python蚁群算法 路径规划_用python实现蚁群算法

用python实现蚁群算法

由于学python不久，写的程序比较冗杂，借鉴了很多网上程序之后写的，有不足之处希望指出，谢谢。

#!/usr/bin/python

import time;

import random;

a=1;

b=2;

n=5

i=0;

aa=1;

j=1;

xigema=0.3;

zuiduanlujing=[0]*n;#最短路径矩阵

zdl=100000.0;#最短路径距离

arf=0.003;#距离影响因子，也是信息素初始矩阵

beta=0.007;#信息素影响因子

M=int(input("请输入蚂蚁数量:"));#M为蚂蚁数量

n=input("请输入城市数量:");

n=int(n) ; # 城市数量

c=[[0,0] for j in range(n)]; # n个城市的坐标

for i in range(n):

print("请输入第",i+1);

heng=int(input("城市的横坐标:"));

print("请输入第",i+1,);

zong=int(input("城市的纵坐标:"));

c[i][0]=heng;

c[i][1]=zong;

cc=int(input("请输入迭代次数;"));#循环次数

d=[[0]*n for j in range(n)];#d矩阵为两城市间距离

xinx=[[0]*n for j in range(n)];#xinx为启发函数，也是信息素初始函数

xxs=[[0]*n for j in range(n)];#信息素初始函数

xx=0.0;

yy=0.0;

dd=0.0;

for i in range(n):

for j in range(n):

xx=c[i][0]-c[j][0];

xx=xx**2;

yy=c[i][1]-c[j][1];

yy=yy**2;

dd=(xx+yy)**0.5;

d[i][j]=dd;

for i in range(n):#信息素初始值赋值为距离函数的倒数

for j in range(n):

if d[i][j]==0:

xinx[i][j]=0;

else:

xinx[i][j]=1/d[i][j];

xxs=xinx;

for ij in range(cc):#主循环，循环模拟次数

kexs=n-1;

kx=[0]*n;#剩余可选择城市根据矩阵

m=[[0]*n for i in range(M)];#m为每只蚂蚁走过的路径

sy=[[0]*n for i in range(M)];#sy矩阵为蚂蚁剩余可走路径，0表示未走过，1表示走过

for i in range(M):

m[i][0]=random.randint(1,n);

ssy=m[i][0]-1;

ssy=int(ssy);

sy[i][ssy]=1;

#print("m",m);

p=n;#辅助p为剩余城市数量

for i in range(n):

if sy[0][i]==1:

p=p-1;

kqd=[[0]*2 for i in range(p)];#概率以及对应城市

gail=[[0]*p];#剩余城市概率

zx=0.0#剩余总信息素

k=0;

pz=0.0;

sj=0.0;

zx=0;

w=0;#循环第到第几次

for w in range(n-1):

for i in range(M):

k=0;pz=0;

kqd=[[0]*2 for j in range(p)];#分类下个城市距离矩阵

for j in range(n):

if sy[i][j]==0:

kqd[k][0]=j+1;

k=k+1;

#print(kqd,m);

for j in range(p):

pz=pz+((xinx[(kqd[j][0]-1)][m[i][w]-1]**arf)*(xxs[(kqd[j][0]-1)][m[i][w]-1]**beta));

#print("pingjun",kqd[j][0]-1,m[0][i]-1,pz);

for j in range(p):

kqd[j][1]=((xinx[(kqd[j][0]-1)][m[i][w]-1]**arf)*(xxs[(kqd[j][0]-1)][m[i][w]-1]**beta))/pz;

for j in range(p-1):

kqd[j+1][1]=kqd[j][1]+kqd[j+1][1];

sj=random.random();

for j in range(p):

if sj<=kqd[j][1]:

zx=kqd[j][0];

#print(kqd,zx);

xf=n-p; #下一次放的地方为

m[i][xf]=zx;

sy[i][zx-1]=1;

#print(m,sy,aa);

p=p-1;

julib=[0.0]*M;#M只蚂蚁走过的总距离

for ii in range(M):

for jj in range(n-1):

julib[ii]=julib[ii]+d[m[ii][jj]-1][m[ii][jj+1]-1];

dj=0;#定义第几条路径最短

for ii in range(M-1):

if julib[dj]>=julib[ii+1]:

dj=ii+1;

if zdl>julib[dj]:

zuiduanlujing=m[dj];

zdl=julib[dj];

zeng=[[0]*n for j in range(n)];#信息素增加中间矩阵

zengjj=[[0]*n for j in range(n)];#信息素增加矩阵

for i in range(M):

for j in range(n-1):

zeng[m[i][j]-1][m[i][j+1]-1]=xinx[m[i][j]-1][m[i][j+1]-1];

zeng[m[i][j+1]-1][m[i][j]-1]=xinx[m[i][j+1]-1][m[i][j]-1];

for i in range(n):

for j in range(n):

zengjj[i][j]=zeng[i][j]+zengjj[i][j];

for i in range(n):

for j in range(n):

zengjj[i][j]=xigema*zengjj[i][j];

for i in range(n):

for j in range(n):

xxs[i][j]=xxs[i][j]+zengjj[i][j];

for i in range(n):#约束信息素区间

for j in range(n):

if xxs[i][j]<=0.05:

xxs[i][j]=0.06;

if xxs[i][j]>=10:

xxs[i][j]=10;

zdlj=zuiduanlujing;

print("经过",cc,"次迭代后，最短距离为：",zdl,"\n路径为",zuiduanlujing);

time.sleep(5)

input();