Short-Term Load Forecasting Using Encoder-Decoder WaveNet: Application to the French Grid

Abstract

The prediction of time series data applied to the energy sector (prediction of renewable energy production, forecasting prosumers’ consumption/generation, forecast of country-level consumption, etc.) has numerous useful applications. Nevertheless, the complexity and non-linear behaviour associated with such kind of energy systems hinder the development of accurate algorithms. In such a context, this paper investigates the use of a state-of-art deep learning architecture in order to perform precise load demand forecasting 24-h-ahead in the whole country of France using RTE data. To this end, the authors propose an encoder-decoder architecture inspired by WaveNet, a deep generative model initially designed by Google DeepMind for raw audio waveforms. WaveNet uses dilated causal convolutions and skip-connection to utilise long-term information. This kind of novel ML architecture presents different advantages regarding other statistical algorithms. On the one hand, the proposed deep learning model’s training process can be parallelized in GPUs, which is an advantage in terms of training times compared to recurrent networks. On the other hand, the model prevents degradations problems (explosions and vanishing gradients) due to the residual connections. In addition, this model can learn from an input sequence to produce a forecast sequence in a one-shot manner. For comparison purposes, a comparative analysis between the most performing state-of-art deep learning models and traditional statistical approaches is presented: Autoregressive-Integrated Moving Average (ARIMA), Long-Short-Term-Memory, Gated-Recurrent-Unit (GRU), Multi-Layer Perceptron (MLP), causal 1D-Convolutional Neural Networks (1D-CNN) and ConvLSTM (Encoder-Decoder). The values of the evaluation indicators reveal that WaveNet exhibits superior performance in both forecasting accuracy and robustness.

Mô hình kết hợp HHO-GCN-LSTM ứng dụng trong dự báo đồ thị phụ tải cho lưới điện nhỏ

“…xét cách tiếp cận khác trong nghiên cứu [4], WaveNet sử dụng phức hợp nhân quả giãn nở và kết nối bỏ qua để sử dụng thông tin dài hạn. Loại kiến trúc ML mới lạ này thể hiện những ưu điểm khác nhau liên quan đến các thuật toán thống kê khác.…”

“…Để đánh giá hiệu suất mô hình, chúng tôi so sánh mô hình này với cać mô hình dựa trên học sâu trước đó hoạt động rất tốt trong trường hợp STLF. Những mô hình đó được tham khaỏ từ [16] mô hình 1 sử dụng mạng ANN thuần tuý, (Mô hình 2), sử dụng LSTM xếp chồng và [10] (Mô hình 3) kết hợp lớp CNN và LSTM; mô hình 4 sử dụng mạng Wavenet [4]. Cấu hình của mỗi mô hình so sánh giống hệt với các bài báo đã xuất bản.…”

Deep learning for time series forecasting: The electric load case

“…More recently, temporal convolutional networks have also started to be used and show very promising results indeed. In [39], a modified encoder-decoder architecture based on WaveNet is proposed. The authors performed careful testing of their model against the state-of-the-art power consumption data coming from the French grid.…”

Short-Term Load Forecasting Using Convolutional Neural Networks in COVID-19 Context: The Romanian Case Study

“…The evaluation of the model's accuracy is performed by using several metrics widely applied in the literature [56][57][58]. Additionally, to the MSE index, which is previously defined, the Mean Absolute Error (MAE), the Mean Absolute Percentage Error (MAPE), and the Root Mean Square Error (RMSE) were used as well.…”