Deep Learning-Based Modeling of Demand Response with Bilateral Uncertainty of Source and Load

With the rapid development of smart grid and renewable energy, source-load bilaterally uncertainty puts higher requirements on the stability and reliability of power systems. In this paper, a demand response model with source-load bilateral uncertainty is constructed based on a deep learning method, aiming to improve the responsiveness of the power system in an uncertain environment. First, a training set containing multi-dimensional data such as historical load, weather and market price is established through data collection and preprocessing. Second, a deep learning model was designed to model the source load uncertainty using its powerful nonlinear fitting capability. Through model training and optimization, a high-precision demand response strategy is obtained. The experimental results show that the proposed model has significant advantages in reducing power system operation cost and improving demand response flexibility. Finally, this paper also discusses the practical application scenarios and promotion value of the model, which provides an important reference for the future power market reform.