Power distribution network interrupt flow optimization method and device

Abstract

The application discloses a power distribution network interrupting power flow optimization method and equipment. The method collects operation data in real time through smart meter terminals arranged at key nodes of the power distribution network, establishes a multi-objective optimization model considering minimization of power failure loss, minimization of network loss and optimization of voltage quality, adopts an improved particle swarm algorithm fusing chaos theory to solve an optimal network reconstruction scheme, and realizes automatic execution and real-time monitoring and adjustment of the reconstruction operation through a power distribution automation system. The application can quickly generate an optimal reconstruction scheme when a fault occurs in the power distribution network, effectively improves power supply reliability, reduces power failure loss, and realizes intelligent operation management of the power distribution network.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of power system automation, and relates to a power distribution network interrupted power flow optimization method and device. BACKGROUND

[0002] When a fault occurs in a power distribution network, the traditional processing method mainly relies on the experience judgment and manual operation of dispatchers. This method not only has a long response time, but also is prone to misjudgment, resulting in the expansion of the power outage range and the prolongation of the recovery time.

[0003] At present, there is a certain research basis in the field of power distribution network fault reconstruction at home and abroad, mainly using intelligent optimization methods such as heuristic algorithm, genetic algorithm, ant colony algorithm, etc. However, the existing technology has obvious deficiencies in dealing with multi-objective optimization problems of large-scale power distribution networks: first, the algorithm convergence speed is slow, which is difficult to meet the real-time requirements under fault conditions; second, it is easy to fall into a local optimal solution and cannot guarantee global optimality; third, when dealing with multiple constraint conditions, the constraint violation degree is high, and the feasibility of the generated reconstruction scheme is poor; fourth, there is a lack of adaptive weight adjustment mechanism for different fault severity, resulting in a mismatch between the optimization target and the actual demand. In addition, the existing power distribution network fault processing system also has limitations in data acquisition and processing. Although the traditional SCADA system can provide basic operation data, the data acquisition accuracy and real-time performance are insufficient, which cannot meet the needs of accurate power flow calculation and optimization decision-making.

[0004] Therefore, it is urgent to develop a power distribution network interrupted power flow optimization method that integrates advanced data acquisition technology, improved optimization algorithm and automatic control system, in order to realize fast, accurate and economic network reconstruction and load transfer under fault conditions. SUMMARY

[0005] The purpose of the present application is to solve the problem in the prior art that in the face of large-scale power distribution networks, multiple fault scenarios and dynamic load changes, there is a lack of effective intelligent optimization algorithm and real-time data support, making it difficult to realize fast, accurate and economic network reconstruction and load transfer under power distribution network fault conditions.

[0006] In the first aspect, the present application provides a power distribution network interrupted power flow optimization method for realizing fast reconstruction and load transfer under power distribution network fault conditions. The power distribution network collects operation data in real time through smart meter terminals, has automatic switch equipment and power flow control capability, and the optimization method comprises the following steps:

[0007] Step S1: Collecting power distribution network operation data in real time through smart meter terminals deployed at key nodes of the power distribution network;

[0008] The operation data includes: the effective value and phase angle of the voltage of each node , and the power flow of each branch branch current effective value and power factor, load power and switching state and network topology, denotes node number, denotes branch number;

[0009] Step S2, a multi-objective distribution network interruption flow optimization model considering minimum power loss, minimum network loss and voltage quality optimization is established;

[0010] Step S3, the optimization model is solved by using an improved particle swarm algorithm combined with chaos theory within the constraint condition range, so as to obtain an optimal network reconstruction scheme and load transfer strategy;

[0011] Step S4, the optimization scheme is issued to each switching device through a distribution automation system, so as to realize automatic execution of network reconstruction;

[0012] Step S5, the reconstruction effect is monitored in real time, a control strategy is dynamically adjusted according to feedback information, and a final optimization scheme is output.

[0013] Further, the data acquisition method of the intelligent electric meter terminal in the step S1 comprises:

[0014] A three-phase intelligent electric meter terminal is deployed at the low-voltage side of a distribution transformer and an important load node, a sampling frequency is not less than 50Hz, and a data uploading period is 1 second;

[0015] Based on the collected distribution network operation data, voltage , current , active power , reactive power and topology identification are calculated;

[0016] , wherein is the voltage effective value of the node , the measurement range is 0-35kV, and the accuracy level is 0.2S, is the voltage phase angle; , wherein is the current effective value of the branch , the measurement range is 1A-3000A, and a current transformer ratio is configured according to the voltage level, is the phase difference of the current;

[0017] active power ; reactive power , the measurement accuracy is ±0.5%;

[0018] The topology identification is obtained through switching state information The real-time network topology is determined according to the node connection relationship, wherein represents a branch closed, represents a branch broken.

[0019] Further, the mathematical expression of the multi-objective optimization model in step S2 is as follows:

[0020]

[0021] is the power loss of node , unit: kW; is the interruption time of node , unit: h;

[0022] is the resistance of branch , unit: Ω; is the current effective value of branch , unit: A; is the voltage per unit of node , unit: p.u.; is the reference voltage, taking 1.0 p.u.; is the total number of nodes, is the total number of branches.

[0023] Further, the constraint conditions include the following conditions that need to be met simultaneously: power balance constraint, voltage constraint, branch capacity constraint, network radiation constraint and switch operation constraint, and the network connectivity is guaranteed, and there is no isolated node.

[0024] Further, the mathematical expression of the constraint condition is as follows:

[0025] ; and are the input branch set and the output branch set of connected node , respectively;

[0026] ; and are the active power and the reactive power of branch , respectively;

[0027] , wherein p.u., p.u.;

[0028] ; is the maximum transmission capacity of branch , is the branch transmission capacity of the transmission line;

[0029] ;

[0030] ; and are the new state and the original state of the branch respectively; is the maximum number of allowed switching operations.

[0031] Further, the improved particle swarm algorithm incorporating chaos theory in step S3 comprises:

[0032] Step S31, initializing the particle swarm population, the population size is set to 30-50 particles, the dimension D of each particle is equal to the number of operable switches in the network, and the particle position is coded in binary representing the switch state combination;

[0033] Step S32, adopting a nonlinear adaptive inertia weight strategy: , is the current iteration number, is the maximum iteration number;

[0034] Step S33, introducing a Logistic chaos mapping mechanism to disturb the particles trapped in local optimum: ; the initial value of and ; when the fitness of the particle has not improved for 10 consecutive generations, chaos disturbance is started;

[0035] Step S34, updating the particle position and velocity: ;

[0036] ; wherein, , is a random number in the interval [0,1], is the velocity vector of the particle in the th generation, is the individual historical optimal position of the particle, is the global optimal position of the population;

[0037] Step S35, calculating the fitness function value, adopting a constraint violation penalty mechanism:

[0038] ; is the penalty factor, is the violation degree of the th constraint;

[0039] Step S36, when the global optimal solution changes less than 0.01% for 20 consecutive generations or reaches the maximum number of iterations , the convergence condition is met, and the optimal solution is output, otherwise, return to step S32.

[0040] Further, , , The weight coefficient of the power supply is valued according to the fault severity grading system;

[0041] When the fault is a single feeder fault, the affected load is the total power of the outage load caused by the fault: , the priority is to recover quickly, , , The weight settings are 0.7, 0.15, and 0.15, respectively;

[0042] When the main transformer fails or multiple feeders fail, the affected load , the recovery speed and economy need to be balanced, , , The weight settings are 0.5, 0.25, and 0.25, respectively;

[0043] When the station power supply fails or a large area is powered off, the affected load ; , , The weight settings are 0.4, 0.3, and 0.3, respectively.

[0044] Further, the execution control of the reconstruction scheme in step S4 includes: sending switch operation instructions to a remote terminal unit RTU through a power distribution automation SCADA system.

[0045] In a second aspect, based on the same inventive concept, the present application provides a power distribution network interrupted power flow optimization device, which comprises, in sequence: a data acquisition module, a power flow calculation module, an optimization solving module, a scheme execution module and a monitoring feedback module.

[0046] The data acquisition module acquires real-time power distribution network operation data through a smart meter terminal, establishes a network topology and load distribution model;

[0047] The power flow calculation module calculates the power flow distribution and voltage level under the current network state based on the Newton-Raphson method;

[0048] The optimization solving module uses an improved particle swarm algorithm to solve the multi-objective optimization model and generates an optimal reconstruction scheme;

[0049] The scheme execution module executes the network reconfiguration operation by controlling the switching device through the SCADA system;

[0050] The monitoring feedback module monitors the reconfiguration effect in real time, and provides system operation state evaluation and parameter adjustment suggestions.

[0051] Further, the optimization solving module comprises an algorithm parameter management unit, a population evolution calculation unit and a solution set evaluation unit.

[0052] The algorithm parameter management unit stores and manages various parameter settings of the particle swarm algorithm, and supports parameter adaptive adjustment.

[0053] The population evolution calculation unit executes the particle swarm iteration process based on the multi-objective optimization function using a parallel computing architecture, and supports optimization calculation of a distribution network with no less than 1000 nodes.

[0054] The solution set evaluation unit performs feasibility verification and performance evaluation on the generated reconfiguration scheme, and screens out a Pareto optimal solution set meeting all constraint conditions.

[0055] Compared with the prior art, the present application has the beneficial effects that:

[0056] The present application collects distribution network operation data in real time through the smart meter terminal, establishes a multi-objective optimization model, solves it using an improved particle swarm algorithm integrating chaos theory, and realizes automatic execution and dynamic adjustment of the scheme in combination with the distribution automation system, so that an optimal network reconfiguration scheme can be quickly and accurately generated when a fault occurs in the distribution network, the fault handling efficiency and power supply reliability are significantly improved, power loss and network loss are effectively reduced, voltage quality is optimized, the intelligentization and automation of distribution network fault handling are realized, and a powerful guarantee is provided for the safe and stable operation of the power system. BRIEF DESCRIPTION OF DRAWINGS

[0057] Figure 1 A flow chart of a distribution network interrupted power flow optimization method of the present application;

[0058] Figure 2 A schematic diagram of a distribution network interrupted power flow optimization device of the present application. DETAILED DESCRIPTION

[0059] In order to make the purpose, technical scheme and advantages of the present application clearer, the technical scheme of the present application is described clearly and completely below. Obviously, the described embodiments are part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments of the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the present application.

[0060] It should be noted that in the present application, the key node refers to an electrical node having an important influence on the system operation state in the power distribution network, mainly including the following several types: 1. The substation outgoing line bus node, which is connected to the main transformer and multiple feeder lines and is the hub of power distribution; 2. The ring network switch station node, which is usually located at the intersection of multiple feeders and has load transfer capability; 3. The sectionalizing switch node, which is located at the middle position of the long distance feeder and is used for fault isolation and load transfer; 4. The tie switch node, which connects different power supply areas and can realize cross-regional power supply in the case of fault; 5. The large-capacity distributed power supply access node, such as the grid connection point of the wind farm and photovoltaic power station. The operation state of these nodes directly affects the power supply reliability and power quality of the power distribution network, so it needs to be monitored.

[0061] The important load node refers to a power distribution network node that undertakes the power supply task of key power consumers or has a large load capacity, specifically including: 1. The first level load node, whose power supply object is hospitals, government agencies, important factories and other users that are not allowed to be powered off; 2. The large industrial user node, which is the access point of an industrial enterprise whose single user power capacity exceeds 1000 kilowatts; 3. The residential intensive area distribution transformer node, which is the distribution transformer node of a residential area serving more than 500 households; 4. The commercial center node, which is the power supply point of large shopping malls, office buildings and other commercial load centers; 5. The public facility node, such as the power supply node of schools, stations, airports and other public places. Once these nodes are powered off, it will have a great impact on the society and economy, and the power supply needs to be prioritized.

[0062] Embodiment 1

[0063] As shown in Figure 1 , for realizing the fast reconstruction and load transfer of the power distribution network in the event of a fault, the power distribution network collects operation data in real time through smart meter terminals, has automatic switch equipment and power flow control capability, and the optimization method includes the following steps:

[0064] Step S1, collecting power distribution network operation data in real time through smart meter terminals deployed at each key node of the power distribution network;

[0065] The operation data includes: the voltage effective value and voltage phase angle of each node , the branch current effective value and power factor of each branch , load power and , switch state and network topology, , wherein the node number is represented by , and the branch number is represented by

[0066] In the practical application of a certain urban distribution network, smart meter terminals are deployed at the 10 kV outgoing bus of a 110 kV substation, ring network switch station, and important user substation, etc. Taking a typical 10 kV feeder as an example, the feeder is about 8 km long, with a supply radius covering an industrial park and two residential areas, and the total load capacity is about 15 MW. Smart meter terminals are installed at the starting end, intermediate sectional switch, and terminal tie switch of the feeder to realize real-time monitoring of the electrical parameters of the key nodes.

[0067] The smart meter terminal uses high-precision current and voltage sensors to accurately measure the voltage effective value and phase angle information of each node. For example, under normal operating conditions, the voltage at the starting end of the 10 kV feeder is 10.2 kV, and the phase angle is 0 degrees; the voltage at the intermediate node is 9.8 kV, and the phase angle lags about 2 degrees; and the voltage at the terminal node is further reduced to 9.5 kV, and the phase angle lags about 4 degrees. These data are updated every second to provide real-time network operating state information for the system.

[0068] Step S2, a multi-objective distribution network interruption flow optimization model considering the minimization of power outage loss, the minimization of network loss, and the optimization of voltage quality is established;

[0069] The minimization of power outage loss mainly quantifies the economic loss caused by the fault to the users, and the power outage loss cost of different types of users differs greatly, for example, the power outage loss of important users such as hospitals can reach tens of thousands of yuan per hour, while the loss of general industrial users is about several thousand yuan per hour. The minimization of network loss focuses on the economy of system operation, and reduces the line loss by optimizing the flow distribution, and selects the path with smaller resistance to transmit power during the reconstruction process.

[0070] The optimization of voltage quality ensures that the voltage at each node after reconstruction is within the qualified range, avoiding damage to electrical equipment caused by excessively high or low voltage. In practical applications, the voltage qualification rate of the distribution network is required to be above 95%, and the node voltage should be controlled within ±5% of the rated voltage. The setting of the weight coefficient reflects the optimization focus under different fault conditions, and the power supply is restored quickly in the case of slight fault, and the economy and voltage quality need to be considered comprehensively in the case of serious fault.

[0071] Step S3, within the constraint condition range, an improved particle swarm algorithm integrating chaos theory is used to solve the optimization model to obtain the optimal network reconstruction scheme and load transfer strategy;

[0072] In a distribution network containing 50 switches and 30 load nodes, the traditional genetic algorithm needs about 500 iterations to converge, while the particle swarm algorithm improved by chaos mapping can usually find a satisfactory solution within 200 iterations. The size of the particle swarm is set to 40 particles, and each particle represents a switch state combination scheme.

[0073] The introduction of the chaotic mapping mechanism effectively avoids the problem of the algorithm falling into a local optimal solution. When the algorithm detects that the particle swarm has not improved for 10 consecutive generations, chaotic disturbance is automatically started to re-explore the solution space by changing the positions of some particles.

[0074] Step S4, the optimization scheme is issued to each switch device through the power distribution automation system to realize automatic execution of network reconstruction;

[0075] When the optimization algorithm generates the optimal reconstruction scheme, the system first performs a safety check to confirm that the operation sequence will not cause device damage or personnel safety risks; then it sends operation instructions to each switch device through the SCADA system, including operations such as opening the fault area switch and closing the standby tie switch.

[0076] Taking a feeder fault as an example, the system completes fault location, reconstruction scheme calculation, and switch operation within 3 minutes after detecting the fault, successfully transferring 70% of the outage load to adjacent feeder power supply, greatly shortening the outage time.

[0077] Step S5, real-time monitoring of reconstruction effect, dynamic adjustment of control strategy and output of final optimization scheme according to feedback information.

[0078] Real-time monitoring of reconstruction effect not only focuses on voltage quality and load distribution after reconstruction, but also monitors key indicators such as network loss and switch action frequency. When the reconstruction effect is found to be unsatisfactory, the system can automatically adjust the control strategy, such as further optimizing load distribution or adjusting the working state of voltage regulation equipment; in actual operation, the system has established a perfect feedback mechanism, including voltage out-of-limit alarm, overload warning, communication interruption prompt, etc.; operation personnel can real-time view network operation state through monitoring interface, and manually intervene optimization process if necessary.

[0079] The data acquisition method of the smart meter terminal in step S1 includes:

[0080] Three-phase smart meter terminals are deployed on the low-voltage side of distribution transformers and important load nodes, with a sampling frequency not less than 50Hz and a data upload period of 1 second;

[0081] Based on the collected distribution network operation data, voltage , current , active power , reactive power and topology identification are calculated;

[0082] , where is the voltage effective value of node , the measurement range is 0-35kV, and the accuracy level is 0.2S, is the voltage phase angle; , where is the effective value of the current of the branch , the measurement range is 1A-3000A, and the current transformer ratio is configured according to the voltage level, is the phase difference of the current;

[0083] active power ; reactive power , the measurement accuracy is ±0.5%;

[0084] The topology identification is to determine the real-time network topology through switch state information and node connection relationship, wherein represents that the branch is closed, represents that the branch is open.

[0085] The mathematical expression of the multi-objective optimization model in the step S2 is:

[0086]

[0087] is the outage power of the node , in units of kW; is the interruption time of the node , in units of h;

[0088] is the resistance of the branch , in units of Ω; is the effective value of the current of the branch , in units of A; is the voltage per unit of the node , in units of p.u.; is the reference voltage, which is 1.0 p.u.; , , is the weight coefficient, which satisfies and ; is the total number of nodes, is the total number of branches.

[0089] The constraint conditions include the following conditions that need to be met simultaneously: power balance constraint, voltage constraint, branch capacity constraint, network radiation constraint and switch operation constraint, and the network connectivity is ensured, and there is no isolated node.

[0090] The mathematical expression of the constraint condition is as follows:

[0091] ; and are the connected nodes input and output branch sets;

[0092] ; and active and reactive power of branch respectively;

[0093] where p.u., p.u.;

[0094] ; maximum transmission capacity of branch , transmission capacity of branch ;

[0095] ;

[0096] ; and new state and original state of branch respectively; maximum allowable number of switching operations.

[0097] The improved particle swarm algorithm incorporating chaos theory in the step S3 comprises:

[0098] Step S31, initializing the particle swarm population, the population size is set to 30-50 particles, the dimension D of each particle is equal to the number of operable switches in the network, and the particle position is coded in binary representing the switch state combination;

[0099] Step S32, adopting a nonlinear adaptive inertia weight strategy: , current iteration number, maximum iteration number;

[0100] Step S33, introducing a Logistic chaos mapping mechanism to disturb the particles trapped in local optimum: ; initial value of and ; when the fitness of the particles has not improved for 10 consecutive generations, chaos disturbance is started;

[0101] Step S34, updating the particle position and velocity: ;

[0102] ; wherein, , is a random number in the interval [0, 1], is a particle In is the velocity vector of the i-th generation, is the individual history optimal position of the i-th particle, is the global optimal position of the population;

[0103] Step S35, calculate the fitness function value, using a constraint violation penalty mechanism:

[0104] ; is a penalty factor, is the violation degree of the i-th constraint; Step S36, when the global optimal solution changes less than

[0105] or reaches the maximum iteration number generations, the convergence condition is met, and the optimal solution is output, otherwise, return to step S32.

[0106] , , , The weight coefficients of the three indexes are assigned according to the fault severity grading system;

[0107] When the fault is a single feeder fault, the affected load is the total power of the outage load caused by the fault: , the priority is to restore quickly, , , The weight settings are 0.7, 0.15, and 0.15 respectively;

[0108] When the main transformer fails or multiple feeders fail, the affected load is , the speed and economy of restoration need to be balanced, , , The weight settings are 0.5, 0.25, and 0.25 respectively;

[0109] When the station power supply fails or a large area is powered off, the affected load is ; , , The weight settings are 0.4, 0.3, and 0.3 respectively.

[0110] The execution control of the reconstruction scheme in the step S4 includes: sending switch operation instructions to a remote terminal unit RTU through a power distribution automation SCADA system.

[0111] Embodiment 2

[0112] As Figure 2As shown, it is a power distribution network interrupt flow optimization equipment composition schematic diagram, the equipment includes sequentially connected: data acquisition module, power flow calculation module, optimization solving module, scheme execution module and monitoring feedback module.

[0113] The data acquisition module acquires power distribution network operation data in real time through the intelligent electric meter terminal, establishes a network topology and load distribution model;

[0114] The power flow calculation module calculates the power flow distribution and voltage level under the current network state based on the Newton-Raphson method;

[0115] The optimization solving module uses an improved particle swarm algorithm to solve the multi-objective optimization model and generates an optimal reconstruction scheme;

[0116] The scheme execution module controls the switch device to perform network reconstruction operation through the SCADA system;

[0117] The monitoring feedback module monitors the reconstruction effect in real time, provides system operation state evaluation and parameter adjustment suggestion.

[0118] The optimization solving module includes: algorithm parameter management unit, population evolution calculation unit and solution set evaluation unit;

[0119] The algorithm parameter management unit stores and manages various parameter settings of the particle swarm algorithm, supports parameter adaptive adjustment;

[0120] The population evolution calculation unit executes the particle swarm iteration process based on the multi-objective optimization function using a parallel computing architecture, and supports optimization calculation of power distribution networks with no less than 1000 nodes;

[0121] The solution set evaluation unit performs feasibility verification and performance evaluation on the generated reconstruction scheme, and selects a Pareto optimal solution set that meets all constraint conditions.

[0122] The above specific embodiments further illustrate the purpose, technical solutions and beneficial effects of the present application. It should be understood that the above description is only a specific embodiment of the present application and is not used to limit the protection scope of the present application. Any modification, equivalent replacement, improvement, etc. within the spirit and principles of the present application should be included in the protection scope of the present application.

Claims

1. A distribution network interruption power flow optimization method for achieving rapid reconfiguration and load transfer during distribution network faults, wherein the distribution network collects operational data in real time through smart meter terminals and possesses automated switching equipment and power flow control capabilities, characterized in that... The optimization method includes the following steps: Step S1: Real-time data collection of power distribution network operation is achieved by smart meter terminals deployed at key nodes of the power distribution network. The operational data includes: each node The effective value of the voltage and the voltage phase angle, each branch The effective value of the branch current and the power factor, and the load power. and Switch status and network topology Indicates the node number. Indicates the branch number; Step S2: Establish a multi-objective distribution network interruption power flow optimization model that considers minimizing power outage losses, minimizing network losses, and optimizing voltage quality; The mathematical expression for the multi-objective optimization model is: ; For nodes Power loss, in kW; For nodes Interruption time, in hours (h); branch road Resistance, in Ω; branch road The effective value of the current, in amperes (A); For nodes Voltage value, unit: pu; For reference voltage, take 1.0 pu; , , These are the weighting coefficients. The total number of nodes. The total number of branch roads; Step S3: Within the constraints, the improved particle swarm optimization algorithm incorporating chaos theory is used to solve the optimization model to obtain the optimal network reconstruction scheme and load transfer strategy. The constraints include the following that must be satisfied simultaneously: power balance constraint, voltage constraint, branch capacity constraint, network radial constraint, and switching operation constraint, while ensuring network connectivity and the absence of isolated nodes; the mathematical expressions for the constraints are as follows: ; and Connecting nodes The set of input and output branches; ; and Branch roads at the time of separation The active and reactive power; , among them oh, pu; ; branch road Maximum transmission capacity branch road The transmission capacity; ； ; and Branch roads The new state and the original state; The maximum number of allowed switching operations; Step S4: The optimization plan is distributed to each switchgear through the power distribution automation system to realize the automatic execution of network reconfiguration; Step S5: Monitor the reconstruction effect in real time, dynamically adjust the control strategy based on feedback information, and output the final optimization solution.

2. The power flow optimization method for power distribution network interruptions according to claim 1, characterized in that, The data acquisition method for the smart meter terminal in step S1 includes: Three-phase smart meter terminals are deployed on the low-voltage side of the distribution transformer and at important load nodes, with a sampling frequency of not less than 50Hz and a data upload cycle of 1 second. Calculate voltage based on collected power distribution network operation data Current Active power reactive power And topology recognition; ,in For nodes The effective voltage value is measured within a range of 0-35kV with an accuracy class of 0.2S. This refers to the voltage phase angle; ,in branch road The effective value of the current is measured, with a range of 1A-3000A. The current transformer ratio is configured according to the voltage level. This represents the phase difference of the current. Active power Reactive power Measurement accuracy ±0.5%; The topology is identified through switch status information. The connection relationships between nodes determine the real-time network topology, where Indicates a branch closure, Indicates a branch disconnect.

3. The power flow optimization method for power distribution network interruptions according to claim 2, characterized in that, The improved particle swarm optimization algorithm that incorporates chaos theory in step S3 includes: Step S31: Initialize the particle swarm population. The population size is set to 30-50 particles. The dimension D of each particle is equal to the number of operable switches in the network. The particle positions are encoded using binary encoding. Indicates the combination of switch states; Step S32, adopt a nonlinear adaptive inertia weighting strategy: , This represents the current iteration number. This represents the maximum number of iterations. Step S33: Introduce a Logistic chaotic mapping mechanism to perturb particles trapped in local optima: ; initial value and Chaotic perturbation is initiated when the fitness of a particle does not improve for 10 consecutive generations. Step S34, update particle position and velocity: ; ;in, , A random number in the interval [0,1]. For particles exist The velocity vector of the generation, This represents the optimal position in the individual history of the particle. The optimal position for the entire population; Step S35: Calculate the fitness function value and apply a constraint violation penalty mechanism. ; As a penalty factor, For the first The degree of violation of a constraint; Step S36, when the change in the global optimal solution is less than 20 consecutive generations Or reach the maximum number of iterations If the convergence condition is met, output the optimal solution; otherwise, return to step S32.

4. The power flow optimization method for power distribution network interruptions according to claim 3, characterized in that, , , The weighting coefficients are assigned values ​​according to the fault severity classification system; When the fault is a single feeder fault, the affected load is the total power of the load that is lost due to the fault: In such cases, prioritize rapid recovery. , , The weights were set to 0.7, 0.15, and 0.15 respectively. When the main transformer or multiple feeder lines fail, affecting the load... A balance needs to be struck between recovery speed and economic efficiency. , , The weights were set to 0.5, 0.25, and 0.25 respectively. When the station power supply fails or there is a large-scale power outage, the load is affected. ; , , The weights were set to 0.4, 0.3, and 0.3 respectively.

5. The power flow optimization method for power distribution network interruptions according to claim 4, characterized in that, The execution control of the reconfiguration scheme in step S4 includes: sending switching operation commands to the remote terminal unit (RTU) through the power distribution automation SCADA system.

6. A power flow optimization device for power distribution network interruptions, used to perform the method according to any one of claims 1-5, characterized in that, The optimization device includes, in sequence, a data acquisition module, a power flow calculation module, an optimization solution module, a scheme execution module, and a monitoring and feedback module; The data acquisition module collects power distribution network operation data in real time through smart meter terminals and establishes network topology and load distribution models. The power flow calculation module calculates the power flow distribution and voltage level under the current network state based on the Newton-Raphson method; The optimization solution module uses an improved particle swarm optimization algorithm to solve the multi-objective optimization model and generate the optimal reconstruction scheme. The scheme execution module controls the switching equipment to perform network reconfiguration operations through the SCADA system; The monitoring and feedback module monitors the reconstruction effect in real time and provides system operation status assessment and parameter adjustment suggestions.

7. The power flow optimization device for power distribution network interruptions according to claim 6, characterized in that, The optimization solution module includes: an algorithm parameter management unit, a population evolution calculation unit, and a solution set evaluation unit; The algorithm parameter management unit stores and manages various parameter settings for the particle swarm algorithm, and supports adaptive parameter adjustment; The population evolution computing unit is based on a multi-objective optimization function and uses a parallel computing architecture to execute the particle swarm iteration process, supporting power distribution network optimization calculations for no less than 1,000 nodes. The solution set evaluation unit verifies the feasibility and evaluates the performance of the generated reconstruction scheme, and selects the Pareto optimal solution set that satisfies all constraints.

Images