噪声方差不确定约束系统的滚动时域估计  

Moving horizon estimation approach to constrained systems with uncertain noise covariance

作　　者：赵海艳[1] 陈虹[1] 

机构地区：[1]吉林大学控制科学与工程系,长春130025

出　　处：《控制与决策》2008年第2期217-220,共4页Control and Decision

基　　金：国家杰出青年科学基金项目(60374027);教育部新世纪优秀人才支持计划项目(2004)

摘　　要：针对噪声方差不确定的约束系统,讨论了一种鲁棒滚动时域估计(MHE)方法.首先,根据噪声方差不确定模型,找到满足所有不确定性的最小方差上界,在线性矩阵不等式(LMI)框架下求解优化问题,得到近似到达代价的表达形式;然后再融合预测控制的滚动优化原理,把系统的硬约束直接表述在优化问题中,在线优化性能指标,估计出当前时刻系统的状态.仿真时与鲁棒卡尔曼滤波方法进行比较,结果表明了该方法的有效性.For the constraint system of covariance matrix being uncertain, a robust moving horizon estimation （MHE） strategy is discussed. The upper bound of minimum error covariance satisfying all the uncertainties is found and the optimal problem is solved in the framework of linear matrix inequality（LMI）. Based on the moving horizon strategy, the state is estimated by minimizing performance object while satisfying the constraints for every possible noise covariance within the given bounds. Simulation and comparison results show with the robust Kalman filter are given. The results the effectiveness of the method.

关 键 词：硬约束 滚动时域估计 不确定噪声方差 线性矩阵不等式 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]