Artificial intelligence-based control of continuous polymerization reactor with input dead-zone

This paper utilizes artificial intelligence (AI) techniques to address the trajectory tracking problem of a continuous polymerization reactor with unknown dynamics and unknown asymmetric input dead-zone nonlinearities. A backstepping fractional sliding mode control (BFOSMC) has been proposed to force the average molecular weight (AMW) and the reactor temperature to track the desired trajectories. Due to the simplicity, easy implementation, and robustness of particle swarm optimization (PSO) technique, it is used to optimize the parameters of the controller. The unknown dynamics of the reactor are approximated with feed-forward neural networks (NN). The input dead-zones are estimated and compensated by adaptive offset terms in the control laws. A Lyapunov theory is employed to prove the stability of the NNBFOSMC. The simulation study verifies that our AI-based algorithm resulted in a better performance.