A list,

1. Ant colony algorithm is proposed

Ant Colony Optimization algorithm (ACO), also known as ant algorithm, is a probabilistic algorithm used to find optimal paths. It was proposed by Marco Dorigo in his PhD thesis in 1992 and was inspired by the behavior of ants in finding their way to food. Genetic algorithm has been applied in pattern recognition, neural network, machine learning, industrial optimal control, adaptive control, biological science, social science and so on.

2. Basic principles of the algorithm





Ii. Source code

%% Clean environment CLC; Clear %% obstacle data position = load('barrier.txt');
plot([0.200], [0.200].'. ');
hold on
B = load('barrier.txt');
xlabel('km'.'fontsize'.12)
ylabel('km'.'fontsize'.12)
title('Two dimensional programming space'.'fontsize'.12) %% describe the starting point and ending point S = [20.180];
T = [160.90];
plot([S(1),T(1)],[S(2),T(2)].'. '); % text(S(1) +2,S(2),'S');
text(T(1) +2,T(2),'T'); Fill (position(1:4.1),position(1:4.2),0.0.0]);
fill(position(5:8.1),position(5:8.2),0.0.0]);
fill(position(9:12.1),position(9:12.2),0.0.0]);
fill(position(13:15.1),position(13:15.2),0.0.0]); % download link endpoint data L = load('lines.txt'); V = zeros(size(L));for i=1:20
    plot([position(L(i,1),1),position(L(i,2),1)],[position(L(i,1),2)...
        ,position(L(i,2),2)].'color'.'black'.'LineStyle'.The '-');
    v(i,:) = (position(L(i,1),:)+position(L(i,2), :)) /2;
    plot(v(i,1),v(i,2),The '*');
    text(v(i,1) +2,v(i,2),strcat('v',num2str(i))); End %% sign = load('matrix.txt');
[n,m]=size(sign);
 
for i=1:n
    
    if i == 1
        for k=1:m- 1
            if sign(i,k) == 1
                plot([S(1),v(k- 1.1)],[S(2),v(k- 1.2)].'color'.'black'.'Linewidth'.2.'LineStyle'.The '-');
            end
        end
        continue;
    end
    
    for j=2:i
        if i == m
            if sign(i,j) == 1
                plot([T(1),v(j- 1.1)],[T(2),v(j- 1.2)].'color'.'black'.'Linewidth'.2.'LineStyle'.The '-');
            end
        else
            if sign(i,j) = =1
                plot([v(i- 1.1),v(j- 1.1)],[v(i- 1.2),v(j- 1.2)],...
                    'color'.'black'.'Linewidth'.2.'LineStyle'.The '-');
            end
        end
    end
end
path = DijkstraPlan(position,sign);
j = path(22);
plot([T(1),v(j- 1.1)],[T(2),v(j- 1.2)].'color'.'yellow'.'LineWidth'.3.'LineStyle'.'-');
i = path(22);
j = path(i);
count = 0;
while true
    plot([v(i- 1.1),v(j- 1.1)],[v(i- 1.2),v(j- 1.2)].'color'.'yellow'.'LineWidth'.3.'LineStyle'.'-');
    count = count + 1;
    i = j;
    j = path(i);
    if i == 1 || j==1
        break;
    end
end
plot([S(1),v(i- 1.1)],[S(2),v(i- 1.2)].'color'.'yellow'.'LineWidth'.3.'LineStyle'.'-');


count = count+3;
pathtemp(count) = 22;
j = 22;
for i=2:count
    pathtemp(count-i+1) = path(j);
    j = path(j);
end
path = pathtemp;
path = [1     9     8     7    13    14    12    22]; %% Ant colony algorithm parameter initialization pathCount = length(path)2 -; % Number of line segments pheCacuPara=2; % Pheromone calculation parameter pheThres =0.8; % Pheromone selection threshold pheUpPara=[0.1 0.0003]; % pheromone update parameter QFZ = zeros(pathCount,10); PhePara = ones(pathCount,10)*pheUpPara(2); % pheromone qfzPara1 = ones(10.1) *0.5; % Inspire information parameter qfzPara2 =1.1; % Inspire information parameter m=10; % Population size NC=500; % number of cycles pathk = zeros(pathCount,m); % Search result records shortestPath = zeros(1,NC); % evolution record %% initial shortest path dijpathlen =0;
vv = zeros(22.2);
vv(1,:) = S;
vv(22,:) = T;
vv(2:21,:) = v;
for i=1:pathCount- 1
dijpathlen = dijpathlen + sqrt((vv(path(i),1)-vv(path(i+1),1)) ^2+(vv(path(i),2)-vv(path(i+1),2)) ^2); end LL = dijpathlen; %% lines = zeros(pathCount,4);
for i = 1:pathCount
    lines(i,1:2) = B(L(path(i+1)- 1.1), :); lines(i,3:4) = B(L(path(i+1)- 1.2), :); endCopy the code

3. Operation results



Fourth, note

2014a