欠观测条件下的增量Kalman滤波方法

INCREMENTAL KALMAN FILTER METHOD UNDER POOR OBSERVATION CONDITION

提出欠观测条件下增量Kalman滤波的概念和定义,建立增量Kalman滤波模型及其分析方法,给出主要的计算步骤。经典的Kalman滤波方法要求量测方程有较高精度,否则在递推过程中会产生较大误差。但是量测方程通常受环境因素影响较大,而且在许多实际情况(如深空探测)中不可能对所有的使用环境逐一校准量测方程。如果某量测方程没有在某一环境条件下进行验证或校准,那么在这一条件(称为欠观测条件)下使用该量测方程往往会产生未知的系统误差,从而导致较大的Kalman滤波误差。提出的欠观测条件下增量Kalman滤波方法能够成功消除这种未知的系统误差,大大提高Kalman滤波的精度。该方法计算简单,便于工程应用。

An incremental Kalman filter method under poor observation condition is put forward,in which its concept,model,basic equations and key calculative steps are given.Classical Kalman filter method requires a high precision measurement equation,or else it will cause great error in the recursive process.However,the measurement equation is usually influenced by environmental factors,and many practical situations are called poor observation conditions in which the measurement equation cannot be verified or calibrated.There usually are unknown system errors of the measurement equation under the poor observation conditions（such as deep space exploration）,which lead to great Kalman filter errors.The presented incremental Kalman filter method can successfully eliminate these unknown system errors and greatly improve the precision of Kalman filter.The method is simple to calculate and easy to apply in engineering.