Research Of Predictive Control Algorithm Based On Soft Constraints For The Cement Mill

Cement mill is the key equipment with high energy consumption and heavy wear in the cement production process. The control level to the cement mill affects the quality, production and energy consumption of the cement products directly. However, for the complex cement mill control system, because cement mill itself has the large time delay, multiple constraints, multi-variables, close coupling among variables and other characteristics, it is commonly difficult to establish a accurate mathematical model for the cement mill system and control accuracy is usually difficult to ensure. With the development of computer technology and control theory, continuous mature of predictive control method based on dynamic prediction model, receding horizon optimization and feedback correction, bring a new research direction for the control to the cement mill system. For cement grinding technics and the control mechanism, the article proposes a model predictive control algorithm based on soft constraints for multivariable time-delay systems and designs cement mill predictive control system based on soft constraints. Specific research efforts are as follows:Firstly, the article elaborates on the research status of cement mill automation control and predictive control, and points out the importance of improvement of cement mill control automation level. Through analyzing cement grinding equipment and technological process, and further studying several typical conditions and control mechanism of cement mill, various factors that affect the running of cement mill system are clear, the main parameters of weighing bin material level remain system and ball mill load optimization system are determined. They provide theoretical principle for the precise control of cement mill.Then, study the predictive control algorithm based on soft constraints for multivariable time-delay systems, derive predicted value of system output based on prediction model, and construct objective function through error feedback. Through designing output soft constraints, relaxing the original constraints, and defining constraint function based on soft constraints, a new objective function is constructed. The optimal control input can be gotten by solving the objective function.Finally, according to the predictive control algorithm based on soft constraints for multivariable time-delay systems, the article studies the cement mill predictive control algorithm based on soft constraints, and designs the structure of cement mill control system. It configures graphic interface and predictive controller parameters based on simulation platform of cement production process control system. The predictive control algorithm based on soft constraints is applied to the cement mill process control system, and control effect is analysed.