|
| Author | Year | Problem | Data set | Techniques | Results | Limit |
|
| [30] | 2020 | Short-term forecasting of electrical demand | 3 US networks | FCRBM, GWDO | Precision time = 58 s, 102 s and 63 s | Complex hybridization |
|
| [31] | 2021 | Short-term electrical demand forecasting | Bandırma Onyedi | LSTM | RMSE = 736.706; MAE = 352.176 and MAPE = 8.145 | Require large volume of data |
| Eylu university buildings |
|
| [33] | 2021 | Short-term electrical demand forecasting | Distribution transformers in US | DNN, LSTM | RMSE = 2.6874 kWh; MAPE = 15.9380%; training time = 10.76 s; execution time = 0.1070 s | Complex model |
|
| [41] | 2022 | Short-term electrical demand forecasting | Cameroon households | ANFIS, grey, PSO | RMSE = 0.20158 and MAPE = 0.6291% | Complex hybridization |
|
| [54] | 2020 | Residential building electrical consumption forecasting | Climate data in US | DRNN-GRU | MAE = 89.36; MSE = 45.28 | Model is only validated in small dataset |
|
| [55] | 2021 | Short-term residential load forecasting | 553 consumers in Ohta, Japan | Deep reservoir architecture | MSE = 2.15; RMSE = 1.466; MAE = 5.42; MAPE = 0.64%; R = 0.8896 | Low convergence |
|
| [56] | 2021 | Electrical load forecasting | Households in China | GC-LSTM | MSE = 3.66; RMSE = 1.913; MAE = 7.85; MAPE = 2.36%; R = 0.8795 | Low performance |
|
| [57] | 2021 | Short-term electrical load forecasting | 02 datasets in England | ADDPG-AEFRIM | MSE = 2.04; RMSE = 1.428; MAE = 2.58; MAPE = 0.94%; R = 0.8996 | Performance coefficients are not effectively evaluated |
|
| [58] | 2021 | Electrical load forecasting | Historical and climate data in China | TgDLF, EnLSTM | MSE = 1.58; RMSE = 1.241; MAE = 1.33; MAPE = 0.89%; R = 0.9654 | Model complexity |
|
| [59] | 2022 | Short-term electrical demand forecasting | 479 buildings in Japan | CNN, DNN, GRU-FCL, LSTM-FCL, Bi-GRU-FCL | MSE = 0.06; RMSE = 0.244; MAE = 0.48; MAPE = 0.75%; R = 0.9788 | Small aggregation |
|
| [60] | 2022 | Forecasting of the electrical load in microgrid | 69 consumers in Australia | k-means, QRLSTM, KDE | MSE = 0.012; RMSE = 0.11; MAE = 0.15; MAPE = 0.47%; R = 0.9979 | Complexity of model |
|
| [61] | 2022 | Electrical load forecasting | 12000 households in korea | CNN-LSTM | MAE = 0.04; MAPE = 0.38%; R = 0.9987 | Gradient problem |
|
| [62] | 2022 | Electrical demand forecasting | Microgrid in China | TCN-DNN | MSE = 0.0035; RMSE = 0.059; R = 0.9995 | Model complexity |
|
