Abstract: This article aims to explore a new method for calculating the spacing between supports and hangers of power plant steam and water pipelines based on deep learning, in order to improve design efficiency and optimize pipeline safety. Extracting pipeline parameter features through CNN, modeling pipeline dynamic characteristics using LSTM, and optimizing support hanger spacing using genetic algorithm. The experimental results show that compared with traditional methods, this method has significant improvements in stress, displacement, and frequency, enhancing the safety and reliability of pipeline systems.