Intelligent metaheuristic algorithm based FOPID controller for CSTR system

Abstract

The purpose of this research is to assess a continuous stirred-tank reactor (CSTR) system's performance. To enhance its performance, a fractional-order proportional-integral-derivative (FOPID) controller was employed, necessitating the tuning of independent control parameters. For this purpose, a sine-cosine algorithm (SCA) was introduced to optimize these parameters. The FOPID controller, tuned using the SCA, provides a powerful combination that addresses the complexities of the CSTR system. The fractional-order nature of the FOPID controller allows for superior tuning and robustness, offering enhanced flexibility in adjusting the system’s response characteristics and improving overall control performance. The SCA, known for its effective exploration of the search space through sine and cosine functions, ensures that the controller parameters are optimally selected to enhance the system’s performance by achieving an optimal fitness function. To showcase the effectiveness of the proposed SCA-tuned FOPID controller, comparisons were drawn with other optimization techniques designed for the CSTR system. The study presents time-domain characteristics and frequency responses of the proposed controller. The simulation results demonstrated that the SCA-FOPID controller significantly outperforms the other designed controllers, achieving a 54.07% reduction in the integral of time absolute error (ITAE) compared to genetic algorithm (GA), an 18.64% reduction compared to grey wolf optimizer (GWO), and a 34.79% reduction compared to differential evolution (DE). These significant reductions in ITAE underscore the effectiveness of this approach, highlighting the superior performance and robustness of the SCA-tuned FOPID controller in optimizing the CSTR system.