Distributed photovoltaic energy storage system optimal configuration method and system

Abstract

The invention relates to a distributed photovoltaic energy storage system optimization configuration method and system, the system is divided into a plurality of clusters, and comprehensive performance indexes of the clusters are established by using the modularity and the active balance degree of the clusters; optimizing the comprehensive performance indexes by adopting an improved genetic algorithm to determine a cluster division result; a double-layer model is established, an outer layer model takes the clusters as minimum units, and optimal configuration is carried out on cluster access photovoltaic total capacity, energy storage total capacity and energy storage time sequence output by taking the minimum annual comprehensive cost of carbon emission of the power distribution network as a target according to load time sequence output in each cluster; the inner-layer model takes each node in the cluster as a minimum unit, and according to an optimization configuration result of the outer-layer model, the photovoltaic access capacity of each node and the access position of energy storage are optimized and configured by taking the minimum deviation value of a node voltage comprehensive expected value in the cluster and the minimum comprehensive operation risk index as targets. According to the method, the stability, the economy and the new energy utilization rate of the system are considered, and the optimal configuration of the distributed photovoltaic energy storage system in the optimal state is comprehensively achieved.

Description

Technical Field

[0001] The present invention belongs to the technical field of distributed energy storage systems, and in particular relates to a distributed photovoltaic energy storage system optimization configuration method and system. Background Art

[0002] The randomness and intermittency of photovoltaic power generation lead to large fluctuations in output power, which poses a serious challenge to the safe and stable operation of the power grid. At the same time, as the peak-to-valley difference of the demand-side load increases, the problem of power supply at peak load moments becomes more prominent. Simply increasing the backup capacity of generators is not only expensive but also has low equipment resource utilization. Energy storage has the characteristics of rapid power control and flexible energy throughput, and is currently one of the effective means to solve the problem of photovoltaic grid connection and consumption.

[0003] In the existing technology, distributed photovoltaic power generation systems are usually connected to medium and low voltage distribution networks. The entire microgrid system may contain multiple distributed power sources (such as photovoltaic power generation and thermal power, etc.), and the load locations and power consumption levels are also different. The location and power of the energy storage system connected to the microgrid as a bidirectional power element not only changes the power flow distribution of the power grid, but also affects the voltage level, network loss and safety stability of the power grid. The photovoltaic access regional energy storage optimization planning method based on the operation strategy, according to the actual operation requirements of the energy storage system for the regional network with high penetration rate access, reasonably selects the energy storage operation strategy to match the operation function of energy storage under high penetration rate photovoltaic access, accurately calculates the real-time charging and discharging power of the energy storage system according to the energy storage operation strategy of the high penetration rate photovoltaic access region, comprehensively and multi-angle evaluates the investment cost of the energy storage system, the low storage and high generation operation benefits of the energy storage system, the benefits of the energy storage system to improve regional network losses, the benefits of regional network load power supply reliability, the benefits of energy storage to delay regional network transformation and upgrading, the environmental benefits of energy storage, the benefits of energy storage to promote photovoltaic consumption, and the benefits of energy storage to improve voltage fluctuations, etc., and reasonably selects different energy storage operation strategies to effectively avoid redundant investment and construction capacity of energy storage systems. However, when implementing the optimization configuration, the existing technology mainly considers the optimal performance of the two aspects of system economy and new energy utilization rate, but does not consider the optimal configuration of the distributed photovoltaic energy storage system in the optimal state of system stability, economy and new energy utilization rate. Summary of the invention

[0004] In order to solve the deficiencies in the prior art, the present invention provides a distributed photovoltaic energy storage system optimization configuration method and system, which comprehensively considers the three aspects of system stability, economy and new energy utilization rate to achieve the optimal configuration of the distributed photovoltaic energy storage system in the optimal state, which is beneficial to the safe and stable operation of the power grid.

[0005] The present invention adopts the following technical solution.

[0006] The present invention proposes a distributed photovoltaic energy storage system optimization configuration method, the distributed photovoltaic energy storage system includes a distributed photovoltaic power generation system, an energy storage system and a distribution network; including: The distributed photovoltaic energy storage system is divided into multiple clusters, and the cluster's modularity and active power balance are used to establish the comprehensive performance index of the cluster; the improved genetic algorithm is used to iteratively optimize the comprehensive performance index to determine the cluster division result corresponding to the optimal value of the comprehensive performance index; According to the cluster division results, a two-layer model for the optimal configuration of distributed photovoltaic energy storage systems is established. The outer model takes the cluster as the smallest unit, optimizes the total photovoltaic capacity, total energy storage capacity and the sequential output of energy storage connected to the cluster according to the sequential output of the loads in each cluster and takes the annual comprehensive cost of carbon emissions of the distribution network as the minimum. The inner model takes each node in the cluster as the smallest unit, and according to the optimization configuration results of the outer model, optimizes the access capacity of photovoltaic power at each node and the access location of energy storage with the goal of minimizing the deviation of the comprehensive expected value of node voltage in the cluster and minimizing the comprehensive operation risk index.

[0007] Preferably, the modularity of each cluster in the distributed photovoltaic energy storage system satisfies the following relationship: , In the formula, For Node i For Node j The voltage-active power sensitivity coefficient, ; , , They are active distance, reactive distance and comprehensive electrical distance within the cluster respectively; t is the weight coefficient; A ij To connect nodes i ,node j The edge weight of the node i and nodes j It takes 1 when directly connected, otherwise it takes 0; is the sum of all edges in the network, satisfying ; For Node i degree, satisfy , For Node j degree, satisfy ; When the node i ,node j In the same cluster , otherwise it takes 0, is the modularity.

[0008] Preferably, the active power balance of each cluster in the distributed photovoltaic energy storage system satisfies the following relationship: , In the formula, is the active balance, m is the total number of clusters; P q,clu Inflow cluster for network q The net power of P q,load,i For cluster q Internal Node i Active load power.

[0009] Preferably, the comprehensive performance index of the cluster is established by the following relationship, satisfying the following relationship: , In the formula, As a comprehensive performance indicator, d is the weight coefficient, is the modularity, It is the active balance degree.

[0010] Preferably, when the improved genetic algorithm is used to iteratively optimize the comprehensive performance index, the fitness values ​​of the individuals are sorted from large to small, and the individuals are assigned mutation probabilities according to the sorting results, wherein the mutation probability assigned to the individuals ranked first is smaller than the mutation probability assigned to the individuals ranked later; In the current iteration, the difference between the fitness value of each individual and the optimal value of the comprehensive performance index determined in the current iteration is used to correct the mutation probability assigned to the individual.

[0011] Preferably, the outer model aims to minimize the annual comprehensive cost of carbon emissions from the distribution network and satisfies the following relationship: , In the formula, is the annual comprehensive cost of carbon emissions from the distribution network, , They are the investment cost and operation and maintenance cost of the distribution network; C CO2 Carbon trading costs for fossil fuel units; The cost of abandoned light.

[0012] Preferably, a multi-objective improved sparrow search algorithm is used to solve the outer model, wherein the fitness value of the individual determined at the end of iterative optimization of the comprehensive performance index using the improved genetic algorithm is used as the fitness value of the population at the beginning of the optimization of the multi-objective improved sparrow algorithm; Based on the initial fitness value of the population, the optimal fitness value of the population is iterated using the following relationship: , In the formula, For the The iteration A sparrow for the The optimal fitness value of a population is is the number of sparrows, is the dimension of the objective function, i.e., the number of populations in the algorithm; The best fitness value of the population determined at the end of the optimization of the multi-objective improved sparrow algorithm is used as the historical fitness value of the individual when the improved genetic algorithm iteratively optimizes the comprehensive performance index; when the improved genetic algorithm iteratively optimizes the comprehensive performance index, when the fitness values ​​of two individuals are the same, the mutation probability assigned to the individual with a larger historical fitness value is smaller than the mutation probability assigned to the individual with a smaller historical fitness value.

[0013] Preferably, the inner model aims to minimize the deviation of the comprehensive expected value of node voltage in the cluster and the comprehensive operation risk index, and satisfies the following relationship: , In the formula, is the deviation of the comprehensive expected value of node voltage in the cluster, is a comprehensive operational risk indicator. N b is the number of clusters, r U,i For Node i The voltage exceeding the limit probability; r L,l For branch l Probability of exceeding the transmission power limit; U i , U Ni Node i The actual voltage and rated voltage; oh i For Node i Voltage quality requirement indicators; l 1. l 2 are the voltage over-limit weight and power over-limit weight respectively, and l 1+ l 2=1.

[0014] Preferably, a non-dominated sorting genetic algorithm NSGA2 with an elite strategy is used to solve the inner model to obtain the optimal solution for optimizing the access capacity of photovoltaic nodes and the access location of energy storage.

[0015] The present invention also proposes a distributed photovoltaic energy storage system optimization configuration system, comprising: The cluster division module is used to divide the distributed photovoltaic energy storage system into multiple clusters, and use the modularity and active power balance of the cluster to establish the comprehensive performance index of the cluster; the improved genetic algorithm is used to iteratively optimize the comprehensive performance index to determine the cluster division result corresponding to the optimal value of the comprehensive performance index; The optimization configuration module is used to establish a two-layer model for the optimal configuration of distributed photovoltaic energy storage systems based on the cluster division results. The outer model takes the cluster as the smallest unit, optimizes the total photovoltaic capacity, total energy storage capacity and energy storage sequential output connected to the cluster according to the time-series output of the loads in each cluster and the goal of minimizing the annual comprehensive cost of carbon emissions from the distribution network; the inner model takes each node in the cluster as the smallest unit, and according to the optimization configuration results of the outer model, optimizes the access capacity of photovoltaics at each node and the access location of energy storage with the goal of minimizing the deviation of the comprehensive expected value of node voltage in the cluster and minimizing the comprehensive operation risk index.

[0016] The present invention is also a terminal, comprising a processor and a storage medium; the storage medium is used to store instructions; the processor is used to operate according to the instructions to execute the steps of the method.

[0017] The present invention is also a computer-readable storage medium having a computer program stored thereon, which implements the steps of the method when executed by a processor.

[0018] The beneficial effects of the present invention are that, compared with the prior art, at least the present invention includes: taking the minimization of the total annual operating cost of the power system as the goal, the present invention establishes a two-layer optimization model considering constraints such as energy storage access location, configuration capacity, and charge state; the outer layer adopts an improved sparrow search algorithm to optimize the energy storage access location, power and capacity, and the inner layer adopts a non-dominated sorting genetic algorithm NSGA2 to optimize the energy storage daily operation strategy, so that the three aspects of power system stability, economy and new energy utilization rate can be comprehensively optimized, making the energy storage capacity configuration safer, more economical and reasonable. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 It is a flow chart of a distributed photovoltaic energy storage system optimization configuration method proposed by the present invention. DETAILED DESCRIPTION

[0020] In order to make the purpose, technical scheme and advantages of the present invention clearer, the technical scheme of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. The embodiments described in this application are only embodiments of a part of the present invention, rather than all embodiments. Based on the spirit of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work belong to the protection scope of the present invention.

[0021] The present invention proposes a distributed photovoltaic energy storage system optimization configuration method, wherein the distributed photovoltaic energy storage system includes a distributed photovoltaic power generation system, an energy storage system and a distribution network; Figure 1 As shown, the configuration method includes: Step 1: divide the distributed photovoltaic energy storage system into multiple clusters, and use the modularity and active power balance of the cluster to establish the comprehensive performance index of the cluster; use the improved genetic algorithm to iteratively optimize the comprehensive performance index to determine the cluster division result corresponding to the optimal value of the comprehensive performance index.

[0022] In the present invention, modularity is introduced to reflect the structural performance of the cluster. e The electrical coupling strength of complex networks is measured so that the cluster division results meet the requirements of close electrical connection between nodes within the cluster and loose connection between clusters, thereby improving the on-site absorption capacity of distributed photovoltaics within the cluster.

[0023] The modularity of each cluster in the distributed photovoltaic energy storage system satisfies the following relationship: , In the formula, For Node i For Node j The voltage-active power sensitivity coefficient, ; , , They are active distance, reactive distance and comprehensive electrical distance within the cluster respectively; t is the weight coefficient; A ij To connect nodes i ,node j The edge weight of the node i and nodes j It takes 1 when directly connected, otherwise it takes 0; is the sum of all edges in the network, satisfying ; For Node i degree, satisfy , For Node j degree, satisfy ; When the node i ,node j In the same cluster , otherwise it takes 0, is the modularity.

[0024] In the present invention, the matching degree of active power in the cluster is taken into consideration. In order to ensure the improvement of the active self-consumption capacity in the cluster and reduce the active power transmission, the active balance is used as the functional indicator of the cluster. Since the inner layer takes voltage quality as the objective function, the reactive balance indicator is ignored here.

[0025] The active power balance of each cluster in the distributed photovoltaic energy storage system satisfies the following relationship: , In the formula, is the active balance, m is the total number of clusters; P q,clu For network inflow cluster q The net power of P q,load,i For cluster q Internal Node i Active load power.

[0026] The comprehensive performance index of the cluster is established with the following relationship, satisfying the following relationship: , In the formula, As a comprehensive performance indicator, d is the weight coefficient, is the modularity, It is the active balance degree.

[0027] In the present invention, the comprehensive performance index of the cluster includes a modularity index and active power balance based on electrical distance. A larger modularity index indicates a closer electrical connection between nodes in the same cluster. A larger active power balance indicates a higher degree of matching between the distributed photovoltaic installation capacity and output power of each cluster and the load.

[0028] An improved genetic algorithm is used to iteratively optimize the comprehensive performance indicators and determine the cluster division results corresponding to the optimal values ​​of the comprehensive performance indicators.

[0029] In the embodiment, when the improved genetic algorithm is used to iteratively optimize the comprehensive performance index, a larger crossover probability is considered to be set in the initial stage of the iteration to enhance the global search capability, and then stabilized at a smaller value in the later stage to ensure that the excellent individuals are not destroyed. According to the exponential model, the crossover probability decreases with the increase of the evolutionary generation, and finally stabilizes at a preset smaller value. In order to suppress the premature maturity of the algorithm, when the improved genetic algorithm is used to iteratively optimize the comprehensive performance index, the fitness values ​​of the individuals are sorted from large to small, and the individuals are assigned mutation probabilities according to the sorting results, wherein the mutation probability assigned to the individuals ranked in front is less than the mutation probability assigned to the individuals ranked in the back; in the current iteration, the difference between the fitness value of each individual and the optimal value of the comprehensive performance index determined in the current iteration is used to correct the mutation probability assigned to the individual. By assigning a smaller mutation probability to individuals with better fitness values, a larger mutation probability is assigned to inferior individuals with poor fitness values, and the mutation probability is adjusted by using the difference between the fitness value and the optimal value of the current population to improve the convergence speed of the algorithm.

[0030] Step 2: Based on the cluster division results, a two-layer model for the optimal configuration of the distributed photovoltaic energy storage system is established. The outer model takes the cluster as the smallest unit, optimizes the total photovoltaic capacity, total energy storage capacity and energy storage sequential output of the cluster according to the sequential output of the load in each cluster and takes the annual comprehensive cost of carbon emissions of the distribution network as the minimum. The inner model takes each node in the cluster as the smallest unit, and optimizes the access capacity of the photovoltaic power generation and the access location of the energy storage according to the optimization configuration results of the outer model.

[0031] In the present invention, based on the result of cluster division, the idea of ​​hierarchical configuration is adopted to divide the photovoltaic and energy storage configuration problems into two layers, the inner and outer layers. The outer configuration strategy takes the cluster as the smallest unit and each cluster as the decision-making unit. According to the time sequence output of each unit load, the total photovoltaic capacity, total energy storage capacity and energy storage time sequence output of the cluster are optimized. Configuring energy storage with cluster as the unit can reduce the degree of mismatch between photovoltaic and load time sequence requirements within the cluster. The inner configuration strategy takes each node within the cluster as the smallest unit when the total energy storage and photovoltaic capacity within the cluster is determined, and coordinates and optimizes the photovoltaic access capacity of each node and the access location of the energy storage. The outer layer of the hierarchical strategy is adopted to improve the self-consumption capacity of the active power within the cluster and realize the dynamic balance of source, load and storage. The inner layer improves the voltage quality level within the distribution network cluster, taking into account the system operation risk constraints while improving the economy of the system configuration.

[0032] Specifically, in the outer optimization model, the total photovoltaic capacity, total energy storage capacity and energy storage timing output are optimized. Considering the economy and environmental protection, the outer model aims to minimize the annual comprehensive cost of carbon emissions of the distribution network, satisfying the following relationship: , In the formula, is the annual comprehensive cost of carbon emissions from the distribution network, , They are the investment cost and operation and maintenance cost of the distribution network; C CO2 Carbon trading costs for fossil fuel units; The cost of abandoned light.

[0033] In the embodiment, the annual comprehensive cost of carbon emissions from the distribution network includes the investment cost and operation and maintenance cost of coal-fired units, gas-fired units, photovoltaic systems, and energy storage systems, carbon trading costs, and abandoned solar costs.

[0034] For traditional thermal power units, since they use fossil fuels as primary energy, they need to be assigned carbon emission costs. The mechanism of the carbon trading market can quantify the carbon emission costs. By allocating carbon quotas to each fossil fuel unit, the part exceeding the carbon quota will be reflected as carbon trading costs, and the remaining part below the carbon quota can be traded through the carbon trading market, reflected as carbon trading income.

[0035] In order to quantify the impact of abandoned photovoltaic power on system cost, the abandoned photovoltaic cost is included in the objective function as the penalty cost of system abandoned photovoltaic power.

[0036] In the embodiment, a multi-objective improved sparrow search algorithm is used to solve the outer model to obtain the optimal solution for the total photovoltaic capacity, total energy storage capacity and energy storage time sequence output of the cluster access. In the multi-objective improved sparrow algorithm, the position of the sparrow in the search space corresponds to the decision variable to be optimized in the model, the optimal fitness value is the optimal value of the objective function, and the installation position of the energy storage system is rounded in the continuous space. Using the multi-objective improved sparrow search algorithm to solve the outer model includes: 1) Obtain distribution network parameters, photovoltaic power generation, and typical daily load data, describe decision variables and impose upper and lower limit constraints, and set algorithm parameters: population size, maximum number of iterations , explorer and warning ratio, warning value, objective function dimension , external population size.

[0037] 2) Initialize the population, use the sin chaos initialization strategy to generate a uniformly distributed and sufficient initialization population in the search space, perform boundary constraints, and use the fitness value of the individual finally determined at the end of the iterative optimization of the comprehensive performance index as the fitness value of the population at the beginning of the optimization of the multi-objective improved sparrow algorithm to obtain the initialized external population; Since the present invention takes into account the comprehensive optimization of three aspects, namely system stability, economy and energy utilization, to determine the optimal configuration scheme of the distributed photovoltaic energy storage system, and the comprehensive performance index takes into account the stability and energy utilization of the system, the optimal solution of the total photovoltaic capacity, total energy storage capacity and time-series output of the cluster accessed by the iterative solution of the outer model ensures the economy and energy utilization of the system, but will have a certain impact on the stability of the distribution network. Therefore, it is necessary to verify the relevant parameters characterizing the stability of the system again, and the verification process should be embedded in the iterative solution process of the outer model, and the fitness value of the individual determined at the end of the iterative optimization of the comprehensive performance index is used as the fitness value of the population at the initial optimization of the multi-objective improved sparrow algorithm.

[0038] 3) Based on the initialized fitness value, the optimal fitness value of the population is iterated using the following relationship: , In the formula, For the The iteration A sparrow for the The optimal fitness value of a population is is the number of sparrows, is the dimension of the objective function, that is, the number of populations in the algorithm.

[0039] In addition, the present invention also proposes to use the optimal fitness value of the population determined at the end of the optimization of the multi-objective improved sparrow algorithm as the historical fitness value of the individual when the improved genetic algorithm iteratively optimizes the comprehensive performance index. When the fitness values ​​of two individuals are the same, the mutation probability assigned to the individual with a larger historical fitness value is smaller than the mutation probability assigned to the individual with a smaller historical fitness value.

[0040] The present invention sets a target that represents the economic performance of the system. , and embed comprehensive performance indicators that characterize system stability and energy utilization in the process of iterative optimization of comprehensive performance indicators and iterative optimization of multi-objective improved sparrow algorithm, so as to achieve the optimal configuration of the total connected photovoltaic capacity, total energy storage capacity and energy storage timing output under the optimal system stability, economy and energy utilization.

[0041] 4) Select sparrows with better fitness as explorers according to the proportion and introduce nonlinear adjustment factors α The idea of ​​sine-cosine algorithm is integrated to increase the global search capability of the explorer in the early stage of sparrow search, expand the local search space in the later stage of search, update the position of the explorer and perform constraint control to satisfy the following relationship: , , In the formula, r 1 is the control step size parameter, which takes a random number uniformly distributed within [0,2π]; r 2 is the weight that controls the influence of the current optimal solution on the explorer's position update, and takes a random number with uniform internal distribution; For the The optimal fitness value of the population at the iteration; The ratio of explorers to early warning is the warning value.

[0042] 5) Sparrows other than explorers are followers. The Levy flight search mechanism is used to enhance the global search capability of followers. The positions of followers are updated and constrained to satisfy the following relationship: , In the formula, is the follower position, is the Levy function under the objective function dimension.

[0043] 6) Randomly select some sparrows in the population as early warning persons, update their positions and perform constraint control to satisfy the following relationship: , In the formula, F i , F Best , F Worst Sparrow i The fitness value, the global optimal fitness value, and the global worst fitness value; is a very small constant; is the set random ratio, The random ratio is set.

[0044] 7) Introduce an external population to record Pareto non-dominated solutions, calculate the congestion degree, add new non-dominated solutions to the external population and remove some solutions from the external population.

[0045] 8) Determine whether the termination condition is met. If so, terminate the process and use the entropy weight method-TOPSIS method to select the best compromise decision plan for energy storage site selection and capacity determination, and output the result. Otherwise, repeat steps 3) to 7).

[0046] Specifically, after the outer model uses the improved sparrow search algorithm to determine the optimal configuration solution, the inner model takes the minimum deviation of the comprehensive expected value of node voltage in the cluster and the minimum comprehensive operation risk index as the goal, satisfying the following relationship: , In the formula, is the deviation of the comprehensive expected value of node voltage in the cluster, is a comprehensive operational risk indicator. N b is the number of clusters, r U,i For Node i The voltage exceeding the limit probability; r L,l For branch l Probability of exceeding the transmission power limit; U i , U Ni Node i The actual voltage and rated voltage; oh i For Node i Voltage quality requirement indicators; l 1. l 2 are the voltage over-limit weight and power over-limit weight respectively, and l 1+ l 2=1.

[0047] The non-dominated sorting genetic algorithm NSGA2 with elite strategy is used to solve the inner model of the distributed photovoltaic energy storage system, including: 1) Input population size, number of iterations, crossover probability, mutation probability, photovoltaic predicted power, load parameters and related parameters; the decision variable energy storage system charging and discharging power P bat Encoding is performed; an initial population P0 is generated according to the operating strategy formulated on demand.

[0048] 2) Using the probabilistic power flow calculation method based on Latin hypercube sampling, the fitness value of the population is obtained according to the algorithm fitness function, that is, the deviation of the comprehensive expected value of the node voltage in the cluster and the comprehensive operation risk index, including: 2.1) Stratified sampling of uncertainty factors. Calculate the cumulative distribution function based on the probability density function of light intensity and load. Divide [0,1] into N sub-intervals and select the middle value of each sub-interval. Y i As a sampling point; using Conclusion X i No. n sampling values, thus obtaining the initial sampling matrix M , the order is 3× N , each row represents a sampled value of a random variable.

[0049] 2.2) Initial sample matrix sorting. Randomly generate a 3× N A positive integer matrix of order L , L Each row of is randomly arranged by positive integers 1, 2, ..., N, and its rank correlation coefficient matrix is ​​calculated r L ;right r L Perform Cholesky decomposition and use the formula L X = Q -1 L Eliminate random integer matrix L The correlations generated by random permutations are obtained as the matrix L X The target correlation coefficient matrix is ​​obtained based on the historical data of light intensity and load. r obj ,according to r obj = PP TPerform Cholesky decomposition on it and get the lower triangular matrix P ,Depend on L S = PL X Can be used L S The rank correlation coefficient matrix and r obj Similar. According to the matrix L S The size of the numbers in the row of the initial sample matrix M The light intensity and load samples in the corresponding order are sorted to obtain the final sample matrix M Z .

[0050] 2.3) The illumination data samples M Z The matrix is ​​converted into a photovoltaic output matrix; , In the formula, is the rated capacity of the photovoltaic, kW; I r is the rated solar irradiance, W / m2; f PVG is the photovoltaic power factor angle.

[0051] 2.4) Conduct power distribution network analysis on photovoltaic power generation D The deterministic power flow calculation adopts the forward-backward power flow algorithm, where the wind power and photovoltaic access points are regarded as PQ nodes. i On each node, i The load is P Li and DG output is P DGi ,when P DGi < P Li , input to the node i The active power is P Li - P DGi ;when P DGi = P Li ,node i The active power input to the distribution network is 0; when P DGi > P Li ,node i The active power input to the distribution network is PDGi - P Li ; 2.5) D The expected values ​​of voltage and active network losses are calculated from the sub-flow calculation results.

[0052] , Where: Y i =( y i1 , y i2 ,⋯, y iD ), P i =( p i1 , p i2 ,⋯, p iD ), P For the corresponding value Y probability.

[0053] 3) Determine the dominance level of individuals based on their dominance relationships r i ; Initialize the crowding value and sort the individuals in ascending order i The fitness values ​​of each sub-objective function are calculated according to the following formula to calculate the crowding distance: ,

[0054] in f max and f min is the maximum value among the individuals in the dominant layer after sorting; like r i ≥ r j And Γ i >Γ j , then the individual i Ranked in individual j Previously, non-dominated hierarchical sorting was performed like this.

[0055] 4) Select the first N individuals in P0 to form a new population P and perform genetic operations. Use the tournament criterion to select P, perform crossover mutation among individuals according to the single-point crossover operator and the basic position crossover operator, and obtain the offspring population.

[0056] 5) According to the elite retention strategy of the non-dominated sorting genetic algorithm, a new population is generated by merging the parent and child populations and selecting the optimal N individuals based on the non-dominated sorting and crowding distance.

[0057] 6) If the maximum number of iterations is reached, proceed to step 7) otherwise jump to step 2).

[0058] 7) Obtain the Pareto optimal solution set, that is, the set of optimal solutions within each range for the two sub-objective functions, and find the membership vector of the Pareto solution. According to the fuzzy operator and importance of each sub-objective in the capacity configuration, the target weight vector is obtained. Finally, the closeness between the target weight and the membership vector of each Pareto solution is calculated. The membership vector with the best closeness is the optimal compromise solution of the corresponding Pareto solution set.

[0059] The present invention also proposes a distributed photovoltaic energy storage system optimization configuration system, comprising: The cluster division module is used to divide the distributed photovoltaic energy storage system into multiple clusters, and use the modularity and active power balance of the cluster to establish the comprehensive performance index of the cluster; the improved genetic algorithm is used to iteratively optimize the comprehensive performance index to determine the cluster division result corresponding to the optimal value of the comprehensive performance index; The optimization configuration module is used to establish a two-layer model for the optimal configuration of distributed photovoltaic energy storage systems based on the cluster division results. The outer model takes the cluster as the smallest unit, optimizes the total photovoltaic capacity, total energy storage capacity and energy storage sequential output connected to the cluster according to the time-series output of the loads in each cluster and the goal of minimizing the annual comprehensive cost of carbon emissions from the distribution network; the inner model takes each node in the cluster as the smallest unit, and according to the optimization configuration results of the outer model, optimizes the access capacity of photovoltaics at each node and the access location of energy storage with the goal of minimizing the deviation of the comprehensive expected value of node voltage in the cluster and minimizing the comprehensive operation risk index.

[0060] The present disclosure may be a system, a method and / or a computer program product. The computer program product may include a computer-readable storage medium carrying computer-readable program instructions for causing a processor to implement various aspects of the present disclosure.

[0061] A computer-readable storage medium may be a tangible device that can hold and store instructions used by an instruction execution device. A computer-readable storage medium may be, for example, but not limited to, an electrical storage device, a magnetic storage device, an optical storage device, an electromagnetic storage device, a semiconductor storage device, or any suitable combination of the above. More specific examples (a non-exhaustive list) of computer-readable storage media include: a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), a static random access memory (SRAM), a portable compact disk read-only memory (CD-ROM), a digital versatile disk (DVD), a memory stick, a floppy disk, a mechanical encoding device, such as a punch card or a raised structure in a groove on which instructions are stored, and any suitable combination of the above. The computer-readable storage medium used herein is not to be interpreted as a transient signal itself, such as a radio wave or other freely propagating electromagnetic wave, an electromagnetic wave propagating through a waveguide or other transmission medium (e.g., a light pulse through a fiber optic cable), or an electrical signal transmitted through a wire.

[0062] The computer-readable program instructions described herein can be downloaded from a computer-readable storage medium to each computing / processing device, or downloaded to an external computer or external storage device via a network, such as the Internet, a local area network, a wide area network, and / or a wireless network. The network can include copper transmission cables, optical fiber transmissions, wireless transmissions, routers, firewalls, switches, gateway computers, and / or edge servers. The network adapter card or network interface in each computing / processing device receives the computer-readable program instructions from the network and forwards the computer-readable program instructions for storage in the computer-readable storage medium in each computing / processing device.

[0063] The computer program instructions for performing the operations of the present disclosure may be assembly instructions, instruction set architecture (ISA) instructions, machine instructions, machine-dependent instructions, microcode, firmware instructions, state setting data, or source code or object code written in any combination of one or more programming languages, including object-oriented programming languages, such as Smalltalk, C++, etc., and conventional procedural programming languages, such as "C" language or similar programming languages. The computer-readable program instructions may be executed entirely on the user's computer, partially on the user's computer, as a separate software package, partially on the user's computer, partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., using an Internet service provider to connect through the Internet). In some embodiments, by using the state information of the computer-readable program instructions to personalize an electronic circuit, such as a programmable logic circuit, a field programmable gate array (FPGA), or a programmable logic array (PLA), the electronic circuit may execute the computer-readable program instructions, thereby implementing various aspects of the present disclosure.

[0064] Finally, 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 above embodiments, ordinary technicians in the relevant field should understand that the specific implementation methods of the present invention can still be modified or replaced by equivalents, and any modifications or equivalent replacements that do not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A distributed photovoltaic energy storage system optimization configuration method, the distributed photovoltaic energy storage system includes a distributed photovoltaic power generation system, an energy storage system and a distribution network; characterized in that: include: The distributed photovoltaic energy storage system is divided into multiple clusters, and the cluster's modularity and active power balance are used to establish the comprehensive performance index of the cluster; the improved genetic algorithm is used to iteratively optimize the comprehensive performance index to determine the cluster division result corresponding to the optimal value of the comprehensive performance index; According to the cluster division results, a two-layer model for the optimal configuration of distributed photovoltaic energy storage systems is established. The outer model takes the cluster as the smallest unit, optimizes the total photovoltaic capacity, total energy storage capacity and the sequential output of energy storage connected to the cluster according to the sequential output of the loads in each cluster and takes the annual comprehensive cost of carbon emissions of the distribution network as the minimum. The inner model takes each node in the cluster as the smallest unit, and according to the optimization configuration results of the outer model, optimizes the access capacity of photovoltaic power at each node and the access location of energy storage with the goal of minimizing the deviation of the comprehensive expected value of node voltage in the cluster and minimizing the comprehensive operation risk index.

2. The distributed photovoltaic energy storage system optimization configuration method according to claim 1 is characterized in that: The modularity of each cluster in the distributed photovoltaic energy storage system satisfies the following relationship: ， In the formula, For Node i For Node j The voltage-active power sensitivity coefficient, ; , , They are active distance, reactive distance and comprehensive electrical distance within the cluster respectively; τ is the weight coefficient; A ij To connect nodes i ,node j The edge weight of the node i and nodes j It takes 1 when directly connected, otherwise it takes 0; is the sum of all edges in the network, satisfying ; For Node i degree, satisfy , For Node j degree, satisfy ; When the node i ,node j In the same cluster , otherwise it takes 0, is the modularity.

3. The distributed photovoltaic energy storage system optimization configuration method according to claim 1 is characterized in that: The active power balance of each cluster in the distributed photovoltaic energy storage system satisfies the following relationship: ， In the formula, is the active balance, m is the total number of clusters; P q,clu For network inflow cluster q The net power of P q,load,i For cluster q Internal Node i Active load power.

4. The distributed photovoltaic energy storage system optimization configuration method according to claim 1, characterized in that: The comprehensive performance index of the cluster is established with the following relationship, satisfying the following relationship: ， In the formula, As a comprehensive performance indicator, is the weight coefficient, is the modularity, It is the active balance degree.

5. The distributed photovoltaic energy storage system optimization configuration method according to claim 1, characterized in that: When the improved genetic algorithm is used to iteratively optimize the comprehensive performance index, the fitness values ​​of the individuals are sorted from large to small, and the individuals are assigned mutation probabilities according to the sorting results. Among them, the mutation probability assigned to the individuals ranked first is smaller than the mutation probability assigned to the individuals ranked later. In the current iteration, the difference between the fitness value of each individual and the optimal value of the comprehensive performance index determined in the current iteration is used to correct the mutation probability assigned to the individual.

6. The distributed photovoltaic energy storage system optimization configuration method according to claim 1, characterized in that: The outer model aims to minimize the annual comprehensive cost of carbon emissions from the distribution network and satisfies the following relationship: ， In the formula, is the annual comprehensive cost of carbon emissions from the distribution network, , They are the investment cost and operation and maintenance cost of the distribution network; Carbon trading costs for fossil fuel units; The cost of abandoned light.

7. The distributed photovoltaic energy storage system optimization configuration method according to claim 6, characterized in that: The outer model is solved by using a multi-objective improved sparrow search algorithm, wherein the fitness value of the individual determined at the end of iterative optimization of the comprehensive performance index by using an improved genetic algorithm is used as the fitness value of the population at the beginning of the optimization of the multi-objective improved sparrow algorithm; Based on the initial fitness value of the population, the optimal fitness value of the population is iterated using the following relationship: ， In the formula, For the The first iteration A sparrow for the The optimal fitness value of a population is is the number of sparrows, is the dimension of the objective function, i.e., the number of populations in the algorithm; The best fitness value of the population determined at the end of the optimization of the multi-objective improved sparrow algorithm is used as the historical fitness value of the individual when the improved genetic algorithm iteratively optimizes the comprehensive performance index; when the improved genetic algorithm iteratively optimizes the comprehensive performance index, when the fitness values ​​of two individuals are the same, the mutation probability assigned to the individual with a larger historical fitness value is smaller than the mutation probability assigned to the individual with a smaller historical fitness value.

8. The distributed photovoltaic energy storage system optimization configuration method according to claim 1, characterized in that: The inner model aims to minimize the deviation of the comprehensive expected value of node voltage in the cluster and the comprehensive operation risk index, satisfying the following relationship: ， In the formula, is the deviation of the comprehensive expected value of node voltage in the cluster, is a comprehensive operational risk indicator. N b is the number of clusters, ρ U,i For Node i The voltage exceeding the limit probability; ρ L,l For branch l Probability of exceeding the transmission power limit; U i , U Ni Node i The actual voltage and rated voltage; ω i For Node i Voltage quality requirement indicators; λ 1. λ 2 are the voltage over-limit weight and power over-limit weight respectively, and λ 1+ λ 2=1.

9. The distributed photovoltaic energy storage system optimization configuration method according to claim 8, characterized in that: The non-dominated sorting genetic algorithm NSGA2 with an elite strategy is used to solve the inner model to obtain the optimal solution for optimizing the access capacity of photovoltaic nodes and the access location of energy storage.

10. A distributed photovoltaic energy storage system optimization configuration system, characterized in that: include: The cluster division module is used to divide the distributed photovoltaic energy storage system into multiple clusters and establish the comprehensive performance index of the clusters by using the modularity and active power balance of the clusters; An improved genetic algorithm is used to iteratively optimize the comprehensive performance index to determine the cluster division result corresponding to the optimal value of the comprehensive performance index; The optimization configuration module is used to establish a two-layer model for the optimal configuration of distributed photovoltaic energy storage systems based on the cluster division results. The outer model takes the cluster as the smallest unit, optimizes the total photovoltaic capacity, total energy storage capacity and energy storage timing output connected to the cluster according to the timing output of the load in each cluster and the goal of minimizing the annual comprehensive cost of carbon emissions of the distribution network; the inner model takes each node in the cluster as the smallest unit, and according to the optimization configuration results of the outer model, optimizes the access capacity of photovoltaic power at each node and the access location of energy storage with the goal of minimizing the deviation of the comprehensive expected value of node voltage in the cluster and minimizing the comprehensive operation risk index.

Images