Data-driven design of two-degree-of-freedom controllers using reinforcement learning techniques
Motivated by the successful application for feedback management, this study extends the study of reinforcement learning techniques to the look of 2-degree-of-freedom controllers in the info-driven environment. Based mostly on the residual generator primarily based form of Youla parameterisation, all stabilising controllers are initial interpreted within the feedback-feedforward situation with a Kalman filter-based residual generator acting because the core part. For the reference tracking downside, any discussions are conducted from the regulatory perspective and using the Q learning, recursive least squares ways and therefore the policy iteration algorithm. The entire design is administered as a 2-stage process that separately achieves the optimal feedback and feedforward controllers. Finally, the effectiveness of the proposed approach is demonstrated with its application within the laboratory continuous stirred tank heater process.
Did you like this research project?
To get this research project Guidelines, Training and Code... Click Here