摘要
借助广义逆矩阵的理论,本文分别就一般加权情形、最优加权情形和指数加权情形,给出了无需事先提供待估计量的任何初始统计知识而能获得严格意义下的所谓完整的最小二乘递推(PRLS)算法。应用这种算法,分别得到了某些线性系统的无差和无偏状态估计以及机动目标多模型跟踪与预报(MMTP)算法。
Based on the theory of generalized inverse matrix,separately in the light of the generalweighted,optimal weighted and exponetial weighted conditions,this thesis advanced an al-gorithm called as the Perfect Recursive Least Square(PRLS),which has no need any prima-ry statistical knowledge to furnish the estimative quantity beforehand and can be obtainedunder the strict sense.Furthermore,making use of the PRLS algorithm,we formed the er-rorless and deabeat state estimation of some linear systems and acquired the multimodel trac-ing and predicting(MMTP)algorithm for the maneuverable targets.
出处
《南昌高专学报》
1994年第3期7-19,共13页
Journal of Nanchang Junior College