模型不确定和噪声互相关下的分布式容积平滑变结构融合算法

Distributed cubature smooth variable structure fusion algorithm with model uncertainties and cross-correlation noise

中文关键词  平滑变结构滤波  模型参数不确定  互相关噪声  分布式融合  系统辨识与建模

英文关键词  smooth variable structure filter  model parameters uncertainties  cross-correlation noise  distributed fusion  system identification and modelling

中文摘要

      针对模型参数不确定和噪声互相关下的多传感器非线性系统状态估计问题,本文首先提出一种新的模型 不确定和噪声互相关下的容积平滑变结构滤波算法.其次,针对多传感器系统,推导了噪声互相关下,任意局部估 计器之间互协方差的通用计算框架,并利用容积规则计算互协方差中的非线性积分.然后,针对模型参数不确定和 噪声互相关下的多传感器非线性系统,提出了基于矩阵加权的分布式容积平滑变结构融合算法.最后,仿真实验结 果表明,本文所提算法能够有效克服模型不确定和互相关噪声的影响,具有更高的状态估计精度.

英文摘要

      The state estimation for multi-sensor nonlinear systems with model parameters uncertainty and correlated noise is considered. Firstly, a new cubature smooth variable structure filter with model uncertainty and correlation noise is proposed. Secondly, a general cross-covariance framework of any local estimators is derived for multi-sensor systems with correlated noise, and the nonlinear integral is calculated by Cubature rule. Then, based on matrix weighting, the distributed Cubature smooth variable structure fusion algorithm is proposed for multi-sensor nonlinear systems with model parameters uncertainty and correlation noise. Finally, the simulation results show that the proposed algorithms can effectively overcome the interference of model parameters uncertainties and correlation noise, and also have higher estimation accuracy.