读书使人充实,讨论使人机智,笔记使人准确…。凡有所学,皆成性格。
———— (英国)培根
文章目录
- 粒子群算法理论
粒子群算法理论
鸟类在捕食过程中,鸟群成员可以通过个体之间的信息交流与共享获得其他成员的发现与飞行经历。在食物源零星分布并且不可预测的条件下,这种协作机制所带来的优势是决定性的,远远大于对食物的竞争所引起的劣势。粒子群算法受鸟类捕食行为的启发并对这种行为进行模仿,将优化问题的搜索空间类比于鸟类的飞行空间,将每只鸟抽象为一个粒子,粒子无质量、无体积,用以表征问题的一个可行解,优化问题所要搜索到的最优解则等同于鸟类寻找的食物源。粒子群算法为每个粒子制定了与鸟类运动类似的简单行为规则,使整个粒子群的运动表现出与鸟类捕食相似的特性,从而可以求解复杂的优化问题
关键参数说明
参数 | 取值 |
粒子种群规模N | 一般设置粒子数为20~50 对于大部分问题,10个粒子已经可以取得很好的结果,比较难的问题或者特定类型的问题取100~200 |
惯性权重w | 一般取值范围为[0.8,1.2] |
加速常数c1和c2 | 一般设置c1=c2,通常可以取c1=c2=1.5 |
粒子的最大速度vmax | |
停止准则 |
[求最小值]计算函数 的最小值,其中个体x的维数n=10。这是一个简单的平方和函数,只有一个极小点x=(0,0,…,0),理论最小值f(0,0,…,0)=0
解:仿真过程如下:
(1)初始化群体粒子个数为N=100,粒子维数为D=10,最大迭代次数为T=200,学习因子c1=c2=1.5,惯性权重为w=0.8,位置最大值为,位置最小值为
,速度最大值为
,速度最小值为
。
(2)初始化种群粒子位置x和速度v,粒子个体最优位置p和最优值pbest,以及粒子群全局最优位置g和最优值gbest。
(3)更新位置x和速度值v,并进行边界条件处理,判断是否替换粒子个体最优位置和最优值
、粒子群全局最优位置
和最优值
。
(4)判断是否满足终止条件:若满足,则结束搜索过程,输出优化值;若不满足,则继续进行迭代优化。
优化结束后,其适应度进化曲线如图所示,优化后的结果为x=[-0.63250.1572 -0.4814 0.1091 -0.3154 0.2236 -0.3991 0.5907 0.0221 -0.1172]×,函数
的最小值为
。
MATLAB源程序如下:
%%%%%%%%%%%%%%%%%粒子群算法求函数极值%%%%%%%%%%%%%%%%%%%%%%%%%%
%%%%%%%%%%%%%%%%%%%%%%%%初始化%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
clear all; %清除所有变量
close all; %清图
clc; %清屏
N=100; %群体粒子个数
D=10; %粒子维数
T=200; %最大迭代次数
c1=1.5; %学习因子1
c2=1.5; %学习因子2
w=0.8; %惯性权重
Xmax=20; %位置最大值
Xmin=-20; %位置最小值
Vmax=10; %速度最大值
Vmin=-10; %速度最小值
%%%%%%%%%%%%%%%%初始化种群个体(限定位置和速度)%%%%%%%%%%%%%%%%
x=rand(N,D) * (Xmax-Xmin)+Xmin;
v=rand(N,D) * (Vmax-Vmin)+Vmin;
%%%%%%%%%%%%%%%%%%初始化个体最优位置和最优值%%%%%%%%%%%%%%%%%%%
p=x;
pbest=ones(N,1);
for i=1:N
pbest(i)=func1(x(i,:));
end
%%%%%%%%%%%%%%%%%%%初始化全局最优位置和最优值%%%%%%%%%%%%%%%%%%
g=ones(1,D);
gbest=inf;
for i=1:N
if(pbest(i)<gbest)
g=p(i,:);
gbest=pbest(i);
end
end
gb=ones(1,T);
%%%%%%%%%%%按照公式依次迭代直到满足精度或者迭代次数%%%%%%%%%%%%%
for i=1:T
for j=1:N
%%%%%%%%%%%%%%更新个体最优位置和最优值%%%%%%%%%%%%%%%%%
if (func1(x(j,:))<pbest(j))
p(j,:)=x(j,:);
pbest(j)=func1(x(j,:));
end
%%%%%%%%%%%%%%%%更新全局最优位置和最优值%%%%%%%%%%%%%%%
if(pbest(j)<gbest)
g=p(j,:);
gbest=pbest(j);
end
%%%%%%%%%%%%%%%%%跟新位置和速度值%%%%%%%%%%%%%%%%%%%%%
v(j,:)=w*v(j,:)+c1*rand*(p(j,:)-x(j,:))...
+c2*rand*(g-x(j,:));
x(j,:)=x(j,:)+v(j,:);
%%%%%%%%%%%%%%%%%%%%边界条件处理%%%%%%%%%%%%%%%%%%%%%%
for ii=1:D
if (v(j,ii)>Vmax) | (v(j,ii)< Vmin)
v(j,ii)=rand * (Vmax-Vmin)+Vmin;
end
if (x(j,ii)>Xmax) | (x(j,ii)< Xmin)
x(j,ii)=rand * (Xmax-Xmin)+Xmin;
end
end
end
%%%%%%%%%%%%%%%%%%%%记录历代全局最优值%%%%%%%%%%%%%%%%%%%%%
gb(i)=gbest;
end
g; %最优个体
gb(end); %最优值
figure
plot(gb)
xlabel('迭代次数');
ylabel('适应度值');
title('适应度进化曲线')
%%%%%%%%%%%%%%%%%%%适应度函数%%%%%%%%%%%%%%%%%%%%
function result=func1(x)
summ=sum(x.^2);
result=summ;
end
[求最小值]函数的最小值,其中x的取值范围为[-4,4],y的取值范围为[-4,4]。这是一个有多个局部极值的函数,其函数值图形如图所示,其MATLAB实现程序如下
%%%%%%%%%f(x,y)=3*cos(x*y)+x+y*y%%%%%%%%%%
clear all; %清除所有变量
close all; %清图
clc; %清屏
x=-4:0.02:4;
y=-4:0.02:4;
N=size(x,2);
for i=1:N
for j=1:N
z(i,j)=3*cos(x(i)*y(j))+x(i)+y(j)*y(j);
end
end
mesh(x,y,z)
xlabel('x')
ylabel('y')
解:仿真过程如下:
(1)初始化群体粒子个数为N=100,粒子维数为D=2,最大迭代次数为T=200,学习因子c1=c2=1.5,惯性权重最大值为,惯性权重最小值为
,位置最大值为
,位置最小值为
,速度最大值为
,速度最小值为
。
(2)初始化种群粒子位置x和速度v,粒子个体最优位置p和最优值pbest,粒子群全局最优位置g和最优值gbest。
(3)计算动态惯性权重值w,更新位置x和速度值v,并进行边界条件处理,判断是否替换粒子个体最优位置和最优值
,以及粒子群全局最优位置
和最优值
。
(4)判断是否满足终止条件:若满足,则结束搜索过程,输出优化值;若不满足,则继续进行迭代优化
优化结束后,其适应度进化曲线如图6.4所示。优化后的结果为:在x=-4.0000,y=-0.7578时,函数取得最小值-6.408
MATLAB源程序如下:
%%%%%%%%%%%%%%%%%粒子群算法求函数极值%%%%%%%%%%%%%%%%%%%%%%%%%%
%%%%%%%%%%%%%%%%%%%%%%%%初始化%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
clear all; %清除所有变量
close all; %清图
clc; %清屏
N=100; %群体粒子个数
D=2; %粒子维数
T=200; %最大迭代次数
c1=1.5; %学习因子1
c2=1.5; %学习因子2
Wmax=0.8; %惯性权重最大值
Wmin=0.4; %惯性权重最小值
Xmax=4; %位置最大值
Xmin=-4; %位置最小值
Vmax=1; %速度最大值
Vmin=-1; %速度最小值
%%%%%%%%%%%%%%%%初始化种群个体(限定位置和速度)%%%%%%%%%%%%%%%%
x=rand(N,D) * (Xmax-Xmin)+Xmin;
v=rand(N,D) * (Vmax-Vmin)+Vmin;
%%%%%%%%%%%%%%%%%%初始化个体最优位置和最优值%%%%%%%%%%%%%%%%%%%
p=x;
pbest=ones(N,1);
for i=1:N
pbest(i)=func2(x(i,:));
end
%%%%%%%%%%%%%%%%%%%初始化全局最优位置和最优值%%%%%%%%%%%%%%%%%%
g=ones(1,D);
gbest=inf;
for i=1:N
if(pbest(i)<gbest)
g=p(i,:);
gbest=pbest(i);
end
end
gb=ones(1,T);
%%%%%%%%%%%按照公式依次迭代直到满足精度或者迭代次数%%%%%%%%%%%%%
for i=1:T
for j=1:N
%%%%%%%%%%%%%%更新个体最优位置和最优值%%%%%%%%%%%%%%%%%
if (func2(x(j,:))<pbest(j))
p(j,:)=x(j,:);
pbest(j)=func2(x(j,:));
end
%%%%%%%%%%%%%%%%更新全局最优位置和最优值%%%%%%%%%%%%%%%
if(pbest(j)<gbest)
g=p(j,:);
gbest=pbest(j);
end
%%%%%%%%%%%%%%%%计算动态惯性权重值%%%%%%%%%%%%%%%%%%%%
w=Wmax-(Wmax-Wmin)*i/T;
%%%%%%%%%%%%%%%%%跟新位置和速度值%%%%%%%%%%%%%%%%%%%%%
v(j,:)=w*v(j,:)+c1*rand*(p(j,:)-x(j,:))...
+c2*rand*(g-x(j,:));
x(j,:)=x(j,:)+v(j,:);
%%%%%%%%%%%%%%%%%%%%边界条件处理%%%%%%%%%%%%%%%%%%%%%%
for ii=1:D
if (v(j,ii)>Vmax) | (v(j,ii)< Vmin)
v(j,ii)=rand * (Vmax-Vmin)+Vmin;
end
if (x(j,ii)>Xmax) | (x(j,ii)< Xmin)
x(j,ii)=rand * (Xmax-Xmin)+Xmin;
end
end
end
%%%%%%%%%%%%%%%%%%%%记录历代全局最优值%%%%%%%%%%%%%%%%%%%%%
gb(i)=gbest;
end
g; %最优个体
gb(end); %最优值
figure
plot(gb)
xlabel('迭代次数');
ylabel('适应度值');
title('适应度进化曲线')
%%%%%%%%%%%%%%%%%%%%%适应度函数%%%%%%%%%%%%%%%%%%%%%%%
function value=func2(x)
value=3*cos(x(1)*x(2))+x(1)+x(2)^2;
end
[求最小值]用离散粒子群算法求函数的最小值,其中x的取值范围为[0,9]。这是一个有多个局部极值的函数,其函数值图形如图所示,其MATLAB实现程序如下:
%%%%%%%%%f(x)=x+6sin(4x)+9cos(5x)%%%%%%%%%%
clear all; %清除所有变量
close all; %清图
clc; %清屏
x=0:0.01:9;
y=x+6*sin(4*x)+9*cos(5*x);
plot(x,y)
xlabel('x')
ylabel('f(x)')
title('f(x)=x+6sin(4x)+9cos(5x)')
解:仿真过程如下:
(1)初始化群体粒子个数为N=100,粒子维数为D=20,最大迭代次数为T=200,学习因子c1=c2=1.5,惯性权重最大值为,惯性权重最小值为
,位置最大值为
,位置最小值为
,速度最大值为
,速度最小值为
。
(2)初始化速度v和二进制编码的种群粒子位置x,计算适应度值,获得粒子个体最优位置p和最优值pbest,以及粒子群全局最优位置g和最优值gbest。
(3)计算动态惯性权重值w,更新位置x和速度值v,并进行边界条件处理,判断是否替换粒子个体最优位置和最优值
,以及粒子群全局最优位置
和最优值
。
(4)判断是否满足终止条件:若满足,则结束搜索过程,输出优化值;若不满足,则继续进行迭代优化。
其适应度进化曲线如图所示。优化后的结果为:当x=4.37时,函数取得最小值-10.42。
MATLAB源程序如下:
%%%%%%%%%%%%%%%%%离散粒子群算法求函数极值%%%%%%%%%%%%%%%%%%%%%%%
%%%%%%%%%%%%%%%%%%%%%%%%初始化%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
clear all; %清除所有变量
close all; %清图
clc; %清屏
N=100; %群体粒子个数
D=20; %粒子维数
T=200; %最大迭代次数
c1=1.5; %学习因子1
c2=1.5; %学习因子2
Wmax=0.8; %惯性权重最大值
Wmin=0.4; %惯性权重最小值
Xs=9; %位置最大值
Xx=0; %位置最小值
Vmax=10; %速度最大值
Vmin=-10; %速度最小值
%%%%%%%%%%%%%%%%初始化种群个体(限定位置和速度)%%%%%%%%%%%%%%%%
x=randi([0,1],N,D); %随机获得二进制编码的初始种群
v=rand(N,D) * (Vmax-Vmin)+Vmin;
%%%%%%%%%%%%%%%%%%初始化个体最优位置和最优值%%%%%%%%%%%%%%%%%%%
p=x;
pbest=ones(N,1);
for i=1:N
pbest(i)= func3(x(i,:),Xs,Xx);
end
%%%%%%%%%%%%%%%%%%%初始化全局最优位置和最优值%%%%%%%%%%%%%%%%%%
g=ones(1,D);
gbest=inf;
for i=1:N
if(pbest(i)<gbest)
g=p(i,:);
gbest=pbest(i);
end
end
gb=ones(1,T);
%%%%%%%%%%%按照公式依次迭代直到满足精度或者迭代次数%%%%%%%%%%%%%
for i=1:T
for j=1:N
%%%%%%%%%%%%%%更新个体最优位置和最优值%%%%%%%%%%%%%%%%%
if (func3(x(j,:),Xs,Xx)<pbest(j))
p(j,:)=x(j,:);
pbest(j)=func3(x(j,:),Xs,Xx);
end
%%%%%%%%%%%%%%%%更新全局最优位置和最优值%%%%%%%%%%%%%%%
if(pbest(j)<gbest)
g=p(j,:);
gbest=pbest(j);
end
%%%%%%%%%%%%%%%%计算动态惯性权重值%%%%%%%%%%%%%%%%%%%%
w=Wmax-(Wmax-Wmin)*i/T;
%%%%%%%%%%%%%%%%%跟新位置和速度值%%%%%%%%%%%%%%%%%%%%%
v(j,:)=w*v(j,:)+c1*rand*(p(j,:)-x(j,:))...
+c2*rand*(g-x(j,:));
%%%%%%%%%%%%%%%%%%%%边界条件处理%%%%%%%%%%%%%%%%%%%%%%
for ii=1:D
if (v(j,ii)>Vmax) | (v(j,ii)< Vmin)
v(j,ii)=rand * (Vmax-Vmin)+Vmin;
end
end
vx(j,:)=1./(1+exp(-v(j,:)));
for jj=1:D
if vx(j,jj)>rand
x(j,jj)=1;
else
x(j,jj)=0;
end
end
end
%%%%%%%%%%%%%%%%%%%%记录历代全局最优值%%%%%%%%%%%%%%%%%%%%%
gb(i)=gbest;
end
g; %最优个体
m=0;
for j=1:D
m=g(j)*2^(j-1)+m;
end
f1=Xx+m*(Xs-Xx)/(2^D-1); %最优值
figure
plot(gb)
xlabel('迭代次数');
ylabel('适应度值');
title('适应度进化曲线')
%%%%%%%%%%%%%%%%%%%%%%%%%适应度函数%%%%%%%%%%%%%%%%%%%%%%%%%%%%
function result=func3(x,Xs,Xx)
m=0;
D=length(x);
for j=1:D
m=x(j)*2^(j-1)+m;
end
f=Xx+m*(Xs-Xx)/(2^D-1); %译码成十进制数
fit= f+6*sin(4*f)+9*cos(5*f);
result=fit;
end