【Transformer回归预测】基于Transformer-BiGRU实现负荷数据回归预测附matlab代码.rar

%% 清空环境变量 warning off % 关闭报警信息 close all % 关闭开启的图窗 clear % 清空变量 clc % 清空命令行 %% 导入数据 res = xlsread('data.xlsx'); %% 数据分析 num_size = 0.8; % 训练集占数据集比例 outdim = 1; % 最后一列为输出 num_samples = size(res, 1); % 样本个数 res = res(randperm(num_samples), :); % 打乱数据集（不希望打乱时，注释该行） num_train_s = round(num_size * num_samples); % 训练集样本个数 f_ = size(res, 2) - outdim; % 输入特征维度 %% 划分训练集和测试集 P_train = res(1: num_train_s, 1: f_)'; T_train = res(1: num_train_s, f_ + 1: end)'; M = size(P_train, 2); P_test = res(num_train_s + 1: end, 1: f_)'; T_test = res(num_train_s + 1: end, f_ + 1: end)'; N = size(P_test, 2); %% 数据归一化 [P_train, ps_input] = mapminmax(P_train, 0, 1); P_test = mapminmax('apply', P_test, ps_input); [t_train, ps_output] = mapminmax(T_train, 0, 1); t_test = mapminmax('apply', T_test, ps_output); %% 数据平铺 % 将数据平铺成1维数据只是一种处理方式 % 也可以平铺成2维数据，以及3维数据，需要修改对应模型结构 % 但是应该始终和输入层数据结构保持一致 P_train = double(reshape(P_train, f_, 1, 1, M)); P_test = double(reshape(P_test , f_, 1, 1, N)); t_train = t_train'; t_test = t_test' ; %% 数据格式转换 for i = 1 : M p_train{i, 1} = P_train(:, :, 1, i); end for i = 1 : N p_test{i, 1} = P_test( :, :, 1, i); end %% 创建模型 %网络搭建 numChannels = f_; maxPosition = 256; numHeads = 10; numKeyChannels = numHeads*32; layers = [ sequenceInputLayer(numChannels,Name="input") positionEmbeddingLayer(numChannels,maxPosition,Name="pos-emb"); additionLayer(2, Name="add") selfAttentionLayer(numHeads,numKeyChannels,'AttentionMask','causal') selfAttentionLayer(numHeads,numKeyChannels) indexing1dLayer("last") % GRU特征学习 gruLayer(128,'Name','GRU1','RecurrentWeightsInitializer','He','InputWeightsInitializer','He') gruLayer(6, 'OutputMode', 'last', "Name", "gru2" ); % 反向GRU fullyConnectedLayer(1) regressionLayer]; % layers = layerGraph(layers); layers= connectLayers( layers,"input","add/in2"); %% 参数设置 options = trainingOptions('adam', ... % Adam 梯度下降算法 'MaxEpochs', 200, ... % 最大训练次数 'InitialLearnRate', 0.001, ... % 初始学习率 'LearnRateSchedule', 'piecewise', ... % 学习率下降 'LearnRateDropFactor', 0.1, ... % 学习率下降因子 'LearnRateDropPeriod', 200, ... % 经过训练后 学习率 'Shuffle', 'every-epoch', ... % 每次训练打乱数据集 'Plots', 'training-progress', ... % 画出曲线 'Verbose', false); %% 训练模型 net = trainNetwork(p_train, t_train, layers, options); %% 仿真预测 t_sim1 = predict(net, p_train); t_sim2 = predict(net, p_test ); %% 数据反归一化 T_sim1 = mapminmax('reverse', t_sim1', ps_output); T_sim2 = mapminmax('reverse', t_sim2', ps_output); %% 查看网络结构 analyzeNetwork(net) %% 均方根误差 error1 = sqrt(sum((T_sim1 - T_train).^2) ./ M); error2 = sqrt(sum((T_sim2 - T_test ).^2) ./ N); %% 相关指标计算 % R2 R1 = 1 - norm(T_train - T_sim1)^2 / norm(T_train - mean(T_train))^2; R2 = 1 - norm(T_test - T_sim2)^2 / norm(T_test - mean(T_test ))^2; disp(['训练集数据的R2为：', num2str(R1)]) disp(['测试集数据的R2为：', num2str(R2)]) % MAE mae1 = sum(abs(T_sim1 - T_train)) ./ M ; mae2 = sum(abs(T_sim2 - T_test )) ./ N ; disp(['训练集数据的MAE为：', num2str(mae1)]) disp(['测试集数据的MAE为：', num2str(mae2)]) %% 平均绝对百分比误差MAPE MAPE1 = mean(abs((T_train - T_sim1)./T_train)); MAPE2 = mean(abs((T_test - T_sim2)./T_test)); disp(['训练集数据的MAPE为：', num2str(MAPE1)]) disp(['测试集数据的MAPE为：', num2str(MAPE2)]) % MBE mbe1 = sum(T_sim1 - T_train) ./ M ; mbe2 = sum(T_sim2 - T_test ) ./ N ; disp(['训练集数据的MBE为：', num2str(mbe1)]) disp(['测试集数据的MBE为：', num2str(mbe2)]) %均方误差 MSE mse1 = sum((T_sim1 - T_train).^2)./M; mse2 = sum((T_sim2 - T_test).^2)./N; disp(['训练集数据的MSE为：', num2str(mse1)]) disp(['测试集数据的MSE为：', num2str(mse2)]) [R1_Transformer,rmse1_Transformer,biaozhuncha1_Transformer,mae1_Transformer,mape1_Transformer]=calc_error(T_train,T_sim1);%训练集 [R2_Transformer,rmse2_Transformer,biaozhuncha2_Transformer,mae2_Transformer,mape2_Transformer]=calc_error(T_test,T_sim2);%测试集 %% 绘图 figure plot(T_sim1,'-p','Color',[255 0 0]./255,'linewidth',1,'Markersize',2,'MarkerFaceColor','b') hold on plot(T_train,'-^','Color',[50 50 50]./255,'linewidth',1,'Markersize',2,'MarkerFaceColor','r') legend('预测值','真实值') xlabel('预测样本') ylabel('预测结果') string = {'训练集预测结果对比'; ['RMSE=' num2str(error1)]}; title(string) xlim([1, M]) grid ax=gca;hold on % ------------------------------------------------------------------------- % 坐标区域修饰 ax.XLabel.FontWeight='bold'; ax.YLabel.FontWeight='bold'; % 设置轴线粗细和刻度朝外 ax.LineWidth=1; ax.TickDir='out'; %% 训练集误差图 figure bar((T_sim1 - T_train)./T_train) legend('模型训练集相对误差','Location','NorthEast','FontName','楷体') title('模型训练集相对误差','fontsize',12,'FontName','楷体') ylabel('误差','fontsize',12,'FontName','楷体') xlabel('样本','fontsize',12,'FontName','楷体') xlim([1 M]); figure plot(T_sim2,'-p','Color',[255 0 0]./255,'linewidth',1,'Markersize',2,'MarkerFaceColor','b') hold on plot(T_test,'-^','Color',[50 50 50]./255,'linewidth',1,'Markersize',2,'MarkerFaceColor','r') legend('预测值','真实值') xlabel('预测样本') ylabel('预测结果') string = {'测试集预测结果对比'; ['RMSE=' num2str(error2)]}; title(string) xlim([1, N]) grid ax=gca;hold on % ------------------------------------------------------------------------- % 坐标区域修饰 ax.XLabel.FontWeight='bold'; ax.YLabel.FontWeight='bold'; % 设置轴线粗细和刻度朝外 ax.LineWidth=1; ax.TickDir='out'; %% 测试集误差图 figure bar((T_sim2 - T_test )./T_test) legend('模型测试集相对误差','Location','NorthEast','FontName','楷体') title('模型测试集相对误差','fontsize',12,'FontName','楷体') ylabel('误差','fontsize',12,'FontName','楷体') xlabel('样本','fontsize',12,'FontName','楷体') xlim([1 N]); %% 绘制线性拟合图 %% 训练集拟合效果图 figure plot(T_train,T_sim1,'s','Markersize',2); xlabel('真实值') ylabel('预测值') string = {'训练集效果图';['R^2_c=' num2str(R1) ' RMSEC=' num2str(error1) ]}; title(string) hold on ;h=lsline; set(h,'LineWidth',2,'LineStyle','-','Color',[1 1 0]) %% 预测集拟合效果图 figure plot(T_test,T_sim2,'s','Markersize',2); xlabel('真实值') ylabel('预测值') string1 = {'测试集效果图';['R^2_p=' num2str(R2) ' RMSEP=' num2str(error2) ]}; title(string1) hold on ;h=lsline(); set(h,'LineWidth',2,'LineStyle','-','Color',[1 1 0])