改进CPSO-SVM在人脸识别中的应用

为使粒子群优化算法初始粒子均匀分布在解空间，增强全局的搜索能力，通过对混沌运动的遍历性和粒子群优化算法中惯性权重的分析，提出了一种改进型混沌粒子群算法。该算法采用Circle映射，产生了分布均匀的混沌变量轨道点，并结合动态调整惯性权重的思想来避免粒子群算法陷入局部最优。同时，给出了应用混沌粒子群算法训练SVM的方法，并将其应用于人脸识别。仿真实验结果表明，改进CPSO-SVM方法比基本粒子群方法能获得更好的识别性能。

Application of improved CPSO-SVM approach in face recognition

To make the particles distribute in the problem search space evenly,and to improve the overall searching ability of the algorithm,an improved algorithm for Chaos Particle Swarm Optimization（CPSO） is proposed based on the analysis of the ergodicity of chaos and inertia weight of the PSO.The improved algorithm uses Circle map to produce a uniform distribution of the chaotic variable track points,and dynamic adjustment of inertia weight is implemented to avoid PSO getting into local optimum.Then a face recognition method using this improved algorithm to train Support Vector Machine（SVM） is presented.The experimental results show that the presented SVM method optimized by CPSO can achieve higher recognition performance.