used to characterize the degree of improvement of the node voltage over-limit and line overload degree of the repair candidate switch combination relative to the current switch combination. It is obtained based on the difference between the degree of deviation of the operating constraints corresponding to the current switch combination and the degree of deviation of the operating constraints corresponding to the repair candidate switch combination. The degree of deviation from the operating constraints is determined based on the degree of node voltage exceeding limits and the degree of line overload. When the deviation of the operating constraints corresponding to the current switch combination is greater than the deviation of the operating constraints corresponding to the candidate switch combination to be repaired, the improvement amount of the operating constraints is greater than zero; When the deviation of the operating constraints corresponding to the current switch combination is not greater than the deviation of the operating constraints corresponding to the candidate switch combination to be repaired, the improvement amount of the operating constraints is not greater than zero.

[0015] Further, in step S8, determining whether to accept the candidate switch combination for repair according to the acceptance criteria of the simulated annealing algorithm includes: The acceptance probability of the candidate switch combination is calculated based on the comprehensive evaluation value of the candidate switch combination, the comprehensive evaluation value of the current switch combination, the current annealing temperature, and the improvement amount of operating constraints. Generate a random number and compare the random number with the acceptance probability; When the random number is less than the acceptance probability, the candidate switch combination for repair is accepted. When the random number is not less than the acceptance probability, the candidate switch combination for repair is discarded; The current annealing temperature is updated according to the preset cooling coefficient, and then the next iteration begins.

[0016] The beneficial effects of this invention are as follows: By using basic loop breakpoint coding, this invention enables candidate switch combinations to directly correspond to the disconnected switches in each basic loop, reducing the large number of infeasible candidate solutions generated by random coding of all switch states; through topology feasibility repair, the updated candidate switch combinations after the cuckoo search can adjust breakpoint coding for closed loops, islanding, node voltage exceedances, and line overloads, improving the engineering feasibility of candidate solutions; by controlling the simulated annealing acceptance process with operational constraint improvement quantities, candidate switch combinations whose comprehensive evaluation values ​​have not improved only enter the acceptance judgment when improving the degree of node voltage exceedances and line overloads, avoiding the acceptance of non-optimal candidate solutions that do not improve operational constraints; by simultaneously considering active power loss, node voltage deviation, line operational constraint deviation, and the number of switch actions in the comprehensive evaluation value, the output target switch combination balances network loss, voltage, line safety, and operating costs. Attached Figure Description

[0017] Figure 1 This is a flowchart illustrating the distribution network reconfiguration method based on the SA and CS hybrid algorithm of this invention. Figure 2 This is a schematic diagram of the combination of basic loop breakpoint coding and candidate switch generation in this invention; Figure 3 This is a schematic diagram of the topology-feasible repair process of the present invention; Figure 4 This is a schematic diagram of the simulated annealing acceptance judgment method controlled by the improved amount of the operation constraint according to the present invention. Detailed Implementation

[0018] The invention will now be described in further detail with reference to the accompanying drawings.

[0019] It should be noted that the terms such as "upper", "lower", "left", "right", "front", and "back" used in the invention are only for clarity of description and are not intended to limit the scope of the invention. Changes or adjustments to their relative relationships, without substantially altering the technical content, should also be considered within the scope of the invention.

[0020] The embodiments described in this invention are merely some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without inventive effort are within the scope of protection of this invention.

[0021] like Figure 1-4 As shown, the embodiments of the present invention are as follows: Input data and its function.

[0022] This embodiment provides a distribution network reconfiguration method based on a hybrid SA and CS algorithm. The target distribution network is a 10kV radial distribution network with distributed generation access. This target distribution network includes multiple load nodes, multiple distribution lines, multiple sectionalizing switches, and multiple tie switches. Distributed generation can include photovoltaic power, wind power, and energy storage devices. This method uses the on / off states of sectionalizing switches and tie switches in the distribution network as the reconfiguration object, and uses the disconnected switches in the basic loop as candidate decoding objects. Candidate switch combinations are generated using the Cuckoo Search algorithm, and the candidate switch combinations are made to meet the distribution network operation structure requirements through topology feasibility repair. Finally, the candidate switch combinations are screened using a simulated annealing acceptance criterion controlled by operational constraint improvement quantities.

[0023] The specific steps are as follows: First, acquire the basic reconfiguration data, operational status data, operational constraint data, and switch status data of the target distribution network. The basic reconfiguration data is used to establish the distribution network topology. In one embodiment, the basic reconfiguration data includes node numbers, line start-up nodes, line end-up nodes, line impedance, sectionalizing switch positions, and tie-line switch positions. Node numbers are used to uniquely identify nodes in the target distribution network; line start-up nodes and line end-up nodes are used to characterize line connection relationships; line impedance is used for power flow calculation; and sectionalizing switch positions and tie-line switch positions are used to determine the set of operable switches participating in the reconfiguration.

[0024] Operational status data is used for power flow calculations. In one embodiment, the operational status data includes load active power, load reactive power, distributed generation (DG) access nodes, and DG output data. DG output data characterizes the power state injected into the target distribution network by DG during the reconfiguration calculation period. For static reconfiguration calculations, the DG output data can be the currently collected output data; for short-term reconfiguration calculations, the DG output data can be the predicted output data corresponding to the reconfiguration calculation period. Regardless of whether current or predicted output data is used, its purpose is to serve as the DG injection amount in the power flow calculation.

[0025] Operational constraint data is used to determine whether candidate switch combinations meet the operational requirements of the distribution network. In one embodiment, the operational constraint data includes an upper limit for node voltage, a lower limit for node voltage, a line capacity limit, and a limit on the number of switch operations. The upper and lower limits for node voltage together form the allowable range for node voltage; the line capacity limit is used to determine the allowable transmission capacity of the line; and the limit on the number of switch operations is used to limit the number of operations during the reconfiguration process, avoiding the impact of frequent switch operations on operational reliability.

[0026] Switch status data is used to represent the open / closed state of the sectionalizing switches and tie switches before reconfiguration. This data is used both to generate initial candidate switch combinations and to calculate the number of switching actions of the candidate switch combinations relative to their pre-reconfiguration state.

[0027] Using the above four types of data, the system can clearly reconstruct the input, operating status, constraint boundaries, and current switch status, providing a data foundation for subsequent basic loop identification, candidate switch combination generation, power flow calculation, and constraint repair.

[0028] Basic loops are identified based on the reconstructed foundation data. In this embodiment, the target distribution network adopts a closed-loop design and an open-loop operation structure. During normal operation of the distribution network, some tie switches are in the open state, and sectional switches are in the open and closed state according to the power supply path. When the tie switches involved in the reconstructing are closed, one or more basic loops will be formed in the distribution network topology.

[0029] The system identifies basic loops based on the reconstruction baseline data. Specifically, the system can close each tie switch involved in the reconstruction one by one and use a graph search method to find the closed path formed by closing each tie switch. Each closed path is considered a basic loop. Alternatively, the system can identify the corresponding set of basic loops after closing all tie switches involved in the reconstruction in the topology graph.

[0030] For each basic loop, the system determines a candidate set of breakpoints within that loop. This candidate set consists of sectionalizing switches and tie switches located within the basic loop that are permitted to participate in reconfiguration operations. Switches permitted to participate in reconfiguration operations are those with remote control opening and closing capabilities, not in a maintenance-locked state, not prohibited from operation by protection coordination rules, and allowed to participate in reconfiguration by the scheduling operation strategy. Switches not permitted for remote control operation, in a maintenance-locked state, or prohibited from operation by the scheduling strategy are not included in the candidate set of breakpoints.

[0031] For each basic loop, the system selects a switch number from the corresponding breakpoint candidate set as the breakpoint code value for that basic loop. The breakpoint code values ​​are arranged in order of basic loop number to obtain the basic loop breakpoint code corresponding to the candidate switch combination.

[0032] For example, the target distribution network identifies three basic loops, denoted as L1, L2, and L3. The candidate breakpoint set for L1 is {S3, S8, S12}, for L2 it is {S6, S14, S18}, and for L3 it is {S9, S20, S25}. If the basic loop breakpoint code corresponding to the current candidate switch combination is [S8, S14, S20], it means that S8 is disconnected in basic loop L1, S14 is disconnected in basic loop L2, and S20 is disconnected in basic loop L3.

[0033] The purpose of this invention in employing basic loop breakpoint coding is to ensure that candidate switch combinations directly correspond to the engineering structure of the distribution network's closed-loop design and open-loop operation. For each basic loop formed by the closure of tie switches, if a switch in that basic loop is selected as a breakpoint, the basic loop can be restored to an open-loop operating state. Compared to binary random coding of all segment switches and tie switches, basic loop breakpoint coding reduces candidate switch combinations that significantly form closed loops or disconnect power supply paths, making the search space of the Cuckoo Search algorithm closer to the executable reconfigurable space of the distribution network.

[0034] Initial candidate switch combination generation and topology feasibility verification.

[0035] The system generates multiple initial candidate switch combinations based on the basic loop breakpoint encoding. Each candidate switch combination corresponds to the open / closed state of a specific sectionalizing switch and tie switch.

[0036] After generating candidate switch combinations, the system performs topology feasibility verification. Topology feasibility verification includes node connectivity verification, radial structure verification, and island-free operation verification.

[0037] Node connectivity verification is used to determine whether all load nodes in the target distribution network have a power supply path to the power source nodes. In this invention, a power source node refers to the main power supply node in the target distribution network that ensures the load nodes maintain a power supply connection with the upstream grid. Distributed power source access nodes are treated as power injection nodes in power flow calculations, and are not automatically considered as main power supply nodes allowed to operate independently in islanded mode. If the target distribution network operation procedures allow specific distributed power sources to form independent microgrids, then such distributed power source access nodes can be marked as allowed power supply nodes in the operation constraint data; distributed power source access nodes not marked as allowed power supply nodes are not used to satisfy the islanded operation verification.

[0038] Radial structure verification is used to determine whether the topology corresponding to candidate switch combinations meets the open-loop operation requirements of the distribution network. Within a connected power supply area, if the number of closed circuits equals the number of nodes within that power supply area minus one, and there are no closed loops, then the power supply area is determined to meet the radial structure requirement. For a target distribution network containing multiple main power supply nodes, radial structure verification can be performed separately for each power supply area corresponding to each main power supply node, ensuring that there are no disallowed closed loops between power supply areas.

[0039] The islanding-free operation check is used to determine whether there are load nodes that are not maintaining a power supply connection with the power supply node. If there are load nodes that cannot be connected to the power supply node through a closed circuit set, the candidate switch combination is determined to form an islanded node.

[0040] Candidate switch combinations that pass node connectivity verification, radial structure verification, and island-free operation verification are written into the feasible candidate switch combination set. Candidate switch combinations that fail the verification can either enter the topology feasibility repair process or be directly deleted in the initial stage. Preferably, to improve the quality of the initial candidate solutions, candidate switch combinations that fail the topology feasibility verification in the initial stage are directly deleted, and new candidate switch combinations are generated to replenish the population size.

[0041] The cuckoo search algorithm updates the basic loop breakpoint code.

[0042] The system uses each candidate switch combination in the set of feasible candidate switch combinations as a nest position in the cuckoo search algorithm. The state of the nest position is represented by the basic loop breakpoint encoding.

[0043] When updating the basic loop breakpoint encoding, the Cuckoo Search algorithm converts the step size factor into the number of basic loops that need to be adjusted. Specifically, the adjustment amount can be determined by multiplying the absolute value of the step size factor by the total number of basic loops, and the adjustment amount is limited to between 1 and the total number of basic loops. When earlier iterations require improving global search capability, the adjustment amount is larger; when later iterations require improving local search accuracy, the adjustment amount is smaller. The purpose of this mapping method is to transform the continuous search step size in the Cuckoo Search algorithm into the number of loop breakpoint adjustments suitable for discrete switch combinations.

[0044] After determining the number of basic loops that need adjustment, the system selects the breakpoint code bits that need adjustment from the basic loop breakpoint codes corresponding to the current candidate switch combinations. From the breakpoint candidate set corresponding to the selected breakpoint code bits, the system reselects switch numbers and writes the reselected switch numbers into the corresponding breakpoint code bits, forming the updated basic loop breakpoint codes.

[0045] When reselecting switch numbers, switches already designated as breakpoints can be excluded, and new breakpoint switches can be chosen from the remaining candidate breakpoint switches. Selection can be random or based on a ranking of the branch's line load rate, adjacent node voltage deviation, and post-break network loss changes. Regardless of whether random or ranked selection is used, it must be confirmed in subsequent topology feasibility repair and power flow verification that the switch meets distribution network operation constraints.

[0046] In this way, the Cuckoo Search algorithm does not blindly search within the high-dimensional binary space composed of all switch states, but rather searches within the basic loop breakpoint encoding space. This preserves the global search capability of the Cuckoo Search algorithm while reducing the probability of generating obviously infeasible topologies.

[0047] The overall sequence of topology-feasible repair.

[0048] For the updated candidate switch combinations generated by the Cuckoo Search algorithm, the system performs topology feasibility repair. Topology feasibility repair is performed in the order of topology structure first, followed by running constraints.

[0049] The first step is to determine whether the candidate switch combinations have formed an island. If islanding has occurred, the priority is to restore the power supply connection between the islanded node and the power supply node. This step has the highest priority because subsequent power flow calculations and voltage improvements are meaningless when the islanded node is not connected to the main power supply.

[0050] The second step is to determine whether the candidate switch combination forms a closed loop. If a closed loop is formed, a disconnect switch is selected from the set of branches forming the closed loop to restore the target distribution network to a radial topology. This step ensures that the candidate switch combination meets the open-loop operation requirements of the distribution network.

[0051] The third step, after updating the candidate switch combinations to meet the requirements of node connectivity, radial structure, and island-free operation, is to determine whether the node voltage is within the allowable range and whether the line load rate meets the allowable transmission capacity of the line. If there is a node voltage exceeding the limit, node voltage exceeding the limit repair is performed; if there is a line overload, line overload repair is performed.

[0052] Through the above processing sequence, the system first ensures that the candidate switch combination has an executable topology for the distribution network, and then further improves the power flow operation status, avoiding meaningless operation optimization of candidate switch combinations whose topology is not executable.

[0053] Island repair and closed-loop repair.

[0054] When an island is formed by updating candidate switch combinations, the system identifies the island node set and searches for a switch that can reconnect the island node set to the power supply node. If closing a single switch can restore the power supply path between the island node set and the power supply node, then the switch is closed, and a switch is selected as the disconnect switch in the newly formed basic loop after closing the switch, to ensure that the target distribution network is restored to a radial topology.

[0055] If no situation exists where power connection can be restored by closing a single switch, the candidate switch combination is marked as an unrepairable candidate switch combination and discarded in the current iteration. This approach avoids the algorithm outputting a reconstructed topology where power connectivity cannot be restored.

[0056] When updating candidate switch combinations to form a closed loop, the system identifies the set of branches forming the closed loop and filters out disconnectable switches from the set that are allowed to participate in the reconfiguration operation. The system simulates the topology state after disconnecting each disconnectable switch and performs power flow calculations. For switches that meet the allowable node voltage range and line transmission capacity after disconnection, the system further compares the corresponding changes in active power loss, changes in node voltage deviation, and switch operation costs, and prioritizes the switch with the smaller comprehensive evaluation value as the disconnecting switch.

[0057] If there is no switch in the set of branches forming a closed loop that can meet the allowable range of node voltage and the allowable transmission capacity of the line after being disconnected, the system marks the updated candidate switch combination as an unrepairable candidate switch combination and discards it in the current iteration.

[0058] Repairing overloaded node voltages and overloaded lines.

[0059] When the updated candidate switch combination has met the requirements of radial structure and islandless operation, but there is a node voltage exceeding the limit, the system determines the power supply path from the voltage exceeding the limit node to the power supply node, and adjusts the breakpoint code value in the basic loop related to the power supply path.

[0060] A basic loop associated with the power supply path of a voltage-over-limit node refers to a basic loop in which at least one branch is located on the power supply path from the voltage-over-limit node to the power supply node, or where at least one node in the basic loop is adjacent to a node on the power supply path. By limiting the scope of the relevant basic loops, breakpoint adjustments can be concentrated in local topology regions that have a direct impact on the voltage-over-limit node.

[0061] When an updated candidate switch combination satisfies a radial topology but has a line overload, the system determines the power supply path of the overloaded line and adjusts the breakpoint coding values ​​in the basic loops related to the upstream and downstream nodes of the overloaded line. The upstream and downstream nodes of the overloaded line are determined based on the power supply direction from the power source node to the load node in the current radial topology. The node closer to the power source node is the upstream node, and the node farther from the power source node is the downstream node. If the target distribution network has multiple main power supply nodes, the upstream and downstream nodes are determined separately within the power supply area corresponding to the current candidate switch combination.

[0062] A basic loop related to the power supply path of an overloaded line refers to a basic loop that includes the overloaded line, its upstream node, its downstream node, or an adjacent connecting branch that can change the power flow distribution of the overloaded line. By limiting the scope of the relevant basic loop, the breakpoint adjustment can be concentrated in a local area that can change the power flow distribution of the overloaded line.

[0063] When a candidate switch combination for repair exhibits both node voltage exceedance and line overload, the system first calculates the degree of each. If the line load rate exceeds the line's allowable transmission capacity, line overload repair is prioritized to avoid safety risks to branch operations. If the line load rate meets the line's allowable transmission capacity but node voltage exceedance exists, node voltage exceedance repair is performed. If both exist and exceed the allowable range, candidate breakpoints for voltage repair and overload repair are generated separately, and the candidate switch combination with the better repair effect is selected based on a comprehensive evaluation value.

[0064] Calculation of comprehensive evaluation value.

[0065] The system performs power flow calculations on the candidate switch combinations for repair, obtaining line active power loss, node voltage, line load rate, and number of switch operations, and calculates a comprehensive evaluation value based on these.

[0066] In one embodiment, the comprehensive evaluation value is calculated according to the following formula: Where J represents the comprehensive evaluation value of the candidate switch combination; This represents the normalized active power loss. This represents the normalized node voltage deviation; This represents the normalized number of switching actions; α represents the degree of deviation from operating constraints; β represents the weight of active power loss; γ represents the weight of node voltage deviation; δ represents the weight of the number of switching actions; and δ represents the weight of the degree of deviation from operating constraints.

[0067] This formula is used to unify active power loss, node voltage deviation, number of switching operations, and deviation from operating constraints into a comparable evaluation metric. The smaller J is, the better the distribution network reconfiguration effect corresponding to the candidate switch combination.

[0068] The evaluation quantities of active power loss, node voltage deviation, line operation constraint deviation, and switch action are normalized before weighting. The purpose of normalization is to eliminate dimensional differences between different evaluation quantities, enabling them to participate in the comprehensive evaluation on the same numerical scale. Normalization can be performed linearly using the maximum and minimum values ​​in the current candidate switch combination set, or proportionally using the pre-reconstruction baseline values. When using either normalization method, the normalization baseline for each candidate switch combination within the same iteration should be kept consistent.

[0069] In one specific embodiment, the following is set: Where α, β, γ, and δ are all non-negative numbers. Preferably, α is 0.40, β is 0.30, γ is 0.10, and δ is 0.20. This set of weights is used to ensure that the comprehensive evaluation value focuses on reducing network losses and improving voltage, while also incorporating the number of switching actions and the degree of deviation from operational constraints into the evaluation. The weights can also be determined based on distribution network operation targets, historical operation data, simulation experiments, or orthogonal experiments.

[0070] The degree of deviation from operating constraints and the amount of improvement in operating constraints.

[0071] The deviation from operating constraints is used to uniformly characterize the degree of deviation of the operating state of a candidate switch combination from the allowable range of node voltage and the allowable transmission capacity of the line. The smaller the deviation from operating constraints, the closer the operating state of the candidate switch combination is to the safe operation requirements.

[0072] In one embodiment, the deviation from the operational constraints is calculated according to the following formula: in, Indicates the degree of deviation from operational constraints; Indicates the degree to which the node voltage exceeds the limit; Indicates the degree of line overload; Indicates the weight of the degree to which node voltage exceeds the limit; This indicates the weight of the line overload level.

[0073] Node voltage over-limit Calculate using the following formula: Where N represents the number of nodes in the target distribution network; i represents the node number; This represents the voltage value of the i-th node; This represents the lower voltage limit of the i-th node; This represents the upper voltage limit of the i-th node; This represents the rated voltage of the i-th node.

[0074] This formula is used to calculate the total deviation of each node voltage when it is below the lower voltage limit and above the upper voltage limit. When the node voltage is between the lower and upper voltage limits, the node... Its contribution is 0.

[0075] Line overload Calculate using the following formula: Where M represents the number of lines in the target distribution network; l represents the line number; This represents the load rate of the l-th line.

[0076] This formula is used to calculate the degree of deviation when the line load rate exceeds the allowable value. When the line load rate does not exceed 1, the line... Its contribution is 0.

[0077] After completing the topology feasibility repair, the system calculates the improvement in operational constraints. The improvement in operational constraints is calculated according to the following formula: Where R represents the improvement amount of the operating constraints; This indicates the degree of deviation from the operating constraints corresponding to the current switch combination; This indicates the degree of deviation from the operating constraints corresponding to the candidate switch combination being repaired. When R is greater than 0, it means that the candidate switch combination being repaired has reduced the degree of node voltage over-limit and line overload; when R is not greater than 0, it means that the candidate switch combination being repaired has not improved the degree of deviation from the operating constraints.

[0078] The purpose of setting the operational constraint improvement amount is to limit the range of suboptimal solutions accepted by the simulated annealing algorithm to candidate switch combinations that improve operational safety. When the operational constraint improvement amount is greater than zero, it indicates that although repairing the candidate switch combination may worsen the overall evaluation value, it reduces the degree of node voltage exceedance or line overload, and has engineering significance for continued retention. When the operational constraint improvement amount is not greater than zero, it indicates that repairing the candidate switch combination neither improves the overall evaluation value nor the degree of operational constraint deviation, and therefore it is not included in the simulated annealing acceptance judgment and is directly discarded.

[0079] Simulated annealing acceptance criteria controlled by operational constraints and improved quantities.

[0080] In this embodiment, the current switch combination refers to the baseline switch combination compared with the candidate switch combinations for repair during the current iteration. For a single nest update process, the current switch combination is the switch combination corresponding to the nest before the update; for a globally optimal record update process, the current switch combination is the switch combination with the best comprehensive evaluation value in the current candidate switch combination set. Simulated annealing acceptance judgment is preferentially used for candidate solution replacement of a single nest, and the globally optimal switch combination is only updated after the candidate solution is accepted and enters the current candidate switch combination set.

[0081] When the comprehensive evaluation value of the candidate switch combination to be repaired Less than the comprehensive evaluation value of the current switch combination At that time, the system directly accepts candidate switch combinations for repair. Not less than The system first determines the improvement amount R of the operating constraints. Only when R is greater than 0 does the system calculate the acceptance probability Pr. The acceptance probability Pr is calculated according to the following formula: , Where Pr represents the acceptance probability of the candidate switch combination for repair; T represents the current annealing temperature; This represents the improvement in operating constraints after normalization.

[0082] The normalized improvement in operating constraints is calculated using the following formula: Here, ε represents a very small positive number to prevent the denominator from being zero. In one specific embodiment, ε is 10. -6 This value is only used to ensure computational stability and does not change the direction of the judgment of the improvement amount of the running constraints.

[0083] The above acceptance probability formula means that, with the same magnitude of variation in the comprehensive evaluation value, the greater the improvement in operating constraints, the higher the probability of the candidate switch combination being accepted; with the same magnitude of improvement in operating constraints, the greater the magnitude of variation in the comprehensive evaluation value, the lower the probability of the candidate switch combination being accepted.

[0084] The system generates random numbers uniformly distributed between 0 and 1. If the random number is less than the acceptance probability Pr, the candidate switch combination for repair is accepted; if the random number is not less than the acceptance probability Pr, the candidate switch combination for repair is discarded. This random acceptance process is used to retain some candidate switch combinations that can improve the running constraints but whose overall evaluation value has not yet improved, thereby reducing the risk of the algorithm getting trapped in a local search region too early.

[0085] The current annealing temperature is updated according to a preset cooling coefficient after each candidate switch combination acceptance judgment. The preset cooling coefficient is a value between 0 and 1. The current annealing temperature is multiplied by the preset cooling coefficient to obtain the annealing temperature used for the next acceptance judgment. By gradually reducing the annealing temperature, the algorithm maintains a certain ability to escape local search regions in the early stages and gradually enhances its convergence ability for candidate switch combinations with better comprehensive evaluation values ​​in the later stages.

[0086] Iteration stops and reconstruction results are output.

[0087] The system continuously executes the cuckoo search update, topology feasibility repair, power flow calculation, runtime constraint improvement judgment, and simulated annealing acceptance process until the preset iteration stop condition is reached.

[0088] The preset iteration stopping conditions include: the current iteration count reaches the maximum iteration count; the current annealing temperature is lower than the termination temperature; and the comprehensive evaluation value has not improved for a preset number of consecutive cycles. The system stops iterating when any of the above conditions are met.

[0089] After stopping the iteration, the system selects the candidate switch combination with the best comprehensive evaluation value from the current candidate switch combination set, which also satisfies the following requirements: node connectivity, radial structure, no islanding operation, allowable node voltage range, allowable line transmission capacity, and switch action restrictions. This selected switch combination is then used as the target switch combination. Based on the target switch combination, the system outputs the reconstructed distribution network topology, the set of open switches, the set of closed switches, the reconstructed active power loss, the reconstructed node voltage, the reconstructed line load rate, and the number of switch actions.

[0090] To verify the effectiveness of the method of this invention, an IEEE 33-node distribution network with distributed generation was selected as the simulation test system, and MATLAB was used for simulation calculations. The test system includes 33 nodes, 32 normally closed branches, and 5 interconnecting branches. The initial load is 3715kW + j2300kvar, the reference voltage is 12.66kV, and the power reference is 10MV·A.

[0091] To compare the reconstruction effects of different methods, distribution network reconstruction calculations were performed using the traditional CS algorithm, the ordinary SA-CS algorithm, and the method of this invention. The traditional CS algorithm uses ordinary switch state encoding and directly deletes infeasible candidate solutions; the ordinary SA-CS algorithm introduces an ordinary simulated annealing acceptance criterion after CS update; the method of this invention, based on the ordinary SA-CS algorithm, further employs basic loop breakpoint encoding, topology feasible repair, and a simulated annealing acceptance criterion controlled by operational constraint improvement.

[0092] During the simulation, the population size was set to 30, the maximum number of iterations to 100, the search dimension to 5, and the probability of finding a foreign egg using the cuckoo search algorithm to 0.25. The initial temperature of the simulated annealing algorithm was set to 1, the cooling coefficient to 0.95, and the termination tolerance to 1.0 × 10⁻⁶. - ³. The allowable range for node voltage is set to 0.95 pu to 1.05 pu, and the allowable line load rate is set to not exceed 1.

[0093] Option 1 involves three distributed power supply access scenarios. The distributed power supply access nodes and parameters are shown in Table 1.

[0094] Table 1. Distributed Power Supply Connection Nodes and Parameters The reconstruction results of different algorithms are shown in Table 2.

[0095] Table 2 shows the reconstruction results of the traditional CS algorithm, the ordinary SA-CS algorithm, and the method of this invention. The results in the table show that, in Scheme 1, the ordinary SA-CS algorithm can reduce active power loss and increase the minimum node voltage compared to the CS algorithm. Compared with the ordinary SA-CS algorithm, the active power loss of the method of this invention is reduced from 32.29 kW to 31.92 kW, a reduction of approximately 1.15%; the minimum node voltage is increased from 0.9840 pu to 0.9851 pu. The switch combination of the method of this invention is consistent with that of the ordinary SA-CS algorithm, but due to the use of basic loop breakpoint coding and topology feasibility repair, the proportion of infeasible candidate solutions is reduced from 21.3% to 7.6%, and the average convergence generation is reduced from 41 generations to 34 generations.

[0096] The results show that, under the same optimal switch combination, the method of the present invention can still improve convergence efficiency by reducing infeasible candidate solutions and increasing the effective search ratio. Option 2 involves five distributed power supply access scenarios. The distributed power supply access nodes and parameters are shown in Table 3.

[0097] Table 3. Distributed Power Supply Connection Nodes and Parameters The reconstruction results of different algorithms are shown in Table 4.

[0098] Table 4 shows the reconstruction results of the traditional CS algorithm, the ordinary SA-CS algorithm, and the method of this invention. As shown in the table above, in Scheme 2, the active power loss of the ordinary SA-CS algorithm is 29.24 kW, and the lowest node voltage is 0.9895 pu; the active power loss of the method of this invention is reduced to 28.86 kW, a reduction of approximately 1.30%, and the lowest node voltage is increased to 0.9902 pu. The proportion of infeasible candidate solutions of the method of this invention is reduced from 22.4% of the ordinary SA-CS algorithm to 8.3%, and the average number of convergence generations is reduced from 43 generations to 35 generations.

[0099] The results demonstrate that the method of this invention can maintain relatively stable search performance even when there are a large number of distributed power sources and more complex power flow distributions. This is because the Cuckoo Search does not directly delete infeasible candidate switch combinations after updating; instead, it first performs topology feasibility repair based on islanding, closed-loop, voltage over-limit, and line overload conditions. Simultaneously, candidate solutions whose overall evaluation value has not improved only enter simulated annealing for evaluation when the improvement in operational constraints is greater than zero, thereby reducing suboptimal solutions that do not improve operational constraints.

[0100] To highlight the difference between this invention and the ordinary SA-CS algorithm, the "proportion of infeasible candidate solutions" and the "proportion of valid acceptance of suboptimal solutions" can be further statistically analyzed, as shown in Table 5. The proportion of infeasible candidate solutions refers to the percentage of candidate switch combinations that fail to meet the requirements of node connectivity, radial structure, island-free operation, node voltage allowable range, and line allowable transmission capacity during the iteration process. The proportion of valid acceptance of suboptimal solutions refers to the percentage of suboptimal candidate switch combinations that are accepted by simulated annealing and whose operational constraint improvement is greater than zero, out of the total number of accepted suboptimal candidate switch combinations.

[0101] Table 5. Differences between this invention and the ordinary SA-CS algorithm Therefore, in the IEEE-33 node distribution network test system with distributed generation, the method of this invention, compared with the ordinary SA-CS algorithm, can further reduce active power loss, increase the minimum node voltage, and reduce the proportion of infeasible candidate solutions in both distributed generation access scenarios. Especially in Scheme 2, when the number of distributed generation access nodes increases from 3 to 5, the ordinary candidate solution update is more likely to generate switch combinations that do not meet the operating constraints, while the method of this invention, through basic loop breakpoint encoding and topology feasibility repair, controls the proportion of infeasible candidate solutions to a low level.

[0102] Therefore, the method of the present invention can not only maintain the effect of SA-CS algorithm in reducing network losses and improving node voltage, but also further improve the topology executability of candidate switch combinations and the effectiveness of algorithm iteration, so that the output reconstructed topology is more suitable for the engineering operation requirements of distribution networks with distributed power sources.

[0103] In summary, this invention is not merely a simple combination of the simulated annealing algorithm and the Cuckoo Search algorithm. Instead, it establishes a candidate solution generation, repair, and acceptance mechanism oriented towards basic loop breakpoints, specifically addressing the characteristics of distribution network reconfiguration where candidate switch combinations must satisfy radial topology, island-free operation, allowable node voltage range, and allowable line transmission capacity. The Cuckoo Search algorithm is used to rapidly generate candidate switch combinations in the basic loop breakpoint coding space; topology feasibility repair is used to adjust candidate switch combinations that form closed loops, islands, exceed node voltage limits, or experience line overloads to the executable reconfiguration space; the operational constraint improvement amount is used to determine whether candidate switch combinations whose comprehensive evaluation value has not improved have operational safety improvement value; the simulated annealing acceptance criterion only performs acceptance judgment on candidate switch combinations whose operational constraint improvement amount is greater than zero. Through the above processing, the ordinary SA-CS hybrid algorithm can avoid accepting non-optimal candidate solutions that do not improve the operational constraints of the distribution network and improve the engineering executability of the output reconfigured topology.

[0104] The above are merely preferred embodiments of the present invention. The scope of protection of the present invention is not limited to the above embodiments. All technical solutions falling within the scope of the present invention's concept are within the scope of protection of the present invention. It should be noted that for those skilled in the art, any improvements and modifications made without departing from the principles of the present invention should be considered within the scope of protection of the present invention.

Claims

1. A power distribution network reconstruction method based on a SA and CS hybrid algorithm, characterized in that, Includes the following steps: S1. Obtain the basic data for reconfiguration of the target distribution network, operating status data, operating constraint data, and switch status data; S2. Identify the basic loops formed in the target distribution network after the tie switch is closed based on the reconstructed basic data, and establish basic loop breakpoint codes; S3. Generate an initial candidate switch combination set based on the basic loop breakpoint encoding, and perform topological feasibility verification on the candidate switch combinations in the initial candidate switch combination set based on the reconstructed basic data and switch state data to obtain a feasible candidate switch combination set. S4. The candidate switch combinations in the feasible candidate switch combination set are used as the nest positions of the cuckoo search algorithm, and the basic loop breakpoint codes corresponding to the candidate switch combinations are updated by the cuckoo search algorithm to obtain the updated candidate switch combinations. S5. Based on the reconstructed basic data, operating status data and operating constraint data, perform topology feasibility repair on the updated candidate switch combination, adjust the basic loop breakpoint coding, and obtain the repaired candidate switch combination. S6. Perform power flow calculation on the candidate switch combination to obtain the comprehensive evaluation value and operational constraint improvement amount corresponding to the candidate switch combination; S7. When the comprehensive evaluation value of the candidate switch combination to be repaired is better than the comprehensive evaluation value of the current switch combination, the candidate switch combination to be repaired is accepted, and the current candidate switch combination set is updated with the candidate switch combination to be repaired. S8. When the comprehensive evaluation value of the candidate switch combination to be repaired is not better than the comprehensive evaluation value of the current switch combination, and the improvement amount of the operating constraints is greater than zero, the candidate switch combination to be repaired is judged according to the acceptance criteria of the simulated annealing algorithm based on the current annealing temperature, the difference in comprehensive evaluation values ​​and the improvement amount of operating constraints. When the judgment result is "accept", the current candidate switch combination set is updated with the repair candidate switch combination; Otherwise, discard the proposed repair candidate switch combinations; S9. When the comprehensive evaluation value of the candidate switch combination to be repaired is not better than the comprehensive evaluation value of the current switch combination, and the improvement amount of the operating constraints is not greater than zero, the candidate switch combination to be repaired is discarded. S10. When the preset iteration stop condition is reached, select a candidate switch combination from the current candidate switch combination set that satisfies the following conditions: node connectivity, radial structure, no islanding operation, node voltage allowable range, line allowable transmission capacity, and switch action limit, as the target switch combination, and output the reconstructed distribution network topology based on the target switch combination.

2. The power distribution network reconfiguration method of claim 1, wherein, In step S1, the reconstructed basic data is used to characterize the connection relationship between nodes, lines, sectionalizing switches and tie switches in the target distribution network, including node number, line start node, line end node, line impedance, sectionalizing switch position and tie switch position; The operational status data is used to characterize the load access status and distributed generation access status of the target distribution network during the reconfiguration calculation period, including load active power, load reactive power, distributed generation access nodes, and distributed generation output data. The operational constraint data is used to characterize the allowable range of node voltage, allowable transmission capacity of lines, and switch action limits of the target distribution network, including the upper limit of node voltage, the lower limit of node voltage, the line capacity limit, and the limit of the number of switch actions; The switch status data is used to characterize the open / closed state of the sectionalizing switch and the tie switch before reconstruction, including the open / closed state of the sectionalizing switch and the open / closed state of the tie switch before reconstruction.

3. The power distribution network reconfiguration method of claim 1, wherein, In step S2, the basic loop breakpoint code is used to represent the disconnected switches in each basic loop, and the establishment process includes: Close the tie switches in the target distribution network and identify the basic loops formed by the closure of the tie switches; The segment switches and tie switches that can serve as breakpoints in each basic loop are combined into a candidate breakpoint set for the corresponding basic loop. The switch number in the candidate set of breakpoints for each basic loop is used as the breakpoint code value for that basic loop; Arrange the breakpoint codes according to the numbering order of the basic loop to form the basic loop breakpoint codes corresponding to the candidate switch combinations.

4. The power distribution network reconfiguration method of claim 1, wherein, Step S3, the topology feasibility verification of the candidate switch combinations includes: Determine the set of closed circuits and the set of open circuits based on the candidate switch combinations; Based on the set of closed circuits, determine whether all load nodes in the target distribution network have a power supply path to the power source nodes; Determine whether the candidate switch combination satisfies the radial structure based on the number of closed circuits, the number of nodes, and the connection relationship of power supply nodes; Determine whether the candidate switch combination forms an island node based on the power supply path of each load node; Candidate switch combinations that simultaneously satisfy the conditions of having a power supply path to the load node, having a radial structure, and not having formed an island node are written into the set of feasible candidate switch combinations.

5. The power distribution network reconfiguration method of claim 1, wherein, In step S4, updating the basic loop breakpoint code corresponding to the candidate switch combination using the cuckoo search algorithm includes: The number of basic loops that need to be adjusted in this update is determined based on the current iteration number and the step size factor. Select the breakpoint code bits that need to be adjusted from the basic loop breakpoint codes corresponding to the candidate switch combinations; Select a new switch number from the breakpoint candidate set corresponding to the selected breakpoint code bit; Write the newly selected switch number into the corresponding breakpoint code bit to form the updated basic loop breakpoint code; An updated candidate switch combination is generated based on the updated basic loop breakpoint code.

6. The power distribution network reconfiguration method of claim 1, wherein, In step S5, the topology feasibility repair adjusts the basic loop breakpoint coding based on the closed-loop state, islanding state, node voltage over-limit state, and line overload state of the updated candidate switch combinations. Adjusting the basic loop breakpoint coding based on the closed-loop state and islanding state includes: When updating the candidate switch combination to form a closed loop, identify the set of branches that form the closed loop, and select the switch that meets the node voltage allowable range and the line allowable transmission capacity after being disconnected from the set of branches as the disconnecting switch; When updating candidate switch combinations to form an island, identify the island node set, close the switch that can reconnect the island node set with the power node, and select a switch as the disconnect switch in the basic loop formed after closing the switch. The basic loop breakpoint code is updated based on the disconnected switch, so that the updated candidate switch combination is restored to a radial topology and the islanded nodes are eliminated.

7. The power distribution network reconfiguration method of claim 6, wherein, In step S5, adjusting the basic loop breakpoint coding based on the node voltage over-limit status and line overload status includes: When the updated candidate switch combination satisfies the radial structure and there is a node voltage over-limit, the power supply path from the voltage over-limit node to the power supply node is determined, and the breakpoint code value is adjusted in the basic loop related to the power supply path. When the updated candidate switch combination satisfies the radial structure and there is a line overload, the power supply path of the overloaded line is determined, and the breakpoint coding value is adjusted in the basic loop related to the upstream and downstream nodes of the overloaded line. Power flow calculations were performed on the adjusted candidate switch combinations, and the candidate switch combinations that met the allowable range of node voltage and the allowable transmission capacity of the line were selected as the repair candidate switch combinations.

8. The power distribution network reconfiguration method of claim 1, wherein, In step S6, the comprehensive evaluation value is determined based on the active power loss evaluation quantity, the node voltage deviation evaluation quantity, the line operation constraint deviation evaluation quantity, and the switch action evaluation quantity. The active power loss evaluation quantity is determined based on the active power loss of the line obtained from the power flow calculation. The node voltage deviation evaluation quantity is determined based on the deviation between each node voltage and the corresponding node voltage allowable range; The deviation evaluation value of the line operation constraint is determined based on the degree of deviation between the load rate of each line and the allowable transmission capacity of the corresponding line. The switching action evaluation quantity is determined based on the amount of change of the candidate switch combination relative to the initial state of the switch.

9. The power distribution network reconfiguration method of claim 1, wherein, In step S6, the improvement amount of the operating constraints is used to characterize the degree of improvement of the node voltage over-limit and line overload degree of the repair candidate switch combination relative to the current switch combination. It is obtained based on the difference between the degree of deviation of the operating constraints corresponding to the current switch combination and the degree of deviation of the operating constraints corresponding to the repair candidate switch combination. The degree of deviation from the operating constraints is determined based on the degree of node voltage exceeding limits and the degree of line overload. When the deviation of the operating constraints corresponding to the current switch combination is greater than the deviation of the operating constraints corresponding to the candidate switch combination to be repaired, the improvement amount of the operating constraints is greater than zero; When the deviation of the operating constraints corresponding to the current switch combination is not greater than the deviation of the operating constraints corresponding to the candidate switch combination to be repaired, the improvement amount of the operating constraints is not greater than zero.

10. The power distribution network reconfiguration method of claim 1, wherein, In step S8, determining whether to accept the candidate switch combination for repair according to the acceptance criteria of the simulated annealing algorithm includes: The acceptance probability of the candidate switch combination is calculated based on the comprehensive evaluation value of the candidate switch combination, the comprehensive evaluation value of the current switch combination, the current annealing temperature, and the improvement amount of operating constraints. Generate a random number and compare the random number with the acceptance probability; When the random number is less than the acceptance probability, the candidate switch combination for repair is accepted. When the random number is not less than the acceptance probability, the candidate switch combination for repair is discarded; The current annealing temperature is updated according to the preset cooling coefficient, and then the next iteration begins.

Images