非高斯噪声下输出残差或状态估计误差的熵研究

基于非Gaussian噪声线性定常控制系统,通过控制滤波器输出残差或状态估计误差的条件概率密度函数形状来建立有效的滤波设计算法,创建滤波器输出残差或状态估计误差的条件概率密度函数的统一表现形式。利用复合概率密度函数的关系对残差或状态估计误差的条件概率密度函数的近似来实现非高斯残差的高斯化或相应的熵最小化。

Study on Entropy of Output Residuals or State Estimation Error for Non-Gaussian Noise

For non-Gaussian noise of linear time-invariant control system,presents an effective filter design algorithm by controlling the shape of conditional probability density function of the output residuals or state estimation error,where a unified expression has been established for the conditional probability density function of the output residuals or state estimation error of the filter.Using relation of compound probability density functions,these conditional probability density functions of the output residuals or state estimation error are approximated so as to either make these probability density functions close to a Gaussian shape or minimize the entropy of non-Gaussian residuals.