Advances in Research

With the increasing construction of large-scale hydropower stations and pumped storage power stations, reinforced concrete branch pipes, as a key part of the water conveyance structure, their stability is crucial to the safety of the power stations. Traditional optimization design of branch pipes mostly relies on engineers' experience or optimizations for specific parts, and the application of intelligent algorithms is relatively limited. However, the development of computer technology has made it possible to optimize structures by combining intelligent algorithms. This paper aims to provide a more effective optimization scheme for reinforced concrete branch pipes. The particle swarm genetic hybrid algorithm is adopted, and an optimization platform is built by combining ANSYS and Matlab software. Parametric modeling of branch pipes is carried out through APDL language, and a fitness function is constructed based on stress conditions to screen the population. Verified by CEC standard test functions, the hybrid algorithm exhibits excellent performance in optimization accuracy, ability to escape local optima, and global optimization capability. In practical engineering applications, the volume of the optimized branch pipe is reduced, the concrete consumption is decreased, and the stress and displacement results meet the design specifications. The research shows that the particle swarm genetic hybrid algorithm effectively solves the problems of premature convergence of genetic algorithms and the tendency of particle swarm algorithms to fall into local optima. It provides new ideas and methods for the design optimization of reinforced concrete branch pipes and has important guiding significance for related engineering practices.