Recursive Least Squares-Based Adaptive Control of a Biodiesel Transesterification Reactor

Abstract

Proper control of biodiesel reactors poses a number of challenges. These arise from the presence of multiple chemical reactions, the complex heat and mass transfer characteristics, and the highly nonlinear dynamics. In this work, a new adaptive control scheme was implemented to automatically tune the controller on the basis of the most recent updated process dynamics. This scheme demonstrates the powerful integration of an online process modeling tool (recursive least squares, RLS) into a renowned and yet simple model-based controller design method (internal model control, IMC). Two adaptive control loops were designed, in which the sampling time of the RLS algorithm and the IMC closed loop time constant were determined by constrained optimization with the genetic algorithm. Comparison with conventional PID controllers revealed the superiority of the new adaptive control scheme in set point tracking. Good disturbance rejection properties were also demonstrated by the new adaptive control scheme. The results attained in this work demonstrate that a good adaptive control scheme can be implemented for the biodiesel transesterification reactors with minimal knowledge about the process model.