Parameter Estimator Design with Application to a Chemical Reactor

Abstract

A new method of on-line parameter estimation for a class of nonlinear systems is presented in this paper. The parameter estimator includes a left inverse of process model and at each time instant calculates least-squared-error estimates of parameters by using readily available on-line measurements. The estimator is computationally efficient and easy to implement. The application and performance of the parameter estimation method are illustrated and compared with those of parameter estimation via state estimation, by using a jacketed chemical reactor example for which the reaction-heat rate and overall heat-transfer coefficient are estimated. The simulation results show that the inversion-based estimator outperforms the state-estimation-based parameter estimator in the presence of measurement noise and plant−model mismatch.