1 简介
2015 年，Hussain Shareef 等基于闪电的机理提出了一种新型的启发式优化算法———闪电搜索算法( lightning search algorithm，LSA) ，该算法具有调节参数少、收敛精度高和全局寻优能力强等优点，已在函数优化、旅行商问题寻优等方面得到应用。
2 部分代码
%{
Lightning Search Algorithm (LSA)
Platform: Matlab 2012a,2013a
Cite this algorithm as;
[1]. Hussain Shareef , Ahmad Asrul Ibrahim, Ammar Hussein Mutlag, Lightning Search Algorithm,
Applied Soft Computing 36 (2015) 315?333.
Copyright Notice
Copyright (c) 2015, Hussain Shareef and Ahmad Asrul Ibrahim
All rights reserved.
Redistribution and use in source and binary forms, with or without
modification, are permitted provided that the following conditions are
met:
* Redistributions of source code must retain the above copyright
notice, this list of conditions and the following disclaimer.
* Redistributions in binary form must reproduce the copyright
notice, this list of conditions and the following disclaimer in
the documentation and/or other materials provided with the distribution
THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE
LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
POSSIBILITY OF SUCH DAMAGE.
%}
clc
clear all
close all
tic;%start stopwatch
F_index =6; % Function 1-6 % Refer to other test function 7-24 in [1].
[low,up,dim]=test_functions_range(F_index);
N = 20; % population size/number of agents
D = dim; % number of dimension
T = 100; % maximum number of iteration
% Set upper bound & lower bound
if size(up,2)==1
UB = ones(1,D).*up;
LB = ones(1,D).*low;
end
if size(up,2)>1
UB = up;
LB = low;
end
%randomly create 1st population
for d = 1:D
Dpoint(:,d)= rand(N,1)*(UB(d)-LB(d))+LB(d);
end
ch_time = 0; % reset
max_ch_time = 10;
fit_old = 10^10*(ones(1,N));
direct = sign(unifrnd(-1,1,1,dim));
for t = 1:T
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
% Evaluate performance
fit = evaluateF(Dpoint,F_index);
Ec = fit;
% update channel
ch_time = ch_time+1;
if ch_time >= max_ch_time
[Ms ds]=sort(Ec,'ascend');
Dpoint(ds(N),:) = Dpoint(ds(1),:); % Eliminate the worst channel
Ec(ds(N)) = Ec(ds(1)); % Update
ch_time = 0; % reset
end
% Rangking the fitness value
[Ms ds]=sort(Ec,'ascend');
best = Ec(ds(1));
worst = Ec(ds(N));
Energy = 2.05 - 2*exp(-5*(T-t)/T);% Update energy
% update direction
for d = 1:D
Dpoint_test = Dpoint(ds(1),:);
Dpoint_test(d) = Dpoint_test(d)+direct(d)*0.005*(UB(d)-LB(d));
fv_test = evaluateF(Dpoint_test,F_index);
if fv_test < best % If better, +ve direction
direct(d) = direct(d);
else
direct(d) = -1*direct(d);
end
end
% update position
for i = 1:N
dist=Dpoint(i,:)-Dpoint(ds(1),:);
for d = 1:D
if Dpoint(i,:)==Dpoint(ds(1),:)
Dpoint_temp(d) = Dpoint(i,d)+direct(d)*abs(normrnd(0,Energy));
else
if dist(d)<0
Dpoint_temp(d) = Dpoint(i,d)+exprnd(abs(dist(d)));
else
Dpoint_temp(d) = Dpoint(i,d)-exprnd(dist(d));
end
end
if (Dpoint_temp(d)>UB(d))||(Dpoint_temp(d)<LB(d))
Dpoint_temp(d) = rand(1)*(UB(d)-LB(d))+LB(d); % Re-initialized
end
end
fv = evaluateF(Dpoint_temp,F_index);
if fv < Ec(i)
Dpoint(i,:) = Dpoint_temp;
Ec(i) = fv;
% Focking procedure
if rand < 0.01
for d = 1:D
Dpoint_fock(d) = UB(d)+LB(d)-Dpoint_temp(d);% Focking
end
fock_fit = evaluateF(Dpoint_fock,F_index); % Evaluate
if fock_fit < Ec(i)
Dpoint(i,:) = Dpoint_fock; % Replace the channel
Ec(i) = fock_fit;
end
end
end
end
if best == worst
break
end
% record the performance
fitness(t) = min(fit);
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
end
elapsed_time=toc; % time taken for this algorithm
% select the optimal value
grandmin = min(fitness);
y = fitness;
x = 1:length(y);
plot(x,y,'Linewidth',2)
xlabel('No of Iteration');
ylabel('Fitness Value');
axis([0 max(x) min(y) max(y)]);
fprintf(1,'Optimal value = %f\n', grandmin);
3 仿真结果
4 参考文献
[1]卢玉婷. 元启发式闪电搜索算法及应用研究[D]. 广西大学, 2019.
博主简介：擅长智能优化算法、神经网络预测、信号处理、元胞自动机、图像处理、路径规划、无人机等多种领域的Matlab仿真，有科研问题可私信交流。
部分理论引用网络文献，若有侵权联系博主删除。
