【优化控制】基于遗传算法实现红绿灯优化管理附matlab代码

1 内容介绍

城市交通拥堵阻碍城市发展:(1)减少市民可用于工作时间;(2)造成环境污染;(3)难以应变道路紧急情况.特种车辆在城市中执行紧急任务时,由于现阶段灯控系统未能对其做出区分,无法动态引导特种车辆到路口之间的交通流,并在其到达路口时设置绿灯,造成特种车辆通行过程中常常遇到阻碍.红绿灯作为城市交通管理的工具,根据感知到的路口周边车辆调整红灯时间和绿灯时间,可以优化交通控制,解决路口交通拥堵以及实现特种车辆执行紧急任务时一路"绿灯"畅行.对于灯控路口拥堵问题的研究​.

2 仿真代码

%% Starting point, clear everything in matlab

tic;

clear all;

close all;

clc;

%% Problem Formulation

FitnessFunction=@(C,g,x,c) TDi(C,g,x,c);     % FitnessFunction

nLights=4;                                   % Number of Traffic Lights

nIntersections=1;                            % Number of Intersections (static as 1 intersection)

VarSize=[1 nIntersections*nLights];          % Decision Chromosome genes based on number of Intersections

greenMin= 10;                                % Lower bound of GREEN LIGHT

greenMax= 60;                                % Upper bound of GREEN LIGHT

Cyclemin=60;                                 % Lower bound of CYCLE

Cyclemax=180 ;

RoadcapacityNSWE=[20,20,20,20];              % Road Capacity for NSWE respectivelly

CarsNSWE=[20,20,11,17];

RoadCongestion1NSWE=RoadcapacityNSWE-CarsNSWE;              % congestion according to free road spaces

RoadCongestionNSWE=RoadCongestion1NSWE./RoadcapacityNSWE;   %  Volume/Capacity RATIO

carpass=5;

%% Genetic Algorithm Parameters

MaxIt=25;                                  % Maximum Number of Iterations

nPop=400;                                     % Population Size

pc=0.5;                                      % Crossover Percentage

nc=2*round(pc*nPop/2);                       % Number of Offsprings (parents)

pm=0.02;                                      % Mutation Percentage

nm=round(pm*nPop);                           % Number of Mutants

mu=0.1;                                      % Mutation Rate

pinv=0.2;

ninv=round(pinv*nPop);

beta=8;                                      % Selection Pressure

   

%% Initialization

% Individual Structure

empty_individual.GreenNSWE=[];

empty_individual.TotalDelay=[];

% Population Structure

pop=repmat(empty_individual,nPop,1);

% Initialize Population

i=1;

current_cycle=160-12; %estw kiklos 160 seconds - 12 seconds gia ;

disp(['FIRST Population..........Best TotalDelay = ' num2str(BestDelay)]);

    fprintf('\n')

    disp('Green Timings in seconds:');

    disp(['  North Green time = ' num2str(BestSol.GreenNSWE(1))]);

    fprintf('\n')

    disp(['  South Green time = ' num2str(BestSol.GreenNSWE(2))]);

    fprintf('\n')

    disp(['  West Green time = ' num2str(BestSol.GreenNSWE(3))]);

    fprintf('\n')

    disp(['  East Green time = ' num2str(BestSol.GreenNSWE(4))]);

    fprintf('\n')

%% Loop For Number of Iterations

count=0;

for it=1:MaxIt

   

    % TERMINATION CRITERIA IF NEEDED

%     if(it>2)

%         if(BestDelay(it-1)==BestDelay(it-2))

%             count=count+1;

%         else

%         count=0;

%         end

%     end

%     if(count>5)

%         disp('5 Generations without evolution. Process Stopped !');

%         break;

%     end

    % Calculate Selection Probabilities

    P=exp(-beta*TotalDelay/WorstDelay);

    P=P/sum(P);

    %% Crossover

    popc=repmat(empty_individual,nc/2,2);

    k=1;

    while k<=nc/2

        

        % Select Parents Indices from roulette wheel

        i1=RouletteWheelSelection(P);

        i2=RouletteWheelSelection(P);

              

        % Select Parents

        p1=pop(i1);

        p2=pop(i2);

        popc(k,1).GreenNSWE=p1.GreenNSWE;

        popc(k,2).GreenNSWE=p2.GreenNSWE;

        popc(k,1).TotalDelay=p1.TotalDelay;

        popc(k,2).TotalDelay=p2.TotalDelay;

        % Select random crossover point

        i=randi([1 3]);

        

        % crossover randomness

         if(i==1)

              

                popc1=popc(k,1).GreenNSWE(2);

                popc(k,1).GreenNSWE(2)= popc(k,2).GreenNSWE(2);

                popc(k,2).GreenNSWE(2)=popc1;

            

                popc1=popc(k,1).GreenNSWE(3);

                popc(k,1).GreenNSWE(3)= popc(k,2).GreenNSWE(3);

                popc(k,2).GreenNSWE(3)=popc1;

        

                popc1=popc(k,1).GreenNSWE(4);

                popc(k,1).GreenNSWE(4)= popc(k,2).GreenNSWE(4);

                popc(k,2).GreenNSWE(4)=popc1;  

        elseif(i==2)

    

                popc1=popc(k,1).GreenNSWE(3);

                popc(k,1).GreenNSWE(3)= popc(k,2).GreenNSWE(3);

                popc(k,2).GreenNSWE(3)=popc1;

        

                popc1=popc(k,1).GreenNSWE(4);

                popc(k,1).GreenNSWE(4)= popc(k,2).GreenNSWE(4);

                popc(k,2).GreenNSWE(4)=popc1;  

         else

             

                popc1=popc(k,1).GreenNSWE(4);

                popc(k,1).GreenNSWE(4)= popc(k,2).GreenNSWE(4);

                popc(k,2).GreenNSWE(4)=popc1;

        end

        

        % check if new green times are out constraints 10-60s. If it is

        % get it to closer min or max

        popc(k,1).GreenNSWE=max(popc(k,1).GreenNSWE,greenMin);

        popc(k,1).GreenNSWE=min(popc(k,1).GreenNSWE,greenMax);

        popc(k,2).GreenNSWE=max(popc(k,2).GreenNSWE,greenMin);

        popc(k,2).GreenNSWE=min(popc(k,2).GreenNSWE,greenMax);

        

        if(sum(popc(k,1).GreenNSWE)>current_cycle || sum(popc(k,2).GreenNSWE)>current_cycle)

            continue;

        end

        

        % Evaluate Generated Offsprings for each traffic light according to

        % the corresponding traffic congestion

        for j=1:nLights

            popc(k,1).TotalDelay(j)=FitnessFunction(current_cycle, popc(k,1).GreenNSWE(j), RoadCongestionNSWE(j),RoadcapacityNSWE(j));

            popc(k,2).TotalDelay(j)=FitnessFunction(current_cycle, popc(k,2).GreenNSWE(j), RoadCongestionNSWE(j),RoadcapacityNSWE(j));

        end

        

        % TOTAL DELAY which correspongs to the summation of of 4 lights quotients

        popc(k,1).TotalDelay= real(sum(popc(k,1).TotalDelay));

        popc(k,2).TotalDelay= real(sum(popc(k,2).TotalDelay));     

        k=k+1; %step

    end

    

    % Make 2 rows 1

    popc=popc(:);

    % Sort popc matrix according to TotalDelay

    TotalDelay=[popc.TotalDelay];

    [TotalDelay, SortOrder]=sort(TotalDelay);

    popc=popc(SortOrder);

    

    %% Mutation

    % Create empty Matrix with length the number of mutants 

    popm=repmat(empty_individual,nm,1);

    k=1;

    while k<=nm

        % Select Parent population

        i=randi([1 nPop]); %nPop value 100

        p=pop(i);

        

        % Apply Mutation   

        nVar=4;

        nmu=ceil(mu*nVar);

        j=randi([1 nVar]);

        prosimo=randi([-1 1]);

        sigma=prosimo*0.02*(greenMax-greenMin);

        

        mutated=p.GreenNSWE(j)+sigma;

        

        popm(k).GreenNSWE = p.GreenNSWE;

        popm(k).GreenNSWE(j)=mutated;

        

        popm(k).GreenNSWE=max(popm(k).GreenNSWE,greenMin);

        popm(k).GreenNSWE=min(popm(k).GreenNSWE,greenMax);

        

        if(sum(popm(k).GreenNSWE)>current_cycle)

            continue;

        end 

        

        for j=1:nLights

            % Evaluate Mutant 

            popm(k).TotalDelay(j)=FitnessFunction(current_cycle, popm(k).GreenNSWE(j), RoadCongestionNSWE(j),RoadcapacityNSWE(j));

        end

        % Summation of delay quotients

        popm(k).TotalDelay=real(sum(popm(k).TotalDelay));

       

        k=k+1; %step

    end

    

    

    %% INVERSION

    % Create empty Matrix 

    popinv=repmat(empty_individual,nm,1);

    k=1;

    while k<=ninv

        

        % Select Parent population

        i=randi([1 nPop]);

        p=pop(i);

        

        % Apply Inversion

        

        nVar=numel(p.GreenNSWE);

        randomgene1=randi([1 4]);

        randomgene2=randi([1 4]);

        y=p.GreenNSWE;

        popinv(k).GreenNSWE=y;

        x=popinv(k).GreenNSWE(randomgene1);

        popinv(k).GreenNSWE(randomgene1)=popinv(k).GreenNSWE(randomgene2);      

        popinv(k).GreenNSWE(randomgene2)=x;

        popinv(k).GreenNSWE=max(popinv(k).GreenNSWE,greenMin);

        popinv(k).GreenNSWE=min(popinv(k).GreenNSWE,greenMax);

        if(sum(popinv(k).GreenNSWE)>current_cycle)

            continue;

        end

        

        for j=1:nLights

            % Evaluate Mutant 

            popinv(k).TotalDelay(j)=FitnessFunction(current_cycle, popinv(k).GreenNSWE(j), RoadCongestionNSWE(j),RoadcapacityNSWE(j));

        end

        % Summation of delay quotients

        popinv(k).TotalDelay=real(sum(popinv(k).TotalDelay));

              

        k=k+1;

    end

    % Make 2 rows 1

    popinv=popinv(:);

    %% Merge Population

    pop=[pop

        popc

        popm

        popinv]; %#ok

     

    % Sort New Population according to TotalDelay

    TotalDelay=[pop.TotalDelay];

    [TotalDelay, SortOrder]=sort(TotalDelay);

    pop=pop(SortOrder);

    

    % Update Worst Cost

    WorstDelay=max(WorstDelay,pop(end).TotalDelay);

    

    % Keep the Best Population from the given number

    pop=pop(1:nPop);

    TotalDelay=TotalDelay(1:nPop);

    

    % Store Best Solution Ever Found

    BestSol=pop(1);

    

    % Store Best Cost Ever Found

    BestDelay(it)=BestSol.TotalDelay;

    

        % Show Iteration Information

    disp(['                   Iteration ' num2str(it) ': Best TotalDelay = ' num2str(BestDelay(it))]);

    fprintf('\n')

    disp('Green Timings:');

    fprintf('\n')

    disp(['  North Green time = ' num2str(BestSol.GreenNSWE(1))'' ' seconds']);

    fprintf('\n')

    disp(['  South Green time = ' num2str(BestSol.GreenNSWE(2))'' ' seconds']);

    fprintf('\n')

    disp(['  West Green time = ' num2str(BestSol.GreenNSWE(3))'' ' seconds']);

    fprintf('\n')

    disp(['  East Green time = ' num2str(BestSol.GreenNSWE(4))'' ' seconds']);

    fprintf('\n')

    %end of generation

    

end

    disp(' ****************************************************************' );

    disp('  CASE: Every 5 seconds 2 vehicles leaves the corresponding road ' );

    disp('  Expected vehicles left through North road' );

    disp(round(2*BestSol.GreenNSWE(1)/carpass));

    disp('  Expected vehicles left through South road' );

    disp(round(2*BestSol.GreenNSWE(2)/carpass));

    disp('  Expected vehicles left through West road' );

    disp(round(2*BestSol.GreenNSWE(3)/carpass));

    disp('  Expected vehicles left through East road' );

    disp(round(2*BestSol.GreenNSWE(4)/carpass));

    fprintf('\n')

    disp(' ****************************************************************' );

disp(['Cycle Time = ' num2str(current_cycle)'' ' seconds']);

%% Results / Plots

figure(1);

semilogy(BestDelay,'LineWidth',2);

% plot(BestCost,'LineWidth',2);

xlabel('Iteration');

ylabel('Total Delay');

grid on;

toc

3 运行结果

4 参考文献

[1]赵功勋, 郭海滨, 苏利. 基于遗传算法的工程项目资源均衡优化及其MATLAB实现[J]. 工程经济, 2016, 26(12):6.

[2]叶文斌. 基于红绿灯优化城市交通控制设计与仿真[D]. 华东师范大学, 2015.

博主简介：擅长智能优化算法、神经网络预测、信号处理、元胞自动机、图像处理、路径规划、无人机等多种领域的Matlab仿真，相关matlab代码问题可私信交流。

部分理论引用网络文献，若有侵权联系博主删除。