中文说明:
描述一个卡尔曼滤波问题需要两个模型,一个是描述系统的状态方程,一个是观测方程,观测量通过观测方程与状态变量建立联系,由观测量估计状态值。与其他频域滤波器不同,卡尔曼滤波器不需要观测和估计的历史记录,可以直接在时域进行设计和使用,是一个时域滤波器,适用于处理实时数据。 对于一个运动模型,建立卡尔曼滤波模型,进行仿真,设已知初始时刻运动目标的真实位置和速度,并已知卡尔曼滤波使用的初始状态值,对该问题给出仿真;进一步分析该问题的稳态卡尔曼解,直接使用稳态卡尔曼滤波(滤波器)仿真该问题。 提供了Matlab源代码,代码中有注释和画图,非常详细。
English Description:
Two models are needed to describe a Kalman filter problem, one is the state equation of the system, and the other is the observation equation. The observation establishes a relationship with the state variables through the observation equation, and the state value is estimated by the observation. Unlike other frequency-domain filters, Kalman filter can be designed and used directly in time domain without the need of observed and estimated historical records. It is a time-domain filter and suitable for processing real-time data.