具有极点约束的无偏H^∞滤波方法及应用

A Method of Unbiased H∞ Filtering with Pole Constraints and It's Application

利用具有极点约束的无偏H∞滤波方法,设计一个稳定的线性无偏滤波器,使由被估计对象和滤波器组成的误差系统满足‖Tew‖∞<γ.在满足误差系统对外部扰动的抑制性能要求的同时,考虑系统的内部动态特性,将极点配置在一个适当的区域内.以连续搅拌反应釜(CSTR)为例,在线性矩阵不等式(LMI)技术下,通过对系统中的可测量状态信息(温度)利用本文方法估计不可测量状态(浓度).仿真结果表明,该方法有较好的估计效果,能逼真地跟随系统中的状态变化,估计误差较小.

The method of unbiased H~∞ filtering with pole constraints is given. A linear unbiased filter is obtained by the method such that the filtering error system satisfies ‖%T-{ew}%‖<%γ%. With the filtering error system satisfying the robust performance within %γ%, the error system dynamics performance is improved by constraining the poles in a suitable region. Taking the CSTR as an example, the unmeasurable vectors (densities) are estimated by the measurable state vectors (temperature) with the Linear Matrix Inequality technique. The result of simulation shows that the method has a better effect. It can follow the tracks of the state changing and the error is smaller.