旅行商问题( traveling salesman problem, TSP),也称货郎担问题,是数学领域中的著名问题之一。TSP 问题已经被证明是一个NP-hard问题,由于TSP问题代表一类组合优化问题,因此对其近似解的研究一直是算法设计的一个重要问题。
TSP问题从描述。上来看是一个非常简单的问题,给定n个城市和各城市之间的距离,寻找一条遍历所有城市且每个城市只被访问一次的路径,并保证总路径距离最短。
其数学描述如下:
设G= (V,E)为赋权图,V ={1,2…,n}为顶点集,E为边集,各顶点间距离为Cij,已知Cij>0,且i,j∈V,并设定
xy=1,最优路径
xy=0,其他情况
那么整个TSP问题的数学模型表示为
minZ = ΣCijXij , i ≠ j
s,t = Σxij= 1,i ≠ j ,j ∈ v
s,t = Σxij < | k | - 1,k ∈ v
xij∈ {0,1}, i ∈ v, j ∈ v
其中,k是v的全部非空子集;| k |是集合k中包含图G的全部顶点的个数。
遗传算法求解TSP的基本步骤如下。
-
种群初始化。个体编码方法有二进制编码和实数编码,在解决TSP问题过程中个体编码方法为实数编码。对于TSP问题,实数编码为1-n的实数的随机排列,初始化 的参数有种群个数M.染色体基因个数N(即城市的个数)、迭代次数C、交叉概率Pc、变 异概率Pmutation。
-
适应度函数。在TSP问题中,对于任意两个城市之间的距离D(i,j)已知,每个染色体(即n个城市的随机排列)可计算出总距离,因此可将-个随机全排列的总距离的倒数作为适应度函数,即距离越短,适应度函数越好,满足TSP要求。
-
选择操作。遗传算法选择操作有轮盘赌法、锦标赛法等多种方法,用户根据实际 情况选择最合适的算法。
-
交叉操作。遗传算法中交叉操作有多种方法。一般对于个体,可以随机选择两
个个体,在对应位置交换若千个基因片段,同时保证每个个体依然是1 - n的随机排列, 防止进人局部收敛。 -
变异操作。对于变异操作,随机选取个体,同时随机选取个体的两个基因进行交 换以实现变异操作。
例1 随机生成一组城市种群,利用遗传算法寻找一条遍历所有城市且每个城市只被访问一次的路径,且总路径距离最短的方法.
主函数
%%
clc;
clear all;
N = 10; %城市的个数
M = 20; %种群的个数
C = 100; %迭代的次数
C_old = C;
m = 2; %适应值归一化淘汰加速指数
Pc = 0.4; %交叉概率
Pmutation = 0.2; %变异概率
%%%%%%%%%%%%%%%%%%%%%%%%%生成城市坐标%%%%%%%%%%%%%%%
pos = randn(N,2);
%%%%%%%%%%%%%%%%%%生成城市之间距离矩阵%%%%%%%%%%%%%%
D = zeros(N,N);
for i = 1:N
for j = i + 1:N
dis = (pos(i,1) - pos(j,1)).^2 + (pos(i,2) - pos(j,2)).^2 ;
D(i,j) = dis.^(0.5);
D(j,i) = D(i,j);
end
end
%%%%%%%%%%%%%%生成初始群体%%%%%%%%%%%%%%%%%%
popm = zeros(M,N);
for i = 1:M
popm(i,:) = randperm(N); %返回1:N,N个的随机数.
end
%%%%%%%%%%%%%%%随机选择一个种群%%%%%%%%%%%%
R = popm(1,:);
figure(1);
scatter(pos(:,1),pos(:,2),'k.'); %画出城市的坐标点
xlabel('横轴');ylabel('纵轴');
title('随机产生的种群');
axis([-3 3 -3 3]);
figure(2);
plot_route(pos,R);
xlabel('横轴');ylabel('纵轴');
title('随机生成种群中城市路径情况');
axis([-3 3 -3 3]);
%%%%%%%%%%初始化种群及其适应函数%%%%%%%%%%%
fitness = zeros(M,1);
len = zeros(M,1);
for i = 1:M
len(i,1) = myLength(D,popm(i,:));
end
maxlen = max(len);
minlen = min(len);
fitness = fit(len,m,maxlen,minlen);
rr = find(len==minlen);
R = popm(rr(1,1),:);
for i = 1:N
fprintf('%d',R(i));
end
fprintf('\n');
fitness = fitness/sum(fitness);
distance_min = zeros(C+1,1); %各次迭代的最小的种群的距离
while C >= 0
fprintf('迭代第%d次\n',C);
%%%%%选择操作%%%%%%%
nn = 0;
for i = 1: size(popm,1)
len_1(i,1) = myLength(D,popm(i,:));
jc = rand * 0.3;
for j = 1: size(popm,1)
if fitness(j,1) >= jc
nn = nn + 1;
popm_sel(nn,:) = popm(j,:);
break;
end
end
end
%%%%%%每次选择都保存最优的种群%%%%%%%%%%%%
popm_sel = popm_sel(1:nn ,:);
[len_m ,len_index] = min(len_1);
popm_sel = [popm_sel;popm(len_index,:)];
%%%%%交叉操作%%%%%%%
nnper = randperm(nn);
A = popm_sel(nnper(1),:);
B = popm_sel(nnper(2),:);
for i = nn*Pc
[A,B] = cross(A,B);
popm_sel(nnper(1),:) = A;
popm_sel(nnper(2),:) = B;
end
%%%%%%%%%%%变异操作%%%%%%%%%%%
for i = 1:nn
pick = rand;
while pick == 0
pick = rand;
end
if pick <= Pmutation
popm_sel(i,:) =Mutation(popm_sel(i,:));
end
end
%%%%%%%%%%%%%%%求适应度函数%%%%%%%%%%%%%
NN = size(popm_sel,1);
len = zeros(NN,1);
for i = 1:NN
len(i,1) = myLength(D, popm_sel(i,:));
end
maxlen= max(len);
minlen= min(len);
distance_min(C+1,1) = minlen;
fitness = fit(len,m,maxlen,minlen);
rr = find(len == minlen);
fprintf('minlen=%d\n',minlen);
R = popm_sel(rr(1,1),:);
for i = 1:N
fprintf('%d ',R(i));
end
fprintf('\n');
popm=[];
popm= popm_sel;
C = C- 1;
%pause(1);
end
figure(3)
plot_route(pos, R);
xlabel('横轴')
ylabel('纵轴')
title('优化后的种群中城市路径情况')
axis([-3 3 -3 3]);
子函数
%%%%%%%%%%%%适应度函数%%%%%%%%%%%%%%%%%%%%%
function fitness = fit(len,m,maxlen, minlen)
fitness = len;
for i = 1:length(len)
fitness(i,1)=(1-(len(i,1)- minlen)/(maxlen - minlen+ 0.0001)).^m;
end
end
%%%%%%%%%%%%%%计算个体距离函数%%%%%%%%%%%%
function len = myLength(D,p)
N = size(D,1);
len = D(p(1,N),p(1,1));
for i = 1 : (N-1)
len = len + D(p(1,i),p(1,i+1));
end
end
%%%%%%%%%%%%%%交叉操作函数%%%%%%%%%%%%%%%%%
function [A,B] = cross(A, B)
L= length(A);
if L<10
W = L;
elseif ((L/10) - floor(L/10)) >= rand&&L > 10
W = ceil(L/10) +8;
else
W = floor(L/10) +8;
end
p = unidrnd(L-W+ 1);
fprintf('p=%d ',p);
for i = 1:W
x = find(A== B(1,p+ i- 1));
y = find(B== A(1,p+ i- 1));
[A(1,p+i-1),B(1,p+i- 1)] = exchange(A(1,p+ i- 1),B(1,p+ i-1));
[A(1,x), B(1,y)] = exchange(A(1,x), B(1,y));
end
end
%%%%%%%%%%对调函数%%%%%%%%%%%%%%%
function [x, y] = exchange(x, y)
temp = x;
x = y;
y= temp;
end
%%%%%%%%%%%%%%变异函数%%%%%%%%%%%%%%%%%%
function a = Mutation(A)
index1 = 0; index2 = 0;
nnper = randperm(size(A,2));
index1 = nnper(1);
index2 = nnper(2);
%fprintf( 'index1=%d ',indexl);
%fprintf( 'index2=%d ',index2);
temp = 0;
temp = A(index1);
A(index1) = A(index2);
A(index2) = temp;
a=A;
end
%%%%%%%%%%%%%%%%%%%%连点画图函数8%%%%%%%%%%%%%%%%
function plot_route(a, R)
scatter(a(:,1),a(:,2), 'rx');
hold on;
plot([a(R(1),1), a(R(length(R)),1)],[a(R(1),2),a(R(length(R)),2)]);
hold on;
for i= 2:length(R)
x0 = a(R(i- 1),1);
y0 = a(R(i- 1),2);
x1 = a(R(i),1);
y1 = a(R(i),2);
xx = [x0,x1];
yy = [y0,y1];
plot(xx, yy);
hold on;
end
end
运行主程序,得到随机产生的城市种群图,随机生成种群中城市路径情况如图所示。
从图中可以看出,速记产生的种群城市点不对称,也没有规律,用一般的方法很难得到其最优路径。从图中可以看出,随机产生的路径长度很长,空行浪费比较多。
运行遗传算法,得到如下图所示的城市路径。从图中可以看出,该路径明显优于上图中的路径,且每个城市只经过一次。