基于多目标粒子群算法的卫星结构动力学优化

针对卫星结构的多目标动力学优化问题,在其优化过程中建立了一种多目标粒子群优化(MOPSO)算法.该算法采用惯性权重递减策略,对违反约束的粒子给予不同惩罚,并在算法后期引入变异算子,增强种群的多样性,使算法更好地进行全局寻优.结合支持向量机近似模型,将MOPSO方法用于卫星结构动力学优化,并与多目标遗传算法(NSGA-II)的结果进行了对比.数值结果表明,MOPSO可以有效地搜寻优化问题的Pareto前沿,具有良好的分散度和均匀性.

Dynamical Optimization of Satellite Structure Based on Multi-Objective Particle Swarm Optimization Algorithm

Abstract： Aimed at the multi-objective and dynamic optimization problem of satellite structure, a method called MOPSO was proposed. A strategy of decreasing the inertia weight was utilized, the particles that violated the constraints were punished respectively, and the mutation operator was introduced to enhance the diversity of swarms, giving this algorithm a better capability of global optimization. Combined with the support vector machine, MOPSO was applied to solve the multi objective optimization problem of satellite structural dynamics. This approach obtained relatively better results compared with the results obtained by using the NSGA-II algorithm. Numerical results show that MOPSO can effectively and efficiently search and converge to the Pareto optimal front, which is dispersed and uniform.