算法在多目标优化问题中的仿真应用

综合了PSO算法应用于求解多目标问题的一些思想,引入罚函数对约束条件进行处理,并利用均值的概念来计算选取PSO算法更新方程中的全局最优值,从而使算法在最优解的引导下,跳出局部最小值,更快地向Pareto最优解前沿收敛。压缩弹簧多目标优化设计实例证明了所提出算法的有效性。

Applications of Particle Swarm Optimization in Multi-objective Problems

In this paper,some ideas about PSO research and applications in multi-objective problem were synthesized.Based on the concept of average,the global best fitness function value in updated formula of PSO was calculated,which benefits to the algorithm escaping from the local minimums and converging quickly to the front of Pareto optimal set.Meanwhile the penalty function method was used as constraint-handling technique.The algorithm was proved effective by the simulation and experimental results for the given multi-objective optimal problem of a compression spring.