拟自适应推广卡尔曼滤波算法

Quasi-Adaptive Extended Kalman-Like Filter Algorithm

生化生产过程经常遇到具有隐含观测方程,噪声特性不能确定的问题,本文通过对状态方程和观测方程进行适当处理,把离散化误差、线性化误差、模型噪声和模型本身不准确所造成的误差都归结为虚拟噪声,利用虚拟噪声补偿技术并结合推广卡尔曼滤波算法,提出了拟自适应推广卡尔曼滤波算法,仿真表明该算法对于解决这类问题是比较有效的.

In the biochemical production process, there often occur the problems whose observation equation is implicit and whose noise property of actual process is undetermined. A quasi-adaftive extended Kalmanlike filter algorithm is put forward. It has been show that the algorithm is efficient in solving the estimation problem by simulation.