CAD Versus Self-Tuning Control of a Clinker Grate Cooler

Abstract

A study has been reported on computer identification and control of a Puller clinker grate cooler pilot plant cooperating with a wet-process clinker rotary kiln. Having briefly described the cooling process, its different experimental models have been analysed, both deterministic and stochastic ones. Offline and on-line versions of popular parameter estimation schemes have been applied, i.e. the least squares, correlation, generalised least squares and maximum likelihood methods. A CAD control algorithm based on a constant process model has been firstly inspected. Next, a simple adaptive control strategy based on self-tuning regulation philosophy has been more successfully applied. Taking finaly into account some emergency situations also, we have found that best solution is to properly compromise between the CAD and self-tuning control.