County power distribution network partition autonomous optimization method and system based on dynamic partition strategy

Abstract

The invention discloses a county power distribution network partition autonomous optimization method and system based on a dynamic partition strategy. The method comprises the steps of collecting power distribution network data, evaluating the comprehensive adjustment capability of various adjustable resources of a power distribution network, and constructing a dynamic partition index system. And under the zoning framework of the county power distribution network, constructing an autonomous zoning optimization model with the minimization of the network loss as an optimization target. And solving the model by adopting an improved genetic algorithm to realize minimization of the loss of the county power distribution network. And a dynamic updating mechanism of the partition structure is designed, and continuous optimization of a partition result is realized. According to the county power distribution network partition autonomous optimization method and system based on the dynamic partition strategy, evaluation and optimization of the comprehensive adjustment capability are achieved, and the operation efficiency and stability of the power distribution network are remarkably improved. The method has the remarkable advantages in the aspects of dealing with voltage fluctuation, power fluctuation and resource adjusting capacity, and the problems which cannot be dealt with in the prior art are solved.

Description

Technical Field

[0001] The present invention relates to the field of smart grid management technology, and in particular to a method and system for autonomous optimization of county distribution network partitions based on a dynamic partitioning strategy. Background Art

[0002] In recent years, with the rapid popularization of distributed energy, especially photovoltaic power generation systems, the operation of power systems faces new challenges. After the distributed energy is connected to the distribution network, the volatility and uncertainty of the power grid have increased significantly, especially the intermittent and instability of photovoltaic power generation has put forward higher requirements for the voltage stability and power quality of the distribution network. In order to meet these challenges, the power grid management technology has been continuously developed, and an intelligent power grid management system including real-time data acquisition, dynamic partition management and comprehensive application of multiple adjustment means has gradually been formed. Although the existing technology has made significant progress in distribution network management, there are still some shortcomings. First, the traditional power grid management method underestimates the regulation capacity of distributed energy, especially the lack of flexibility in the regulation capacity of photovoltaic power generation, resulting in serious abandonment of light and affecting the effective utilization of new energy. Secondly, the existing energy storage system regulation strategy design is relatively rough, and the charging and discharging upper limit and capacity limit of the energy storage equipment are not fully considered, which may cause the system to operate in a non-optimal state and affect the regulation effect. In addition, the regulation potential of flexible loads has not been fully utilized, resulting in low overall efficiency of power grid regulation resources. Finally, when dealing with voltage stability issues, the existing technology usually adopts a fixed reactive compensation equipment configuration, lacks dynamic adjustment capabilities, and cannot adapt to rapidly changing power grid 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 the present invention is: In the current distribution network partitioning technology, clustering algorithms are widely used. However, this method has some limitations. First, it mainly ensures the weak structural coupling between partitions, but does not fully consider the coupling of energy within each partition. If the energy coupling between partitions is too high, this may mean that there is a significant difference between the power supply capacity and load demand of the partition. Secondly, the traditional partitioning method tends to statically divide the power grid, which may not adapt to the volatility brought about by the access of renewable energy, which may lead to insufficient power reserves under fixed partitioning.

[0005] To solve the above technical problems, the present invention provides the following technical solutions: a method for autonomous optimization of county distribution network zoning based on dynamic zoning strategy, comprising: collecting distribution network data, evaluating the comprehensive regulation capabilities of various adjustable resources of the distribution network, and building a dynamic zoning indicator system.

[0006] Under the partition framework of county distribution network, an autonomous partition optimization model is constructed with minimizing network loss as the optimization goal.

[0007] An improved genetic algorithm is used to solve the model to minimize the losses in the county distribution network.

[0008] Design a dynamic update mechanism for the partition structure to achieve continuous optimization of the partition results.

[0009] As a preferred scheme of the county distribution network partition autonomous optimization method based on dynamic partition strategy described in the present invention, wherein: the collection of distribution network data and the evaluation of the comprehensive regulation capability of various adjustable resources of the distribution network include, the adjustable resources include distributed photovoltaic power generation, energy storage equipment and flexible loads with regulation functions.

[0010] Distributed photovoltaics need to have regulatory flexibility, expressed as:

[0011]

[0012] Among them, α represents the proportion of photovoltaic power that can be abandoned. It represents the active power regulation capability of the i-th photovoltaic at time t. Represents the predicted value of distributed photovoltaics. Represents the expected value of distributed photovoltaic prediction error.

[0013] When designing the regulation strategy of the energy storage system, the upper limit of its charging and discharging power and the limitation of energy storage capacity are considered, which can be expressed as:

[0014]

[0015] in, represents the minimum capacity of the i-th energy storage unit, represents the maximum capacity of the i-th energy storage unit, represents the maximum charging power of the i-th energy storage, and Represents the maximum discharge power of the i-th energy storage.

[0016] The maximum adjustment capacity of the flexible load is expressed as:

[0017]

[0018] Among them, μ i,t Represents the response ratio of the i-th flexible load. Represents the power of the i-th flexible load at the current moment.

[0019] The adjustable capacity of the reactive power compensation device is expressed as:

[0020]

[0021] in, represents the upper limit of the supplementary capacity of the i-th reactive power compensation device, Represents the lower limit of the supplementary capacity of the i-th reactive compensation device.

[0022] As a preferred solution of the autonomous optimization method for county distribution network partition based on dynamic partition strategy described in the present invention, the construction of dynamic partition index system includes: the regional partition index system is determined by the division objectives and principles of the county distribution network. With the goal of reducing the peak value of voltage fluctuation and the fluctuation amplitude of net load, a comprehensive index system is constructed taking into account the functional needs of the region.

[0023] In the architecture design of the distribution network, the electrical connection between nodes is tight within the region, while the connection between regions is relatively loose. For the county distribution network, the quantitative index of electrical coupling is expressed as:

[0024]

[0025] Among them, e i,j It represents the weight of the edge between node i and node j, expressed by the electrical distance between the two nodes. Represents 1 / 2 of the sum of the weights of all edges in the region. Represents the sum of the weights of all edges connected to node i.

[0026] When node i and node j are located in the same region, δ(i,j)=1, otherwise δ(i,j)=0.

[0027] The voltage regulation capability of the distribution network is evaluated by calculating the weighted average of the regulation capabilities of each region, expressed as:

[0028]

[0029] in, Indicates the voltage regulation capability index of region n, where N represents the total number of divided regions. ΔV i represents the voltage deviation of the node i with the largest voltage deviation in the region, It represents the maximum voltage regulation of node i when considering the active and reactive power margin in the area.

[0030] The quantitative expression of the new energy consumption rate indicator is:

[0031]

[0032] in, represents the adjustable capacity of distributed photovoltaic in the region at time t, represents the adjustable capacity of energy storage in the region at time t, represents the adjustable capacity of the flexible load in the area at time t, Represents the net load in the area at time t.

[0033] As a preferred solution of the autonomous optimization method for county distribution network zoning based on dynamic zoning strategy described in the present invention, the construction of a dynamic zoning index system also includes combining various regional division indicators and considering the regulation capacity of source-load-storage in the region. The comprehensive zoning index system of the county distribution network is expressed as:

[0034]

[0035] Among them, γ1 represents the weight coefficient, γ2 represents the weight coefficient, and γ3 represents the weight coefficient.

[0036] As a preferred solution of the autonomous optimization method for county distribution network partition based on dynamic partition strategy described in the present invention, wherein: the autonomous partition optimization model with minimizing network loss as the optimization goal includes minimizing active network loss, and the objective function is expressed as:

[0037]

[0038] The constraints include power constraints of each distributed power source, distributed power source ramp constraints, and energy storage capacity constraints.

[0039] As a preferred solution of the autonomous optimization method of county distribution network partitioning based on dynamic partitioning strategy described in the present invention, the use of an improved genetic algorithm to solve the model includes first defining a fitness function to evaluate the quality of each individual in the population.

[0040] Initialize the population and generate the initial population.

[0041] Select individuals with excellent performance and apply the simulated annealing algorithm for local optimization. In the genetic algorithm, the random selection mechanism based on genetic factors is disabled to reduce randomness and improve the predictability of the algorithm. The simulated annealing algorithm is implemented on the selected excellent individuals for local search and fine-tuning, which is expressed as:

[0042]

[0043] Where P represents probability density. exp represents exponential function with natural number e as base. k represents Boltzmann constant. T represents temperature. E i represents the energy in the i-th state. j represents the energy in the jth state.

[0044] The fitness value of the new individuals generated by simulated annealing is calculated and merged with the previous generation population to form a new generation population. For the newly generated solution, if its fitness function value is lower than the best solution recorded previously, the probability acceptance function is used to decide whether to accept the new solution.

[0045] The temperature T in the simulated annealing algorithm is appropriately increased after each iteration to control the probability of accepting a degraded solution during the search process, expressed as:

[0046]

[0047] In the framework of genetic algorithm, a roulette wheel selection mechanism is used to determine which parent individuals will participate in crossover to form a new generation of individuals.

[0048] By comparing the absolute error between the solution variable and the objective function, we can determine whether the algorithm has converged, which can be expressed as:

[0049]

[0050] Where ε1 represents a very small constant, ε2 represents a very small constant, and ε3 represents a very small constant. When the solution variable and the objective function meet the convergence conditions at the same time, the algorithm converges.

[0051] As a preferred scheme of the autonomous optimization method of county distribution network zoning based on dynamic zoning strategy described in the present invention, wherein: the dynamic update mechanism of the designed zoning structure realizes the continuous optimization process of zoning results, including the construction of dynamic partitioning indicators, initial zoning setting, preliminary branch merging, zoning result output, autonomous optimization model establishment and optimization model solution.

[0052] The construction of dynamic partitioning indicators includes comprehensive consideration of the regulation capabilities of photovoltaic power generation, energy storage systems, flexible loads and reactive compensation equipment, and the construction of a dynamic partitioning indicator system including electrical coupling degree, voltage safety margin and new energy consumption rate.

[0053] The initial partition setting includes that in the county distribution network, each independent node is set as an initial independent sub-partition.

[0054] The preliminary branch merging includes performing preliminary merging operations on the branches in the distribution network to form larger partition units.

[0055] The zoning result output includes using the above-mentioned dynamic division indicators to calculate and output the zoning results of the county distribution network.

[0056] The establishment of the autonomous optimization model includes minimizing the active network loss as the goal and establishing a regional autonomous optimization model for the county distribution network.

[0057] The optimization model solution includes solving the autonomous optimization model using an improved genetic algorithm. The partitions are adjusted based on the comparison results with the optimization indicators obtained in the previous solution.

[0058] A county distribution network partition autonomous optimization system based on dynamic partition strategy, characterized by: including:

[0059] Build an indicator system module, collect distribution network data, evaluate the comprehensive regulation capabilities of various adjustable resources in the distribution network, and build a dynamic zoning indicator system.

[0060] Construct a model module and, under the partition framework of the county distribution network, build an autonomous partition optimization model with minimizing network losses as the optimization goal.

[0061] The solution module uses an improved genetic algorithm to solve the model and minimize the losses of the county distribution network.

[0062] Optimize the module and design a dynamic update mechanism for the partition structure to achieve continuous optimization of the partition results.

[0063] A computer device comprises a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of the above method when executing the computer program.

[0064] A computer-readable storage medium stores a computer program, which implements the steps of the method described above when executed by a processor.

[0065] The beneficial effects of the present invention are as follows: by evaluating the comprehensive regulation capabilities of various adjustable resources in the distribution network, a dynamic zoning index system is constructed to achieve the autonomous management capabilities of each region in different operating periods. In this way, voltage fluctuations can be effectively suppressed and power fluctuations can be smoothed to ensure the stability of the power grid operation. The regulation capability of distributed photovoltaic power generation is evaluated in detail, and the prediction error and regulation elasticity of its output are considered. When designing the regulation strategy, the prediction error of photovoltaic power generation is taken into consideration to ensure that the voltage is controlled within a safe range, reduce the phenomenon of abandoned light, and improve the utilization efficiency of new energy. When designing the regulation strategy of the energy storage system, the upper limit and capacity limit of the charging and discharging power are fully considered to achieve two-way regulation of the active power of the power grid. This not only ensures the safety and effectiveness of the system operation, but also improves the regulation efficiency of the energy storage system. The maximum regulation capability of the flexible load is analyzed, its regulation strategy is optimized, and the overall efficiency of the power grid regulation resources is improved. By making full use of the regulation capability of the flexible load, the power fluctuations of the power grid can be significantly smoothed and the operation stability of the system can be improved. Since the distributed photovoltaic access may cause voltage stability problems, the present invention is equipped with reactive compensation equipment with adjustable capacity in some distribution network areas. These devices can dynamically adjust according to real-time grid conditions, optimize regional voltage levels, and ensure stable operation of the grid. BRIEF DESCRIPTION OF THE DRAWINGS

[0066] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for describing the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work. Among them:

[0067] Figure 1 An overall flow chart of a method and system for autonomous optimization of county distribution network partitioning based on a dynamic partitioning strategy provided in the first embodiment of the present invention. DETAILED DESCRIPTION

[0068] In order to make the above-mentioned purposes, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the drawings of the specification. Obviously, the described embodiments are part of the embodiments of the present invention, but not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary persons in the art without creative work should fall within the scope of protection of the present invention.

[0069] Example 1, reference Figure 1 , is an embodiment of the present invention, and provides a method for autonomous optimization of county distribution network partitions based on a dynamic partition strategy, comprising:

[0070] S1: Collect distribution network data, evaluate the comprehensive regulation capabilities of various adjustable resources in the distribution network, and build a dynamic zoning indicator system.

[0071] It should be noted that each region has the ability to manage itself in different operating periods. This involves the ability to suppress voltage fluctuations and smooth power fluctuations. This step will first conduct a detailed assessment of the adjustable power resources in the region, including distributed photovoltaic power generation, energy storage equipment, and flexible loads with regulation functions.

[0072] Furthermore, regarding the regulation capacity of photovoltaic power generation, although in principle it is not used as the main means of regulation to reduce the phenomenon of abandoned light, considering the possible shortage of other regulation resources, distributed photovoltaic power generation still needs to have a certain degree of regulation flexibility when necessary. In addition, since there is a prediction error in the output of distributed photovoltaic power generation, this uncertainty needs to be taken into account when designing the regulation strategy to ensure that the voltage is controlled within a safe range.

[0073] Distributed photovoltaics need to have regulatory flexibility, expressed as:

[0074]

[0075] Among them, α represents the proportion of photovoltaic power that can be abandoned. It represents the active power regulation capability of the i-th photovoltaic at time t. Represents the predicted value of distributed photovoltaics. Represents the expected value of distributed photovoltaic prediction error.

[0076] It should be noted that energy storage devices can achieve bidirectional regulation of the active power of the power grid by performing charging and discharging operations. However, when designing the regulation strategy of the energy storage system, the upper limit of its charging and discharging power and the limitation of energy storage capacity must be fully considered to ensure that the system operates within a safe and effective range.

[0077] When designing the regulation strategy of the energy storage system, the upper limit of its charging and discharging power and the limitation of energy storage capacity are considered, which can be expressed as:

[0078]

[0079] in, represents the minimum capacity of the i-th energy storage unit, represents the maximum capacity of the i-th energy storage unit, represents the maximum charging power of the i-th energy storage, and Represents the maximum discharge power of the i-th energy storage.

[0080] The maximum adjustment capacity of the flexible load is expressed as:

[0081]

[0082] Among them, μ i,t Represents the response ratio of the i-th flexible load. Represents the power of the i-th flexible load at the current moment.

[0083] It should be noted that, in view of the voltage stability problems that may be caused by the access of distributed photovoltaic systems to the distribution network, some distribution network substations have been equipped with reactive power compensation equipment to optimize the voltage level in these areas. These reactive power compensation devices have adjustable capacity, and the adjustable capacity of the reactive power compensation device is shown in the following formula:

[0084] The adjustable capacity of the reactive power compensation device is expressed as:

[0085]

[0086] in, represents the upper limit of the supplementary capacity of the i-th reactive power compensation device, Represents the lower limit of the supplementary capacity of the i-th reactive compensation device.

[0087] The regional division index system is determined by the division objectives and principles of the county distribution network. With the goal of reducing the peak value of voltage fluctuation and the fluctuation range of net load, a comprehensive index system is constructed taking into account the functional needs of the region.

[0088] Furthermore, the regional division index system is mainly determined by the division objectives and principles of the county distribution network. This patent aims to reduce the peak value of voltage fluctuations and reduce the fluctuation amplitude of net load. On this basis, while taking into account the functional needs of the region, a comprehensive index system is constructed. This system aims to comprehensively evaluate and optimize the regional division of the distribution network, including the electrical coupling degree that measures the electrical distance between different nodes, the voltage safety margin that measures the overall deviation of the node voltage when the distribution network participates in the optimization and regulation, and the new energy consumption rate index that takes into account the new energy consumption rate index to measure the self-regulation ability of each zone under the zoning situation, thereby maximizing the regional autonomy of the distribution network.

[0089] In the architecture design of the distribution network, the electrical connection between nodes is tight within the region, while the connection between regions is relatively loose. For the county distribution network, the quantitative index of electrical coupling is expressed as:

[0090]

[0091] Among them, e i,j It represents the weight of the edge between node i and node j, expressed by the electrical distance between the two nodes. Represents 1 / 2 of the sum of the weights of all edges in the region. Represents the sum of the weights of all edges connected to node i.

[0092] When node i and node j are located in the same region, δ(i,j)=1, otherwise δ(i,j)=0.

[0093] It should be noted that in county distribution networks, a high proportion of distributed resource access may cause the voltage to exceed the normal operating range, which is a key issue. To meet this challenge, the voltage safety margin indicator is defined as the ability of the adjustable active and reactive power of photovoltaic, energy storage and load resources in the region to control voltage deviations. Ideally, each region should use the regulation capacity of its internal resources to correct voltage over-limit problems while reducing power transmission losses between regions. The voltage regulation capability of the distribution network can be evaluated by calculating the weighted average of the regulation capabilities of each region, expressed as:

[0094]

[0095] in, Indicates the voltage regulation capability index of region n, where N represents the total number of divided regions. ΔV irepresents the voltage deviation of the node i with the largest voltage deviation in the region, It represents the maximum voltage regulation of node i when considering the active and reactive power margin in the area.

[0096] It should be noted that in the county distribution network, due to the uncertainty and volatility of photovoltaic power generation and load demand, the fluctuation range of net load increases, which poses a challenge to the stability of the power grid. The new energy consumption rate indicator is defined as the adjustable power capacity of various regulating resources in the region to deal with net load fluctuations. This indicator reflects the response and regulation effect of resources in the region to fluctuations.

[0097] Since regional division based only on a specific time point may not ensure the best power fluctuation control effect within the dispatch cycle, the new energy consumption rate indicator should consider the ability of adjustable resources in the region to smooth net load fluctuations within the dispatch cycle. This method helps to evaluate and optimize the new energy consumption and load regulation capabilities of the distribution network in different time periods. The quantitative expression of this indicator is as follows:

[0098]

[0099]

[0100] in, represents the adjustable capacity of distributed photovoltaic in the region at time t, represents the adjustable capacity of energy storage in the region at time t, represents the adjustable capacity of the flexible load in the area at time t, Represents the net load in the area at time t.

[0101] Combining the indicators of each region division and considering the regulation capacity of source-load-storage in the region, the comprehensive zoning indicator system of county distribution network is expressed as:

[0102]

[0103] Among them, γ1 represents the weight coefficient, γ2 represents the weight coefficient, and γ3 represents the weight coefficient.

[0104] S2: Under the partition framework of the county distribution network, an autonomous partition optimization model is constructed with the minimization of network loss as the optimization goal.

[0105] It should be noted that this patent proposes a data-driven dynamic partitioning technology for county distribution networks. The method first determines the preliminary partitioning between loads and compensation points based on the grid structure and impedance distance, then calculates the required reactive reserve capacity in each area through real-time measurement data, and performs local optimization of the partitioning accordingly. Finally, the final partitioning scheme is formed by applying the partition agglomeration algorithm.

[0106] Assume that the distribution network contains N load nodes, g photovoltaic grid-connected points, and the target number of partitions is g0. The partitioning process based on local measurement information can be summarized as follows:

[0107] Data input and structure construction: Input the relevant data of the distribution network and establish the node association matrix. The local measurement information mainly includes the voltage and power data of the photovoltaic grid-connected point, and the load power data obtained by the smart meter. The grid structure is abstracted into an undirected graph model using graph theory methods, and the impedance value of the branch is used as the weight of the edge in the graph.

[0108] Initial partitioning: Following the principle of power supply to the nearest load, the electrical distance between the photovoltaic grid-connected point and the load node is measured according to the mutual impedance parameter between the nodes, and the N load nodes are allocated to the g photovoltaic grid-connected points with the closest electrical distance to form g0 preliminary partitions.

[0109] Verification and adjustment of reactive power reserve capacity: Verify the reactive power reserve capacity in each preliminary zone and make local adjustments as needed. Merge isolated PV nodes into adjacent areas with insufficient reactive power reserves to optimize the distribution of reactive power.

[0110] Execute the partitioning algorithm to form the final partition. The partitioning algorithm takes the maximum WARD distance between regions and the minimum WARD distance within a region as the goal, and adjusts the g initial partitions to g0 final partitions. During the execution of the partitioning agglomeration algorithm, C i The internal distance of the region is denoted as d i , Region C i , C j The distance between them is denoted as l ij , calculated by the following formula:

[0111]

[0112] Among them, x rs represents the electrical distance between photovoltaic nodes r and s. If r = s, then x rs =0; C i , C j Corresponding to the sub-regions numbered i and j. Assume that the regions are aggregated into C i , C j The WARD distance inside the new area is denoted as d q , the WARD distance between q and other regions p is recorded as l dp , then d q and l dp Calculated by the following formula:

[0113] d q =d i +d j +l ij

[0114] l dp = l ip +l jp

[0115] In a county-level distribution network environment with a well-developed communication infrastructure, the power data of photovoltaic power generation and loads can be acquired and exchanged in real time. Using a data-driven approach, the distribution network can achieve rapid adaptive dynamic partitioning to optimize the operation and management of the power grid. This approach ensures that when the grid conditions change, the dynamic partitioning strategy can be adjusted in time to adapt to new operating requirements.

[0116] The goal is to minimize the active network loss, and the objective function is expressed as:

[0117]

[0118] The constraints include power constraints of each distributed power source, distributed power source ramp constraints, and energy storage capacity constraints.

[0119] The power constraints of each distributed generation are expressed as:

[0120] P i,min ≤P i ≤P i,max

[0121] Among them, P i P represents the power output of the i-th distributed generation. i,min P represents the minimum power output of the i-th distributed generation. i,max Represents the maximum power output of the i-th distributed generation.

[0122] The distributed generation ramp constraint is expressed as:

[0123] R i,down Δt≤P i,t -P i,t-1 ≤R i,up Δt

[0124] Among them, P i,t , P i,t-1 Respectively represent the output power of the i-th distributed generation at time t and time t-1. i,down Indicates the lower limit of output active power. i,up It is the upper limit of output active power.

[0125] The energy storage capacity constraint is expressed as:

[0126] SOC min ≤SOC≤SOC max

[0127] SOC start =SOCend

[0128] Among them, SOC represents the energy storage charge state SOC min It is the minimum state of charge for energy storage. max Indicates the maximum state of charge of energy storage. SOC start Indicates the initial state of charge of energy storage. SOC end Indicates the energy storage termination charging state.

[0129] S3: An improved genetic algorithm is used to solve the model to minimize the losses in the county distribution network.

[0130] It should be noted that intelligent optimization algorithms usually use random variations of individuals in a population to introduce mutation mechanisms, which helps to improve the algorithm's exploration capabilities during the global search process. In this patent, the genetic diversity of the population is increased by introducing new random individuals in each generation, and local search is performed using simulated annealing technology to improve the algorithm's solution efficiency. The specific steps are summarized as follows:

[0131] First, define a fitness function to evaluate the quality of each individual in the population.

[0132] Initialize the population and generate the initial population.

[0133] Select individuals with excellent performance and apply the simulated annealing algorithm for local optimization. In the genetic algorithm, the random selection mechanism based on genetic factors is disabled to reduce randomness and improve the predictability of the algorithm. The simulated annealing algorithm is implemented on the selected excellent individuals for local search and fine-tuning, which is expressed as:

[0134]

[0135] Where P represents probability density. exp represents exponential function with natural number e as base. k represents Boltzmann constant. T represents temperature. E i represents the energy in the i-th state. j represents the energy in the jth state.

[0136] The fitness value of the new individuals generated by simulated annealing is calculated and merged with the previous generation population to form a new generation population. For the newly generated solution, if its fitness function value is lower than the best solution recorded previously, the probability acceptance function is used to decide whether to accept the new solution.

[0137] The temperature T in the simulated annealing algorithm is appropriately increased after each iteration to control the probability of accepting a degraded solution during the search process, expressed as:

[0138]

[0139] In the framework of genetic algorithm, a roulette wheel selection mechanism is used to determine which parent individuals will participate in crossover to form a new generation of individuals.

[0140] By comparing the absolute error between the solution variable and the objective function, we can determine whether the algorithm has converged, which can be expressed as:

[0141]

[0142] Where ε1 represents a very small constant, ε2 represents a very small constant, and ε3 represents a very small constant. When the solution variable and the objective function meet the convergence conditions at the same time, the algorithm converges.

[0143] S4: Design a dynamic update mechanism for the partition structure to achieve continuous optimization of the partition results.

[0144] The dynamic update mechanism of the partition structure is designed to achieve continuous optimization of the partition results. The process includes the construction of dynamic partition indicators, initial partition setting, preliminary branch merging, partition result output, autonomous optimization model establishment and optimization model solution.

[0145] The construction of dynamic partitioning indicators includes comprehensive consideration of the regulation capabilities of photovoltaic power generation, energy storage systems, flexible loads and reactive compensation equipment, and the construction of a dynamic partitioning indicator system including electrical coupling degree, voltage safety margin and new energy consumption rate.

[0146] The initial partition setting includes that in the county distribution network, each independent node is set as an initial independent sub-partition.

[0147] The preliminary branch merging includes performing preliminary merging operations on the branches in the distribution network to form larger partition units.

[0148] The zoning result output includes using the above-mentioned dynamic division indicators to calculate and output the zoning results of the county distribution network.

[0149] The establishment of the autonomous optimization model includes minimizing the active network loss as the goal and establishing a regional autonomous optimization model for the county distribution network.

[0150] The optimization model solution includes solving the autonomous optimization model using an improved genetic algorithm. The partitions are adjusted based on the comparison results with the optimization indicators obtained in the previous solution.

[0151] The above embodiments also include a county distribution network partition autonomous optimization system based on a dynamic partition strategy, specifically:

[0152] Build an indicator system module, collect distribution network data, evaluate the comprehensive regulation capabilities of various adjustable resources in the distribution network, and build a dynamic zoning indicator system.

[0153] Construct a model module and, under the partition framework of the county distribution network, build an autonomous partition optimization model with minimizing network losses as the optimization goal.

[0154] The solution module uses an improved genetic algorithm to solve the model and minimize the losses of the county distribution network.

[0155] Optimize the module and design a dynamic update mechanism for the partition structure to achieve continuous optimization of the partition results.

[0156] The computer device may be a server. The computer device includes a processor, a memory, an input / output interface (I / O for short) and a communication interface. The processor, the memory and the input / output interface are connected via a system bus, and the communication interface is connected to the system bus via the input / output interface. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store data cluster data of a power monitoring system. The input / output interface of the computer device is used to exchange information between the processor and an external device. The communication interface of the computer device is used to communicate with an external terminal via a network connection. When the computer program is executed by the processor, a method is implemented.

[0157] A person of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment method can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to the memory, database or other medium used in the embodiments provided in the present application can include at least one of non-volatile and volatile memory. Non-volatile memory may include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory may include random access memory (RAM) or external cache memory, etc. As an illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM). The database involved in each embodiment provided in this application may include at least one of a relational database and a non-relational database. Non-relational databases may include distributed databases based on blockchains, etc., but are not limited to this. The processor involved in each embodiment provided in this application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, etc., but are not limited to this.

[0158] Example 2, an embodiment of the present invention, provides a method and system for autonomous optimization of county distribution network partitions based on a dynamic partitioning strategy. In order to verify the beneficial effects of the present invention, scientific demonstration is carried out through simulation experiments.

[0159] In order to verify the effectiveness of the distribution network optimization method based on dynamic partitioning proposed in the present invention, a series of experiments and data analysis were carried out. This embodiment aims to evaluate and optimize the operation effect of the county distribution network through dynamic partitioning technology, focusing on suppressing voltage fluctuations, smoothing power fluctuations and improving the ability to absorb new energy.

[0160] First, the experiment collected actual data from a county distribution network, including the operation data of distributed photovoltaic power generation, energy storage equipment, flexible loads and reactive power compensation equipment. In order to ensure the authenticity and accuracy of the experimental data, high-precision power monitoring equipment was used in the data collection process, covering the main power nodes and load points in the area.

[0161] During the data evaluation process, the adjustable power resources in each area were evaluated in detail, including:

[0162] Distributed photovoltaic power generation: Although the main function of photovoltaic power generation is not regulation, it still needs to have a certain degree of regulation flexibility when other resources are insufficient. The regulation capability of photovoltaic power generation takes into account the forecast error of its output to ensure that the voltage is controlled within a safe range.

[0163] Energy storage equipment: bidirectional regulation of the active power of the power grid is achieved through charging and discharging operations. When designing the regulation strategy of the energy storage system, the upper limit of its charging and discharging power and the limitation of energy storage capacity should be fully considered.

[0164] Flexible load: The response ratio of the flexible load and the power at the current moment are evaluated to determine its maximum regulation capacity.

[0165] Reactive power compensation equipment: Due to the voltage stability issues that may be caused by the access of distributed photovoltaic systems to the distribution network, some substations are equipped with reactive power compensation equipment to optimize the voltage level. The regulation capacity of reactive power compensation equipment is fully considered in its capacity range when it is designed.

[0166] Furthermore, a regional division index system was constructed, with the goal of reducing the peak value of voltage fluctuations and the fluctuation amplitude of net load, covering indicators such as electrical coupling, voltage safety margin and new energy consumption rate. This index system evaluates the electrical distance between different nodes, measures the overall deviation of node voltage when the distribution network participates in optimization and regulation, and the autonomous management capabilities of each zone under the zone situation.

[0167] During the experiment, an improved genetic algorithm was used to solve the autonomous partition optimization model, with the goal of minimizing the distribution network loss. The intelligent optimization algorithm introduces a mutation mechanism through random mutation of individuals in the population to improve the exploration ability in the global search process, and performs local search through simulated annealing technology to optimize the algorithm's solution efficiency.

[0168] In the dynamic partition design of the experiment, the calculation of reactive power reserve capacity in the area based on real-time measurement data was considered, and the final partition scheme was formed through the partition aggregation algorithm. The whole experiment is divided into the following steps:

[0169] Data input and structure construction: Establish a node association matrix and use graph theory methods to abstract the power grid structure into an undirected graph model.

[0170] Initial partitioning: Based on the mutual impedance parameters between nodes, the load nodes are assigned to the photovoltaic grid-connected points with the closest electrical distance to form preliminary partitions.

[0171] Verification and adjustment of reactive power reserve capacity: Verify the reactive power reserve capacity within the preliminary zone and make local adjustments as needed.

[0172] Execute the partitioning algorithm: take the minimum WARD distance between regions as the goal, and finally form an optimized partitioning result.

[0173] Dynamic update of partition structure: A dynamic update mechanism for partition structure is designed to achieve continuous optimization of partition results. The experimental results are shown in Table 1.

[0174] Table 1 Experimental results

[0175]

[0176] It can be seen from the table data that photovoltaic power generation, energy storage equipment and flexible loads have certain adjustment capabilities in different operating periods. The comprehensive evaluation of various adjustment resources in the present invention shows that compared with the prior art, our invention has significant innovation and advantages in the following aspects:

[0177] Photovoltaic power generation regulation flexibility: By considering the prediction error, photovoltaic power generation has a certain regulation capability. For example, photovoltaic power generation 1 and photovoltaic power generation 2 can be adjusted to 270kW and 230kW respectively when necessary, which can effectively smooth out power fluctuations compared to the current power of 300kW and 250kW.

[0178] Charge and discharge regulation of energy storage devices: Energy storage devices 1 and 2 have regulation capacities of 480kWh and 430kWh respectively. Compared with their maximum capacities of 500kWh and 450kWh, they fully utilize the regulation space of the energy storage devices to ensure that the system operates within a safe and effective range.

[0179] Responsiveness of flexible loads: Flexible load 1 and flexible load 2 can respond to powers of 190kW and 170kW respectively during adjustment, which demonstrates their strong adjustment flexibility compared to the current powers of 200kW and 180kW.

[0180] Optimization effect of reactive power compensation: The adjustment range of reactive power compensation equipment is from the upper limit of 50kW (photovoltaic power generation 1) to the lower limit of 10kW, which effectively optimizes the regional voltage level.

[0181] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A method for autonomous optimization of county distribution network partitions based on dynamic partition strategy, characterized in that: include: Collect distribution network data, evaluate the comprehensive regulation capabilities of various adjustable resources in the distribution network, and build a dynamic zoning indicator system; Under the partition framework of county distribution network, an autonomous partition optimization model is constructed with the goal of minimizing network loss. An improved genetic algorithm is used to solve the model to minimize the losses of the county distribution network. Design a dynamic update mechanism for the partition structure to achieve continuous optimization of the partition results.

2. The autonomous optimization method for county distribution network partitioning based on dynamic partitioning strategy according to claim 1 is characterized in that: The collecting of distribution network data and the evaluation of the comprehensive regulation capability of various adjustable resources of the distribution network include, the adjustable resources include distributed photovoltaic power generation, energy storage equipment and flexible loads with regulation function; Distributed photovoltaics need to have regulatory flexibility, expressed as: Among them, α represents the proportion of photovoltaic power abandonment allowed; It represents the active power regulation capability of the i-th photovoltaic at time t; represents the predicted value of distributed photovoltaics; represents the expected value of distributed photovoltaic prediction error; When designing the regulation strategy of the energy storage system, the upper limit of its charging and discharging power and the limitation of energy storage capacity are considered, which can be expressed as: in, represents the minimum capacity of the i-th energy storage unit, represents the maximum capacity of the i-th energy storage unit, represents the maximum charging power of the i-th energy storage, and represents the maximum discharge power of the i-th energy storage; The maximum adjustment capacity of the flexible load is expressed as: Among them, μ i,t represents the response ratio of the i-th flexible load; represents the power of the i-th flexible load at the current moment; The adjustable capacity of the reactive power compensation device is expressed as: in, represents the upper limit of the supplementary capacity of the i-th reactive power compensation device, Represents the lower limit of the supplementary capacity of the i-th reactive compensation device.

3. The autonomous optimization method for county distribution network partitioning based on dynamic partitioning strategy according to claim 2 is characterized in that: The construction of a dynamic zoning index system includes: the regional division index system is determined by the division objectives and principles of the county distribution network; With the goal of reducing the peak value of voltage fluctuation and the fluctuation range of net load, a comprehensive indicator system is constructed taking into account the functional needs of the region; In the architecture design of the distribution network, the electrical connection between nodes is tight within the region, while the connection between regions is relatively loose. For the county distribution network, the quantitative index of electrical coupling is expressed as: Among them, e i,j represents the weight of the edge between node i and node j, expressed as the electrical distance between the two nodes; Represents 1 / 2 of the sum of the weights of all edges in the region; represents the sum of the weights of all edges connected to node i; When nodes i and j are in the same region, δ(i,j) = 1, otherwise δ(i,j) = 0; The voltage regulation capability of the distribution network is evaluated by calculating the weighted average of the regulation capabilities of each region, expressed as: in, Indicates the voltage regulation capability index of region n, where N represents the total number of divided regions; ΔV i represents the voltage deviation of the node i with the largest voltage deviation in the region, It represents the maximum voltage regulation of node i when considering the active and reactive power margin in the area; The quantitative expression of the new energy consumption rate indicator is: in, represents the adjustable capacity of distributed photovoltaic in the region at time t, represents the adjustable capacity of energy storage in the region at time t, represents the adjustable capacity of the flexible load in the area at time t, Represents the net load in the area at time t.

4. The autonomous optimization method for county distribution network partitioning based on dynamic partitioning strategy as claimed in claim 3 is characterized by: The construction of a dynamic zoning index system also includes combining the various regional division indicators and considering the source-load-storage regulation capacity in the region. The county distribution network comprehensive zoning index system is expressed as: Among them, γ1 represents the weight coefficient, γ2 represents the weight coefficient, and γ3 represents the weight coefficient.

5. The autonomous optimization method for county distribution network partitioning based on dynamic partitioning strategy according to claim 4 is characterized in that: The autonomous partition optimization model with minimizing network loss as the optimization goal is constructed, wherein the goal is to minimize active network loss, and the objective function is expressed as: The constraints include power constraints of each distributed power source, distributed power source ramp constraints, and energy storage capacity constraints.

6. The autonomous optimization method for county distribution network partitioning based on dynamic partitioning strategy according to claim 5 is characterized in that: The method of solving the model by using an improved genetic algorithm includes first defining a fitness function to evaluate the quality of each individual in the population; Initialize the population and generate the initial population; Select individuals with excellent performance and apply the simulated annealing algorithm for local optimization; in the genetic algorithm, disable the random selection mechanism based on genetic factors to reduce randomness and improve the predictability of the algorithm; implement the simulated annealing algorithm on the selected excellent individuals for local search and fine-tuning, which is expressed as: Where P represents probability density; exp represents exponential function with natural number e as base; k represents Boltzmann constant; T represents temperature; E i represents the energy in the i-th state; E j represents the energy in the jth state; The fitness value of the new individuals generated based on simulated annealing is calculated and merged with the previous generation population to form a new generation population; for the newly generated solution, if its fitness function value is lower than the best solution recorded previously, the probability acceptance function is used to decide whether to accept this new solution; The temperature T in the simulated annealing algorithm is appropriately increased after each iteration to control the probability of accepting a degraded solution during the search process, expressed as: In the framework of genetic algorithm, a roulette wheel selection mechanism is used to determine which parent individuals will participate in crossover to form a new generation of individuals; By comparing the absolute error between the solution variable and the objective function, we can determine whether the algorithm has converged, which can be expressed as: Among them, ε1 represents a very small constant, ε2 represents a very small constant, and ε3 represents a very small constant; when the solution variable and the objective function meet the convergence conditions at the same time, the algorithm converges.

7. The autonomous optimization method for county distribution network partitioning based on dynamic partitioning strategy according to claim 6 is characterized in that: The dynamic update mechanism of the designed partition structure realizes the continuous optimization process of the partition results, including the construction of dynamic partition indicators, initial partition setting, preliminary branch merging, partition result output, autonomous optimization model establishment and optimization model solution; The construction of dynamic partitioning indicators includes comprehensive consideration of the regulation capabilities of photovoltaic power generation, energy storage systems, flexible loads, and reactive power compensation equipment, and the construction of a dynamic partitioning indicator system including electrical coupling, voltage safety margin, and new energy consumption rate; The initial partition setting includes that in the county distribution network, each independent node is set as an initial independent sub-partition; The preliminary branch merging includes performing preliminary merging operations on the branches in the distribution network to form a larger partition unit; The partition result output includes calculating and outputting the partition result of the county distribution network by using the above dynamic partition index; The establishment of autonomous optimization model includes minimizing active network loss as the goal, and establishing a district autonomous optimization model of county distribution network; Solving the optimization model includes solving the autonomous optimization model using an improved genetic algorithm; The partitions are adjusted based on the comparison results with the optimization indicators obtained in the previous solution.

8. A county distribution network partition autonomous optimization system based on dynamic partition strategy using the method according to any one of claims 1 to 7, characterized in that: Construct an indicator system module, collect distribution network data, evaluate the comprehensive regulation capabilities of various adjustable resources in the distribution network, and build a dynamic zoning indicator system; Construct a model module, and build an autonomous partition optimization model with minimizing network loss as the optimization goal under the partition framework of the county distribution network; The solution module uses an improved genetic algorithm to solve the model and minimize the losses of the county distribution network; Optimize the module and design a dynamic update mechanism for the partition structure to achieve continuous optimization of the partition results.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images