EKF-based data-driven backstepping control of continuous stirred tank reactor

This paper proposes a data-driven backstepping control (DDBC) method based on Extended Kalman Filter (EKF) for a class of nonlinear chemical process with Gaussian noise. Firstly, the states are observed by EKF to reduce the influence of stochastic interference such as Gaussian noise. Secondly, DDBC framework is designed. In this framework, nonlinear terms of system without model dynamic are estimated from states for solving the influence of model uncertainty. Then, based on Lyapunov theory, the controller is designed recursively. Finally, EKF-based DDBC is used to continuous stirred tank reactor (CSTR) with Gaussian noise and the effectiveness of this design is verified through simulation.