matlab常用算法大全Word文档格式.docx
《matlab常用算法大全Word文档格式.docx》由会员分享，可在线阅读，更多相关《matlab常用算法大全Word文档格式.docx（20页珍藏版）》请在冰豆网上搜索。
a'
b'
x0'
},{u
（1）,u
（2）,x1
（1）}）;
yuce1=subs（x,'
t'
[0:
n-1]）;
digits（6）,y=vpa（x）%为提高预测精度，先计算预测值，再显示微分方程的解
yuce=[x0
（1）,diff（yuce1）]
epsilon=x0-yuce%计算残差
delta=abs（epsilon./x0）%计算相对误差
rho=1-（1-0.5*u
（1））/（1+0.5*u
（1））*lamda%计算级比偏差值
%以深圳人口数据得到预测模型及预测误差相关数据
lamda=
Columns1through8
0.97410.96110.94190.87490.93110.90930.93020.9254
Columns9through16
0.92450.92780.94420.93760.91270.91480.93320.9477
Columns17through24
0.95920.94450.95510.95620.95940.94610.94690.9239
Columns25through31
0.91400.90770.92430.92680.93120.94460.9618
range=
0.87490.9741
x1=
1.0e+003*
0.03130.06340.09670.13220.17270.21620.26410.3155
0.37110.43130.49610.56470.63800.71820.80590.8999
0.99901.10241.21191.32651.44631.57121.70331.8427
Columns25through32
1.99362.15882.34072.53752.74992.97803.21943.4705
u=
-0.0665
31.3737
y=
-472.117+503.377*exp（.664533e-1*t）
yuce=
31.260034.587636.964139.504042.218345.119248.219451.5326
55.073458.857662.901767.223871.842876.779282.054887.6928
93.7183100.1578107.0397114.3945122.2547130.6550139.6324149.2267
159.4802170.4382182.1492194.6649208.0405222.3352237.6121253.9386
epsilon=
0-2.4976-3.5741-4.0540-1.6983-1.5992-0.3594-0.0826
0.52661.28241.91831.42621.37723.44085.63526.2772
5.44173.22222.42030.2055-2.4047-5.7350-7.5924-9.7767
-8.5502-5.3082-0.21922.16514.33955.73483.8379-2.9086
delta=
00.07780.10700.11440.04190.03670.00750.0016
0.00950.02130.02960.02080.01880.04290.06430.0668
0.05490.03120.02210.00180.02010.04590.05750.0701
0.05670.03210.00120.01100.02040.02510.01590.0116
rho=
-0.0411-0.0271-0.00660.06500.00490.02820.00580.0110
0.01190.0084-0.0091-0.00200.02450.02230.0027-0.0128
-0.0251-0.0094-0.0208-0.0219-0.0254-0.0111-0.01190.0126
0.02320.03000.01220.00950.0048-0.0095-0.0280
二、遗传算法程序代码
%Optimizingafunction
usingSimpleGeneticAlgorithmwithelitistpreserved
%Maxf（x1,x2）=100*（x1*x1-x2）.^2+（1-x1）.^2;
-2.0480<
=x1,x2<
=2.0480
%Author:
WangYonglin（wylin77@）
clc;
clearall;
formatlong;
%设定数据显示格式
%初始化参数
T=100;
%仿真代数
N=80;
%群体规模
pm=0.05;
pc=0.8;
%交叉变异概率
umax=2.048;
umin=-2.048;
%参数取值范围
L=10;
%单个参数字串长度，总编码长度2L
bval=round（rand（N,2*L））;
%初始种群
bestv=-inf;
%最优适应度初值
%迭代开始
forii=1:
T
%解码，计算适应度
fori=1:
N
y1=0;
y2=0;
forj=1:
1:
L
y1=y1+bval（i,L-j+1）*2^（j-1）;
x1=（umax-umin）*y1/（2^L-1）+umin;
y2=y2+bval（i,2*L-j+1）*2^（j-1）;
x2=（umax-umin）*y2/（2^L-1）+umin;
obj（i）=100*（x1*x1-x2）.^2+（1-x1）.^2;
%目标函数
xx（i,:
）=[x1,x2];
func=obj;
%目标函数转换为适应度函数
p=func./sum（func）;
q=cumsum（p）;
%累加
[fmax,indmax]=max（func）;
%求当代最佳个体
iffmax>
=bestv
bestv=fmax;
%到目前为止最优适应度值
bvalxx=bval（indmax,:
%到目前为止最佳位串
optxx=xx（indmax,:
%到目前为止最优参数
end
Bfit1（ii）=bestv;
%存储每代的最优适应度
%%%%遗传操作开始
%轮盘赌选择
（N-1）
r=rand;
tmp=find（r<
=q）;
newbval（i,:
）=bval（tmp
（1）,:
end
newbval（N,:
）=bvalxx;
%最优保留
bval=newbval;
%单点交叉
2:
cc=rand;
ifcc<
pc
point=ceil（rand*（2*L-1））;
%取得一个1到2L-1的整数
ch=bval（i,:
bval（i,point+1:
2*L）=bval（i+1,point+1:
2*L）;
bval（i+1,point+1:
2*L）=ch（1,point+1:
bval（N,:
%位点变异
mm=rand（N,2*L）<
pm;
%N行
mm（N,:
）=zeros（1,2*L）;
%最后一行不变异，强制赋0
bval（mm）=1-bval（mm）;
%输出
plot（Bfit1）;
%绘制最优适应度进化曲线
bestv
%输出最优适应度值
optxx
%输出最优参数
三、种子群算法程序代码
%declaretheparametersoftheoptimization
max_iterations=1000;
no_of_particles=50;
dimensions=1;
delta_min=-0.003;
delta_max=0.003;
c1=1.3;
c2=1.3;
%initialisetheparticlesandteirvelocitycomponents
forcount_x=1:
no_of_particles
forcount_y=1:
dimensions
particle_position（count_x,count_y）=rand*10;
particle_velocity（count_x,count_y）=rand;
p_best（count_x,count_y）=particle_position（count_x,count_y）;
end
%initializethep_best_fitnessarray
forcount=1:
p_best_fitness（count）=-1000;
%particle_position
%particle_velocity
%mainparticleswrmroutine
max_iterations
%findthefitnessofeachparticle
%changefitnessfunctionasperequationrequiresdanddimensions
%x=particle_position（count_x,1）;
%y=particle_position（count_x,2）;
%z=particle_position（count_x,3）;
%soln=x^2-3*y*x+z;
%x=particle_position（count_x）;
%soln=x^2-2*x+1;
x=particle_position（count_x）;
soln=x-7;
ifsoln~=0
current_fitness（count_x）=1/abs（soln）;
else
current_fitness=1000;
%decideonp_bestetcforeachparticle
ifcurrent_fitness（count_x）>
p_best_fitness（count_x）
p_best_fitness（count_x）=current_fitness（count_x）;
%decideontheglobalbestamongalltheparticles
[g_best_val,g_best_index]=max（current_fitness）;
%g_bestcontainsthepositionoftehglobalbest
g_best（count_y）=particle_position（g_best_index,count_y）;
%updatethepositionandvelocitycompponents
p_current（count_y）=particle_position（count_x,count_y）;
particle_velocity（count_y）=particle_velocity（count_y）+c1*rand*（p_best（count_y）-p_current（count_y））+c2*rand*（g_best（count_y）-p_current（count_y））;
particle_positon（count_x,count_y）=p_current（count_y）+particle_velocity（count_y）;
g_best
current_fitness（g_best_index）
clearall,clc%psoexample
iter=1000;
%numberofalgorithmiterations
np=2;
%numberofmodelparameters
ns=10;
%numberofsetsofmodelparameters
Wmax=0.9;
%maximuminertialweight
Wmin=0.4;
%minimuminertialweight
c1=2.0;
%parameterinPSOmethodology
c2=2.0;
Pmax=[1010];
%maximummodelparametervalue
Pmin=[-10-10];
%minimummodelparametervalue
Vmax=[11];
%maximumchangeinmodelparameter
Vmin=[-1-1];
%minimumchangeinmodelparameter
modelparameters（1:
np,1:
ns）=0;
%setallmodelparameterestimatesforallmodelparametersetstozero
modelparameterchanges（1:
%setallchangeinmodelparameterestimatesforallmodelparametersetstozero
bestmodelparameters（1:
%setbestmodelparameterestimatesforallmodelparametersetstozero
setbestcostfunction（1:
ns）=1e6;
%setbestcostfunctionofeachmodelparametersettoalargenumber
globalbestparameters（1:
np）=0;
%setbestmodelparametervaluesforallmodelparametersetstozero
bestparameters=globalbestparameters'
%bestmodelparametervaluesforallmodelparametersets（toplot）
globalbestcostfunction=1e6;
%setbestcostfunctionforallmodelparametersetstoalargenumber
i=0;
%indicatesithalgorithmiteration
j=0;
%indicatesjthsetofmodelparameters
k=0;
%indicateskthmodelparameter
fork=1:
np%initialization
forj=1:
ns
modelparameters（k,j）=（Pmax（k）-Pmin（k））*rand
（1）+Pmin（k）;
%randomlydistributemodelparameters
modelparameterchanges（k,j）=（Vmax（k）-Vmin（k））*rand
（1）+Vmin（k）;
%randomlydistributechangeinmodelparameters
fori=2:
iter
x=modelparameters（:
j）;
%calculatecostfunction
costfunction=105*（x
（2）-x
（1）^2）^2+（1-x
（1））^2;
ifcostfunction<
setbestcostfunction（j）%bestcostfunctionforjthsetofmodelparameters
bestmodelparameters（:
j）=modelparameters（:
setbestcostfunction（j）=costfunction;
四、模拟退火算法
%ford=1:
50%循环10次发现最小路径为4.115，循环50次有3次出现4.115
T_max=80;
%input（'
pleaseinputthestarttemprature'
T_min=0.001;
pleaseinputtheendtemprature'
iter_max=100;
%input（'
pleaseinputthemostinterpstepsonthefittemp'
s_max=100;
pleaseinputthemoststeadystepsontthefittemp'
T=T_max;
load.\address.txt;
order1=randperm（size（address,1））'
%生成初始解。
figure
（1）;
plot（address（order1,1）,address（order1,2）,'
*r-'
）
title（'
随机产生的路径'
totaldis1=distance（address,order1）;
forn=1:
size（address,1）
text（address（n,1）+0.01,address（n,2）,num2str（n））%标号
end
text（0.9,0.9,num2str（totaldis1））
figure
（2）;
whileT>
=T_min
iter_num=1;
s_num=1;
plot（T,totaldis1,'
r.'
holdon
whileiter_num<
iter_max&
s_num<
s_max;
order2=exhgpath（order1）;
%随机交换两个城市位置
totaldis2=distance（address,order2）;
R=rand;
DeltaDis=totaldis2-totaldis1;
%新的距离-原来的距离
ifDeltaDis<
0;
order1=order2;
totaldis1=totaldis2;
%新距离小，无条件接受
elseif（exp（（totaldis1-totaldis2）/（T））>
R）%%%%本算法最核心的思想：
以一定概率接受坏的结果，防止局部最优
elses_num=s_num+1;
iter_num=iter_num+1;
T=T*0.99;
set（gca,'
xscale'
log'
%或者使用semilogx，有相同效果
xlabel（'
退火温度'
ylabel（'
总距离'
order1
totaldis1
figure（3）
plot（address（order1,1）
升级会员 