Optimization algorithm-based Elman neural network controller for continuous stirred tank reactor process model

Abstract

A continuous stirred tank reactor (CSTR) is a standout nonlinear system among the most essential units of chemical industries. In this article, an Elman neural network is designed to analyse the characteristics of nonlinear behaviour of the CSTR system. The data generated employing the state-space model of CSTR are used to train the designed Elman neural network controller and the controller parameters are optimally tuned by the proposed hybrid swarm intelligencebased optimization algorithm. Two different hybridizations have been developed, including DPSO, DGSA and hybrid DPSO-DGSA and successfully employed in controller tuning. The significance of the proposed controller is validated by a comparative analysis made with conventional methods and the performance is experimentally demonstrated using MATLAB software.