摘要
提出一种基于改进PSO的优化滤波算法,构造多指标均衡的适应度函数,把滤波增益作为PSO的粒子进行优化求解,同时将最小方差鲁棒滤波增益和H∞滤波增益以及它们的组合平均值作为PSO的初始粒子,赋予粒子一定的认知能力,大大提高收敛速度。仿真表明新的优化滤波算法滤波精度高,鲁棒性强,实时性好。
A new optimization filtering algorithm is proposed based on modified PSO algorithm.Balanced multi-capability adaptability function is constructed.Optimization filtering gain is answered as particle of PSO algorithm.Least variance robust filtering gain and H∞ filtering gain,and average value of them are initialized as particles.So particles have cognitive capability and convergence speed of algorithm will be quick.Simulation results prove that new algorithm accords with the real time control need,also filtering precision and robustness is better.
出处
《计算机工程与应用》
CSCD
北大核心
2011年第23期133-136,共4页
Computer Engineering and Applications
基金
国家自然科学基金No.60874040~~
关键词
最小方差鲁棒滤波
H∞滤波
改进PSO算法
滤波增益
least variance robust filtering
H∞ filtering
modified Particle Swarm Optimization（PSO）algorithm
filtering gain
