Information Technology Reference
In-Depth Information
Fig. 5.7. Linearization of a controlled system by variable change
modeling errors, which is not the case for the inverse-model based control laws.
Training is performed as indicated in the diagram of Fig. 5.5. Nevertheless,
note that the controller input is not the state of the process, but the state
of the internal model . Therefore, during training, it is advantageous to use
relevant values of the state variable of that model.
Fig. 5.8. Internal Model control law
Search WWH ::




Custom Search