摘要： 针对标准粒子群算法易陷入局部最优的问题，提出选择粒子视野范围内具有最优适应度值的粒子作为该粒子本次迭代所需的全局极值，测试结果证明改进算法的全局收敛能力明显提高。将该算法用于群体动画中。仿真实验表明个体具有良好的人工智能性，能够真实模拟群体行为。

关键词: 粒子群优化算法, 群体动画, 视野

Abstract: According to the problem that the standard Particle Swarm Optimization(PSO) easily traps in a local optimum, this paper presents a method that the current particle chooses the one with the best fitness in its visual as the global best particle. Experiment shows the advantaged algorithm has better global search ability. When it is applied to the group animation, the particle has good artificial intelligence performance, and can simulate group behavior in an actual way.

Key words: Particle Swarm Optimization(PSO), group animation, visual field

中图分类号: 

• TP311