Abstract: Particle swarm optimization is recognized as a classic algorithm simulating swarm intelligence.A new algorithm based on particle swarm optimization is discussed，which uses geometrical Pareto selection algorithm as archiving algorithm for improving the speed and uses multiple-direction search for seeking extreme points.The experimental results show that this algorithm can obtain many enough solutions and is insensitive to steep fronts，fast and more approximated to the true Pareto front.

摘要： 粒子群优化算法是一种典型的仿真群智能的算法。探讨了利用粒子群算法求解多目标优化问题，为了提高算法速度，采用了几何Pareto选择算法作为文档算法，用多方向搜索的办法寻找极端点。实验表明：该算法得到的解的数量多，速度快并且近似前沿的程度比较高。