Using "kalman" with a Zero B matrix

5 次查看（过去 30 天）

John F 2014-5-20

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/130414-using-kalman-with-a-zero-b-matrix

评论： Arkadiy Turevskiy 2014-5-29

I am trying to use the "kalman" command to compute the steady-state gains for a Kalman filter in a signal tracking problem. There are no inputs. So, the dynamic equations are:

X[k+1] = A*X[k]; Y[k] = C*X[k];

I feel like there should be a way to do this. Any ideas? Thanks!

0 个评论
显示 -2更早的评论隐藏 -2更早的评论

请先登录，再进行评论。

请先登录，再回答此问题。

回答（1 个）

Arkadiy Turevskiy 2014-5-22

0
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/130414-using-kalman-with-a-zero-b-matrix#answer_137851

Yes, you can use kalman to design a Kalman filter for a system with no input u . Did you try it? Did you run into problems?

2 个评论
显示无隐藏无

John F 2014-5-28

在 MATLAB Online 中打开

I get an error that an internal matrix in the calculation is not positive definite.

Error using ss/kalman (line 167)
In the "kalman(SYS,QN,RN,NN,...)" command, the covariance matrix
    E{(H*w+v)(H*w+v)'} = [H,I]*[QN NN;NN' RN]*[H';I]
must be positive definite.
 I'm assuming I am framing the problem incorrectly.