Error using Kalman function - cannot compute stabilizing Riccati solution

8 次查看(过去 30 天)
Please see below a snippet of my code:
sys=ss(A_ae_control,B_ae_control,C_ae_control,0);
Qn=0.0001;
Rn=0.01*eye(size(C_ae_control,1));
[kest,L,P]=kalman(sys,Qn,Rn);
I can't seem to figure out why this won't run. I keep getting the following error:
Cannot compute the stabilizing Riccati solution P for the Kalman filter.
This could be because:
* RN is singular,
* [QN NN;NN' RN] needs to be positive definite,
* The E matrix in the state equation is singular.
Any insight would be appreciated. I have attached my A, B and C matrices.
Also, is it possible to somehow create a Kalman estimator without any measurement noise (only a disturbance). I tried setting Rn to 0 but not sure why this does not work.
  4 个评论
Emir Acevedo
Emir Acevedo 2022-9-13
I had designing a LQR controller and after an Kalman Filter. For mi satellite control, I used the same Q and R matrix in Ricatti to solve Kest. May that can fix your problem.
Regards.

请先登录,再进行评论。

回答(1 个)

Ivo Houtzager
Ivo Houtzager 2023-4-17
Your state-space model is probably not a minimal realization, in other words not fully obervable or not fully controllable. Perform the function minreal to remove unobservable or uncontrollable states.
sys=ss(A_ae_control,B_ae_control,C_ae_control,0);
sys=minreal(sys);
Qn=0.0001;
Rn=0.01*eye(size(C_ae_control,1));
[kest,L,P]=kalman(sys,Qn,Rn);
  2 个评论
rajesh r
rajesh r 2023-4-17
My model is:
A = [1 0.01; 0 1];
B = [0 ; 0];
C= [1 0];
D = 0;
after minreal,
the complete A matrix is zero
Ivo Houtzager
Ivo Houtzager 2023-4-17
You have no inputs defined, because both B and D are zero. Thus your model is fully uncontrollable. So you have review your model inputs for correctness.

请先登录,再进行评论。

类别

Help CenterFile Exchange 中查找有关 Matrix Computations 的更多信息

产品


版本

R2020a

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by