In the case of renewable energy generation such as photovoltaic power generation, the problem of supply-demand balance capacity should be evaluated and solved in the future power system. Improving existing balancing measures and new technologies such as power supply and demand synergy and energy storage will solve this problem. In this case, the remote power system supply and demand analysis should have the ability to evaluate the equilibrium countermeasures. Compared with time series analysis, this method has some limitations. But there is a big advantage in the maintenance of various electrical equipment can be supply and demand assessment. This paper uses machine learning and deep learning models to provide a new solution for predicting consumer demand and distributed renewable energy. The proposed power system supply and demand analysis model ESPRIT provides a new solution for power supply and demand balance.