(188e) Controller Design for CSTR Process Output Using a Combination of GA, PSO, Fuzzy and PID Algorithms for Quick Rejection of Process Disturbances

Research in methods for process control design has been reinvigorated due to the advancement of computational capabilities and development of optimization and machine learning algorithms e.g. Genetic Algorithm, Neural network, Particle Swarm Optimization, Fuzzy Logic. As a result, a significant increase in the development of new control design algorithms has been observed.

Efficient controller design is a vital component of chemical process design. The aim of the control system is to address any process upset as quickly as possible to maintain the process within the specified operating parameters. As such, fast response time for the controller is an important design criteria. One option for achieving this is to develop an adaptive controller that exhibits the lowest possible variable dependence.

In this work, a CSTR system output has been used to study and compare responses of controllers developed using GA-PID, GA-Fuzzy, and PSO-PID algorithms. The aim of this study is to evaluate which combination of algorithm is the quickest in rejecting disturbances in the system. In the case of GA-PID and PSO-PID, GA and PSO were used to determine best control weighting factor (𝝀) and rise-time parameter (𝝈). However, in the case of GA-Fuzzy, GA was used to determine the best applicable Fuzzy membership functions for the operating parameters obtained from literature.

Selected References:

1. Rejane de B. Araújo, Antonio A.R. Coelho, Filtered predictive control design usingmulti-objective optimization based on geneticalgorithm for handling offset in chemical processes, Chemical engineering research and design, Vol. 117, 2017
2. Turker Tekin Erguzel, Fuzzy Controller Parameter optimization Using Genetic Algorithm for a Real Time Controlled System, World Congress on Engineering, Vol II, 2013
3. Yasue Mitsukura, Toru Yamamoto and Masahiro Kaneda, A Design of Self-Tuning PID Controllers Using a Genetic Algorithm, American Control Conference, 1999
4. Zwe-Lee Gaing, A Particle Swarm Optimization Approach for Optimum Design of PID Controller in AVR System, IEEE Transactions on Energy Conversation, Vol. 19, No. 2, 2004