Wind Speed Forecasting Using Attention-Based Causal Convolutional Network and Wind Energy Conversion

As one of the effective renewable energy sources, wind has received attention because it is sustainable energy. Accurate speed forecasting can pave way to goal development. However, current methods ignore temporal characteristics speed, which leads inaccurate results. In this paper, we propose a novel SSA-CCN-ATT model forecast speed. Specifically, singular spectrum analysis (SSA) first applied decompose original into several sub-signals. Secondly, build new deep learning CNN-ATT that combines causal convolutional network (CNN) and mechanism (ATT). The used extract information in time series. After that, employed focus on important information. Finally, fully connected neural layer get Three experiments four datasets show proposed performs better than other comparative models. Compared with different models, maximum improvement percentages MAPE reaches up 26.279%, minimum 5.7210%. Moreover, conversion curve was established by simulating historical data.