Disturbance Rejection using PID Controller and Input Estimation

Abstract

— Disturbance rejection and accounting for the uncertainties associated with the system in real time
have always been a challenging task in industries. Closed loop control systems are employed in industries
for rejection of such disturbance inputs. Sometimes observers are used for estimating the unmeasurable
input disturbances entering the process, but are limited to deterministic systems. In the proposed work, for a
stochastic system, the states are estimated using Kalman filter (KF). The disturbance is assumed to be added
at the input side and recursive least square estimation (RLSE) with forgetting factor is used for estimating
the unknown disturbances affecting the process and thereby the rejection of those disturbances. The
validation of the proposed work is observed with the simulation results showing the comparison of PID
controller alone and PID controller plus input disturbance estimation for eliminating the disturbances in a
benchmark process control system- Continuous Stirred Tank reactor (CSTR)