Abstract
This paper presents practical aspects and an application of the decentralized partial state reference model adaptive control (DPSRMAC) to a grinding circuit. This control algorithm belongs to the class of long-range predictive controllers. Having in mind the pole placement approach, the quadratic control objective is expressed in terms of input and output tracking errors. The regulation and servo performances can be specified independently. A robust parameter estimation algorithm is used for on-line identification of the single-input/single-output (SISO) models. Simulations have been carried out using a phenomenological model of a grinding circuit. Comparisons are made between the DPSRMAC and decentralized SISO extended PI controllers, showing the efficiency and robustness of the adaptive control algorithm. © 1997 John Wiley & Sons, Ltd.
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call