時序預(yù)測 | MATLAB實現(xiàn)CNN-LSTM卷積長短期記憶神經(jīng)網(wǎng)絡(luò)時間序列預(yù)測（風(fēng)電功率預(yù)測）

預(yù)測效果

基本介紹

1.MATLAB實現(xiàn)CNN-LSTM卷積長短期記憶神經(jīng)網(wǎng)絡(luò)時間序列預(yù)測（風(fēng)電功率預(yù)測）；
2.運行環(huán)境為Matlab2021b；
3.單個變量時間序列預(yù)測；
4.data為數(shù)據(jù)集，單個變量excel數(shù)據(jù)，MainCNN_LSTMTS.m為主程序，運行即可,所有文件放在一個文件夾；
5.命令窗口輸出R2、MSE、RMSE、MAE、MAPE多指標(biāo)評價；

程序設(shè)計

• 完整源碼和數(shù)據(jù)獲取方式：私信博主回復(fù)MATLAB實現(xiàn)CNN-LSTM卷積長短期記憶神經(jīng)網(wǎng)絡(luò)時間序列預(yù)測（風(fēng)電功率預(yù)測）；

%% 預(yù)測
t_sim1 = predict(net, p_train); 
t_sim2 = predict(net, p_test ); %%  數(shù)據(jù)反歸一化
T_sim1 = mapminmax('reverse', t_sim1, ps_output);
T_sim2 = mapminmax('reverse', t_sim2, ps_output);%%  均方根誤差
error1 = sqrt(sum((T_sim1' - T_train).^2) ./ M);
error2 = sqrt(sum((T_sim2' - T_test ).^2) ./ N);%%  相關(guān)指標(biāo)計算%  MAE
mae1 = sum(abs(T_sim1' - T_train)) ./ M ;
mae2 = sum(abs(T_sim2' - T_test )) ./ N ;disp(['訓(xùn)練集數(shù)據(jù)的MAE為：', num2str(mae1)])
disp(['測試集數(shù)據(jù)的MAE為：', num2str(mae2)])%% 平均絕對百分比誤差MAPE
MAPE1 = mean(abs((T_train - T_sim1')./T_train));
MAPE2 = mean(abs((T_test - T_sim2')./T_test));disp(['訓(xùn)練集數(shù)據(jù)的MAPE為：', num2str(MAPE1)])
disp(['測試集數(shù)據(jù)的MAPE為：', num2str(MAPE2)])%  MBE
mbe1 = sum(abs(T_sim1' - T_train)) ./ M ;
mbe2 = sum(abs(T_sim1' - T_train)) ./ N ;disp(['訓(xùn)練集數(shù)據(jù)的MBE為：', num2str(mbe1)])
disp(['測試集數(shù)據(jù)的MBE為：', num2str(mbe2)])%均方誤差 MSE
mse1 = sum((T_sim1' - T_train).^2)./M;
mse2 = sum((T_sim2' - T_test).^2)./N;disp(['訓(xùn)練集數(shù)據(jù)的MSE為：', num2str(mse1)])
disp(['測試集數(shù)據(jù)的MSE為：', num2str(mse2)])

參考資料

[1] https://blog.csdn.net/kjm13182345320/article/details/128577926?spm=1001.2014.3001.5501
[2] https://blog.csdn.net/kjm13182345320/article/details/128573597?spm=1001.2014.3001.5501