State Estimation for Tennessee Eastman Challenge Process (TECP) Using Time Varying Discrete Kalman Filter

Abstract

This paper describes a linearized model of an industrial chemical process for the purpose of developing the solutions of some practical problems like sensor validation by continuous state estimation of the process. The Tennessee Eastman Challenge Process (TECP) is proposed here as a benchmark problem to evaluate the behavior of process variables. State Estimation is a key factor in design of Fault tolerant system. The proposed approach for state estimation is Time Varying Discrete Kalman Filter. The proposed Kalman observer provides continuous monitoring of some unmeasured states of TECP reactor. The ability of Kalman Filter for nonlinear TECP chemical process is tested with random noise.