模拟退火算法应用java_模拟退火算法（Simulated Annealing，SA）MATLAB案例详细解析...

一、理论

二、模拟退火算法应用于函数优化

1.流程图

SAToFunction流程图

2.代码解析

clear all; %清除所有变量

close all; %清图

clc ; %清屏

D = 10;

Xs = 20;%%函数上限

Xx =-20;%%函数下限

%%%%%%%%%%%%%%%%%%%%冷却表参数%%%%%%%%%%%%%%%%%%%%%%%%%%

L= 200; %马尔可夫链长度

K = 0.998; %衰减参数

S = 0.01; %步长因子，就是每次解变化的大小

T = 100; % 初始温度

YZ = 1e-8; %容差

P = 0; %Metropolis过程接受点

%%%%%%%%%%%%%%%%%随机选点初值设定%%%%%%%%%%%%%%%%%%%5

PreX = rand(D,1)*(Xs-Xx)+Xx;%初始化矩阵D × 1的解

PreBestX = PreX;%用于保存上一个最优的解

PreX = rand(D,1)*(Xs-Xx)+Xx;

BestX = PreX;%用于保存当前最优解

%%%%%%%%%%%%%%%%%每迭代一次退火一次(降温)，直到满足迭代条件为止%%%%%%%%%

deta = abs(func1(BestX)-func1(PreBestX));%能量差值，适应度差值

while (deta>YZ) && (T>0.001)

T = K*T;%%降温

%%%%%%%%%%%%在当前温度T下迭代次数%%%%%%%%%%%%

for i = 1:L

%%%%%%%%%%%%在此点附近随机选择下一点%%%%%%%%%%%%%%%

%%%%%%%%%%%%根据当前条件产生一个新解%%%%%%%%%%%%%%%%%

NextX = PreX + S*(rand(D,1) * (Xs-Xx)+Xx);

%%%%%%%%%%%%%%%边界条件处理%%%%%%%%%%%%%