基于双向长短期神经网络的回归分析，基于GRU神经网络的回归分析（代码完整，数据齐全）

%% BILSTM network % Programmed by 红松科技工作室,2027771338 % Updated 6 Apr. 2024. clc clear close all %读取double格式数据 [num1,ax,ay] = xlsread('负荷数据.xlsx',1); num2 = []; for ii = 1:720 a1 = ay{ii+1,13}; % a1(1)=[]; a1(end)=[]; a2 = ay{ii+1,14}; % a2(1)=[]; a2(end)=[]; num2 = [num2;str2double(a1) str2double(a2)]; end num1(:,9) = []; num1(:,1:4) = []; xpan =3; num = []; for ii = 1:720-xpan num = [num; reshape(num1(ii:ii+xpan-1,:),1,18) reshape(num2(ii:ii+xpan-1,:),1,6) num1(ii+xpan,1:3)]; end n = randperm(length(num)); m=693; input_train =num(n(1:m),1:24);%训练数据输出数据 output_train = num(n(1:m),25:27);%训练数据输入数据 input_test = num((m+1:end),1:24);%测试数据输出数据 output_test = num((m+1:end),25:27);%测试数据输入数据 [inputn,inputps]=mapminmax(input_train',-1,1);%训练数据的输入数据的归一化 [outputn,outputps]=mapminmax(output_train',-1,1);%训练数据的输出数据的归一化de inputn_test=mapminmax('apply',input_test',inputps); %% Define Network Architecture % Define the network architecture. numFeatures = 24;%输入层维度 numResponses = 3;%输出维度 % 200 hidden units numHiddenUnits = 70;%第一层维度 % a fully connected layer of size 50 & a dropout layer with dropout probability 0.5 fungru % Specify the training options. % Train for 60 epochs with mini-batches of size 20 using the solver 'adam' maxEpochs = 60;%最大迭代次数 miniBatchSize =2;%最小批量 % the learning rate == 0.01 % set the gradient threshold to 1 % set 'Shuffle' to 'never' options = trainingOptions('adam', ... %解算器 'MaxEpochs',maxEpochs, ... %最大迭代次数 'MiniBatchSize',miniBatchSize, ... %最小批次 'InitialLearnRate',0.01, ... %初始学习率 'GradientThreshold',0.03, ... %梯度阈值 'Shuffle','every-epoch', ... %打乱顺序 'Plots','training-progress',... %画图 'Verbose',1); %不输出训练过程 %% Train the Network net = trainNetwork(inputn,outputn,layers,options);%开始训练 % save maydata.mat %% Test the Network y_pred = predict(net,inputn_test,'MiniBatchSize',miniBatchSize)';%测试仿真输出 y_pred=(mapminmax('reverse',y_pred',outputps))'; % y_pred0 = predict(net,inputn,'MiniBatchSize',1)';%训练拟合值 % y_pred0=(mapminmax('reverse',y_pred0',outputps))'; y_pred=(double(y_pred)); figure%打开一个图像窗口 plot(y_pred(:,1),'k-')%黑色实线，点的形状为* hold on%继续画图 plot(output_test(:,1),'r--')%红色实线，点的形状为o hold off%停止画图 title('测试图')%标题 ylabel('电负荷')%Y轴名称 legend('测试值','实际值')%标签 ylim([0 32]) figure%打开一个图像窗口 plot(y_pred(:,2),'k-')%黑色实线，点的形状为* hold on%继续画图 plot(output_test(:,2),'r--')%红色实线，点的形状为o hold off%停止画图 title('测试图')%标题 ylabel('冷负荷')%Y轴名称 legend('测试值','实际值')%标签 ylim([0 40]) figure%打开一个图像窗口 plot(y_pred(:,3),'k-')%黑色实线，点的形状为* hold on%继续画图 plot(output_test(:,3),'r--')%红色实线，点的形状为o hold off%停止画图 title('测试图')%标题 ylabel('热负荷')%Y轴名称 legend('测试值','实际值')%标签 ylim([0 10]) error1 = y_pred-output_test;%误差 figure plot(error1(:,1),'k-') title('测试误差图') ylabel('电负荷误差') ylim([-15 15]) figure plot(error1(:,2),'k-') title('测试误差图') ylabel('冷负荷误差') ylim([-15 15]) figure plot(error1(:,2),'k-') title('测试误差图') ylabel('热负荷误差') ylim([-15 15]) [MSE,RMSE,MBE,MAE ] =MSE_RMSE_MBE_MAE(output_test(:,1),y_pred(:,1)); result_table = table; result_table.simd = y_pred(:,1); result_table.trued = output_test(:,1); result_table.siml = y_pred(:,2); result_table.truel = output_test(:,2); result_table.simr = y_pred(:,3); result_table.truer = output_test(:,3); writetable(result_table,'./结果.csv')