Research On Multi-Objective Optimization And Model Predictive Control Of Cement Calcining System

The calcining subsystem is one of the key aspects of the cement production process and its production equipment has been updated with the progress of modern science and technology.However,the optimization and control methods of the calcining system have not been improved along with the upgrading of the production equipment.The backward optimization and control methods have led to the problems of low resource utilization,poor product quality and difficult control of production equipment in China’s cement manufacturing industry.Therefore,it is important to study the intelligent optimization and control of cement calcining system,which will reduce the energy consumption of production,maintain the stable operation of production and guarantee the quality of production.In this paper,the research on optimization and model prediction control of cement calcining system operation parameters is carried out to address the above problems,and the main research contents are as follows:Firstly,starting from the process flow of cement calcining system,we analyze the operation mechanism of the system in detail and analyze the production index relationship,determine the system operation parameters optimization variables and control variables of calcining equipment operation parameters related to the production index,and finally propose the optimization and control framework scheme of cement calcining system.Secondly,to address the problems of many production equipment,complex mechanism and difficult to model,high operating energy consumption and difficult to set operating parameters of cement calcining system,it is proposed to establish the production index model of cement calcining system using deep GRU network to achieve accurate prediction of energy consumption and quality,based on which a multi-objective optimization model is established to solve the problems of high energy consumption and difficult to control production quality of calcining system,and to use The MOEA/D-DE algorithm is used to solve the Pareto solution set for the operating parameters,and then the VIKOR algorithm is used to select the optimal solution in the Pareto solution set and send it to the actual operating system as the set value to finally realize the intelligent optimization of the cement calcining system operating parameters.Then,in order to solve the problem that cement calcining system equipment is difficult to control due to time lag,nonlinearity,and large disturbance,the method of combining deep learning and dynamic equations is proposed to construct the Hammerstein model.The CNNGRU network is used to extract the operating state of the calcining system and use the attention mechanism to reinforce the important time point information,and then the ARX dynamic model is constructed based on it to predict the operating state of the calcining equipment at the future time.The Hammerstein model predictive control system(HMPC)is constructed based on the prediction model,and the complex nonlinear model solution is composed of linear control and nonlinear optimal solution links,and the generalized predictive control(GPC)is used for linear control and the improved sparrow search algorithm(ISSA)for search solution,respectively,to finally achieve stable control of the calcining system equipment.