利用粒子滤波原理求解函数优化问题

Application of Particle Filter Algorithm in Function Optimization Problems

针对多维函数优化容易陷入局部极小值问题,提出一种基于粒子滤波原理的优化算法.首先简要地介绍粒子滤波算法的基本原理;再针对函数优化问题,阐述利用粒子滤波实现优化算法的基本思想,并介绍了其具体的实现步骤,同时为了丰富粒子的多样性,引入了遗传算法的交叉和变异操作;最后为了验证新算法的有效性,采用30维的Benchmark函数进行仿真实验.仿真实验结果表明:基于粒子滤波的优化算法在解决多维函数优化问题方面较其他优化算法具有更强的全局搜索能力和求解精度,这也为优化算法的研究提供一种新的途径和手段.

To overcome the shortcomings of slowly converging and easily immerging in partial minimum in the multi-dimensional function optimization problems,a new optimization algorithm based on the particle filter is brought forward in this paper.Firstly,the basic principle of the particle filter algorithm is introduced.Secondly,a new optimization method based on particle filter is enunciated and its design idea and steps are given.Besides,the crossover and mutation operator of genetic algorithm is combined in the new optimization method to enhance the variety of particles.Finally,the simulation results of 30-dimensional Benchmark functions optimization are performed to prove the validity of the new algorithm.The simulation results have shown that the new optimization method based on particle filter has stronger searching ability and higher precision than other optimization methods in the multi-dimensional function optimization,which provides a new method and means for the optimization research.