Signals and Telecommunication Journal Fei Li 14 1 2025 https://doi.org/10.6025/stj/2025/14/1/1-8 https://www.dline.info/stj/fulltext/v14n1/stjv14n1_1.pdf By combining various renewable energy sources, new hybrid power generation technologies can significantly improve energy utilization efficiency and reduce environmental impact. However, due to the uncertainty and intermittent nature of this technology, signal optimization and control become particularly important. In this study, we use advanced techniques to enhance the performance of hybrid power generation systems. Our work involves utilizing deep learning to analyze and predict various factors, thereby finding the optimal operating mode. The conclusions of this work will have a positive impact on future energy management and dispatching. Additionally, the methods proposed in this article can provide valuable references for research in other related fields.