Research On Short Term Power Load Forecasting Based On Deep Learning

Load forecasting is one of the crucial methods for ensuring the safe and reliable operation of power systems and the scientific dispatching of electricity,providing a reliable guarantee for the smooth development of smart grids.Short-term load forecasting,in particular,has strong randomness and can directly regulate electricity prices,making it more important and requiring higher prediction accuracy.With the continuous development of deep learning in recent years,this technology has been widely applied to load forecasting,further improving prediction performance.This paper presents a detailed introduction to the research background and significance of this topic,and based on the current research status at home and abroad,discusses the commonly used methods and feasible development directions for short-term load forecasting.The focus is on the following research on short-term load forecasting based on deep learning algorithms in power systems and integrated energy systems:A predictive method based on cluster identifica-tion and convolutional neural network-bi-directional long short-term memory-temporal pattern attention(CNN-BiLSTM-TPA)is proposed to solve the problem of excessive input features and strong load periodicity in regional short-term power load forecasting.Firstly,load nodes within the region are identified as clusters based on second-order clustering algorithm with consideration of power consumption mode and weather as the influence factors.And then,the representative features are selected from each cluster as inputs of the deep learning model,which can not only reduce the input feature dimension and decrease the computational complexity,but also comprehensively consider the overall characteristics of the prediction region to improve the predic-tion accuracy.Thereafter,aiming at the strong load periodicity of regional power load,the CNN-BiLSTM-TPA model is trained and applied for prediction,extracting the bi-directional information from the input data to generate the hidden state matrix and weighing the important features of the hidden state matrix,while capturing the bi-directional time series infor-mation on multiple time steps for prediction.Finally,the effectiveness of proposed method is verified using the actual load data in California,USA.As integrated energy systems(IES)are increasingly applied in various scenarios,power loads often do not appear in isolation but occur simultaneously with other loads,mutually influencing each other.Therefore,focusing solely on power loads while ignoring other loads in IES cannot accurately predict power loads.This paper proposes a short-term load forecasting model for integrated energy systems based on multi-task learning and CNN-BiLSTM-TPA.Faced with highly volatile and temporally strong integrated energy loads,multi-task learning is first employed,using CNN-BiLSTM as a shared layer to mine the coupling relationship among multiple loads and to extract bidirectional features from load data,improving prediction accuracy.Then,TPA is used to weight feature row vectors,focusing on important features for different loads,allowing multiple loads to learn from each other while retaining their individual characteristics.Finally,a comparative experiment is conducted on the IES dataset of Arizona State University,demonstrating the superior prediction performance of the proposed model.