A Virtual Power Plant Load Prediction Method Based on Improved K-Means Clustering and Bi LSTM

There are many types of loads within a virtual power plant, and the load characteristics are often very different, so it is particularly important to accurately predict the loads of a virtual power plant including base loads, controllable loads, and so on. In this paper, a virtual power plant load forecasting method based on improved K-means clustering and bidirectional long and short-term memory neural network (Bi LSTM) is proposed. Firstly, the related complex load data are dimensionality reduced by principal component analysis (PCA); secondly, the data are clustered by improved K-means clustering. Finally, the load prediction model is constructed and trained according to the method of this paper to achieve the accurate output of virtual power plant load prediction results. Through comparison experiments with the prediction models such as Long Short-Term Memory Neural Network (LSTM), Bidirectional Long Short-Term Memory Neural Network (Bi LSTM), K-means Bi LSTM, etc., the experimental results show that the prediction model proposed in this paper is more effective, and it can play a better role in the virtual power plant load forecasting.