In recent years, under the guidance of new energy development policies, the new energy power generation industry in Jiangsu Province, especially wind and solar power generation, has shown a rapid growth trend. In the new energy power system, the power generation methods of wind power and photovoltaic power are greatly influenced by regional and meteorological conditions, but the current accuracy of meteorological forecasting in terms of frequency, time, resolution, etc. cannot meet the requirements. This article proposes a method for selecting similar days based on meteorological influencing factors, and studies the technology for improving the effectiveness of the prediction model based on the best similar day. Finally, the new energy generation prediction model is used to obtain the best similar day as the basis, and validated and analyzed with 2022 data. The results show that the technology proposed in this article has a high degree of accuracy and feasibility.