CN114329783A - Multi-target electric vehicle charging network planning method - Google Patents
Multi-target electric vehicle charging network planning method Download PDFInfo
- Publication number
- CN114329783A CN114329783A CN202111653174.4A CN202111653174A CN114329783A CN 114329783 A CN114329783 A CN 114329783A CN 202111653174 A CN202111653174 A CN 202111653174A CN 114329783 A CN114329783 A CN 114329783A
- Authority
- CN
- China
- Prior art keywords
- charging
- chromosome
- chromosomes
- network
- population
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Withdrawn
Links
Images
Landscapes
- Electric Propulsion And Braking For Vehicles (AREA)
- Charge And Discharge Circuits For Batteries Or The Like (AREA)
Abstract
The invention relates to the technical field of electric vehicle charging networks, and discloses a multi-target electric vehicle charging network planning method. The method comprises the following steps: s1: setting a planning boundary condition; s2: establishing a charging network multi-target opportunity constraint planning model considering the shortest average charging travel distance of the electric automobile and the lowest construction cost of a charging network; s3: designing a chromosome coding scheme and crossover and mutation operators, solving a charging network multi-target opportunity constraint planning model by adopting a non-dominated genetic algorithm based on a feasibility rule, and providing a Pareto solution set of the charging network multi-target opportunity constraint planning model; s4: and determining an optimal charging network planning scheme according to a principle of maximum marginal investment income from a solution with lowest investment cost in a Pareto solution set. The invention reduces the average running distance from the electric automobile to the charging station and the construction cost of the charging network.
Description
Technical Field
The invention relates to the technical field of electric vehicle charging network planning, in particular to a multi-target electric vehicle charging network planning method.
Background
In recent years, with the overuse of fossil energy and the increasing increase of environmental pollution, the development of green vehicles represented by electric vehicles becomes a consensus of countries in the world, and China also sets a series of policies to encourage the development of the electric vehicle industry. At present, electric vehicles are mainly supplemented with electric energy from a power grid, and a charging station is one of important charging places. The reasonable-layout electric automobile charging network can remarkably improve the charging convenience of the automobile owner, stimulate the purchasing desire of the electric automobiles in the whole society and promote the further development of the electric automobile industry.
In the literature, "electric vehicle charging network opportunity constraint planning under multiple stochastic characteristics" (power system protection and control, 2021, volume 49, phase 6, pages 30 to 39), a charging network planning model is established, which takes into account the stochastic characteristics of distributed photovoltaic output and electric vehicle charging load, and by optimizing the access positions and capacities of charging stations in a power distribution system, the network loss of the power distribution system is minimized, and the voltage deviation of nodes of the power distribution system and the branch load current violation meet the given opportunity constraint. The average charging running distance of the electric automobile is one of important indexes for measuring the convenience of the charging service, the problem is not considered, meanwhile, the construction cost of the optimized charging network is not considered, and certain limitation is realized. In document two, urban electric vehicle charging station multi-target planning (protection and control of electric power system, 2021, volume 49, phase 5, page 67 to 80) considering multi-demand scenarios adopts a K-means clustering algorithm to calculate electric vehicle charging demands under multiple scenarios, and on the basis, a multi-target planning model considering user charging total cost and charging station total construction operation and maintenance cost is established, so as to optimize charging station construction addresses and capacity. However, this document does not consider the charging convenience of the owner of the electric vehicle, and has certain limitations. In the third document, "charging station planning taking user travel characteristics and distribution network line availability margin into consideration" (power system automation, 2018, volume 42, phase 23, pages 48 to 56), an electric vehicle user annual time consumption total cost model is established, an electric vehicle charging network planning model with optimization objectives of annual time consumption total cost of electric vehicle users, minimum sum of charging station construction operation and maintenance cost and distribution network annual construction operation and maintenance cost is provided, and site selection, capacity and access feeder path construction of a charging station are optimized. Like the second document, the second document also does not consider the construction cost of the charging station, and has certain limitations.
The electric automobile charging network is an important infrastructure for supporting the development of the electric automobile industry, and the unreasonable charging network can not only reduce the charging convenience of an electric automobile owner, but also hinder the further development of the electric automobile industry. Therefore, a method for optimizing and planning a charging network of an electric vehicle is urgently needed, and the average charging running distance of an electric vehicle owner and the construction cost of the charging network are reduced by optimizing the access position of a charging station. However, the existing methods do not consider the average charging travel distance and the charging network construction cost at the same time, and have certain limitations.
Disclosure of Invention
Aiming at the defects in the prior art, the invention aims to provide a multi-target electric vehicle charging network planning method, which is used for optimizing the construction address of an electric vehicle charging station under the condition that the construction number of the charging station, the construction cost of candidate sites and the construction cost of each candidate site are given, and reducing the average charging running distance of an electric vehicle owner and the total construction cost of a charging network on the premise of ensuring that the charging running distance of the electric vehicle meets the given opportunity constraint
In order to achieve the purpose, the invention adopts the following technical scheme:
a multi-target electric vehicle charging network planning method comprises the following steps:
s1: setting planning boundary conditions, wherein the planning boundary conditions comprise: the method comprises the following steps that (1) a traffic network topological structure and parameters, charging station candidate addresses, charging station construction cost of each candidate site, charging station construction total number, charging mileage threshold value and confidence coefficient are obtained; the candidate addresses of the charging stations are all traffic nodes in a traffic network;
s2: establishing a charging network multi-target opportunity constraint planning model considering the shortest average charging running distance of the electric vehicle and the lowest construction cost of a charging network at the same time, wherein opportunity constraints of the charging network multi-target opportunity constraint planning model comprise charging running distance opportunity constraints and charging station construction total number constraints;
the shortest average charging travel distance of the electric automobile can be represented by the following formula:
wherein D isaveAverage charging travel distance, omega, for electric vehicles in the entire traffic networkRFor a set of roads in a traffic network, i is a road index in the traffic network, TiTraffic flow for road i, dav,iThe average charging running distance of the electric automobile running on the road i is obtained;
the lowest construction cost of the charging network can be expressed by the following formula:
wherein, CtotalCost of construction of charging network for electric vehicle, Ccon,jBuilding cost for building the charging station at the candidate address j, wherein N is the total number of the candidate addresses of the charging station;
the charging travel distance opportunity constraint is the probability that the charging travel distance is lower than a given mileage threshold value does not exceed a given confidence coefficient, and can be expressed by the following formula:
wherein p isiThe probability that the EV charging travel mileage on the road i does not exceed the charging travel mileage threshold value is given, and beta is a confidence coefficient;
the total number of charging station construction constraints are shown as follows:
wherein M is the number of charging station constructions; n is a charging station candidate address in the traffic network and is positioned at a traffic network node; y isjFor the optimization variables in the charging network planning model, taking "1" indicates that a charging station is built at the candidate address j, and taking "0" indicates that a charging station is not built at the candidate address j.
S3: solving the multi-target opportunity constraint planning model of the charging network by adopting a non-dominated genetic algorithm based on a feasibility rule, and giving a Pareto solution set of the multi-target opportunity constraint planning model of the charging network;
the solving of the multi-target opportunity constraint planning model of the charging network by adopting the non-dominated genetic algorithm based on the feasibility rule specifically comprises the following steps:
s31, setting genetic algorithm parameters, including: population size NpopCross rate PcThe rate of variation PmAnd maximum evolution algebra Gmax；
S32, randomly generating a random number NpopParent population Q consisting of chromosome bars1,1(ii) a Parent population Q1,1The chromosome in the site is a binary code string consisting of N code bits, the j code bit value of 1 indicates that a charging station is built in the candidate site j, otherwise, a charging station is not built in the candidate site j; each chromosome has and only has M code bits to take the value of '1';
s33, making g equal to 0, wherein g is an evolution algebraic index;
s34, enabling g to be g +1, starting the evolution of the g generation, and adopting a binary tournament method to carry out the evolution of the parent generation Q1,gPerforming copy operation, crossover operation and mutation operation to generate a temporary population Q2,gTemporary population Q2,gPopulation size of NpopAnd a temporary population Q2,gAnd the parent population Q1,gCombining to form a population Q to be evolved3,gPopulation Q to be evolved3,gThe population size of (2N)pop；
S35, calculating an evolutionary population Q3,gThe average driving distance D from the electric vehicle to the nearest charging station in the planning scheme represented by each chromosomeave,kAnd the comprehensive construction cost C of the charging networktotal,kProbability P that the charging driving distance from the electric automobile does not exceed a given mileage threshold valueev,k；
S36, determining a population Q to be evolved based on feasibility rule3,gThe priorities of all chromosomes in the population Q are determined, the chromosomes are sorted according to the priorities of the chromosomes, and then the population Q to be evolved is determined according to the priorities of the chromosomes3,gDesigning chromosome fitness by the order in (1);
determining a population Q to be evolved based on a feasibility rule3,gThe priorities of all chromosomes in the chromosome library are specifically:
in the population Q to be evolved3,gClassifying chromosomes meeting the opportunity constraint into a feasible solution set, layering all chromosomes in the feasible solution set according to a Pareto domination principle, and establishing the cost C of all chromosomes of each layer according to a charging networktotalCalculating the virtual fitness of each chromosome in each level after sequencing from small to large; classifying chromosomes which do not meet the opportunity constraint into a non-feasible solution set, calculating the constraint violation degree of each chromosome in the non-feasible solution set, and determining the chromosome according to the following principleA priority;
(1) chromosomes in the feasible solution set have a higher priority than chromosomes in the infeasible solution set;
(2) in the feasible solution set, the lower the level of the chromosome is, the higher the priority is; when the levels of chromosomes are the same, the chromosome with large virtual fitness has high priority, and the calculation method of the virtual fitness is shown as the following formula:
wherein, Fr,nIs the virtual fitness of the nth chromosome of the r layer; r ismaxFor the population Q to evolve3,gThe number of chromosome layers of the feasible solution set; n isrThe number of the chromosome of the r layer;
the comprehensive construction cost of the charging network corresponding to the r-th chromosome and the (n + 1) th chromosome is saved,
the average charging travel distance corresponding to the average charging travel distance;
the comprehensive construction cost of the charging network corresponding to the r-th chromosome and the n-1 th chromosome is saved,
the average charging travel distance corresponding to the average charging travel distance; h is a given large number;
(4) in the infeasible solution set, the smaller the constraint violation degree of the chromosome is, the higher the priority is, and the constraint violation degree calculation formula is as follows:
wherein, CVRepresenting the degree of constraint violation of the planning scheme corresponding to a chromosome, PevThe probability that the charging travel distance of the electric automobile in a planning scheme corresponding to a certain chromosome does not exceed a given mileage threshold value is given;
s37, judging whether the evolution algebra index G is equal to the maximum evolution algebra Gmax(ii) a If G is GmaxThen, go on to step S38; otherwise, adopting the strategy of 'elite selection' to take the first NpopThe chromosome is taken as a parent population, and the step S34 is skipped;
s38, outputting a chromosome of a first level in the feasible solution set as a Pareto optimal solution set of the multi-target opportunity constraint planning model of the electric vehicle charging network;
and S4, determining an optimal charging network planning scheme according to the principle of maximum marginal investment income by taking the charging network planning scheme with the lowest comprehensive construction cost of the Pareto solution centralized charging network as a reference.
Further, the calculation formula of the average charging travel distance of the travel EV on the road i is as follows:
wherein ld,iIs the length of road i; f. ofi(x) The charging travel distance is the EV on the road i at a distance x from the starting point.
Further, the traffic flow of the road i is set to be omega by the shortest path of the traffic networkqThe sum of the traffic flow of the shortest path passing through the road i is obtained; the calculation formula of the shortest path traffic flow is as follows:
in the formula, WS,qAnd WE,qRespectively are the weight coefficients of the starting point and the end point of the shortest path q; omegaqIs a traffic network shortest path set; dqIs the length of the shortest path q.
Further, the charging travel distance f of the EV at the distance x from the starting point on the road ii(x) Is calculated byThe formula is as follows:
fi(x)＝min[x+l1,ld,i-x+l2]0≤x≤ld,i；
wherein x is a random variable representing the distance between the EV driving on the road i and the starting point of the road i; ld,iRepresents the length of road i; l1Representing the distance between the energy storage charging station closest to the starting point of the road i and the starting point of the road i; l2Is the distance from the nearest charging station to the end of the road i.
Further, the probability that the charging travel distance of the EV on the road i does not exceed the charging travel range threshold may be calculated by:
wherein, gi(x) D auxiliary function for judging whether the charging travel distance of EV from the end point x on the road i does not exceed the charging travel mileage threshold valuecha-limAnd the charging mileage threshold value.
Further, the layering is performed on all chromosomes in the feasible solution set according to the Pareto domination principle, specifically:
s361, initializing a chromosome level index r to be 1;
s362, traversing unclassified chromosomes in the feasible solution set, searching all unclassified chromosomes in the unclassified chromosomes, and classifying the unclassified chromosomes into an r level; wherein the unopposed chromosome is a chromosome in which two optimization objectives are absent from the unorganized chromosomes and which is superior to the unopposed chromosome; the two optimization targets are that the average charging running distance of the electric automobile is shortest and the construction cost of a charging network is lowest;
s363, determining whether there is an unclassified chromosome in the feasible solution set, if yes, making r ═ r +1, and jumping to step S362; otherwise, ending the chromosome layering work in the feasible solution set.
Further, in step S34, the specific steps of the interleaving operation are as follows:
(1) randomly selecting two chromosomes from the current chromosome population as chromosomes to be crossed;
(2) repeatedly and randomly generating cross bit N to be selectedcro(1<Ncro<N) until a feasible cross point N is foundavI.e. the Nth of two chromosomes to be crossedcroThe number of code bits with the value of 1 after the code bits is consistent;
(3) with a cross probability PcExchange of the crossover site N of two chromosomesavThe subsequent binary code string completes the cross operation;
the mutation operation comprises the following specific steps:
(1) randomly selecting a chromosome from the current chromosome population as a chromosome to be mutated;
(2) randomly generating two code bits N to be variedmut1And Nmut2(1≤Nmut1≤N,1≤Nmut2N) or less), two code bits to be varied Nmut1And Nmut2The value cannot be simultaneously '1' or '0';
(3) with a mutation probability PmTreating simultaneously the variant code bit Nmut1And Nmut2And performing variation operation, wherein the code bit to be varied with the value of 1 is varied into 0, and the code bit to be varied with the value of 0 is varied into 1.
Further, step S4 specifically includes the following steps:
s41, according to the construction cost C of the charging network of the electric automobiletotalArranging all chromosomes in the Pareto optimal solution set from small to large;
s42, calculating the marginal investment profit of all chromosomes except the 1 st chromosome according to the following formula:
wherein E ismConcentration of the mth stain for Pareto optimal solutionMarginal investment profit for the body; ctotal,mAnd Ctotal,1The construction cost D of the electric vehicle charging network corresponding to the mth chromosome and the 1 st chromosome in the Pareto optimal solution set respectivelyave,mAnd Dave,1Respectively the average charging travel distances corresponding thereto; m ismaxThe number of chromosomes in the set for Pareto optimal solution;
and S43, outputting a planning scheme corresponding to the chromosome with the maximum marginal investment income in the Pareto optimal solution set as an optimal charging network planning scheme.
Compared with the prior art, the multi-target electric vehicle charging network planning method provided by the invention can optimize the construction position of the electric vehicle charging station under the condition that the total construction number of the charging station and the candidate address of the charging station are given, reduce the average charging running distance of the electric vehicle and the construction cost of the charging network, keep the probability that the running distance of the electric vehicle running to the charging station does not exceed the given mileage threshold value at a higher level, and provide the optimal construction scheme of the electric vehicle charging network.
Drawings
FIG. 1 is a schematic diagram illustrating a calculation of a charging travel distance of an electric vehicle according to an embodiment of the present invention;
FIG. 2 is a schematic diagram of a solving step of the electric vehicle charging network optimization model provided by the invention;
fig. 3 is a topology structure diagram of a node traffic network according to an embodiment of the present invention;
fig. 4 is a charging travel distance cumulative probability density curve under the optimal planning scheme provided by the embodiment of the present invention;
FIG. 5 is a probability density curve under the optimal planning scheme provided by the embodiment of the present invention;
fig. 6 is an optimal planning scheme provided by an embodiment of the present invention.
Detailed Description
The present invention will be further explained with reference to the drawings so that those skilled in the art can more thoroughly understand the present invention and can smoothly carry out the present invention, but the following reference examples are only for explaining the present invention and are not to be construed as limiting the present invention.
The embodiment of the invention provides a multi-target electric vehicle charging network planning method, which comprises the following steps:
s1: setting planning boundary conditions, wherein the planning boundary conditions comprise: the method comprises the following steps that (1) a traffic network topological structure and parameters, charging station candidate addresses, charging station construction cost of each candidate site, charging station construction total number, charging mileage threshold value and confidence coefficient are obtained; the charging station candidate addresses are all traffic nodes in a traffic network.
S2: and establishing a charging network multi-target opportunity constraint planning model considering the shortest average charging travel distance of the electric automobile and the lowest construction cost of the charging network.
1. First optimization objective of multi-objective opportunity constraint planning model of charging network
The first optimization target of the multi-target opportunity constraint planning model of the charging network is that the average charging travel distance of the electric automobile is shortest, and can be represented by formula (1):
in the formula (1), wherein DaveThe average driving distance from all EVs in the traffic network to the nearest charging station; omegaRIs a set of roads in a traffic network; t isiThe traffic flow of the road i; dav,iThe average charging travel distance of the travel EV on the road i.
(1) Calculation of traffic flow for road i
The traffic flow of the road i is integrated into omega by the shortest path of the traffic networkqThe sum of the traffic flow of the shortest path passing through the road i.
It is assumed that EVs always travel along the shortest path between traffic nodes (important nodes in the traffic network). The shortest path traffic flow can be calculated by a gravity space interaction model, which is shown in formula (2):
in the formula (2), WS,qAnd WE,qRespectively are the weight coefficients of the starting point and the end point of the shortest path q; omegaqThe shortest path set of the traffic network can be obtained through a Flyod algorithm; dqThe length of the shortest path q can be calculated by the set of roads passed by the shortest path q and the length of the roads.
(2) Calculation of EV charging travel distance on road i
In a traffic network, after an EV generates a charging demand, a vehicle owner drives to a charging station closest to the EV under guidance of navigation software represented by a Baidu map to charge the vehicle, and fig. 1 is a schematic diagram illustrating calculation of an EV charging travel distance.
In FIG. 1, AB denotes a road in the traffic network and has a length lAB(ii) a The charging station closest to the traffic node A is located at the traffic node C, AC represents the shortest path from the traffic node A to the traffic node C, and the length of the charging station is lAC(ii) a The charging station closest to the traffic node B is positioned at the traffic node D, BD represents the shortest path from the traffic node B to the traffic node D, and the charging station also comprises a plurality of roads and traffic nodes, and has the length of lBD. On the road AB, the distance between a certain EV and the traffic node A is x (x is more than or equal to 0 and less than or equal to lAB) If it generates a charging demand, there are two charging schemes to be selected: 1) passing through the traffic node A and reaching a charging station built at the traffic node C for charging, wherein the charging travel distance is x + lAC(ii) a 2) Passing through the traffic node B, and arriving at a charging station built at a traffic node D for charging, wherein the charging travel distance is lAB-x+lBD. Obviously, the owner of the vehicle will select the nearest charging station to charge, and the EV charging travel distance f (x) on the road can be calculated according to the following formula:
f(x)＝min[x+lAC,lAB-x+lBD]0≤x≤lAB (3)
as can be seen from equation (3), the charging travel distance is related to various factors such as the topology of the traffic network, the construction conditions of the charging station, and the position of the EV on the road.
According to the formula (3), the charging travel distance f of the EV at the distance x from the starting point on the road i can be set toi(x) Expressed as shown in equation (4):
fi(x)＝min[x+l1,ld,i-x+l1]0≤x≤ld,i (4)
in the formula (4), x represents the distance between the EV running on the road i and the end point traffic node A of the road i, x is a random variable, and the obedient section [0, ld,i]Uniform distribution of the components; ld,iRepresents the length of road i; l1Representing the distance between the energy storage charging station closest to the starting point of the road i and the starting point of the road i; l2Is the distance from the nearest charging station to the end of the road i.
(3) Average charging travel distance d of electric vehicle traveling on road iav,iIs calculated by
Average charging travel distance d of electric vehicle traveling on road iav,iThe calculation of (2) is specifically shown in formula (5).
In the formula (5), ld,iIs the length of road i; x is the distance between the EV to be charged and the end point of the road i; f. ofi(x) The charging travel distance is the EV on the road i at a distance x from the starting point.
2. Second optimization objective of multi-objective opportunity constraint planning model of charging network
The second optimization objective of the multi-objective opportunity constraint planning model of the charging network is that the construction cost of the charging network is the lowest, and is specifically shown in formula (6):
in the formula (6), CtotalCost is built for the electric vehicle charging network; ccon,jThe construction cost for constructing the charging station at the candidate address j consists of land cost, equipment cost, access power system cost, operation and maintenance cost and network loss cost; n is the total number of charging station candidate sites.
Opportunity constraints for EV energy storage charging network planning model
(1) Charged travel distance opportunity constraints
In the traffic network, the charging travel distance of the electric automobile is a random variable, and in order to avoid overlong charging travel distance, a charging travel distance opportunity constraint is set in a multi-target opportunity constraint planning model of the electric automobile charging network, as shown in formula (7),
wherein p isiAnd the probability that the EV charging travel distance on the road i does not exceed the charging travel distance threshold value is determined, and beta is the confidence coefficient that the EV charging travel distance constraint is met.
Probability p that EV charging mileage on road i does not exceed charging mileage threshold valueiThe calculation formula of (a) is as follows:
wherein, gi(x) In order to judge whether the charging travel distance of the EV on the road i from the endpoint x does not exceed an auxiliary function of a mileage threshold, two values of '0' and '1' are provided, and the two values are as follows:
(2) energy storage charging station construction number constraint
In the energy storage charging network planning, a planner determines the construction number and the address to be selected of a charging station in a traffic network according to boundary conditions such as the planned investment amount of charging network construction, municipal planning and EV permeability, so that the model has the following constraints:
wherein M is the number of charging station constructions; n is a charging station candidate in a traffic networkAn address located at a traffic network node; y isjFor the optimization variables in the charging network planning model, taking "1" indicates that a charging station is built at the candidate address j, and taking "0" indicates that a charging station is not built at the candidate address j.
S3: solving the multi-target opportunity constraint planning model of the charging network by adopting a non-dominated genetic algorithm based on a feasibility rule, and providing a Pareto solution set of the multi-target opportunity constraint planning model of the charging network.
A flowchart for solving the multi-objective opportunity constraint programming model of the charging network by using a non-dominated genetic algorithm based on the feasibility rule is shown in fig. 2, and specifically comprises the following steps:
s3.1: setting genetic algorithm parameters, including: population size NpopCross rate PcThe rate of variation PmAnd maximum evolution algebra Gmax；
S3.2: random generation of NpopParent population Q consisting of chromosome bars1,1(ii) a Parent population Q1,1The chromosome in the site is a binary code string consisting of N code bits, the j code bit value of 1 indicates that a charging station is built in the candidate site j, otherwise, the charging station is not built in the site; the constraint conditions given by the formula (8) indicate that the total number of charging stations in the electric vehicle charging network is M, so that in the initial parent population, each chromosome has and only has M code bits with the value of 1;
the generation steps of each chromosome in the parent population of the initial species described in step S3.2 are as follows:
(1) assigning all code bits of the chromosome to be 0;
(2) randomly picking M code bits from the chromosome, and changing the assignment from '0' to '1'.
S3.3: the evolution algebra index g is initialized to 0, namely, g is 0;
s3.4: starting the evolution of the g-th generation when g is g +1, and adopting a binary championship method to carry out the evolution of the parent population Q1,gPerforming duplication, crossing and mutation operations to generate the same number of N populationspopTemporary population Q of2,gAnd then it is mixed with the parent population Q1,gCombined to form the evolution to be proceededGroup Q3,gThe size of the population is 2Npop。
The crossover operator of step S3.4 specifically includes the following steps:
(1) randomly selecting two chromosomes from the current chromosome population as chromosomes to be crossed;
(2) repeatedly and randomly generating cross bit N to be selectedcro(1<Ncro<N) until a feasible cross point N is foundavI.e. the Nth of two chromosomes to be crossedcroThe number of code bits with the value of 1 after the code bits is consistent;
(3) with a cross probability PcExchange of the crossover site N of two chromosomesavAnd the subsequent binary code string completes the cross operation.
The mutation operator in step S3.4 specifically includes the following steps:
(1) randomly selecting a chromosome from the current chromosome population as a chromosome to be mutated;
(2) randomly generating two code bits N to be variedmut1And Nmut2(1≤Nmut1≤N,1≤Nmut2N) or less), two code bits to be varied Nmut1And Nmut2The value cannot be simultaneously '1' or '0';
(3) with a mutation probability PmTreating simultaneously the variant code bit Nmut1And Nmut2And performing variation operation, wherein the code bit to be varied with the value of 1 is varied into 0, and the code bit to be varied with the value of 0 is varied into 1.
S3.5: the chromosome index k is initialized to 1, that is, k is 1;
s3.6: population Q to be evolved3,gThe k chromosome in (a) is decoded to determine the construction positions of the M electric vehicle charging stations, and the average driving distance D from the electric vehicle to the nearest charging station under the planning scheme represented by the k chromosome is calculated according to the method introduced in step S2ave,kAnd the comprehensive construction cost C of the charging networktotal,kProbability P that the charging driving distance from the electric automobile does not exceed a given mileage threshold valueev,k；
S3.7: judging whether to calculate the population Q to be evolved3,gD corresponding to all chromosomes inave,k、Ctotal,kAnd Pev,kNamely, whether the chromosome index k is equal to the population Q to be evolved3,gPopulation size of 2Npop(ii) a If k is<2NpopIf yes, let k be k +1, and jump to step S3.6, otherwise, continue to execute the next step S3.8;
s3.8: determination of a population Q to be evolved on the basis of a feasibility rule3,gThe priority of all chromosomes in the chromosome library is determined, the chromosomes are sorted according to the priority of each chromosome, the higher the priority is, the more the chromosomes are sorted, and the specific steps are as follows:
s3.8.1: the population Q to be evolved3,gThe chromosome in (1) is divided into two parts: one part is chromosomes satisfying opportunity constraints and is called a feasible solution set; the other part is chromosomes which do not meet the opportunity constraint and are called a non-feasible solution set; when the priority is determined, the chromosomes in the feasible solution set are superior to the chromosomes in the unreliable solution set;
s3.8.2: determining the priority of all chromosomes in the feasible solution set, and the specific steps are as follows:
s3.8.2.1, layering all chromosomes in the feasible solution set according to the Pareto domination principle, wherein the lower the hierarchy the chromosomes belong to, the higher the priority, the specific steps are as follows:
s3.8.2.1.1 chromosome level index r is initialized to 1;
s3.8.2.1.2 traversing unclassified chromosomes in the feasible solution set, searching all unclassified chromosomes in the unclassified chromosomes, and classifying the unclassified chromosomes into the r level; wherein the unopposed chromosome is a chromosome in which two optimization objectives are absent from the unorganized chromosomes and which is superior to the unopposed chromosome; the two optimization targets are that the average charging running distance of the electric automobile is shortest and the construction cost of a charging network is lowest;
s3.8.2.1.3, judging whether there is any un-classified chromosome in the feasible solution set, if yes, making r equal to r +1 and jumping to step S3.8.2.1.2; otherwise, ending the chromosome layering work in the feasible solution set.
S3.8.2.2 cost C of charging network construction for all chromosomes in each leveltotalSequencing from small to large, and calculating the virtual fitness of each chromosome in the same level on the basis; in the same level, the chromosome with large virtual fitness has high priority, and the calculation method is as follows:
wherein, Fr,nIs the virtual fitness of the nth chromosome of the r layer; r ismaxFor the population Q to evolve3,gThe number of chromosome layers of the feasible solution set; n isrThe number of the chromosome of the r layer;
the comprehensive construction cost of the charging network corresponding to the r-th chromosome and the n-th chromosome is saved,
the average charging travel distance corresponding to the average charging travel distance; h is given as a large number.
S3.8.3: determining the priority of all chromosomes in the infeasible solution set according to the violation degree of the constraint solution, wherein the smaller the violation degree of the constraint solution is, the higher the priority is, and the violation degree of the chromosome constraint is calculated according to the formula (12):
wherein, CVRepresenting the degree of constraint violation of the planning scheme corresponding to a chromosome, PevThe probability that the charging travel distance of the electric automobile in a planning scheme corresponding to a certain chromosome does not exceed a given mileage threshold value is given;
s3.8.4: the chromosomes are sorted according to priority, and the higher the priority, the earlier the sorting.
S3.9: judging whether the maximum evolution algebra is reached, namely judging whether the evolution algebra index G is equal to the maximum evolution algebra Gmax(ii) a If G is GmaxIf yes, continuing to execute step S3.10; otherwise, selecting by EliteChoose strategy, take NpopTaking the chromosome as a parent population, and skipping to the step S3.4;
s3.10: and (5) outputting the chromosome of the first level in the feasible solution set as a Pareto optimal solution set of the multi-target opportunity constraint planning model of the electric vehicle charging network, and ending the algorithm process.
And S4, determining an optimal charging network planning scheme according to the principle of maximum marginal investment income by taking the charging network planning scheme with the lowest comprehensive construction cost of the Pareto solution centralized charging network as a reference.
The method comprises the following specific steps:
s4.1: according to the construction cost C of the charging network of the electric automobiletotalArranging all chromosomes in the Pareto optimal solution set from small to large;
s4.2: the marginal return on investment for all but the 1 st chromosome was calculated according to equation (13):
wherein E ismMarginal investment income of the mth chromosome in the Pareto optimal solution set; ctotal,mAnd Ctotal,1The construction cost D of the electric vehicle charging network corresponding to the mth chromosome and the 1 st chromosome in the Pareto optimal solution set respectivelyave,mAnd Dave,1Respectively the average charging travel distances corresponding thereto; m ismaxThe number of chromosomes in the set for Pareto optimal solution;
s4.3: and outputting the planning scheme corresponding to the chromosome with the maximum marginal investment income in the Pareto optimal solution set as the optimal charging network planning scheme.
In order to verify the effectiveness of the model and the solving method, simulation analysis is performed in this section by taking a 25-node traffic system as an example. The topology of the 25-node traffic system is shown in fig. 3, and the topology is composed of 25 traffic nodes and 43 roads, and the weight coefficient of each traffic node and the corresponding construction cost of the charging station are shown in table 1.
TABLE 1 traffic node weight coefficients and corresponding charging station construction costs
In fig. 3, the roads only represent the topological relations among the traffic nodes, and do not represent the actual trends of the roads. In the example, it is assumed that each road can be passed in both directions, that is, the shortest path from the starting point to the end point is the same as the shortest path from the end point back to the starting point. At this time, the shortest path set Ω of the traffic networkqConsisting of 300 shortest paths, i.e. [ 25X (25-1)]And 300 traffic nodes and roads which are passed by each shortest path can be obtained by a Floyd algorithm. The sum of the traffic flow of all shortest paths is 0.307, and the sum of the traffic flow on all roads is 0.84. In the example, the number of charging stations to be built is 4. Currently, the range of the existing full-charge EV is mostly between 300 and 500km, and considering that most EV owners charge the residual electric quantity by about 20%, the charging range threshold value d is calculatedcha-limSet to 80 km; the confidence level β of the charging travel distance opportunity constraint is set to 95%.
Adopting GA to solve the EV charging network planning model, and setting algorithm parameters as follows: population size NpopIs 100, the crossing rate Pc0.5, the mutation rate Pm0.2, maximum evolution algebra GmaxIs 150. The obtained Pareto optimal solution set is according to DaveThe final Pareto optimal solution set and the planning scheme represented by the solution set are shown in table 2 by sorting from big to small:
TABLE 2 Pareto optimal solution set and planning results represented thereby
Table 2 shows 7 planning schemes, EV to EV under the 1 st planning schemeAverage travel distance D of charging stationaveThe maximum is 43.42km, and the investment cost is 1250 ten thousand yuan. And calculating the marginal investment profit of other 6 planning schemes by taking the 1 st planning scheme as a reference, and comparing the marginal investment profit of the 2 nd planning scheme to be the maximum value of 0.22 km/ten thousand yuan. Thus, the 2 nd plan in the table is the optimal plan, with the charging station building nodes 3, 11, 21, 23, as shown in figure 6. Fig. 4 and 5 show an accumulated probability density curve and a probability density curve of the EV charging travel distance when the charging station is constructed according to the plan.
Specifically, the invention establishes a charging network multi-target opportunity constraint planning model considering the shortest average charging travel distance of the electric vehicle and the lowest construction cost of the charging network at the same time, the optimization variables are the construction positions of each charging station, the optimization target is the shortest average charging travel distance of the electric vehicle and the lowest construction cost of the charging network, and the specific constraint conditions are as follows: the total construction number of the charging stations is equality constraint, and the probability that the driving distance from the electric vehicle to the charging stations does not exceed a given mileage threshold value is opportunity constraint. The electric vehicle charging network optimization planning model is a 0-1 integer planning problem considering opportunity constraint, clear analytical expressions are lacked among optimization targets, constraint conditions and optimization variables, and a non-dominated genetic algorithm based on a feasibility rule can be adopted for solving. In order to solve the electric vehicle charging network optimization planning model by adopting a genetic algorithm, a chromosome coding scheme and mutation and crossover operators are designed according to the characteristics of the model to be optimized. After the chromosomes are decoded, the chromosomes are sorted based on a feasibility rule, a Pareto optimal solution set is output, and a final charging network planning scheme, namely the construction position of a charging station, is determined according to the principle of maximum marginal investment income. In the solution, a maximum evolution algebra is preset, once the genetic algorithm is evolved to the maximum algebra, the algorithm is considered to be converged, and a final planning result is output.
The above-mentioned embodiments, objects, technical solutions and advantages of the present invention are further described in detail, it should be understood that the above-mentioned embodiments are only illustrative of the present invention, and are not intended to limit the scope of the present invention, and any person skilled in the art should understand that equivalent changes and modifications made without departing from the concept and principle of the present invention should fall within the protection scope of the present invention.
Claims (8)
1. A multi-target electric vehicle charging network planning method is characterized by comprising the following steps:
s1: setting planning boundary conditions, wherein the planning boundary conditions comprise: the method comprises the following steps that (1) a traffic network topological structure and parameters, charging station candidate addresses, charging station construction cost of each candidate site, charging station construction total number, charging mileage threshold value and confidence coefficient are obtained; the candidate addresses of the charging stations are all traffic nodes in a traffic network;
s2: establishing a charging network multi-target opportunity constraint planning model considering the shortest average charging running distance of the electric vehicle and the lowest construction cost of a charging network at the same time, wherein opportunity constraints of the charging network multi-target opportunity constraint planning model comprise charging running distance opportunity constraints and charging station construction total number constraints;
the shortest average charging travel distance of the electric automobile can be represented by the following formula:
wherein D isaveAverage charging travel distance, omega, for electric vehicles in the entire traffic networkRFor a set of roads in a traffic network, i is a road index in the traffic network, TiTraffic flow for road i, dav,iThe average charging running distance of the electric automobile running on the road i is obtained;
the lowest construction cost of the charging network can be expressed by the following formula:
wherein, CtotalCost of construction of charging network for electric vehicle, Ccon,jBuilding cost for building the charging station at the candidate address j, wherein N is the total number of the candidate addresses of the charging station;
the charging travel distance opportunity constraint is the probability that the charging travel distance is lower than a given mileage threshold value does not exceed a given confidence coefficient, and can be expressed by the following formula:
wherein p isiThe probability that the EV charging travel mileage on the road i does not exceed the charging travel mileage threshold value is given, and beta is a confidence coefficient;
s3: solving the multi-target opportunity constraint planning model of the charging network by adopting a non-dominated genetic algorithm based on a feasibility rule, and giving a Pareto solution set of the multi-target opportunity constraint planning model of the charging network;
the solving of the multi-target opportunity constraint planning model of the charging network by adopting the non-dominated genetic algorithm based on the feasibility rule specifically comprises the following steps:
s31, setting genetic algorithm parameters, including: population size NpopCross rate PcThe rate of variation PmAnd maximum evolution algebra Gmax；
S32, randomly generating a random number NpopParent population Q consisting of chromosome bars1,1(ii) a Parent population Q1,1The chromosome in the site is a binary code string consisting of N code bits, the j code bit value of 1 indicates that a charging station is built in the candidate site j, otherwise, a charging station is not built in the candidate site j; each chromosome has and only has M code bits to take the value of '1';
s33, making g equal to 0, wherein g is an evolution algebraic index;
s34, enabling g to be g +1, starting the evolution of the g generation, and adopting a binary tournament method to carry out the evolution of the parent generation Q1,gPerforming copy operation, crossover operation and mutation operation to generate a temporary population Q2,gTemporary population Q2,gPopulation size of NpopAnd a temporary population Q2,gAnd the parent population Q1,gCombining to form a population Q to be evolved3,gPopulation Q to be evolved3,gThe population size of (2N)pop；
S35, calculating an evolutionary population Q3,gThe average driving distance D from the electric vehicle to the nearest charging station in the planning scheme represented by each chromosomeave,kAnd the comprehensive construction cost C of the charging networktotal,kProbability P that the charging driving distance from the electric automobile does not exceed a given mileage threshold valueev,k；
S36, determining a population Q to be evolved based on feasibility rule3,gThe priorities of all chromosomes in the population Q are determined, the chromosomes are sorted according to the priorities of the chromosomes, and then the population Q to be evolved is determined according to the priorities of the chromosomes3,gDesigning chromosome fitness by the order in (1);
determining a population Q to be evolved based on a feasibility rule3,gThe priorities of all chromosomes in the chromosome library are specifically:
in the population Q to be evolved3,gClassifying chromosomes meeting the opportunity constraint into a feasible solution set, layering all chromosomes in the feasible solution set according to a Pareto domination principle, and establishing the cost C of all chromosomes of each layer according to a charging networktotalCalculating the virtual fitness of each chromosome in each level after sequencing from small to large; classifying chromosomes which do not meet the opportunity constraint into a non-feasible solution set, calculating the constraint violation degree of each chromosome in the non-feasible solution set, and determining the priority of each chromosome according to the following principle;
(1) chromosomes in the feasible solution set have a higher priority than chromosomes in the infeasible solution set;
(2) in the feasible solution set, the lower the level of the chromosome is, the higher the priority is; when the levels of chromosomes are the same, the chromosome with large virtual fitness has high priority, and the calculation method of the virtual fitness is shown as the following formula:
wherein, Fr,nDyeing the r-th layer and the n-th layerVirtual fitness of the volume; r ismaxFor the population Q to evolve3,gThe number of chromosome layers of the feasible solution set; n isrThe number of the chromosome of the r layer;
the comprehensive construction cost of the charging network corresponding to the r-th chromosome and the (n + 1) th chromosome is saved,
the average charging travel distance corresponding to the average charging travel distance;
the comprehensive construction cost of the charging network corresponding to the r-th chromosome and the n-1 th chromosome is saved,
the average charging travel distance corresponding to the average charging travel distance; h is a given large number;
(4) in the infeasible solution set, the smaller the constraint violation degree of the chromosome is, the higher the priority is, and the constraint violation degree calculation formula is as follows:
wherein, CVRepresenting the degree of constraint violation of the planning scheme corresponding to a chromosome, PevThe probability that the charging travel distance of the electric automobile in a planning scheme corresponding to a certain chromosome does not exceed a given mileage threshold value is given;
s37, judging whether the evolution algebra index G is equal to the maximum evolution algebra Gmax(ii) a If G is GmaxThen, go on to step S38; otherwise, adopting the strategy of 'elite selection' to take the first NpopThe chromosome is taken as a parent population, and the step S34 is skipped;
s38, outputting a chromosome of a first level in the feasible solution set as a Pareto optimal solution set of the multi-target opportunity constraint planning model of the electric vehicle charging network;
and S4, determining an optimal charging network planning scheme according to the principle of maximum marginal investment income by taking the charging network planning scheme with the lowest comprehensive construction cost of the Pareto solution centralized charging network as a reference.
2. The multi-target electric vehicle charging network planning method according to claim 1, wherein the calculation formula of the average charging travel distance of the EV traveling on the road i is as follows:
wherein ld,iIs the length of road i; f. ofi(x) The charging travel distance is the EV on the road i at a distance x from the starting point.
3. The method for planning multi-objective electric vehicle charging network according to claim 1, wherein the traffic flow of the road i is set to be Ω according to the shortest path of the traffic networkqThe sum of the traffic flow of the shortest path passing through the road i is obtained;
the calculation formula of the shortest path traffic flow is as follows:
in the formula, WS,qAnd WE,qRespectively are the weight coefficients of the starting point and the end point of the shortest path q; omegaqIs a traffic network shortest path set; dqIs the length of the shortest path q.
4. The method of claim 2, wherein the charging travel distance f of the EV on the road i at a distance x from the starting point is the charging travel distance f of the EVi(x) The calculation formula of (a) is as follows:
fi(x)＝min[x+l1,ld,i-x+l2] 0≤x≤ld,i；
wherein x is a random variable representing the distance between the EV driving on the road i and the starting point of the road i; ld,iRepresents the length of road i; l1Representing the distance between the energy storage charging station closest to the starting point of the road i and the starting point of the road i; l2Is the distance from the nearest charging station to the end of the road i.
5. The method for planning the multi-target electric vehicle charging network according to claim 1, wherein the probability that the charging travel distance of the EV on the road i does not exceed the charging travel mileage threshold is calculated by the following formula:
wherein, gi(x) D auxiliary function for judging whether the charging travel distance of EV from the end point x on the road i does not exceed the charging travel mileage threshold valuecha-limAnd the charging mileage threshold value.
6. The multi-target electric vehicle charging network planning method according to claim 1, wherein the layering is performed on all chromosomes in the feasible solution set according to a Pareto governing principle, specifically:
s361, initializing a chromosome level index r to be 1;
s362, traversing unclassified chromosomes in the feasible solution set, searching all unclassified chromosomes in the unclassified chromosomes, and classifying the unclassified chromosomes into an r level; wherein the unopposed chromosome is a chromosome in which two optimization objectives are absent from the unorganized chromosomes and which is superior to the unopposed chromosome; the two optimization targets are that the average charging running distance of the electric automobile is shortest and the construction cost of a charging network is lowest;
s363, determining whether there is an unclassified chromosome in the feasible solution set, if yes, making r ═ r +1, and jumping to step S362; otherwise, ending the chromosome layering work in the feasible solution set.
7. The multi-target electric vehicle charging network planning method according to claim 1, wherein in step S34, the specific steps of the crossover operation are as follows:
(1) randomly selecting two chromosomes from the current chromosome population as chromosomes to be crossed;
(2) repeatedly and randomly generating cross bit N to be selectedcro(1<Ncro<N) until a feasible cross point N is foundavI.e. the Nth of two chromosomes to be crossedcroThe number of code bits with the value of 1 after the code bits is consistent;
(3) with a cross probability PcExchange of the crossover site N of two chromosomesavThe subsequent binary code string completes the cross operation;
the mutation operation comprises the following specific steps:
(1) randomly selecting a chromosome from the current chromosome population as a chromosome to be mutated;
(2) randomly generating two code bits N to be variedmut1And Nmut2(1≤Nmut1≤N,1≤Nmut2N) or less), two code bits to be varied Nmut1And Nmut2The value cannot be simultaneously '1' or '0';
(3) with a mutation probability PmTreating simultaneously the variant code bit Nmut1And Nmut2And performing variation operation, wherein the code bit to be varied with the value of 1 is varied into 0, and the code bit to be varied with the value of 0 is varied into 1.
8. The multi-target electric vehicle charging network planning method according to claim 1, wherein the step S4 specifically comprises the following steps:
s41, according to the construction cost C of the charging network of the electric automobiletotalArranging all chromosomes in the Pareto optimal solution set from small to large;
s42, calculating the marginal investment profit of all chromosomes except the 1 st chromosome according to the following formula:
wherein E ismMarginal investment income of the mth chromosome in the Pareto optimal solution set; ctotal,mAnd Ctotal,1The construction cost D of the electric vehicle charging network corresponding to the mth chromosome and the 1 st chromosome in the Pareto optimal solution set respectivelyave,mAnd Dave,1Respectively the average charging travel distances corresponding thereto; m ismaxThe number of chromosomes in the set for Pareto optimal solution;
and S43, outputting a planning scheme corresponding to the chromosome with the maximum marginal investment income in the Pareto optimal solution set as an optimal charging network planning scheme.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111653174.4A CN114329783A (en) | 2021-12-30 | 2021-12-30 | Multi-target electric vehicle charging network planning method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111653174.4A CN114329783A (en) | 2021-12-30 | 2021-12-30 | Multi-target electric vehicle charging network planning method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114329783A true CN114329783A (en) | 2022-04-12 |
Family
ID=81019237
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111653174.4A Withdrawn CN114329783A (en) | 2021-12-30 | 2021-12-30 | Multi-target electric vehicle charging network planning method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114329783A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114117703A (en) * | 2021-11-30 | 2022-03-01 | 南方电网能源发展研究院有限责任公司 | Charging station configuration method and device based on coupling of traffic network and power distribution network |
-
2021
- 2021-12-30 CN CN202111653174.4A patent/CN114329783A/en not_active Withdrawn
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114117703A (en) * | 2021-11-30 | 2022-03-01 | 南方电网能源发展研究院有限责任公司 | Charging station configuration method and device based on coupling of traffic network and power distribution network |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109034465B (en) | Charging station two-layer planning method considering coupling of charging station site selection and travel path | |
| CN108921472B (en) | Multi-vehicle-type two-stage vehicle and goods matching method | |
| CN108764555B (en) | Shared bicycle parking point site selection method based on Hadoop | |
| CN109886468B (en) | Charging station planning method based on improved adaptive genetic algorithm | |
| CN111311158B (en) | Electric logistics vehicle path planning method under limited charging facility condition | |
| CN107025518A (en) | A kind of electric automobile charging station method and device for planning | |
| CN113222241B (en) | Taxi quick-charging station planning method considering charging service guide and customer requirements | |
| CN111626492B (en) | Fuzzy multi-target opportunity constraint planning method for electric vehicle charging network | |
| CN114285063B (en) | Short-term carbon emission factor-based intelligent electric vehicle carbon-reducing charging method | |
| CN112668129B (en) | Space load clustering-based intelligent grid dividing method for power distribution network | |
| CN111626493B (en) | Charging network planning method considering charging service capacity and operation efficiency | |
| CN108256969A (en) | A kind of public bicycles lease point dispatcher-controlled territory division methods | |
| CN110189025A (en) | Consider the electric automobile charging station programme acquisition methods that different load increases | |
| CN117688968B (en) | Tramcar layout method based on particle swarm optimization | |
| CN115470600A (en) | Electric vehicle charging station planning method based on multi-objective optimization | |
| CN115689237A (en) | Electric vehicle charging station double-view-angle site selection method based on multi-target particle swarm optimization | |
| CN112378415A (en) | Path planning method, device and equipment | |
| CN112734041A (en) | Opportunity constraint planning method for electric vehicle charging network considering random charging load | |
| Xiong et al. | Route network design of community shuttle for metro stations through genetic algorithm optimization | |
| CN110034596A (en) | More base station charging methods in WRSNs based on SOM neural network | |
| CN115907258A (en) | GVRTW-MDHV optimal scheduling method based on environment and cost consideration view angle | |
| CN114742340B (en) | Optimal layout solving method for intelligent network sharing electric vehicle charging station in large-scale road network | |
| CN114329783A (en) | Multi-target electric vehicle charging network planning method | |
| CN114491882B (en) | EV energy storage charging network planning method considering battery endurance capacity | |
| Zhao et al. | Robust Optimization of Mixed-Load School Bus Route Based on Multi-Objective Genetic Algorithm. |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| WW01 | Invention patent application withdrawn after publication | ||
| WW01 | Invention patent application withdrawn after publication |
Application publication date: 20220412 |