CN112036655A - Opportunity constraint-based planning method for photovoltaic power station and electric vehicle charging network - Google Patents

Abstract

The invention provides a photovoltaic power station and electric vehicle charging network planning method based on opportunity constraint, which comprises the following steps: s10, setting planning boundary conditions; s20, establishing a photovoltaic power station and electric vehicle charging network joint random planning model based on opportunity constraint by using the planning boundary conditions; and S30, designing a chromosome coding scheme and crossover and mutation operators, solving a photovoltaic power station and electric vehicle charging network combined random planning model by adopting a genetic algorithm, and giving an optimal planning scheme for the photovoltaic power station and the electric vehicle charging network. The photovoltaic power station and electric vehicle charging network planning method based on opportunity constraint reduces the power loss of a power distribution system as much as possible, ensures that node voltage deviation and line power flow exceed the limit and meets the opportunity constraint, and can provide a reasonable photovoltaic power station/charging station construction scheme and provide reference for engineering technicians.

Description

Opportunity constraint-based planning method for photovoltaic power station and electric vehicle charging network

technical field

The invention relates to the technical field of electric vehicle charging networks, in particular to a photovoltaic power station and an electric vehicle charging network planning method based on chance constraints.

Background technique

Increasing the proportion of renewable energy power generation in the power grid and the penetration rate of electric vehicles in the transportation system is one of the important ways to transform the form of energy utilization and achieve sustainable development. In recent years, distributed power sources represented by photovoltaic power plants have emerged in the power distribution system, and the power distribution system has gradually evolved from a simple passive system to a complex active system. In addition, with the increasing popularity of electric vehicles, electric vehicle charging stations have also become an important power load in the power distribution system, and their impact on the operating conditions of the power distribution system has become increasingly prominent. For the power distribution system, the unreasonable layout of photovoltaic power stations and electric vehicle charging stations will deteriorate their operating conditions and affect the normal power supply to users. The specific manifestations are the increase of power loss in the network, the excessive node voltage deviation and the line power flow exceeding the limit, etc. . In this context, it is necessary to jointly plan the photovoltaic power station and electric vehicle charging network in the power distribution system to improve the operating conditions of the power distribution system as much as possible. Affected by uncertain factors such as vehicle owners' traffic behavior and charging habits, the charging load of electric vehicle charging stations has random characteristics. In addition, the output of photovoltaic power stations also has random characteristics. Under the combined action of these two stochastic factors, the operating conditions of the power distribution system exhibit significant stochastic characteristics, and the joint planning problem of photovoltaic power stations and electric vehicle charging networks will inevitably become a stochastic programming problem. To sum up, it is urgent to propose a joint stochastic programming model and a corresponding solution method of photovoltaic power station and electric vehicle charging network considering the stochastic characteristics of power distribution system operating conditions.

Literature 1 "Comprehensive optimization model for sizing and sitting of DGunits, EV charging stations, and energy storage systems" (IEEE Transactions on Smart Grid, 2018, Vol. 9, No. 4, pp. 3871-3882) established a pairing The second-order cone programming model for the comprehensive optimization of the construction address and construction capacity of distributed photovoltaic power stations, electric vehicle charging stations and energy storage systems in the electrical system is solved by GAMS software. The model proposed in this paper considers the time-varying characteristics of the output of distributed photovoltaic power plants and the charging load of electric vehicles, but does not consider the random characteristics, which has certain limitations. Under the premise that distributed power sources including photovoltaic power plants can supply power to distribution network loads and charging stations at the same time, Literature 2 "Research on Multi-objective Planning of Distribution Networks with Distributed Power Sources and Electric Vehicle Charging Stations" (Grid Technology , 2015, Vol. 39, No. 2, pp. 450 to 456) established a multi-objective planning model for simultaneously optimizing the construction address and capacity of distributed power and electric vehicle charging stations, and adopted a multi-objective free search algorithm The Pareto solution set of the model is given. However, the literature does not consider the random characteristics of distributed power output and electric vehicle charging station charging load in the research, and the planning results given have certain limitations. Document 3 "Opportunity Constrained Planning of Electric Vehicle Charging Stations with Photovoltaic Distributed Power Distribution Networks" (Journal of Electric Power Systems and Automation, 2017, Vol. 29, No. 6, pp. 45-52) establishes a framework that considers photovoltaics. The location optimization model of electric vehicle charging station and photovoltaic power station based on the random characteristics of power station output and charging station charging load, and the bat algorithm is used to solve it. This document does not sufficiently consider the random characteristics of photovoltaic power station output and charging station charging load, and mainly focuses on optimizing the construction addresses of electric vehicle charging stations and photovoltaic power stations, which has certain limitations.

Photovoltaic power stations and electric vehicle charging stations are distributed power sources and electricity loads with random characteristics in the power distribution system, and their penetration in the power distribution system is increasing. For the power distribution system, the unreasonable layout of photovoltaic power stations and electric vehicle charging stations will deteriorate their operating conditions and affect the normal power supply to users. The specific manifestations are the increase of power loss in the network, the excessive node voltage deviation and the line power flow exceeding the limit, etc. . Therefore, it is necessary to jointly plan the photovoltaic power station and electric vehicle charging network in the power distribution system to improve the operating conditions of the power distribution system as much as possible. However, the prior art method does not fully consider the random characteristics of the output and charging load of the distributed photovoltaic power station, and has certain limitations.

SUMMARY OF THE INVENTION

In order to solve the above problems, the present invention provides a photovoltaic power station and electric vehicle charging network planning method based on chance constraints, under the condition that the number of photovoltaic power stations/charging stations and the total construction capacity are given, the construction address and To build capacity, minimize the power loss of the power distribution system, and ensure that node voltage offset and line power flow exceed the opportunity constraints scheme, to provide reference for engineers and technicians.

In order to realize the above purpose, a kind of technical scheme that the present invention adopts is:

The method for planning a photovoltaic power station and an electric vehicle charging network based on chance constraints includes the following steps: S10, a planning boundary condition is given, and the planning boundary conditions include: distribution system topology parameters, distribution system load, and charging station daily charging load probability scenario set, daily power generation output probability scenario set of photovoltaic power plants, total number of candidate sites for charging stations, total number and total capacity of charging stations under construction, type and capacity of charging stations to be built, total number of candidate addresses for photovoltaic power plants, total number and total capacity of photovoltaic power plants under construction, photovoltaic power plants to be built Power station type and capacity, maximum allowable deviation percentage of node voltage, and confidence level of node voltage over-limit and line power flow over-limit; S20 using the planning boundary conditions to establish a joint stochastic programming model for photovoltaic power plants and electric vehicle charging networks based on chance constraints; S30 According to the characteristics of the joint stochastic programming model of the photovoltaic power station and the electric vehicle charging network, the chromosome coding scheme and the crossover and mutation operators are designed, and the genetic algorithm is used to solve the joint stochastic programming model of the photovoltaic power station and the electric vehicle charging network. Optimal planning scheme of vehicle charging network.

Further, the step S20 includes: the optimization goal of the joint stochastic programming model of the photovoltaic power station and the electric vehicle charging network is to reduce the grid loss power in a typical day of the distribution system planning, as shown in formula (1),

Among them, F _loss is the expected power loss of the power distribution system in a typical day; t is the power flow analysis period index, T _f is the number of power flow analysis periods in a typical day; l is the distribution line index; Ω _br is the distribution line index set ; ΔP _loss,l,t is the power loss of distribution line l in the power flow analysis period t, which is a random variable; E( ) is an operator to find the expectation of the random variable; the chance constraints include respectively formula (2) ~(7) Obtained opportunity constraints representing the total number of charging station constructions, opportunity constraints representing the total number of photovoltaic power station constructions, opportunity constraints representing the total construction capacity of charging stations, opportunity constraints representing the total construction capacity of photovoltaic power stations, and node voltage offsets. Opportunity constraints and opportunity constraints indicating line flow limit violations,

Among them, M _ch is the total number of charging station construction; N _ch is the total number of candidate addresses of charging stations; i is the index of candidate addresses; _xi is the 0-1 optimization variable in the joint stochastic programming model of photovoltaic power station and electric vehicle charging network, take "1"" means to build a charging station at the candidate address i, taking "0" means not to build a charging station at the candidate address i, i=1, 2, 3, ···, N _ch ;

Among them, M _pv is the total number of photovoltaic power plants constructed; N _pv is the total number of candidate addresses of photovoltaic power plants; j is the index of candidate addresses; y _j is the 0-1 optimization variable in the joint stochastic programming model of photovoltaic power plants and electric vehicle charging network, take "1"" means constructing photovoltaic power station at candidate address j, taking "0" means not constructing photovoltaic power station at candidate address j, j=1, 2, 3, ···, N _pv ;

Among them, _zi is the construction capacity of the charging station at the candidate address i. For the electric vehicle charging network, the electric vehicle charging stations to be built are divided into Q _ev categories, and _zi is the joint stochastic programming model of the photovoltaic power station and the electric vehicle charging network. Discrete optimization variable; C _ch is the total construction capacity of the charging station;

Among them, w _j is the construction capacity of the photovoltaic power station at the candidate address j. For the photovoltaic power station, the photovoltaic power station to be built is divided into Q _pv categories, and w _j is the discrete optimization variable in the joint stochastic programming model of the photovoltaic power station and the electric vehicle charging network; C _pv is the total construction capacity of photovoltaic power plants;

Among them, P _r {·} represents the probability of random events in parentheses; k is the distribution node index; Ω _bus is the distribution node set; U _k is the voltage of node k, which is a random variable, and the probability distribution characteristics are analyzed by the probability power flow The results are given; U _N is the rated voltage of the distribution system; α% is the maximum allowable deviation percentage of the node voltage; β ₁ is the confidence level of the voltage exceeding the limit;

P _r {I _l >I _l,max }≤β ₂ l∈Ω _br (7)

Among them, I _l is the load current in the distribution line l, which is a random variable, and the probability distribution characteristics are given by the probabilistic power flow analysis results; I _l,max is the maximum allowable current of the distribution line l; β ₂ is the power flow over-limit confidence Spend.

Further, the step S30 includes the following steps: S31 sets genetic algorithm parameters, the genetic algorithm parameters include population size N _pop , crossover rate P _c , mutation rate P _m and maximum evolutionary algebra G _max ; S32 randomly generated by N The initial population composed of _pop chromosomes, according to the characteristics of the joint stochastic programming model of the photovoltaic power station and the electric vehicle charging network, uses an integer coding scheme to encode the optimization problem; S33 The evolutionary algebra index g is initialized to 0, that is, g=0; S34 makes g =g+1, start the g-th generation evolution, the chromosome index n is initialized to 1, that is, n=1; S35 decodes the n-th chromosome in the current population, determines the construction positions of M _ch electric vehicle charging stations, and builds Capacity and total construction capacity C _t-ev , determine the construction locations of M _ev photovoltaic power stations, construction capacity and total construction capacity C _t-pv ; use the scenario probability method to calculate the probability power flow of the power distribution system, and determine the network loss in typical planned days The power expectation F _loss , the probability distribution characteristics of the voltage amplitude of each node and the power flow of each line, and the fitness V _fit,n of the nth chromosome is calculated according to formulas (8) to (12):

V _fit,n =F _max -F _loss -η ₁ ×V _p1 -η ₂ ×V _p2 -η ₃ ×V _p3 -η ₄ ×V _p4 (8)

V _p1 =|C _ch -C _t-ev | (9)

V _p2 =|C _pv -C _t-pv | (10)

表示取

中较大的数，采用罚函数法分别处理公式(4)～(7)给出的约束，η₁、η₂、η₃以及η₄为罚系数；V_p1、V_p2、V_p3以及V_p4分别表示公式(4)～(7)给出约束的违背程度；S36判断是否计算完当前种群中所有染色体的适应度，即判断染色体索引n是否等于种群规模N_pop，若n<N_pop，则令n＝n+1，并跳转至步骤S35，继续进行染色体适应度计算；否则，继续执行步骤S37；S37判断是否到达最大进化代数，即判断进化代数索引g是否等于最大进化代数G_max，若g＝G_max，则继续执行步骤S38；否则，以适应度为依据，对当前染色体种群进行复制、交叉与变异操作，更新染色体种群，并跳转至步骤S34；S38将当前种群最优秀染色体对应的光伏电站与充电网络建设方案作为光伏电站与电动汽车充电网络联合随机规划模型的最优解输出，结束算法流程。Among them, _Fmax is a relatively large positive number given in advance to ensure that the chromosome fitness is non-negative, and the operator

means to take

For the larger number in , the penalty function method is used to deal with the constraints given by equations (4) to (7), respectively, η ₁ , η ₂ , η ₃ and η ₄ are penalty coefficients; V _p1 , V _p2 , V _p3 and V _p4 represents the degree of violation of the constraints given by formulas (4) to (7) respectively; S36 judges whether the fitness of all chromosomes in the current population has been calculated, that is, judges whether the chromosome index n is equal to the population size N _pop , if n<N _pop , Then let n=n+1, and jump to step S35, and continue to perform chromosome fitness calculation; otherwise, continue to perform step S37; S37 judges whether the maximum evolutionary algebra is reached, that is, it is judged whether the evolutionary algebra index g is equal to the maximum evolutionary algebra G _max , if g=G _max , then proceed to step S38; otherwise, based on fitness, perform replication, crossover and mutation operations on the current chromosome population, update the chromosome population, and jump to step S34; S38 selects the current population as the best The photovoltaic power station and charging network construction scheme corresponding to the chromosome is output as the optimal solution of the joint stochastic programming model of the photovoltaic power station and the electric vehicle charging network, and the algorithm process ends.

Further, the crossover operation includes the following steps: S41 randomly selects two chromosomes from the current population as the chromosomes to be crossed; _S42 exchanges the chromosome code string after the _Nth code position of the two chromosomes with the crossover probability Pc, and completes the first step. A crossover operation; S43 randomly generates feasible crossover bits N _cr1 , and exchanges the N _cr1 +1 to N _ch code bits in the two chromosomes with the crossover probability P _c to complete the second crossover operation; S44 randomly generates feasible crossover bits N _cr2 , exchange the code string after the N _cr2 code point in the two chromosomes with the crossover probability P _c to complete the third crossover operation.

Further, the mutation operation operator includes the following steps: S51 randomly selects a chromosome from the current population as the chromosome to be mutated; S52 randomly generates two code positions to be mutated N _mu1 and N _mu2 to ensure that the two code positions to be mutated are One of the values is "0", and the other is an integer other than "0"; S53 performs the mutation operation on the mutation bits N _mu1 and N _mu2 at the same time with the mutation probability P _m , that is, the mutation of the to-be-mutated bits whose value is "0" is not greater than Q _ev is a non-"0" random integer, and the mutated bit whose value is not "0" is mutated to "0", and the first mutation operation is completed; S54 randomly generates two to-be-mutated code bits N _mu3 and N _mu4 to ensure that the two One of the values of the code bits to be mutated is "0", and the other is an integer other than "0"; S55 simultaneously treats the mutated bits N _mu3 and N _mu4 with the mutation probability P _m and performs the second mutation operation, that is, the value is "0" ” is mutated into a random integer that is not greater than Q _pv and not “0”, and the value of the mutated bit that is not “0” is mutated to “0”, and the second mutation operation is completed.

The above-mentioned technical scheme of the present invention has the following advantages compared with the prior art:

The method for planning a photovoltaic power station and an electric vehicle charging network based on the chance constraint of the present invention optimizes the construction address and construction capacity of the photovoltaic power station/charging station under the given conditions of the number of photovoltaic power stations/charging stations and the total construction capacity, as far as possible. To reduce the power loss of the power distribution system, and to ensure that the node voltage offset and the line power flow exceed the limit to meet the opportunity constraints, the photovoltaic power station and electric vehicle charging network planning method can consider the impact of the charging power and the random characteristics of distributed photovoltaics on the planning scheme at the same time , to give a more reasonable construction plan of photovoltaic power station/charging station, and provide reference for engineering and technical personnel.

Description of drawings

The technical solutions of the present invention and its beneficial effects will be apparent through the detailed description of the specific embodiments of the present invention below in conjunction with the accompanying drawings.

FIG. 1 is a flowchart of a method for planning a photovoltaic power station and an electric vehicle charging network based on chance constraints according to an embodiment of the present invention;

FIG. 2 is a flowchart of the step S30 according to an embodiment of the present invention;

FIG. 3 is a flowchart of a crossover operation according to an embodiment of the present invention;

FIG. 4 is a flowchart of a mutation operation according to an embodiment of the present invention.

Detailed ways

The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative efforts shall fall within the protection scope of the present invention.

This embodiment provides a method for planning a photovoltaic power station and an electric vehicle charging network based on chance constraints. As shown in FIG. 1 , the method includes the following steps: S10, a planning boundary condition is given. S20 uses the planning boundary conditions to establish a joint stochastic programming model for photovoltaic power plants and electric vehicle charging networks based on chance constraints. And S30, according to the characteristics of the photovoltaic power station and electric vehicle charging network joint stochastic programming model, the chromosome coding scheme and the crossover and mutation operators are designed, the genetic algorithm is used to solve the photovoltaic power station and the electric vehicle charging network joint stochastic programming model, and the photovoltaic power station is given. Optimal planning scheme with electric vehicle charging network.

The planning boundary conditions include: distribution system topology parameters, distribution system load, probability scenario set of charging station daily charging load, probability scenario set of daily power generation output of photovoltaic power station, total number of candidate addresses of charging station, total number and total capacity of charging station construction, Type and capacity of charging stations to be built, total number of candidate addresses for photovoltaic power plants, total number and total capacity of photovoltaic power plants under construction, type and capacity of photovoltaic power plants to be built, maximum allowable deviation percentage of node voltage, and confidence level of node voltage over-limit and line power flow over-limit, etc. .

The step S20 includes the following steps: the optimization goal of the joint stochastic programming model of the photovoltaic power station and the electric vehicle charging network is to reduce the grid loss power in a typical day of the distribution system planning, as shown in formula (1),

Among them, F _loss is the expected power loss of the power distribution system in a typical day; t is the power flow analysis period index, T _f is the number of power flow analysis periods in a typical day; l is the distribution line index; Ω _br is the distribution line index set ; ΔP _{loss, l, t} is the power loss of distribution line l in the power flow analysis period t, which is a random variable; E(·) is the operator that calculates the expectation of the random variable. The photovoltaic power station is a distributed power source in the power distribution system, and the electric vehicle charging station is an important load in the power distribution system, both of which have random characteristics. After the photovoltaic power station and the electric vehicle charging station are connected at the same time, the operating conditions of the power distribution system will show significant random characteristics, and the power loss of the network may change. Therefore, the optimization goal of the joint stochastic programming model of the photovoltaic power station and the electric vehicle charging network is to minimize the grid loss in a typical day of the distribution system planning.

The opportunity constraints include the opportunity constraints representing the total number of charging station constructions, the opportunity constraints representing the total number of photovoltaic power station constructions, the opportunity constraints representing the total construction capacity of charging stations, and the total number of photovoltaic power station construction obtained through formulas (2) to (7). The opportunity constraint of capacity, the opportunity constraint of node voltage offset, and the opportunity constraint of line power flow limit,

Among them, M _ch is the total number of charging stations built; N _ch is the total number of candidate addresses of charging stations, which is the number of nodes in the power distribution system that can be connected to electric vehicle charging stations; i is the candidate address index; _xi is the photovoltaic power station and electric vehicle charging station The 0-1 optimization variable in the network joint stochastic programming model, taking "1" means building a charging station at candidate address i, taking "0" means not building a charging station at candidate address i, i=1,2,3,... , _Nch ;

Among them, M _pv is the total number of photovoltaic power stations constructed; N _pv is the total number of candidate addresses of photovoltaic power stations, which is the number of nodes in the power distribution system that can be connected to photovoltaic power stations; j is the candidate address index; y _j is the combination of photovoltaic power stations and electric vehicle charging networks The 0-1 optimization variable in the stochastic programming model, taking "1" means to build a photovoltaic power station at the candidate address j, taking "0" means not to build a photovoltaic power station at the candidate address j, j=1,2,3,...,N _pv ;

Among them, _zi is the construction capacity of the charging station at candidate address i. For the electric vehicle charging network, the electric vehicle charging stations to be built are divided into Q _ev categories, and the construction capacity of different types of charging stations is different. Therefore, _zi is the difference between the photovoltaic power station and the Discrete optimization variables in the joint stochastic programming model of the electric vehicle charging network; _Cch is the total construction capacity of the charging station, which is determined by the number of electric vehicles in the area to be planned, the construction cost of the charging station and the total investment to be invested in the construction of the charging station;

Among them, w _j is the photovoltaic power station construction capacity of candidate address j. For photovoltaic power stations, the photovoltaic power stations to be built are divided into Q _pv categories, and the construction capacity of different types of photovoltaic power stations is different. Therefore, w _j is the photovoltaic power station and electric vehicle charging network. Discrete optimization variables in the joint stochastic programming model; C _pv is the total construction capacity of the photovoltaic power station, which is jointly determined by factors such as the construction cost of the photovoltaic power station and the total investment to be invested in the construction of the photovoltaic power station;

Among them, P _r {·} represents the probability of random events in parentheses; k is the distribution node index; Ω _bus is the distribution node set; U _k is the voltage of node k, which is a random variable, and the probability distribution characteristics are analyzed by the probability power flow The results are given; U _N is the rated voltage of the distribution system; α% is the maximum allowable deviation percentage of the node voltage; β ₁ is the confidence level of the voltage exceeding the limit;

P _r {I _l >I _l,max }≤β ₂ l∈Ω _br (7)

Among them, I _l is the load current in the distribution line l, which is a random variable, and the probability distribution characteristics are given by the probabilistic power flow analysis results; I _l,max is the maximum allowable current of the distribution line l; β ₂ is the power flow over-limit confidence Spend.

As shown in FIG. 2 , the step S30 includes the following steps: S31 sets genetic algorithm parameters, the genetic algorithm parameters include population size N _pop , crossover rate P _c , mutation rate P _m and maximum evolutionary generation G _max .

S32 randomly generates an initial population consisting of N _pop chromosomes. According to the characteristics of the joint stochastic programming model of the photovoltaic power station and the electric vehicle charging network, an integer coding scheme is used to code the optimization problem. That is, each chromosome is composed of (N _ch +N _pv ) code bits, the first N _ch code bits represent the charging network construction plan, and the last N _pv code bits represent the photovoltaic power station construction plan. The i-th code bit represents the charging station construction status of the i-th candidate address (i=1, 2, 3, ···, N _ch ), and the value is "0", indicating that no charging station is to be built at the candidate address i, and the value is "0". The value is "q", indicating that the q-th type of charging station (q=1, 2, 3, ···, Q _ev ) is built at the candidate address i, and the corresponding construction capacity is C _ev,q . The N _ch +j code bit represents the photovoltaic power station construction status of the _jth candidate address (j=1, 2, 3, . Power station, the value is "m", indicating that the m-th type photovoltaic power station (m=1, 2, 3, ···, Q _pv ) is to be constructed at the candidate address j, and the corresponding construction capacity is C _pv,m . In order to satisfy the equality constraint given by formula (2), among the first N _ch code points of the chromosome, there are and only M _ch code points whose value is an integer other than "0"; Constrained by the formula, among the N _pv code points at the rear of the chromosome, there are and only M _pv code points whose value is an integer other than "0". Therefore, the chromosomes in the initial population are generated as follows: first, assign all code points of the chromosome to "0"; then, randomly select M _ch code points from the first N _ch code points, and change the assignment from "0" to "0" A random integer not greater than Q _ev ; finally, M _pv code bits are randomly selected from the last N _pv code bits, and the assignment is changed from "0" to a random integer not greater than Q _pv .

S33 The evolutionary algebra index g is initialized to 0, that is, g=0.

S34 set g=g+1, start the g-th generation evolution, and initialize the chromosome index n to 1, that is, set n=1.

S35 decodes the nth chromosome in the current population, determines the construction positions of M _ch electric vehicle charging stations, the construction capacity and the total construction capacity C _t-ev , determines the construction positions of M _ev photovoltaic power stations, and the construction capacity and the total construction capacity Construction capacity C _t-pv ; use the scenario probability method to calculate the probability power flow of the power distribution system, determine the expected power loss F _loss in the typical planned day, the probability distribution characteristics of the voltage amplitude of each node and the power flow of each line, and according to the formula (8 )～(12) Calculate the fitness V _fit,n of the nth chromosome:

V _fit,n =F _max -F _loss -η ₁ ×V _p1 -η ₂ ×V _p2 -η ₃ ×V _p3 -η ₄ ×V _p4 (8)

V _p1 =|C _ch -C _t-ev | (9)

V _p2 =|C _pv -C _t-pv | (10)

表示取

中较大的数，采用罚函数法分别处理公式(4)～(7)给出的约束，η₁、η₂、η₃以及η₄为罚系数；V_p1、V_p2、V_p3以及V_p4分别表示公式(4)～(7)给出约束的违背程度。可分别由公式(9)～(12)计算。Among them, _Fmax is a relatively large positive number given in advance to ensure that the chromosome fitness is non-negative, and the better the chromosome corresponds to the higher fitness value; the operator

means to take

For the larger number in , the penalty function method is used to deal with the constraints given by equations (4) to (7), respectively, η ₁ , η ₂ , η ₃ and η ₄ are penalty coefficients; V _p1 , V _p2 , V _p3 and V _p4 represents the degree of violation of the constraints given by formulas (4) to (7), respectively. It can be calculated by formulas (9) to (12) respectively.

S36 judges whether the fitness of all chromosomes in the current population has been calculated, that is, judges whether the chromosome index n is equal to the population size N _pop , if n<N _pop , then set n=n+1, and jump to step S35 to continue the chromosome Fitness calculation; otherwise, continue to step S37.

S37 judges whether the maximum evolutionary algebra is reached, that is, judges whether the evolutionary algebra index g is equal to the maximum evolutionary algebra G _max , if g=G _max , then continue to perform step S38; otherwise, based on the fitness, the current chromosome population is copied and crossed With the mutation operation, update the chromosome population, and jump to step S34.

In order to improve the optimization efficiency of the genetic algorithm, according to the characteristics of the joint stochastic programming model of the photovoltaic power station and the electric vehicle charging network, the crossover operator and mutation operator in the genetic algorithm are designed, as shown in Figure 3. At the same time, the equality constraints given by formula (2) and formula (3) are satisfied, and the crossover operation is performed according to the following steps: S41 randomly selects two chromosomes from the current population as the chromosomes to be crossed.

_S42 exchanges the chromosome code string after the _Nth code point of the two chromosomes with the crossover probability Pc, and completes the first crossover operation.

S43 randomly generates feasible crossover bits N _cr1 , exchanges the N _cr1 +1 to N _ch code bits in the two chromosomes with the crossover probability P _c , and completes the second crossover operation. Repeatedly randomly generate the candidate crossover bits N _can1 , where 1<N _can1 <N _ch , until a feasible crossover bit N _cr1 is found, and exchange the N _cr1 +1 to N _ch code bits in the two chromosomes to be crossed with the crossover probability P _c , to complete the second crossover operation. The criterion for judging whether N _can1 is a feasible crossover bit is as follows: in the code bits N _can1 +1 to N _ch , the code bits whose values are not "0" for the two chromosomes to be crossed are the same.

S44 randomly generates a feasible crossover bit N _cr2 , and exchanges the code string after the N _cr2 th code bit in the two chromosomes with the crossover probability P _c to complete the third crossover operation. Repeatedly and randomly generate the candidate crossover bit N _can2 , where N _ch <N _can2 <N _ch +N _pv , until a feasible crossover bit N _cr2 is found, exchange the N _cr2 th code point of the two chromosomes to be crossed with the crossover probability P _c . Chromosome code string, complete the third crossover operation. The criterion for judging whether N _can2 is a feasible crossover bit is as follows: after the N _can2th code bit, the two code bits whose values are not "0" for the two chromosomes to be crossed are the same.

As shown in Figure 4, in order to ensure that the mutated chromosome satisfies the equality constraints given by formula (2) and formula (3) at the same time, the mutation operation is performed as follows:

S51 randomly selects a chromosome from the current population as the chromosome to be mutated.

S52 randomly generates two code bits to be mutated N _mu1 and N _mu2 to ensure that one of the two code bits to be mutated is "0" and the other is an integer other than "0", where 1<N _mu1 <N _ch , 1< N _mu2 < N _ch .

S53 performs mutation operation on the mutation bits N _mu1 and N _mu2 at the same time with the mutation probability P _m , that is, the mutation bit whose value is "0" is mutated into a non-"0" random integer not greater than Q _ev , and the value is not "0" The to-be-mutated bit is mutated to "0", and the first mutation operation is completed.

S54 randomly generates two code bits to be mutated N _mu3 and N _mu4 to ensure that one of the values of the two code bits to be mutated is "0" and the other is an integer other than "0", where N _ch <N _mu3 <N _ch +N _pv , N _ch <N _mu4 <N _ch +N _pv .

S55 performs the second mutation operation on the mutation bits N _mu3 and N _mu4 at the same time with the mutation probability P _m , that is, the mutation bit whose value is "0" is mutated into a non-"0" random integer not greater than Q _pv , and the value is not The to-be-mutated bit of "0" is mutated to "0", and the second mutation operation is completed.

S38 outputs the photovoltaic power station and charging network construction scheme corresponding to the best chromosome of the current population as the optimal solution of the joint stochastic programming model of photovoltaic power station and electric vehicle charging network, and ends the algorithm process.

The method for planning a photovoltaic power station and an electric vehicle charging network based on the chance constraint of the present invention optimizes the construction address and construction capacity of the photovoltaic power station/charging station under the given conditions of the number of photovoltaic power stations/charging stations and the total construction capacity. On the premise of ensuring that the voltage offset of the distribution node and the line power flow exceed the limit to meet the opportunity constraints, the grid loss power in a typical day of the distribution system planning is minimized. The charging load of the electric vehicle charging station and the power generation output of the photovoltaic power station have random characteristics, which constitute the random disturbance faced by the joint planning of the photovoltaic power station and the electric vehicle charging station. In order to deal with the above random disturbances, a joint stochastic programming model of photovoltaic power station and electric vehicle charging network based on chance constraints is established: the optimization variables are the construction address and capacity of the charging station, and the construction address and capacity of the photovoltaic power station; the optimization goal is the distribution system planning within a typical day. The expected minimum power loss of the network; constraints include: the number of charging stations, the total capacity of charging stations, the number of photovoltaic power stations, the total capacity of photovoltaic power stations, the node voltage offset opportunity constraints and the line power flow overrun opportunity constraints. In order to use genetic algorithm to solve the joint stochastic programming model of photovoltaic power station and electric vehicle charging network, a special chromosome coding scheme and crossover and mutation operators are designed. The crossover operator includes three crossover operations, and the mutation operator includes two mutation operation. After the chromosome is decoded, the construction address and capacity of each photovoltaic power station and charging station can be determined. On this basis, through the probabilistic power flow analysis, the expected power loss and power flow probability distribution characteristics of the power distribution system in a typical day of the distribution system planning are determined, and the chromosome evaluation is completed. . According to the chromosome evaluation results, the genetic operation is used to update the algorithm population until the maximum evolutionary algebra is reached, and the planning results are output.

The above descriptions are only exemplary embodiments of the present invention, and are not intended to limit the scope of patent protection of the present invention. Any equivalent structure or equivalent process transformation made by using the contents of the description and drawings of the present invention, or directly or indirectly applied to other related The technical field of the present invention is similarly included in the scope of patent protection of the present invention.

Claims (5)

1. The photovoltaic power station and electric vehicle charging network planning method based on opportunity constraint is characterized by comprising the following steps:

s10, giving planning boundary conditions, the planning boundary conditions including: the method comprises the following steps that topological parameters of a power distribution system, loads of the power distribution system, daily charging load probability scene sets of charging stations, daily generation output probability scene sets of photovoltaic power stations, the total number of candidate addresses of the charging stations, the total number and the total capacity of construction of the charging stations, the type and the capacity of the charging stations to be built, the total number of the candidate addresses of the photovoltaic power stations, the total number and the total capacity of construction of the photovoltaic power stations, the type and the capacity of the photovoltaic power stations to be built, the maximum allowable deviation percentage of node voltages and the confidence coefficients of node;

s20, establishing a photovoltaic power station and electric vehicle charging network joint random planning model based on opportunity constraint by using the planning boundary conditions;

s30, designing a chromosome coding scheme and crossover and mutation operators according to the characteristics of the photovoltaic power station and electric vehicle charging network combined stochastic programming model, solving the photovoltaic power station and electric vehicle charging network combined stochastic programming model by adopting a genetic algorithm, and giving an optimal programming scheme of the photovoltaic power station and electric vehicle charging network.

2. The opportunity constraint-based photovoltaic power plant and electric vehicle charging network planning method according to claim 1, wherein the step S20 includes:

the optimization goal of the photovoltaic power station and electric vehicle charging network combined stochastic programming model is to reduce the network loss electric quantity within a typical day of power distribution system programming, as shown in formula (1),

wherein, F_lossPlanning the expected grid loss capacity in a typical day for the power distribution system; t is a tide analysis time interval index, T_fThe number of time periods for power flow analysis in a typical day; l is distribution line index; omega_brIndexing a set for a distribution line; delta P_loss,l,tThe loss power of the distribution line l in the power flow analysis time period t is a random variable; e (-) is an operator expecting a random variable;

the opportunity constraints include an opportunity constraint representing the total number of construction of the charging station, an opportunity constraint representing the total number of construction of the photovoltaic power station, an opportunity constraint representing the total construction capacity of the charging station, an opportunity constraint representing the total construction capacity of the photovoltaic power station, an opportunity constraint representing the voltage deviation of the node, and an opportunity constraint representing the line power flow out-of-limit, which are obtained through equations (2) to (7), respectively,

wherein M is_chFor chargingTotal station construction; n is a radical of_chA total number of candidate addresses for the charging station; i is a candidate address index; x is the number of_iThe optimization variables are 0-1 optimization variables in the photovoltaic power station and electric vehicle charging network combined stochastic programming model, wherein the condition that a charging station is built at a candidate address i is taken as '1', the condition that a charging station is not built at the candidate address i is taken as '0', and i is 1,2,3, …, N_ch；

Wherein M is_pvThe total number of the photovoltaic power station is built; n is a radical of_pvThe total number of the candidate addresses of the photovoltaic power station is; j is a candidate address index; y is_jThe optimization variables are 0-1 optimization variables in a combined stochastic programming model of the photovoltaic power station and the electric vehicle charging network, wherein 1 is taken to represent that the photovoltaic power station is built at a candidate address j, 0 is taken to represent that the photovoltaic power station is not built at the candidate address j, j is 1,2,3, …, and N_pv；

Wherein z is_iEstablishing capacity for charging stations of the candidate address i, and dividing the to-be-established electric vehicle charging station into Q for an electric vehicle charging network_evClass z_iThe method comprises the following steps of (1) obtaining a discrete optimization variable in a photovoltaic power station and electric vehicle charging network combined random planning model; c_chTotal construction capacity for the charging station;

wherein, w_jFor the photovoltaic power station construction capacity of the candidate address j, for the photovoltaic power station, the photovoltaic power station to be constructed is divided into Q_pvClass w_jThe method comprises the following steps of (1) obtaining a discrete optimization variable in a photovoltaic power station and electric vehicle charging network combined random planning model; c_pvThe total construction capacity of the photovoltaic power station;

wherein, P_r{. denotes the probability of a random event occurring in parentheses; k is the distribution node index; omega_busIs a power distribution node set; u shape_kThe voltage of the node k is a random variable, and the probability distribution characteristic is given by a probability load flow analysis result; u shape_NRated voltage for the distribution system; alpha% is the maximum voltage allowed deviation percentage of the node; beta is a₁Is the voltage out-of-limit confidence;

P_r{I_l>I_l,max}≤β₂ l∈Ω_br (7)

wherein, I_lThe load current in the distribution line l is a random variable, and the probability distribution characteristic is given by a probability load flow analysis result; i is_l,maxThe maximum allowable current of the distribution line l; beta is a₂And the confidence of the power flow out-of-limit.

3. The opportunity constraint-based photovoltaic power plant and electric vehicle charging network planning method according to claim 2, wherein the step S30 comprises the steps of:

s31 setting genetic algorithm parameters including a population size N_popCross over ratio P_cThe rate of variation P_mAnd maximum evolution algebra G_max；

S32 random generation of N_popAn initial population consisting of the chromosome bars is used for coding an optimization problem by adopting an integer coding scheme according to the characteristics of a photovoltaic power station and electric vehicle charging network combined random programming model;

the S33 evolution algebra index g is initialized to 0, that is, g is made to be 0;

s34 sets g to g +1, and starts the evolution of the g-th generation, where the chromosome index n is initialized to 1, that is, n is set to 1;

s35, decoding the nth chromosome in the current population to determine M_chConstruction position of electric vehicle charging stationSet capacity and total construction capacity C_t-evDetermining M_evConstruction position, construction capacity and total construction capacity C of individual photovoltaic power station_t-pv(ii) a Calculating the probability load flow of the power distribution system by adopting a scene probability method, and determining the expected F of the network loss electric quantity in a planning typical day_lossCalculating the fitness V of the nth chromosome according to the formulas (8) to (12) based on the probability distribution characteristics of the voltage amplitude of each node and the power flow of each line_fit,n：

V_fit,n＝F_max-F_loss-η₁×V_p1-η₂×V_p2-η₃×V_p3-η₄×V_p4 (8)

V_p1＝|C_ch-C_t-ev| (9)

V_p2＝|C_pv-C_t-pv| (10)

Wherein, F_maxOperators for a predetermined relatively large positive number to ensure non-negative fitness of the chromosome

Express get

The larger number in the middle is treated by a penalty function method according to the constraints eta given by the formulas (4) to (7)₁、η₂、η₃And η₄Is a penalty factor; v_p1、V_p2、V_p3And V_p4Respectively expressing the violation degrees of the constraints given by formulas (4) to (7);

judgment at S36Whether the fitness of all chromosomes in the current population is calculated or not is judged, namely whether the chromosome index N is equal to the population scale N or not is judged_popIf n is<N_popIf so, let n be n +1, and go to step S35 to continue the chromosome fitness calculation; otherwise, continuing to execute step S37;

s37 judges whether the maximum evolution algebra is reached, i.e. whether the evolution algebra index G is equal to the maximum evolution algebra G_maxIf G is G_maxThen, go on to step S38; otherwise, based on the fitness, the current chromosome population is copied, crossed and mutated, the chromosome population is updated, and the step S34 is skipped;

s38, outputting the photovoltaic power station and charging network construction scheme corresponding to the current population top-off chromosome as the optimal solution of the photovoltaic power station and electric vehicle charging network combined stochastic programming model, and ending the algorithm process.

4. The opportunity constraint-based photovoltaic power plant and electric vehicle charging network planning method according to claim 3, characterized in that said interleaving operation comprises the following steps:

s41 randomly selecting two chromosomes from the current population as chromosomes to be crossed;

s42 with cross probability P_cExchanging the Nth of two chromosomes_chThe chromosome code string after the code bit completes the first cross operation; s43 random generation of feasible cross bit N_cr1With a cross probability P_cCrossover of the Nth of the two chromosomes_cr1+1 to N_chCode bits, completing the second crossing operation;

s44 random generation of feasible cross bit N_cr2With a cross probability P_cCrossover of the Nth of the two chromosomes_cr2And (5) completing the third cross operation by the code string after the code bit.

5. The opportunity constraint-based photovoltaic power plant and electric vehicle charging network planning method of claim 3, wherein the mutation operator comprises the steps of:

s51 randomly selecting a chromosome from the current population as a chromosome to be mutated;

s52 randomly generating two code bits N to be varied_mu1And N_mu2Ensuring that one of the values of the two code bits to be varied is '0' and the other is a non-0 integer;

s53 mutation probability P_mTreating simultaneously ectopic N_mu1And N_mu2Performing mutation operation, namely changing the position to be mutated with the value of 0 to be not more than Q_evThe random integer of not 0, the bit to be mutated which takes the value of not 0 is mutated into 0, and the first mutation operation is completed;

s54 randomly generating two code bits N to be varied_mu3And N_mu4Ensuring that one of the values of the two code bits to be varied is '0' and the other is a non-0 integer;

s55 mutation probability P_mTreating simultaneously ectopic N_mu3And N_mu4Performing a second mutation operation to obtain a mutation position with a value of 0 and a value of not more than Q_pvThe random integer of not 0, the mutation position to be varied which takes the value of not 0 is changed into 0, and the second mutation operation is completed.

Images