Cascaded model predictive control of a rotary lime kiln

A well-operated kiln should produce high-quality lime in sufficient quantity to meet the needs of the causticizing operation, while minimizing operating costs and environmental emissions. From a control perspective, the kiln may be viewed as a multivariable distributed parameter process. Although control strategies based on univariate methods have been shown to work, what is needed is a truly multivariable strategy. In this report, we propose a cascaded model predictive control strategy that meets this need.