基于改进PSO的多传感器数据自适应加权融合算法  

Adaptive weighted fusion algorithm of multi-sensor data based on improved particle swarm optimization

作　　者：杨晓燕[1] 

机构地区：[1]闽江学院计算机科学系,福建福州350108

出　　处：《闽江学院学报》2011年第5期67-71,共5页Journal of Minjiang University

基　　金：福建省自然科学基金资助项目(2009J01284)

摘　　要：加权融合算法是多传感器数据融合中的常用方法,但加权因子的确定非常困难并直接影响算法的性能.文章提出利用改进的粒子群优化算法对各个传感器的加权因子进行自适应优化,引入种群进化度、聚合度来反映种群的多样性,当种群多样性低于阈值时执行变异操作,并交替使用基于聚合度、进化度的自适应惯性权重函数,从而避免算法陷入局部最优解.通过UCI数据集测例表明本文算法是一种较有效的多传感器数据融合方法,相对其它算法具有较高的融合精度.The weighted fusion algorithm is a common method of multi-sensor data fusion,but it is hard to determine the suitable weighting factors,which directly influence the performance of the algorithm.Therefore,an improved particle swarm optimization（PSO） algorithm is proposed to self-adaptively optimize the weighting factor of each sensor.Evolution degree and aggregation degree are introduced to reflect the population ＇ s diversity.When the diversity declines to some degree,a mutation operation is used and the adaptive inertia weighting function based on the evolution degree or the aggregation degree is used alternately,which can effectively avoid the particle swarm optimization algorithm falling into local optimization.The experiments on UCI datasets show that the proposed algorithm is a kind of effective data fusion method and it has higher fusion precision than other algorithms.

关 键 词：多传感器 加权融合 粒子群优化 加权因子 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]