A list,
The future large-scale development of electric vehicles requires the services of numerous public charging stations, which should be rationally arranged according to the distribution of electric vehicles. The prediction method of electric vehicle distribution is given, the charging device configuration method based on queuing theory is adopted, and the mathematical model of optimal planning of public charging station is proposed. Mathematical characteristics similar to charging station layout has a Voronoi diagram of charging station service area, service in the electric car considering quick charge randomness, using queuing theory model of M/M/s, on the basis of the electric car waiting time determine charging station scale, establish public charging stations layout model of optimal planning, using the particle swarm algorithm.
Second, the source
clear all; clc; Close all %% Basic data % Charging demand point coordinates b =1.0 e+003. * /0.1092 0.1348
0.1197 0.2399
0.2578 0.1724
0.4259 0.1739
0.1257 0.3420
0.2803 0.3375
0.4439 0.3360
0.5505 0.1108
0.5610 0.2024
0.5700 0.3075
0.1332 0.4591
0.3013 0.4455
0.4559 0.4380
0.5850 0.4260
0.1452 0.6092
0.3163 0.5341
0.4739 0.5341
0.5880 0.5341
0.3193 0.6407
0.4784 0.6452
0.6015 0.6467
0.6736 0.1574
0.8657 0.1649
1.0308 0.1634
0.6781 0.2924
0.8327 0.2909
1.0188 0.2939
0.6811 0.4425
0.8192 0.4576
1.0083 0.4546
0.6691 0.5431
0.6976 0.6392
0.8191 0.6377
1.0098 0.6347]; % charging demand point Conventional power load point load b(:,3) (kW) b(:,3) = [2480;2480;8680;11400;890;2340;4160;560;1670;5010;2670;8280;7400;1430;7500;4840;3400;4290;3840;3680;2560;7000;14800;8960;3160;7000;5000;2280;10360;10000;760;6000;7040;5600]; % Coordinates of centralized charging station BCS =[937.7296 379.5010
310.3141 238.4076];
na=4500;
alp=0.1;
b(:,4)=round(alp.*b(:,3)./sum(b(:,3)).*na); Charging demand point Conventional power load point load b(:,3() b(23.4) =37;
ns=3;
mui=0.6;
Nchz=round(mui.*sum(b(:,4))./ns);
bm=1.0 e+003* [0.0086.0.0088;1.1734.0.0088;1.1734.0.7412;0.0086.0.7412;0.0086.0.0088];
BL=sqrt(8.2*1.0 the e6./((max(bm(:,1))-min(bm(:,1)))*(max(bm(:,2))-min(bm(:,2))))); %BL is the ratio of graph coordinates to actual coordinates, and is divided into two regions Area2= for the fixed parameter %%1.0 e+003* [0.0086 0.0088
0.9377 1.0860
0.3103 1.7040
0.0086 0.7412
0.0086 0.0088];
Area2=[Area2,2.*ones(size(Area2,1),1)];
Area1=1.0 e+003* [0.9377 1.0860
1.1734 0.0088
1.1734 0.7412
0.3103 1.7040
0.9377 1.0860];
Area1=[Area1,1.*ones(size(Area2,1),1)];
vv=[Area1;Area2];
for k=1:size(bcs,1)
Ai=find(vv(:,3)==k);
xx=vv(Ai,1);
yy=vv(Ai,2);
kk=convhull(xx,yy);
in=inpolygon(b(:,1),b(:,2),xx(kk),yy(kk));
b(in,5)=k;
end
Ep=[];
for i=1:size(bcs,1)
gb=b(b(:,5)==i,:);
Ep=[Ep;[sum(gb(:,4)),round(mui.*sum(gb(:,4))./ns),i]];
end
Tn=6; % Number of optimal charging stations PopSize=100; % Population MaxIter=300; % Number of iterations c1s=2.5; % Maximum individual learning rate c2s=0.5; % Maximum social learning rate c1e=0.5; % Minimum individual learning rate c2e=2.5; % Minimum social learning rate w_start=0.9; % Weight maximum w_end=0.4; % Minimum weight Iter=1; % xmax= Max (Area2(:,1)); % area2The biggest long xmin = min (Area2 (:1)); % area2Minimum length ymax = Max (Area2 (:,2)); % area2The largest wide ymin = min (Area2 (:2)); % area2X = x + (xmax-xmin).*rand(Tn,PopSize); Y = ymin + (ymax-ymin).*rand(Tn,PopSize); % generate y direction particle X=[X;y]; % merge xy particles to form a complete individual V=rand(Tn*2,PopSize); % Random particle speed Vmax=0.01*max((xmax-xmin),(ymax-ymin)); % Max particle velocity inAr2=find(b(:,5) = =2);
bb=[b(inAr2,1:2),b(inAr2,4)];
for pk=1:1:PopSize
[FX(pk),~,~,~,~,~,~,~,~,~]=VorCostCDEV(X(1:Tn,pk),X(Tn+1:end,pk),bb,bcs(2,:),BL); % find the fitness value endCopy the code
3. Operation results
Fourth, note
Version: 2014 a