You can test the estimator by setting the control input to 0 and testing a step response. The estimator value \(\hat{x}\) should converge to the state value \(x\). After this you can test various dynamic responses for correlation between estimated and actual states.
- bevlyMECH4420Lectureb