灰狼算法优化BP神经网络（GWO-BP）

%灰狼优化多输入多输出BP神经网络代码 clear clc tic global SamIn SamOut HiddenUnitNum InDim OutDim TrainSamNum %% 导入训练数据 data = xlsread('test_data1.xlsx'); [data_m,data_n] = size(data);%获取数据维度 P = 80; %百分之P的数据用于训练，其余测试 Ind = floor(P * data_m / 100); train_data = data(1:Ind,1:end-1)'; train_result = data(1:Ind,end)'; test_data = data(Ind+1:end,1:end-1)';% 利用训练好的网络进行预测 test_result = data(Ind+1:end,end)'; %% 初始化参数 [InDim,TrainSamNum] = size(train_data);% 学习样本数量 [OutDim,TrainSamNum] = size(train_result); HiddenUnitNum = 7; % 隐含层神经元个数 [SamIn,PS_i] = mapminmax(train_data,0,1); % 原始样本对（输入和输出）初始化 [SamOut,PS_o] = mapminmax(train_result,0,1); W1 = HiddenUnitNum*InDim; % 初始化输入层与隐含层之间的权值 B1 = HiddenUnitNum; % 初始化输入层与隐含层之间的阈值 W2 = OutDim*HiddenUnitNum; % 初始化输出层与隐含层之间的权值 B2 = OutDim; % 初始化输出层与隐含层之间的阈值 L = W1+B1+W2+B2; %粒子维度 %%优化参数的设定 dim=L; % 优化的参数 number of your variables for j=1:L lb(1,j)=-3.5; % 参数取值下界 ub(1,j)=3.5; end% 参数取值上界 %%GWO算法初始化 SearchAgents_no=150; % 狼群数量，Number of search agents Max_iteration=3000; % 最大迭代次数，Maximum numbef of iterations % initialize alpha, beta, and delta_posAlpha_pos=zeros(1,dim); % 初始化Alpha狼的位置 Alpha_score=inf; % 初始化Alpha狼的目标函数值，change this to -inf for maximization problems Beta_pos=zeros(1,dim); % 初始化Beta狼的位置 Beta_score=inf; % 初始化Beta狼的目标函数值，change this to -inf for maximization problems Delta_pos=zeros(1,dim); % 初始化Delta狼的位置 Delta_score=inf; % 初始化Delta狼的目标函数值，change this to -inf for maximization problems %Initialize the positions of search agents Positions=initialization(SearchAgents_no,dim,ub,lb); Convergence_curve=zeros(1,Max_iteration); l=0; % Loop counter循环计数器 % Main loop主循环 while l<Max_iteration % 对迭代次数循环 for i=1:size(Positions,1) % 遍历每个狼 % Return back the search agents that go beyond the boundaries of the search space7 % 若搜索位置超过了搜索空间，需要重新回到搜索空间 Flag4ub=Positions(i,:)>ub; Flag4lb=Positions(i,:)<lb; % 若狼的位置在最大值和最小值之间，则位置不需要调整，若超出最大值，最回到最大值边界； % 若超出最小值，最回答最小值边界 Positions(i,:)=(Positions(i,:).*(~(Flag4ub+Flag4lb)))+ub.*Flag4ub+lb.*Flag4lb; % ~表示取反 % 计算适应度函数值 fitness=f(Positions(i,:)); % Update Alpha, Beta, and Delta if fitness<Alpha_score % 如果目标函数值小于Alpha狼的目标函数值 Alpha_score=fitness; % 则将Alpha狼的目标函数值更新为最优目标函数值，Update alpha Alpha_pos=Positions(i,:); % 同时将Alpha狼的位置更新为最优位置 end if fitness>Alpha_score && fitness<Beta_score % 如果目标函数值介于于Alpha狼和Beta狼的目标函数值之间 Beta_score=fitness; % 则将Beta狼的目标函数值更新为最优目标函数值，Update beta Beta_pos=Positions(i,:); % 同时更新Beta狼的位置 end if fitness>Alpha_score && fitness>Beta_score && fitness<Delta_score % 如果目标函数值介于于Beta狼和Delta狼的目标函数值之间 Delta_score=fitness; % 则将Delta狼的目标函数值更新为最优目标函数值，Update delta Delta_pos=Positions(i,:); % 同时更新Delta狼的位置 end end a=2-l*((2)/Max_iteration); % 对每一次迭代，计算相应的a值，a decreases linearly fron 2 to 0 % Update the Position of search agents including omegas for i=1:size(Positions,1) % 遍历每个狼 for j=1:size(Positions,2) % 遍历每个维度 % 包围猎物，位置更新 r1=rand(); % r1 is a random number in [0,1] r2=rand(); % r2 is a random number in [0,1] A1=2*a*r1-a; % 计算系数A，Equation (3.3) C1=2*r2; % 计算系数C，Equation (3.4) % Alpha狼位置更新 D_alpha=abs(C1*Alpha_pos(j)-Positions(i,j)); % Equation (3.5)-part 1 X1=Alpha_pos(j)-A1*D_alpha; % Equation (3.6)-part 1 r1=rand(); r2=rand(); A2=2*a*r1-a; % 计算系数A，Equation (3.3) C2=2*r2; % 计算系数C，Equation (3.4) % Beta狼位置更新 D_beta=abs(C2*Beta_pos(j)-Positions(i,j)); % Equation (3.5)-part 2 X2=Beta_pos(j)-A2*D_beta; % Equation (3.6)-part 2 r1=rand(); r2=rand(); A3=2*a*r1-a; % 计算系数A，Equation (3.3) C3=2*r2; % 计算系数C，Equation (3.4) % Delta狼位置更新 D_delta=abs(C3*Delta_pos(j)-Positions(i,j)); % Equation (3.5)-part 3 X3=Delta_pos(j)-A3*D_delta; % Equation (3.5)-part 3 % 位置更新 Positions(i,j)=(X1+X2+X3)/3;% Equation (3.7) end end l=l+1 Convergence_curve(l)=Alpha_score; end x=Alpha_pos; %% % x = gb; W1 = x(1:HiddenUnitNum*InDim); L1 = length(W1); W1 = reshape(W1,[HiddenUnitNum, InDim]); B1 = x(L1+1:L1+HiddenUnitNum)'; L2 = L1 + length(B1); W2 = x(L2+1:L2+OutDim*HiddenUnitNum); L3 = L2 + length(W2); W2 = reshape(W2,[OutDim, HiddenUnitNum]); B2 = x(L3+1:L3+OutDim)'; HiddenOut = logsig(W1 * SamIn + repmat(B1, 1, TrainSamNum)); % 隐含层网络输出 NetworkOut = W2 * HiddenOut + repmat(B2, 1, TrainSamNum); % 输出层网络输出 Error = SamOut - NetworkOut; % 实际输出与网络输出之差 Forcast_data = mapminmax('reverse',NetworkOut,PS_o); [OutDim,ForcastSamNum] = size(test_result); SamIn_test= mapminmax('apply',test_data,PS_i); % 原始样本对（输入和输出）初始化 HiddenOut_test = logsig(W1 * SamIn_test + repmat(B1, 1, ForcastSamNum)); % 隐含层输出预测结果 NetworkOut = W2 * HiddenOut_test + repmat(B2, 1, ForcastSamNum); % 输出层输出预测结果 Forcast_data_test = mapminmax('reverse',NetworkOut,PS_o); test_error=test_result(1,:)-Forcast_data_test(1,:); mean_error=mean(abs(test_error)/test_result) % test_mse=mean(test_error.^2) test_mse=sqrt(mean(test_error.^2)) %% 绘制结果 figure plot(Convergence_curve,'r') xlabel('迭代次数') ylabel('适应度') title('收敛曲线') figure subplot(2,2,1); plot(train_result(1,:), 'r-*') hold on plot(Forcast_data(1,:), 'b-o'); legend('真实值','拟合值') title('输出训练集拟合效果') subplot(2,2,2); plot(test_result(1,:), 'r-*') hold on plot(Forcast_data_test(1,:), 'b-o'); legend('真实值','预测值') title('输出测试集预测效果') subplot(2,2,3); stem(train_result(1,:) - Forcast_data(1,:)) title('输出训练集误差') subplot(2,2,4); stem(test_result(1,:) - Forcast_data_test(1,:)) title('输出测试集误差') toc %save('灰狼算法预测2-4')