A multi-period power distribution network robust reconfiguration method, system, device and storage medium

Abstract

The application discloses a kind of multi-period power distribution network robust reconfiguration method, system, equipment and storage medium, it is related to power distribution network operation optimization technical field, including obtaining the operation state of power distribution network containing renewable distributed power supply, form multi-period reconfiguration object;Based on the multi-period reconfiguration object, obtain the reconfiguration fitness representation containing robust reconfiguration target and operation safety constraint;With switch reconfiguration state as reconfiguration variable, the reconfiguration fitness representation is optimized to the reconfiguration optimization algorithm, obtain the robust reconfiguration solution matched with different operation period;According to the robust reconfiguration solution, output multi-period power distribution network robust reconfiguration scheme.The method disclosed in the application achieves better results in wind and light output uncertainty adaptation, multi-period switch reconfiguration optimization, power supply topology recovery after failure and voltage risk control.

Description

Technical Field

[0001] This invention relates to the field of distribution network operation optimization technology, specifically to a robust reconfiguration method, system, device, and storage medium for multi-time period distribution networks. Background Technology

[0002] Existing distribution network reconfiguration technologies still struggle to adequately adapt to the coupled changes in output randomness, time-of-day variations, and fault recovery requirements in scenarios involving renewable distributed power sources such as wind and solar power. Many reconfiguration schemes rely on single typical operating conditions or deterministic output data to determine switch combinations, easily overlooking the differences in power supply capacity, power flow distribution, and voltage levels between solar power output periods and non-solar power output periods. This results in a lack of stable adaptation between reconfiguration results across different operating periods. While existing probabilistic scenario modeling methods can describe wind and solar power output fluctuations, increasing the number of scenarios expands the reconfiguration computation scale, and insufficient scenario compression weakens the constraining effect of representative output scenarios on reconfiguration decisions, making it difficult to effectively transform probabilistic output information into executable time-of-day network topologies. Existing multi-objective reconfiguration methods often focus on network losses or the number of switch actions, failing to unify the relationships between renewable energy curtailment, operational safety boundaries, and power recovery after fault isolation into a single reconfiguration evaluation process. This can easily lead to situations where candidate switch states meet local economic objectives but not voltage, capacity, or topology safety requirements. If the islanding in fault conditions lacks a matching relationship between load level, distributed generation capacity, and branch connectivity, it can easily lead to unclear recovery paths for critical loads. Traditional genetic algorithms also suffer from premature convergence and insufficient local search when processing discrete switch combinations, resulting in limited search direction and convergence quality for multi-time-period reconstructed solutions. Furthermore, if only conventional power flow verification is performed after the reconstruction results are output, it is difficult to fully reflect the voltage over-limit risk state under random output conditions. Summary of the Invention

[0003] In view of the above-mentioned problems, the present invention is proposed.

[0004] Therefore, the technical problem solved by this invention is that existing power distribution network reconfiguration methods have the following problems in scenarios with access to renewable distributed power sources such as wind power and photovoltaics: insufficient adaptation to uncertain output and multi-time period operation differences of distributed power sources, low matching degree between candidate switch reconfiguration state and operation safety constraints, unclear network power supply topology restoration relationship after fault isolation, and how to obtain robust reconfiguration solutions that match different operating periods in complex switch combination space.

[0005] To address the aforementioned technical problems, this invention provides the following technical solution: a robust reconfiguration method for a multi-time distribution network, comprising, as a preferred embodiment of the robust reconfiguration method for a multi-time distribution network described in this invention: acquiring the operating state of a distribution network containing renewable distributed power sources to form a multi-time reconfiguration object; acquiring a reconfiguration fitness representation containing robust reconfiguration objectives and operational safety constraints based on the multi-time reconfiguration object; optimizing the reconfiguration fitness representation using a reconfiguration optimization algorithm with switch reconfiguration state as the reconfiguration variable to obtain robust reconfiguration solutions matching different operating periods; and outputting a robust reconfiguration scheme for the multi-time distribution network based on the robust reconfiguration solution; the robust reconfiguration scheme for the multi-time distribution network includes the switch reconfiguration state and network power supply topology state corresponding to the operating period.

[0006] As a preferred embodiment of the robust reconfiguration method for multi-time period distribution networks described in this invention, the formation of the multi-time period reconfiguration object includes: forming reconfiguration object units corresponding to the operating period according to the output status of renewable distributed power sources, the fault status of the distribution network, and the reconfiguration status of switches; combining the reconfiguration object units of multiple operating periods according to their temporal relationships to form the multi-time period reconfiguration object; the multi-time period reconfiguration object is used to limit the matching relationship between the output changes of renewable distributed power sources, the isolation relationship of fault branches, and the reconfiguration status of switches under different operating periods.

[0007] As a preferred embodiment of the robust reconfiguration method for multi-time distribution networks described in this invention, the method further includes: generating a set of probabilistic output scenarios for distributed generation sources based on Latin hypercube sampling; reducing the set of probabilistic output scenarios for distributed generation sources based on synchronous back-substitution scenario reduction to obtain representative output scenarios; dividing the corresponding operating period into photovoltaic output periods and photovoltaic non-output periods according to the photovoltaic output status in the representative output scenarios; and writing the representative output scenarios, photovoltaic output periods, and photovoltaic non-output periods into the multi-time reconfiguration object to form a multi-time reconfiguration object with time period labels.

[0008] As a preferred embodiment of the robust reconfiguration method for multi-time distribution networks described in this invention, the step of obtaining a reconfiguration fitness representation containing robust reconfiguration objectives and operational safety constraints based on the multi-time reconfiguration object includes: for the switch reconfiguration state under the corresponding operating period, obtaining the network loss state, switch action state, and distributed generation curtailment state corresponding to the current switch reconfiguration state, and generating a reconfiguration operation cost; the distributed generation curtailment state includes wind curtailment state and solar curtailment state; obtaining the operational safety boundary state corresponding to the current switch reconfiguration state, and generating a reconfiguration safety state; when the reconfiguration safety state meets the operational safety constraints, determining the reconfiguration operation cost as the fitness evaluation value of the corresponding switch reconfiguration state; when the reconfiguration safety state does not meet the operational safety constraints, marking the corresponding switch reconfiguration state as an infeasible reconfiguration state, updating the reconfiguration fitness representation based on the infeasible reconfiguration state; and generating a penalized fitness evaluation value based on the infeasible reconfiguration state; the reconfiguration fitness representation is formed by the fitness evaluation value and the penalized fitness evaluation value.

[0009] As a preferred embodiment of the robust reconfiguration method for multi-time distribution networks described in this invention, the following is included: the operational safety boundary state includes a network topology boundary; the network topology boundary is configured with islanded power supply constraints; the formation of the islanded power supply constraints includes: determining the set of load nodes to be restored after fault isolation based on the distribution network fault state, and generating the load level state of the set of load nodes to be restored according to the load importance; matching the set of load nodes to be restored with the output state of renewable distributed power sources under the corresponding operating period to generate islanded power supply candidate relationships; when the islanded power supply candidate relationship satisfies the distributed power source output constraints, load level priority constraints, and branch connectivity constraints, the islanded power supply candidate relationship is determined as an islanded power supply relationship; the islanded power supply relationship is written into the network power supply topology state, so that the network power supply topology state characterizes the power supply restoration results of load nodes of different levels after fault branch isolation.

[0010] As a preferred embodiment of the robust reconfiguration method for multi-time distribution networks described in this invention, the step of optimizing the reconfiguration fitness representation using a reconfiguration optimization algorithm includes: encoding the switch reconfiguration state as candidate reconfiguration solutions and forming a reconfiguration population; obtaining the fitness value of each candidate reconfiguration solution based on the reconfiguration fitness representation; performing a genetic iteration operation on the reconfiguration population to form an updated reconfiguration population; correcting the search direction of the candidate reconfiguration solutions in the updated reconfiguration population using a quasi-Newton update matrix; and determining the candidate reconfiguration solution that satisfies the convergence condition and has the optimal fitness value as the robust reconfiguration solution.

[0011] As a preferred embodiment of the robust reconfiguration method for multi-time period distribution networks described in this invention, the method further includes: generating voltage over-limit risk states for corresponding operating periods based on the robust reconfiguration solution; determining the robust reconfiguration solution for the corresponding operating period as a valid reconfiguration solution when the voltage over-limit risk state meets preset risk constraints; binding the valid reconfiguration solutions corresponding to photovoltaic power output periods and photovoltaic non-power output periods according to operating periods to generate the robust reconfiguration scheme for multi-time period distribution networks; the robust reconfiguration scheme for multi-time period distribution networks is used to characterize the time-segmented network reconfiguration results that match the switch reconfiguration state, network power supply topology state, and voltage over-limit risk state under different operating periods.

[0012] Another objective of this invention is to provide a robust reconfiguration system for multi-time distribution networks. This system can form a robust reconfiguration scheme for multi-time distribution networks that matches different operating periods through data collaboration between the object generation module, the fitness representation module, and the reconfiguration output module. This solves the problems in current distribution network reconfiguration technologies, such as the difficulty in matching the uncertain output of distributed power sources with the reconfiguration state of switches, insufficient expression of multi-time operation differences, low correlation between reconfiguration solutions and operational safety constraints, and unclear output of network power supply topology state.

[0013] As a preferred embodiment of the multi-time period distribution network robust reconfiguration system described in this invention, it includes: an object generation module, a fitness characterization module, and a reconfiguration output module; the object generation module is used to obtain the operating status of the distribution network containing renewable distributed power sources, associate the output status of renewable distributed power sources, the fault status of the distribution network, and the reconfiguration status of switches according to the operating time period, form representative output scenarios based on Latin hypercube sampling and synchronous back-substitution scenario reduction, and write the photovoltaic output time period and the photovoltaic non-output time period into the multi-time period reconfiguration object; the fitness characterization module is used to receive the multi-time period reconfiguration object, and based on the network loss status and switch action status... The reconfiguration operation cost is generated based on the state of the distributed power generation abandonment state and the state of the operation safety boundary state. The reconfiguration safety state is generated based on the state of the operation safety boundary state. The island power supply relationship formed under the island power supply constraint is written into the network power supply topology state to form the reconfiguration fitness representation. The reconfiguration output module is used to optimize the reconfiguration fitness representation with the switch reconfiguration state as the reconfiguration variable through the reconfiguration optimization algorithm to obtain a robust reconfiguration solution that matches different operating periods. Based on the robust reconfiguration solution, the voltage over-limit risk state is generated. The effective reconfiguration solutions corresponding to the photovoltaic output period and the photovoltaic non-output period are bound according to the operating period to output a multi-period distribution network robust reconfiguration scheme.

[0014] Another object of the present invention is to provide a robust reconfiguration device for a multi-time distribution network, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of a robust reconfiguration method for a multi-time distribution network.

[0015] Another object of the present invention is to provide a multi-time distribution network robust reconfiguration storage medium storing a computer program thereon, which, when executed by a processor, implements the steps of a multi-time distribution network robust reconfiguration method.

[0016] The beneficial effects of this invention are as follows: The multi-period robust reconfiguration method for distribution networks provided by this invention forms a multi-period reconfiguration object by associating the output status of renewable distributed power sources, the fault status of the distribution network, and the reconfiguration status of switches according to the operating period. This allows the random changes in wind and solar power output, fault isolation relationships, and switch reconfiguration results to be co-expressed in the same reconfiguration object. By constructing a reconfiguration fitness representation that includes robust reconfiguration objectives and operational safety constraints, network losses, switch actions, wind and solar curtailment, operational safety boundaries, and islanded power supply relationships can all participate in the evaluation of switch reconfiguration status. By optimizing the reconfiguration fitness representation through a reconfiguration optimization algorithm and binding effective reconfiguration solutions under different operating periods with voltage over-limit risk states, the output multi-period robust reconfiguration scheme for distribution networks can simultaneously match the power supply topology requirements of both photovoltaic output periods and photovoltaic non-output periods. This invention achieves better results in terms of distributed power source uncertain output adaptation, complex switch combination optimization, post-fault network power supply topology recovery, and time-period voltage risk constraints. Attached Figure Description

[0017] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 The present invention provides an overall flowchart of a robust reconfiguration method for a multi-time period distribution network.

[0019] Figure 2 The IEEE-33 node system diagram is provided for a robust reconfiguration method for multi-time distribution networks according to the present invention.

[0020] Figure 3 The node voltage probability density curve is provided by the present invention for a robust reconfiguration method for multi-time distribution networks.

[0021] Figure 4 The present invention provides a robust reconfiguration method for multi-time distribution networks using Latin hypercube sampling of 10 random scenarios.

[0022] Figure 5 The diagram shows the scenario with the highest probability after scenario reduction for the robust reconfiguration method for multi-time distribution networks provided by this invention.

[0023] Figure 6 The IEEE-33 node system diagram for simultaneous faults in branches 8-9 and 2-22, which is provided by the present invention for a robust reconfiguration method for multi-time distribution networks.

[0024] Figure 7 The diagram shows the node voltage distribution before and after the reconstruction of a photovoltaic-powered distribution network, which is part of a robust multi-time distribution network reconstruction method provided by this invention.

[0025] Figure 8 The diagram shows the node voltage distribution before and after the reconfiguration of a distribution network without photovoltaic output, which is provided by the present invention as a robust reconfiguration method for multi-time distribution networks.

[0026] Figure 9 The diagram shows the IEEE-33 node system after the reconstruction of a photovoltaic power distribution network, which is part of a robust reconfiguration method for multi-time distribution networks provided by this invention.

[0027] Figure 10 The diagram shows the IEEE-33 node system after the reconfiguration of a distribution network without photovoltaic output, which is provided by the present invention as a robust reconfiguration method for multi-time distribution networks.

[0028] Figure 11 The voltage over-limit risk value diagram for a multi-time period distribution network robust reconfiguration method provided by the present invention is shown in the figure.

[0029] Figure 12 The voltage over-limit risk value diagram is a robustly optimized version of the robust reconfiguration method for multi-time distribution networks provided by this invention. Detailed Implementation

[0030] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the protection scope of the present invention.

[0031] Example 1, referring to Figures 1-3 This is the first embodiment of the present invention, which provides a robust reconfiguration method for a multi-time distribution network. This embodiment illustrates how the present invention transforms the operating state of a distribution network containing renewable distributed power sources into a multi-time reconfiguration object 100 that can participate in reconfiguration optimization during overall operation, and forms a reconfiguration fitness representation 200, a robust reconfiguration solution 400, and a final multi-time distribution network robust reconfiguration scheme 500 based on this object.

[0032] In this embodiment, the operating status of a distribution network containing renewable distributed power sources refers to the set of operating information that can affect the reconfiguration results of the distribution network before and after a fault occurs and under different operating conditions. This set of operating information includes the connection relationship of distribution network nodes, the connection relationship of branches, the initial opening and closing status of sectionalizing switches and tie switches, the location of the faulty branch, the node load, wind power output, photovoltaic power output, and distributed power source access nodes.

[0033] Taking the IEEE-33 node distribution network as an example, refer to Figure 2 The system can contain multiple sectionalizing switches and multiple tie switches (not shown in the figure, but can be set at the dashed line position of nodes 8-14). In the initial state, the sectionalizing switches are closed and the tie switches are open. When a short circuit fault occurs in some branches, the system needs to re-establish the power supply path by changing the opening and closing state of some switches.

[0034] Furthermore, in this embodiment, renewable distributed power sources mainly include wind power and photovoltaic power.

[0035] Reference Figure 3 The area of ​​the shaded region represents the probability of the node voltage exceeding the limit. The larger the shaded area, the greater the probability of the node exceeding the limit, which is more worthy of attention from operators.

[0036] Wind power source, represented as: , in, The probability density function representing wind power generation. Indicates the actual wind speed. Indicates shape parameters, This represents the scale parameter. The wind speed scenario is converted into a wind power output scenario using the wind speed-wind power conversion formula: , in, This indicates the output power of the wind turbine. This refers to the cut-in wind speed, which is the minimum wind speed required for the wind turbine to begin generating electricity from the grid. This indicates the cut-off wind speed, which is the highest wind speed at which the wind turbine will stop generating electricity to protect the turbine's safety. Indicates the rated wind speed. This indicates the rated power of the wind turbine.

[0037] Photovoltaic power sources are represented as: , in, Indicates solar irradiance The probability density function when it follows a beta distribution. The shape parameter represents the Beta distribution. The obtained solar irradiance is then converted into a photovoltaic power output scenario. , in, This refers to photovoltaic output, specifically the power output of photovoltaic power generation equipment under current solar irradiance conditions. This refers to the radiation area of ​​a photovoltaic module, that is, the effective area of ​​the photovoltaic module that receives solar radiation. It represents the photoelectric conversion efficiency, which is the efficiency with which a photovoltaic module converts solar radiation energy into electrical energy.

[0038] The output power of wind power is affected by the actual wind speed. When the wind speed is lower than the cut-in wind speed or higher than the cut-out wind speed, the wind turbine does not participate in effective power output. When the wind speed is between the cut-in wind speed and the rated wind speed, the wind power output changes with the wind speed. When the wind speed is between the rated wind speed and the cut-out wind speed, the wind turbine can output according to the rated power. The output power of photovoltaic power is affected by solar irradiance, photovoltaic module area and photoelectric conversion efficiency. It forms effective power output when the daytime irradiance meets the power generation conditions, and forms a photovoltaic non-output state at night or when the irradiance is insufficient.

[0039] In this embodiment, the multi-time-period reconfiguration object 100 is a reconfiguration input object formed by organizing the distribution network operating status according to operating segments. This object is not a power flow section at a single moment, nor a single fault scenario, but rather a set of objects formed by associating wind power output, photovoltaic output, fault branch isolation relationships, load status, and candidate switch reconfiguration status under different time periods. For example, within a daily operating cycle, the time period when photovoltaics have effective output can be divided into photovoltaic output periods, and the nighttime period when photovoltaics have no effective output can be divided into photovoltaic non-output periods, and different reconfiguration object units can be formed for each. Each reconfiguration object unit stores the wind and solar power output status, fault isolation status, set of switches participating in reconfiguration, and load nodes to be restored under that time period.

[0040] Furthermore, the reconfiguration fitness representation 200 is used to evaluate whether each set of candidate switch reconfiguration states is suitable as the reconfiguration result for the corresponding operating period. This representation is not simply a calculation of network losses, but simultaneously includes robust reconfiguration objectives and operational safety constraints. Robust reconfiguration objectives include network losses, the number of switch actions, and wind and solar power curtailment; operational safety constraints include node voltage upper and lower limits, line capacity upper limits, branch allowable current carrying capacity, and network topology boundaries. When evaluating a candidate switch reconfiguration state, the system first substitutes the switch state into the distribution network topology to form a candidate power supply network, and then performs power flow calculations and constraint judgments, thereby obtaining the corresponding fitness evaluation result.

[0041] In this embodiment, the switch reconfiguration state refers to the combination of opening and closing of sectionalizing switches and tie switches participating in the reconfiguration in the distribution network. Each set of switch reconfiguration states corresponds to a candidate network topology. The system uses the switch reconfiguration state as the reconfiguration variable and performs optimization through the reconfiguration optimization algorithm 300.

[0042] The reconstruction optimization algorithm 300 is a quasi-Newton genetic algorithm. The optimization of the reconstruction fitness representation 200 by the reconstruction optimization algorithm includes the following: the genetic algorithm is responsible for forming candidate reconstruction solutions from a large number of switch combinations and updating the population; the quasi-Newton algorithm is responsible for correcting the search direction of the candidate solutions during the iteration process, so that the candidate reconstruction solutions can combine global search and local adjustment.

[0043] After obtaining robust reconfiguration solution 400, the system outputs a multi-time period robust reconfiguration scheme 500 for the distribution network. This scheme includes the switch reconfiguration state and network power supply topology state for the corresponding operating period. The network power supply topology state refers to the comprehensive topology result of the main grid power supply path, fault branch isolation area 102, distributed generation power supply area, and islanded power supply relationship in the reconfigured distribution network. For example, during the photovoltaic output period, some load nodes can be locally islanded by photovoltaic distributed generation; during the photovoltaic non-output period, the same load node may need to connect to other power supply paths through tie switch closure. The system binds the switch states and power supply topologies under these different time periods to form a reconfiguration scheme that can be called according to the operating period.

[0044] Example 2, refer to Figures 4-5 This is the second embodiment of the present invention. Based on embodiment 1, this embodiment further explains the formation process of the multi-time period reconstruction object 100, and, in conjunction with the content of wind power model, photovoltaic model, Latin hypercube sampling and synchronous back-substitution scenario reduction, refines how the distributed power source probabilistic output scenario enters the multi-time period reconstruction object 100.

[0045] In this embodiment, the system first associates the output status of renewable distributed power sources, the fault status of the distribution network, and the reconfiguration status of switches according to the runtime segment. The output status of renewable distributed power sources includes wind power output status and photovoltaic power output status; the fault status of the distribution network includes the location of the faulty branch, the nodes at both ends of the faulty branch, the power outage area after fault isolation, and the load nodes that need to restore power; the reconfiguration status of switches includes the opening and closing status of sectionalizing switches and tie switches in the candidate reconfiguration schemes. The system binds the above status data within the same runtime segment to form reconfiguration object units corresponding to the runtime segment, and combines multiple reconfiguration object units according to the time sequence to form a multi-time-segment reconfiguration object 100.

[0046] Reference Figure 4The dotted line at the top represents wind power generation, and the solid line at the bottom represents photovoltaic power generation. The sampling points of the Latin hypercube sampling method are discretely distributed throughout the sampling space, so not too many are needed to roughly characterize the probability distribution of the random variable.

[0047] However, generating too many scenarios slows down the computation speed, severely impacting efficiency. Therefore, a synchronous back-substitution scenario reduction method is needed to calculate the probability of each scenario occurring, thereby selecting the scenario with the highest probability for distribution network reconfiguration. This method ensures that the computation speed does not decrease while maintaining accuracy. The scenario reduction results are referenced... Figure 5 In the diagram, the blue dashed line represents wind power generation, and the red solid line represents photovoltaic power generation. It can be seen that photovoltaic power generation did not generate any power between 0:00 and 6:00 and between 18:00 and 24:00.

[0048] Furthermore, the wind power output status can be determined through wind speed probability distribution and the wind speed-power conversion relationship. After acquiring the wind speed random variable, the system determines the probability of different wind speeds occurring based on the wind speed distribution characteristics, and then combines this with the wind turbine's cut-in wind speed, rated wind speed, and cut-out wind speed to determine the wind power output. When the actual wind speed is lower than the cut-in wind speed, the wind turbine does not meet the grid connection conditions; when the actual wind speed is within the normal power generation range, the wind power output varies with the wind speed; when the actual wind speed exceeds the cut-out wind speed, the wind turbine is taken out of operation. Through this method, wind speed uncertainty is converted into wind power output uncertainty.

[0049] Furthermore, the photovoltaic (PV) power output status can be determined through the probability distribution of solar irradiance and the PV power output conversion relationship. Solar irradiance is affected by time, weather, and the location of PV modules. The system generates PV power output scenarios based on the random variable of irradiance and determines the PV output power by combining the PV module's radiation area and photoelectric conversion efficiency. During the day when irradiance is high, the PV power source generates effective output; at night or when irradiance is insufficient, the PV output is close to zero. Thus, the system can distinguish between PV power output periods and non-PV power output periods in the multi-time period reconstruction object 100.

[0050] Preferably, the system employs Latin hypercube sampling to generate a set of distributed power generation probabilistic output scenarios. This sampling method first determines the number of random variables participating in the sampling and the number of samples required for each random variable. Then, it divides the probability distribution interval of each random variable into multiple sub-intervals and draws samples within each sub-interval. In this way, the sampling points can cover the main probability space of the random variables, avoiding the problem of samples being concentrated in local areas in simple random sampling. For example, the system can simultaneously perform stratified sampling on wind speed and solar irradiance random variables, combining the resulting wind power output samples and photovoltaic power output samples into a wind-solar combined output scenario.

[0051] Furthermore, since too many output scenarios would increase the computational load of subsequent distribution network reconfiguration, the system obtains representative output scenarios based on synchronous back-substitution scenario reduction. During synchronous back-substitution scenario reduction, the system first calculates the scenario distance between any two initial scenarios and forms a scenario distance matrix. Then, for each initial scenario, it finds the nearest other scenario and calculates the probability distance based on the occurrence probability of that scenario. Next, scenarios with weak representativeness and close proximity to other scenarios are deleted, and the probability of the deleted scenarios is transferred to the nearest retained scenario. This process is repeated until the number of remaining scenarios meets the preset requirements, resulting in representative output scenarios that reflect the main probability distribution.

[0052] In one specific implementation, after performing Latin hypercube sampling on wind and solar power output, the system obtains multiple random output scenarios. Through synchronous back-substitution and scenario reduction, scenarios with a high probability of occurrence and representative of major output changes are retained. If a representative output scenario shows that solar power has effective output during the daytime, the system marks that period as a solar power output period; if a representative output scenario shows that solar power has no effective output at night, the system marks that period as a solar power non-output period. The system writes the representative output scenarios, solar power output periods, and solar power non-output periods into a multi-period reconstruction object 100, thereby forming a multi-period reconstruction object 100 with period labels.

[0053] Example 3 is the third embodiment of the present invention. Based on Examples 1 and 2, this embodiment refines the formation process of the reconstructed fitness characterization 200 and explains how each constraint and parameter enters the reconstructed fitness characterization 200.

[0054] In this embodiment, the system acquires the network loss state, switch operation state, and distributed generation abandonment state for the current switch reconfiguration state under the corresponding operating segment, and generates a reconfiguration operation cost. The current switch reconfiguration state is a set of switch switching combinations currently being evaluated by the reconfiguration optimization algorithm 300. After substituting this switch switching combination into the distribution network topology, the system performs power flow calculations to obtain the current of each branch, the power of each branch, and the voltage of each node. The network loss state can be formed based on the branch current and the line resistance, and is used to represent the active power loss of the distribution network due to the power flow distribution under the current switch reconfiguration state.

[0055] Furthermore, the switch action status is used to indicate the number of changes in the current switch reconfiguration status relative to the switch status before reconfiguration. The more switch actions there are, the more complex the reconfiguration operation is, and the higher the requirements for operation scheduling and equipment action. The system can compare the candidate switch status with the initial switch status before or after fault isolation, count the number of switches that have changed from open to closed, and thus generate the switch action status.

[0056] Furthermore, the curtailment status of distributed generation includes wind curtailment and solar curtailment. Wind curtailment can be determined by the difference between the available wind power and the actual injected power; solar curtailment can be determined by the difference between the available photovoltaic power and the actual injected power. When a candidate switch reconfiguration state leads to limited local line capacity or node voltage constraints, it may be unable to fully absorb wind or photovoltaic output, resulting in wind or solar curtailment. The system incorporates network loss status, switch operation status, and distributed generation curtailment status into the reconfiguration operation cost, ensuring that candidate reconfiguration solutions are no longer evaluated solely based on a single network loss objective.

[0057] This distribution network reconfiguration selects minimizing network losses, minimizing switch operations, and minimizing wind and solar power curtailment as the objective function, expressed as: , , , , in, This represents the overall optimization target value. This indicates the active power loss of the distribution network after reconfiguration. This represents the network active power loss when a system failure occurs. This indicates the number of switch operations required to complete this power distribution network reconfiguration. This indicates the number of switching actions required for the system to complete the power distribution network reconfiguration or to reach the baseline state. This represents the target item for distributed power curtailment. Indicates node voltage. Indicates line capacity or line transmission power. Indicates the first The actual current carrying capacity or branch current of the branch circuit. Indicates the first The maximum allowable flow rate of each branch road.

[0058] Preferably, the system also obtains the operational safety boundary state corresponding to the current switch reconfiguration state, and generates a reconfiguration safety state. The operational safety boundary state includes the electrical operational boundary and the network topology boundary. The electrical operational boundary is used to determine whether the node voltage is within the allowable voltage range, whether the line transmission capacity exceeds the maximum capacity, and whether the branch current exceeds the allowable current carrying capacity. The network topology boundary is used to determine whether the reconfigured power supply structure meets the fault isolation requirements, whether invalid loops are avoided, and whether an executable power supply path can be formed.

[0059] In one specific implementation, after the system performs power flow calculations on a candidate switch reconfiguration state, if the voltage of each node is within the allowable upper and lower limits, the line capacity does not exceed the upper limit, the branch current carrying capacity does not exceed the allowable value, and the network topology meets the power supply requirements, then a reconfiguration safety state that meets the operational safety constraints is generated. At this time, the system determines the reconfiguration operation cost as the fitness evaluation value of the switch reconfiguration state. If the voltage of a node exceeds the limit, the capacity of a line exceeds the limit, the current carrying capacity of a branch exceeds the limit, or the network topology does not meet the power supply requirements, then the system marks the switch reconfiguration state as an infeasible reconfiguration state and updates the reconfiguration fitness representation 200 accordingly.

[0060] Voltage over-limit index is defined as: , in, This represents a comprehensive risk indicator for system-level voltage over-limit. Represents a node Comprehensive risk assessment indicators for voltage over-limit, Represents a node The probability density function of voltage is used to represent the probability distribution of node voltages under different values. Represents a node The severity function after a voltage limit violation occurs. Represents a node Voltage value.

[0061] It should be noted that infeasible reconfiguration states can trigger penalized fitness evaluation values. Penalized fitness evaluation values ​​are used to reduce the retention probability of candidate reconfiguration solutions that do not meet operational safety constraints in the genetic iteration. For example, if a certain switch combination reduces network losses but causes node voltages to fall below the allowable lower limit, the system does not directly use this switch combination as the preferred solution. Instead, it marks this switch combination as an infeasible reconfiguration state and reflects this state through the penalized fitness evaluation value. The fitness evaluation value and the penalized fitness evaluation value together form the reconfiguration fitness representation 200, ensuring that the candidate switch reconfiguration state is simultaneously controlled by operational costs and operational safety constraints.

[0062] Example 4, refer to Figure 6 This is the fourth embodiment of the present invention. Based on embodiment 3, this embodiment further explains the network topology boundary in the operational safety boundary state, and combines the contents of load classification, islanding, and the fact that the output of distributed power sources in the island is not less than the load demand of the island to explain how to form an island power supply relationship after fault isolation.

[0063] In this embodiment, the operational safety boundary state includes the network topology boundary, which is configured with islanded power supply constraints. These constraints determine which load nodes can be restored to power by distributed generation after fault isolation, and whether these load nodes can form an executable local power supply area based on branch connectivity. The system first determines the set of load nodes to be restored after fault isolation based on the distribution network fault state. The distribution network fault state includes the location of the faulty branch, the nodes on both sides of the faulty branch, the fault isolation range, and the load nodes that have lost their original power supply path due to the fault.

[0064] Furthermore, the system generates load level statuses for the set of load nodes to be restored based on load importance. Load level statuses represent the priority of different load nodes in fault recovery, and loads can be divided into high-level, medium-level, and low-level loads. High-level loads correspond to load nodes with significant power outage losses, medium-level loads correspond to load nodes requiring rapid power restoration, and low-level loads correspond to load nodes without special restoration requirements. The system can calculate load importance based on load weight and node power demand; the higher the load weight and the greater the power demand, the higher the restoration priority of the corresponding node in the islanded power supply relationship.

[0065] Reference Figure 6 When branch lines 8-9 and 2-22 experience short-circuit faults simultaneously, nodes 5, 6, 12, 13, 23, 24, 29, and 31 are considered low-level loads; nodes 7, 11, 15, 22, 26, 30, and 32 are considered medium-level loads; and nodes 1, 2, 3, 4, 8, 9, 10, 14, 16, 17, 18, 19, 20, 21, 25, 27, and 28 are considered high-level loads.

[0066] Load level status is represented as follows: , in, This value represents the importance of the load; the higher the value, the more important the corresponding load node is, and the more priority should be given to ensuring power supply after a fault. Indicates the first The weighting coefficient of each load node, Indicates the first The active power required by each load node.

[0067] If two nodes appear If the values ​​are the same, then line loss will be considered, and the line with lower loss will be assigned to this island: , in, This indicates the total number of load nodes that are unable to continue supplying power normally due to a fault. This represents the impedance value between nodes in an isolated system, or the impedance of the power supply path.

[0068] The total power generation of renewable distributed generation (DG) within the island should not be less than the power required by the load within the island. , in, This indicates the number of distributed power sources within the isolated island. This indicates the power consumed by the lines during island power supply. This indicates the power that the distributed power source can provide to the isolated island.

[0069] Ideally, when different load nodes to be restored are of equal or similar importance, the system further considers power supply path loss or branch impedance. If two load nodes are of the same load level, the system prioritizes the node with lower power supply path loss and lower branch impedance for inclusion in the islanded power supply candidate relationship. In this way, the islanded power supply relationship not only reflects the priority of load level but also the line operation cost.

[0070] Furthermore, the system matches the set of load nodes to be restored with the output status of renewable distributed power sources under the corresponding operating period, generating islanded power supply candidate relationships. These islanded power supply candidate relationships must simultaneously satisfy distributed power source output constraints, load level priority constraints, and branch connectivity constraints. Distributed power source output constraints require that the total output of renewable distributed power sources within the island is not less than the load demand within the island and the corresponding line losses; load level priority constraints require priority restoration of higher-level loads; and branch connectivity constraints require that there be a conductive power supply path between the distributed power source and the load nodes to be restored.

[0071] In one specific implementation, when a fault occurs in some branches of the distribution network, the system first disconnects the faulty branch 101, and then determines whether an islanded power supply area can be formed around the distributed power source. If a photovoltaic distributed power source has effective output during the photovoltaic output period, and the surrounding load nodes to be restored can be connected to the photovoltaic power source through a conductive branch, and the total output within the island can cover the load demand and line losses, then the system determines the candidate power supply relationship as an islanded power supply relationship. If it is during the photovoltaic non-output period and the same photovoltaic power source has no effective output, the system will not use the photovoltaic power source as an islanded power source, but will re-establish the power supply path through a tie switch or other distributed power sources. The system finally writes the islanded power supply relationship into the network power supply topology state, so that the network power supply topology state can characterize the power supply restoration results of different levels of load nodes after the faulty branch 101 is isolated.

[0072] Example 5, refer to Figures 7-12This is the fifth embodiment of the present invention. Based on embodiments 3 and 4, this embodiment further explains the optimization process of the reconstruction optimization algorithm 300 and the output process of the multi-time distribution network robust reconstruction scheme 500. It also explains how candidate reconstruction solutions are updated, converged, and bound to the voltage over-limit risk state in conjunction with the reconstruction optimization algorithm 300, the BFGS update idea, and the voltage over-limit risk verification content.

[0073] Reconstruction optimization algorithm 300 is represented as: The Hessian matrix must satisfy the following conditions: , in, Indicates the first The gradient vector of the objective function at the nth iteration Indicates the first The Hessian matrix corresponding to the objective function at the nth iteration. Indicates the first The variable vector or parameter vector at the next iteration; substitute: , in, Describes the inverse of the Hessian matrix. This represents the gradient difference between two adjacent iterations of the same value s. This represents the difference in variables or the step size vector between two adjacent iterations.

[0074] , in, This represents the approximate matrix of the Hessian inverse matrix in the quasi-Newton method.

[0075] , The reconstruction optimization algorithm 300 can adopt the BFGS algorithm, where the BFGS algorithm is used to update the quasi-Newton update matrix according to the change of candidate reconstruction solutions and the change of fitness, and to correct the search direction of candidate reconstruction solutions in the updated reconstruction population based on the quasi-Newton update matrix.

[0076] in, This represents the approximate matrix of the Hessian inverse matrix in the BFGS algorithm.

[0077]

[0078] in, Indicates a correction term. This represents the adjustment matrix.

[0079]

[0080] In this embodiment, the system encodes the switch reconstruction states as candidate reconstruction solutions and forms a reconstruction population. A candidate reconstruction solution represents a set of switch on / off combinations participating in the optimization process. The system can arrange all switches participating in the reconstruction according to their switch numbers and use an encoded value to represent the on / off state of each switch. One candidate reconstruction solution corresponds to a complete set of switch reconstruction states, and multiple candidate reconstruction solutions together form the reconstruction population. The system substitutes each candidate reconstruction solution into the reconstruction fitness representation 200 to obtain the corresponding fitness value.

[0081] Furthermore, the system performs genetic iteration operations on the reconstructed population. Genetic iteration can include selection, crossover, and mutation. Selection is used to retain candidate reconstructed solutions with better fitness, crossover is used to exchange some switch state information among candidate reconstructed solutions, and mutation is used to change some switch states to generate new topology combinations. After genetic iteration, an updated reconstructed population is formed. Each candidate reconstructed solution in the updated reconstructed population is substituted back into the distribution network reconstructed model for comparison, and the population is updated according to the fitness value.

[0082] Preferably, the system introduces a quasi-Newton update matrix on top of the genetic iteration to correct the search direction of candidate reconstructed solutions. The quasi-Newton update matrix can adopt a BFGS-like update approach, replacing the second-order derivative information with an approximate matrix, and updating the matrix based on changes in candidate solutions and fitness during adjacent iterations. This matrix is ​​used to correct and update candidate reconstructed solutions in the reconstructed population, ensuring that changes in candidate reconstructed solutions not only rely on random crossover and mutation but can also be locally adjusted along the fitness improvement direction.

[0083] Furthermore, after each population update, the system determines whether the convergence condition is met. The convergence condition can be reaching a preset number of iterations, the change in the optimal fitness value being less than a preset threshold over multiple consecutive rounds, or the fitness of candidate reconstruction solutions in the updated reconstruction population tending to stabilize. When the convergence condition is met, the system determines the candidate reconstruction solution with the optimal fitness value as robust reconstruction solution 400. This robust reconstruction solution 400 corresponds to a specific switch reconstruction state and can generate the network power supply topology state for the corresponding runtime segment.

[0084] After obtaining the robust reconfiguration solution 400, the system generates the voltage over-limit risk state for the corresponding operating segment based on the robust reconfiguration solution 400. The voltage over-limit risk state is used to describe the risk of node voltage over-limits occurring in the reconfigured distribution network under random output conditions. The system can substitute the robust reconfiguration solution 400 into multiple random output scenarios for power flow verification, count the cases where the node voltage exceeds the upper limit or falls below the lower limit, and combine the over-limit probability and over-limit severity to form the voltage over-limit risk state. If the voltage over-limit risk state meets the preset risk constraints, the robust reconfiguration solution 400 for the corresponding operating segment is determined as a valid reconfiguration solution.

[0085] In one specific implementation, the system reconfigures the photovoltaic (PV) power output periods and non-PV power output periods separately. The effective reconfiguration solution for the PV power output periods can include a set of switch reconfiguration states adapted to PV participation in power supply during that period, and generates a corresponding network power supply topology state. The effective reconfiguration solution for the non-PV power output periods can include another set of switch reconfiguration states adapted to PV non-PV participation in power supply during that period, and generates a corresponding network power supply topology state. The system binds the effective reconfiguration solution, switch reconfiguration states, network power supply topology state, and voltage limit exceedance risk state for the PV power output periods, and also binds the effective reconfiguration solution, switch reconfiguration states, network power supply topology state, and voltage limit exceedance risk state for the non-PV power output periods, generating a multi-period distribution network robust reconfiguration scheme 500.

[0086] For details on the node voltage distribution before and after the reconstruction, please refer to... Figures 7-8 The blue dotted line represents the distribution network before reconstruction, the green dashed line represents the distribution network reconstruction without robust optimization, and the red solid line in the two line graphs represents robust reconstruction of the distribution network during photovoltaic periods and robust reconstruction of the distribution network during non-photovoltaic periods, respectively. The vertical axis represents the per-unit voltage value.

[0087] pass Figures 7-8 It can be seen that the solid red line is higher than the dashed blue line and the dashed green line, indicating that during periods of photovoltaic (PV) output, the node voltage distribution is higher than the comparison curve. When there is no PV output, although the reconfiguration effect is not as significant as when there is PV output, the voltage exceedance risk is lower than the risk value without robust optimization compared to network reconfiguration. The reconfigured system is as follows: Figures 9-10 As shown, when a fault occurs, the distribution network can be divided into several islands by switching actions, thereby ensuring the normal operation of the load points.

[0088] In terms of output format, the multi-time period distribution network robust reconfiguration scheme 500 can be represented as a running segment table, a switch reconfiguration status table, a network power supply topology status table, and a voltage over-limit risk status table. Operators or the dispatching system can directly call the corresponding switch reconfiguration status and network power supply topology status according to the running segment. When photovoltaic output conditions change, fault conditions change, or voltage over-limit risk status changes, the system can re-call the multi-time period reconfiguration object 100 and the reconfiguration fitness representation 200 to perform reconfiguration optimization.

[0089] Two thousand random scenarios were generated using the Monte Carlo method. The two reconstructed optimization results were then substituted into these scenarios for voltage limit testing. Specific results are as follows: Figures 11-12As shown in the figure, the x-axis represents the node number, the y-axis represents time, and the z-axis represents the voltage over-limit risk value. By comparing the voltage over-limit risk values ​​of each node in the distribution network at different time periods after the two reconfiguration methods, it can be clearly seen that robust optimization can effectively reduce the risk of voltage over-limit occurrences in the distribution network. The main reason is that deterministic reconfiguration is only for a single scenario, and this scenario does not have randomness. Therefore, when deterministic reconfiguration faces different scenarios, the risk of voltage over-limit will inevitably occur.

[0090] Example 6 is an embodiment of the present invention, which provides a robust reconfiguration system for a multi-time distribution network, including an object generation module, a fitness characterization module, and a reconfiguration output module.

[0091] The object generation module is used to obtain the operating status of the distribution network containing renewable distributed power sources. It associates the output status of renewable distributed power sources, the fault status of the distribution network, and the reconfiguration status of switches according to the operating time period. It forms representative output scenarios based on Latin hypercube sampling and synchronous back-substitution scenario reduction, and writes the photovoltaic output period and photovoltaic non-output period into the multi-time period reconfiguration object 100.

[0092] The fitness characterization module is used to receive the multi-time period reconfiguration object 100, generate the reconfiguration operation cost based on the network loss state, switching action state and distributed power abandonment state, generate the reconfiguration safety state based on the operation safety boundary state, and write the island power supply relationship formed under the island power supply constraint into the network power supply topology state to form the reconfiguration fitness characterization 200.

[0093] The reconfiguration output module is used to optimize the reconfiguration fitness representation 200 by using the switch reconfiguration state as the reconfiguration variable and the reconfiguration optimization algorithm 300 to obtain a robust reconfiguration solution 400 that matches different operating periods. Based on the robust reconfiguration solution 400, a voltage over-limit risk state is generated, and the effective reconfiguration solutions corresponding to the photovoltaic output period and the photovoltaic non-output period are bound according to the operating period to output a multi-period distribution network robust reconfiguration scheme 500.

[0094] This embodiment also provides a computer device, including a memory and a processor. The memory stores a computer program, and when the processor executes the computer program, it implements a robust reconfiguration method for a multi-time distribution network as proposed in the above embodiment.

[0095] This embodiment also provides a computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements a multi-time period distribution network robust reconfiguration method as proposed in the above embodiment.

[0096] If a function is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0097] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (such as a computer-based system, a processor-including system, or other system that can fetch and execute instructions from, an instruction execution system, apparatus, or device). For the purposes of this specification, "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transmit programs for use by, or in conjunction with, an instruction execution system, apparatus, or device.

[0098] More specific examples of computer-readable media (a non-exhaustive list) include: electrical connections (electronic devices) having one or more wires, portable computer disk drives (magnetic devices), random access memory (RAM), read-only memory (ROM), erasable and editable read-only memory (EPROM or flash memory), fiber optic devices, and portable optical disc read-only memory (CDROM). Furthermore, computer-readable media can even be paper or other suitable media on which the program can be printed, because the program can be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, interpreting, or otherwise processing as necessary, and then stored in computer memory.

[0099] It should be understood that various parts of the present invention can be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented in software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, it can be implemented using any one or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0100] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A robust reconfiguration method for multi-time distribution networks, characterized in that, include: Obtain the operating status of the distribution network containing renewable distributed power sources and form multi-time period reconfiguration objects; Based on multi-period reconstruction objects, obtain a reconstruction fitness representation that includes robust reconstruction goals and runtime safety constraints; Using the switch reconstruction state as the reconstruction variable, the reconstruction fitness representation is optimized by a reconstruction optimization algorithm that includes iterative updates of candidate reconstruction solutions and correction of search direction, so as to obtain a robust reconstruction solution that matches different runtime stages. Based on the robust reconfiguration solution, output a multi-time period distribution network robust reconfiguration scheme; The robust reconfiguration scheme for multi-time distribution networks includes the switch reconfiguration status and network power supply topology status for the corresponding operating time period; Based on the photovoltaic output status in representative power output scenarios, the corresponding operating periods are divided into photovoltaic power output periods and photovoltaic non-power output periods; Representative power output scenarios, photovoltaic power output periods, and photovoltaic non-power output periods are written into a multi-period reconstruction object to form a multi-period reconstruction object with period labels; Obtain the operational safety boundary state corresponding to the current switch reconfiguration state, and generate the reconfiguration safety state; The operational security boundary state includes the network topology boundary; The network topology boundary is configured with islanded power supply constraints; The formation of islanded power supply constraints includes determining the set of load nodes to be restored after fault isolation based on the fault status of the distribution network, and generating the load level status of the set of load nodes to be restored according to the importance of the loads; The set of load nodes to be restored is matched with the output status of renewable distributed power sources under the corresponding operating segment to generate candidate relationships for islanded power supply. When the islanded power supply candidate relationship satisfies the distributed generation output constraint, load level priority constraint and branch connectivity constraint, the islanded power supply candidate relationship is determined as an islanded power supply relationship; Write the islanded power supply relationship into the network power supply topology state so that the network power supply topology state can represent the power supply recovery results of different levels of load nodes after the fault branch is isolated; For the switch reconfiguration state under the corresponding runtime segment, obtain the network loss state, switch action state and distributed power abandonment state corresponding to the current switch reconfiguration state, and generate the reconfiguration operation cost. Optimizing the reconstructed fitness representation through a reconstruction optimization algorithm includes... The switch reconstruction state is encoded as a candidate reconstruction solution and a reconstruction population is formed. The fitness value of each candidate reconstruction solution is obtained based on the reconstruction fitness representation; Perform genetic iteration operations on the reconstructed population to form an updated reconstructed population; The search direction of candidate reconstruction solutions in the update and reconstruction population is corrected by using a quasi-Newton update matrix; The candidate reconstruction solution that satisfies the convergence condition and has the best fitness value is determined as the robust reconstruction solution; It also includes generating voltage over-limit risk states for the corresponding runtime segment based on robust reconstructed solutions; When the voltage over-limit risk state meets the preset risk constraints, the robust reconstruction solution of the corresponding running segment will be determined as the valid reconstruction solution. The effective reconfiguration solutions corresponding to the photovoltaic power output period and the effective reconfiguration solutions corresponding to the photovoltaic non-power output period are bound according to the operating period to generate a robust reconfiguration scheme for the multi-period distribution network.

2. The robust reconfiguration method for multi-time distribution networks as described in claim 1, characterized in that: The objects formed by multi-time period reconstruction include, Based on the output status of renewable distributed power sources, the fault status of the distribution network, and the reconfiguration status of switches associated with the runtime segment, reconfiguration object units corresponding to the runtime segment are formed. Multiple runtime refactoring objects are combined according to their temporal relationship to form multi-time period refactoring objects; Multi-period reconfiguration objects are used to define the matching relationship between the output changes of renewable distributed power sources, the isolation relationship of fault branches, and the reconfiguration state of switches under different operating periods.

3. The robust reconfiguration method for multi-time distribution networks as described in claim 1 or 2, characterized in that: It also includes, A set of distributed power source probabilistic output scenarios is generated based on Latin hypercube sampling; Based on the reduction of synchronous back-generation scenarios, the set of probabilistic output scenarios of distributed power sources is reduced to obtain representative output scenarios.

4. The robust reconfiguration method for multi-time distribution networks as described in claim 2, characterized in that: Based on multi-period reconstruction objects, a reconstruction fitness representation including robust reconstruction goals and runtime safety constraints is obtained, including... Distributed power curtailment includes wind curtailment and solar curtailment. When the reconfiguration safety state meets the operational safety constraints, the reconfiguration operation cost is determined as the fitness evaluation value of the corresponding switch reconfiguration state; When the reconstructed safety state does not meet the operational safety constraints, the corresponding switch reconstructed state is marked as an infeasible reconstructed state. The reconstructed fitness representation is updated based on the infeasible reconstructed state, and a penalty fitness evaluation value is generated based on the infeasible reconstructed state. The fitness representation is formed by the fitness evaluation value and the penalty fitness evaluation value.

5. The robust reconfiguration method for multi-time distribution networks as described in claim 4, characterized in that: The multi-time period distribution network robust reconfiguration scheme is used to characterize the time-period network reconfiguration results that match the switch reconfiguration state, network power supply topology state, and voltage over-limit risk state under different operating periods.

6. A robust reconfiguration system for a multi-time distribution network, employing the robust reconfiguration method for a multi-time distribution network as described in any one of claims 1 to 5, characterized in that: It includes an object generation module, a fitness representation module, and a reconstruction output module; The object generation module is used to obtain the operating status of the distribution network containing renewable distributed power sources. It associates the output status of renewable distributed power sources, the fault status of the distribution network, and the reconfiguration status of switches according to the operating time period. It forms representative output scenarios based on Latin hypercube sampling and synchronous back-substitution scenario reduction, and writes the photovoltaic output period and photovoltaic non-output period into the multi-time period reconfiguration object. The fitness characterization module is used to receive multi-time period reconfiguration objects, generate reconfiguration operation costs based on network loss status, switching action status and distributed power abandonment status, generate reconfiguration safety status based on operation safety boundary status, and write the island power supply relationship formed under island power supply constraints into the network power supply topology status to form a reconfiguration fitness characterization. The reconfiguration output module is used to optimize the reconfiguration fitness representation by using the switch reconfiguration state as the reconfiguration variable and the reconfiguration optimization algorithm to obtain robust reconfiguration solutions that match different operating periods. Based on the robust reconfiguration solutions, the voltage over-limit risk state is generated, and the effective reconfiguration solutions corresponding to the photovoltaic output period and the photovoltaic non-output period are bound to the operating period to output a multi-period distribution network robust reconfiguration scheme.

7. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, The steps of implementing the multi-time distribution network robust reconfiguration method of any one of claims 1 to 5 when the processor executes the computer program.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that, When a computer program is executed by a processor, it implements the steps of the multi-time distribution network robust reconfiguration method according to any one of claims 1 to 5.

Images