Hi Benjamin,
The process noise covariance Q and the sensor noise covariance R are values greater than zero that you typically obtain from studies or measurements of your system. These values should be chosen based on the characteristics of your system and the noise present in your measurements. It is possible that the values you are using for the R and Q matrices are not appropriate for your specific problem.
Here are a few workarounds you can try to troubleshoot the issue:
- If the values are too small, the filter may not be able to effectively correct for noise. On the other hand, if the values are too large, the filter may overcorrect and introduce instability. Ensure that the values you are using for the R and Q matrices are appropriate for your problem.
- The performance of the Kalman filter can be influenced by the dynamics of your system. If the dynamics are highly nonlinear or if there are significant uncertainties, the filter may not be able to accurately estimate the state. It often requires some trial and error and a good understanding of the system dynamics and noise characteristics.
Further, one thing to note is that the Kalman filter assumes that the system is linear and that the noise is Gaussian. If these assumptions do not hold for your problem, you may need to consider using a different filtering technique.
To know more about the assumptions and algorithm of Kalman Filtering you can refer to its documentation here: https://in.mathworks.com/help/control/ug/kalman-filtering.html
Hope this helps!
