New energy generation has characteristics such as volatility and intermittency, making it difficult to meet the needs of power grid operation in the short term. At the same time, due to the randomness of its power generation output, there are system safety hazards in the operation of the power grid. At present, most regions mainly regulate the output of new energy generation through the dispatch of the large power grid. With the large-scale integration of renewable energy, the issue of how to ensure the consumption of new energy in the power grid is becoming increasingly prominent. The article took the power system of a certain region as an example, established output models for wind power generation (WPG) and photovoltaic power generation (PPG), and used particle swarm optimization (PSO) algorithm to simulate and optimize the output of WPG and PPG, thereby obtaining the output prediction results of WPG and PPG. The former had the highest prediction precision of 87.8%, while the latter had a prediction precision of 93.3%. Finally, by analyzing and comparing the predicted results, differences in the consumption of new energy between different seasons and regions can be obtained.