并行混沌与和声搜索的多目标混合优化算法

针对传统的混沌优化算法对初始值敏感、搜索精度低和收敛速度慢,以及和声搜索收敛不稳定、处理多目标优化问题时适应性差等不足,研究了一种多目标并行混沌与和声搜索混合优化算法(MOCOHSA).MOCOHSA利用并行混沌优化的全局搜索能力与和声搜索算法的局部搜索能力,并在和声搜索中引入自适应操作,在解决多目标优化问题时表现出良好的搜索速度和收敛性能.对8个多目标优化测试函数的优化计算中,该算法表现出比其它多目标优化算法更好的性能.算法最后用于解决卫星热管设计问题.

A Multi-objective Hybrid Optimization Algorithm Based on Parallel Chaos and Harmony Search

Conventional chaotic optimization is a sensitive initial values algorithm with low search accuracy and slowness of convergence. Besides, convergence of harmony search algorithm is not stable and its adaptability for the multi-objective optimization problems is poor. Therefore, a multi-objective hybrid algorithm based on parallel chaos optimization and harmony search algorithm (MOCOHSA) is proposed in this paper. In the MOCOHSA, the global search ability of parallel chaos optimization and local search ability of harmony research are utilized, and adaptive operation is introduced in the harmony search algorithm. In this way, the MOCOHSA has good search speed and convergence performance for multi-objective optimization problems. The proposed algorithm performs better than other multi-objective optimization algorithms for the optimization of 8 multi-objective optimization test functions. Finally, the algorithm is used to solve the satellite heat pipe design problem.