Recursive orthogonal least squares based adaptive control of a polymerisation reactor

Abstract

ABSTRACTA Nonlinear Internal Model Control (NIMC) scheme, based on a recursively updated Radial Basis Function Network (RBFN), is applied to control the nonlinear Polymerisation process. Recursive Orthogonal Least Squares (ROLS) algorithm recursively updates the weighting matrix of RBFN in real time, such that modelling errors are minimised. In addition to the ROLS based adaptive IMC, a non-adaptive IMC, based on the fixed Orthogonal Least Squares (OLS) algorithm and a normal PID controller are also applied to control the polymerisation process. The simulated results show the effectiveness of the ROLS based controller in terms of response time to step changes and disturbance rejection capabilities over the classical PID controller and the fixed OLS based IMC. Also, the proposed algorithm fits well in the IMC framework.