多目标优化算法：多目标红鸢优化算法（MultiobjectiveRedkiteOptimizationAlgorithm，MOROA）提供Matlab代码

红鸢优化算法&＃xff08;Red kite Optimization Algorithm&＃xff0c;ROA&＃xff09;由Jalal Raeisi Gahruei和Zahra Beheshti于2022年提出。ROA详细信息见&＃xff1a;点击查看红鸢优化算法ROA原理

参考文献&＃xff1a;Jalal Raeisi Gahruei, Zahra Beheshti, The Electricity Consumption Prediction using Hybrid Red kite Optimization Algorithm with Multi-Layer Perceptron Neural Network, Journal of Intelligent Procedures in Electrical Technology, (2022), Vol. 15 (60), pp. 1-22.

多目标红鸢优化算法&＃xff08;Multi-objective Red kite Optimization Algorithm&＃xff0c;MOROA&＃xff09;由红鸢优化算法的优良策略与多目标优化思想融合而成。为了验证所提的MOROA的有效性&＃xff0c;将其在46个多目标测试函数&＃xff08;ZDT1、ZDT2、ZDT3、ZDT4、ZDT6、DTLZ1-DTLZ7、WFG1-WFG10、UF1-UF10、CF1-CF10、Kursawe、Poloni、Viennet2、Viennet3&＃xff09;以及1个工程应用&＃xff08;盘式制动器设计&＃xff09;上实验&＃xff0c;并采用IGD、GD、HV、SP进行指标评价。

close all;
clear ;
clc;
% Multi-objective Red kite Optimization Algorithm (MOROA)
%%
% TestProblem测试问题说明&＃xff1a;
%一共46个多目标测试函数&＃xff0c;详情如下&＃xff1a;
%1-5:ZDT1、ZDT2、ZDT3、ZDT4、ZDT6
%6-12&＃xff1a;DTLZ1-DTLZ7
%13-22&＃xff1a;wfg1-wfg10
%23-32&＃xff1a;uf1-uf10
%33-42&＃xff1a;cf1-cf10
%43-46:Kursawe、Poloni、Viennet2、Viennet3
%47 盘式制动器设计&＃xff08;工程应用&＃xff09; 温泽宇,谢珺,谢刚,续欣莹.基于新型拥挤度距离的多目标麻雀搜索算法[J].计算机工程与应用,2021,57(22):102-109.
%%
TestProblem&＃61;1;
MultiObj &＃61; GetFunInfo(TestProblem);
MultiObjFnc&＃61;MultiObj.name;%问题名
% Parameters
params.Np &＃61; 300; % Population size
params.Nr &＃61; 300; % Repository size
params.maxgen &＃61;100; % Maximum number of generations
name&＃61;MultiObj.name;
% MOROA
REP &＃61; MOROA(params,MultiObj);
Xbest&＃61;REP.pos;%Repository particles positions
Fbest&＃61;REP.pos_fit;%Repository fitness values
%% 画结果图
figure(2)
if(size(Fbest,2)&＃61;&＃61;2)h_rep &＃61; plot(Fbest(:,1),Fbest(:,2),&＃39;ok&＃39;); hold on;if(isfield(MultiObj,&＃39;truePF&＃39;))h_pf &＃61; plot(MultiObj.truePF(:,1),MultiObj.truePF(:,2),&＃39;.r&＃39;); hold on;legend(&＃39;MOROA&＃39;,&＃39;TruePF&＃39;);elselegend(&＃39;MOROA&＃39;);endgrid on; xlabel(&＃39;f1&＃39;); ylabel(&＃39;f2&＃39;);
end
if(size(Fbest,2)&＃61;&＃61;3)h_rep &＃61; plot3(Fbest(:,1),Fbest(:,2),Fbest(:,3),&＃39;ok&＃39;); hold on;if(isfield(MultiObj,&＃39;truePF&＃39;))h_pf &＃61; plot3(MultiObj.truePF(:,1),MultiObj.truePF(:,2),MultiObj.truePF(:,3),&＃39;.r&＃39;); hold on;legend(&＃39;MOROA&＃39;,&＃39;TruePF&＃39;);elselegend(&＃39;MOROA&＃39;);endgrid on; xlabel(&＃39;f1&＃39;); ylabel(&＃39;f2&＃39;);zlabel(&＃39;f3&＃39;);
end
title(name)%% Metric Value
Obtained_Pareto&＃61;REP.pos_fit;
if(isfield(MultiObj,&＃39;truePF&＃39;))%判断是否有参考的PF
True_Pareto&＃61;MultiObj.truePF;
% ResultData的值分别是IGD、GD、HV、Spacing (HV越大越好&＃xff0c;其他指标越小越好)
ResultData&＃61;[IGD(Obtained_Pareto,True_Pareto),GD(Obtained_Pareto,True_Pareto),HV(Obtained_Pareto,True_Pareto),Spacing(Obtained_Pareto)];
disp([&＃39;IGD、GD、HV、SP&＃xff1a;&＃39;,num2str(ResultData)]);
else%计算每个算法的Spacing&＃xff0c;Spacing越小说明解集分布越均匀ResultData&＃61;Spacing(Obtained_Pareto);%计算的Spacingdisp([&＃39;SP&＃xff1a;&＃39;,num2str(ResultData)]);
end


部分结果&＃xff1a;

ZDT1&＃xff1a;

ZDT2&＃xff1a;

ZDT3&＃xff1a;

Viennet3:

盘式制动器设计:

文件夹内包含所有代码&＃xff0c;可以直接点击main.m运行。