On-Line Control Of Rawmix Quality In Cement Plants

Abstract

A new control strategy is proposed to continuously control the quality of the raw mix in cement plants. The proposed control strategy is based on the gain-adaptive model based predictive control, some heuristics, and the soft sensor. The soft sensor provides the continuous feedback of the quality parameters. The features of the proposed control strategy are -- avoiding the need of costly on-line sensors, handling of non-linearity by means of gain adaptations, use of heuristics based computational block to handle feed composition uncertainties, handing of the constraints on various process variables, consideration of the interaction among various process variables, and optimization in terms of the usage of the raw materials. Based on the proposed control strategy, the software based supervisory controller is implemented on the actual plant and has shown encouraging results.