Extended Kalman Filter Approach for Model Based State Estimation of a Continuous Stirred Tank Reactor with Model and State Uncertainties

Abstract

Accurate state estimation is crucial in various fields, particularly in control systems and signal processing, as it directly influences system performance and reliability. In environments where measurements are susceptible to noise and external disturbances, the ability to derive precise state estimations enables effective real-time decision-making.Traditional linear estimation techniques often fail to capture the nuances of nonlinear dynamics, leading to a demand for more sophisticated methods. A non-isothermal continuous stirred tank reactor operating in three modes is considered where the mode transition is triggered whenever a higher yield is desired. To evaluate the efficaciousness of state estimation scheme simulation studies have been carried out on the simulated model of the chemical reactor. Simulation results under initial-state and parameter-mismatch conditions demonstrate that the proposed EKF-based estimator provides accurate and robust estimation of reactor concentration and temperature, making it suitable for monitoring and control of hybrid nonlinear chemical processes.