(346q) Use of Multistage Optimal Control Principles for Novel Design and Implementation of Classical Controllers

Classical controllers, such as Proportional-Integral (PI) and Proportional-Integral-Derivative (PID) controllers, are the most long-established and widely used in industry. Various methods for tuning these types of controllers exist (Ziegler et al., 1942; Blondin et al., 2018; Do et al., 2021), and up to this point, there is no fruitful avenue to improve their performance.

In this contribution, a new approach for PI and PID controller implementation, based on a Multistage Optimal Control (MSOCP) approach is introduced. Our approach incorporates path and end-point constraints during its controller tuning phase, as well as parameter and disturbance uncertainty.

The proposed framework is applied for different case studies and is able to reject any disturbances introduced to the examined systems, with or without uncertainty, satisfies end-point constraints and exhibits quicker response for switching steady states, compared to classical methods.

Other aspects of controller design and incorporation within industrial process models, as related to using rigorous optimization methodologies and implementations, will further be highlighted within the context of the PI and PID controllers.

References

Blondin, M.J., Sanchis, J., Sicard, P. and Herrero, J.M., 2018. New optimal controller tuning method for an AVR system using a simplified Ant Colony Optimization with a new constrained Nelder–Mead algorithm. Applied soft computing, 62, pp.216-229.

Do, H.T., Van Bach, N., Tran, H.T. and Nguyen, M.T., 2021. A design of higher-level control based genetic algorithms for wastewater treatment plants. Engineering Science and Technology, an International Journal.

Ziegler, J. G. and Nichols, N. B., 1942. Optimum settings for automatic controllers. trans. ASME, 64(11).