Real Time Optimization Of Cement Precalciner Based On Multi-objective Predictive Control

Fuel combustion and raw calcium carbonate pre-decomposition are conducted by air suspension in cement precalciner.As the core equipment of cement production line,the operation regulation of precalciner is the key to ensuring the quality of cement and the economic and environmental protection of the production process.However,the complex coupling process of gas-solid turbulence and chemical reaction is carried out in precalciner.The control of operating parameters such as coal injection and ammonia injection mainly depends on experience.This will cause problems such as high production energy consumption and large pollutant emissions.In order to realize the high-efficiency and low-pollution regulation of cement precalciner,a mathematical model of the precalciner for optimal control is constructed in this paper,and the multi-objective optimization strategy and real-time control scheme of the production process are studied to provide an effective solution for the control and optimization of the precalciner.Firstly,a one-dimensional mathematical model of the precalciner is established and the steady-state and dynamic characteristics of the system are studied.In this paper,a precalciner of a 5500 t/d cement production line is taken as the research object.On the basis of analyzing the process characteristics of the precalciner production process,a cell model is established to describe the gas-solid flow,pulverized coal combustion,raw meal decomposition and nitrogen oxide removal in the furnace along the direction of flue gas.The idea of hybrid modeling is introduced in the modeling process,and the complex gas-solid flow formula is approximated by the neural network method.By solving the mass balance equation and the heat balance equation in each cell,the rate of change of the state parameters in the furnace is obtained.The relative error between the model calculation results and the field operation data is basically within 10%.Based on the mathematical model of precalciner,the one-dimensional distribution characteristics of parameters in the furnace,the steady-state relationship between parameters and the sensitivity of key variables are discussed respectively.The dynamic response characteristics of the system under the step disturbance of raw meal feeding amount,coal injection amount,ammonia injection amount and high temperature fan speed are studied in turn.On the basis of the above modeling and operating characteristics research,a multiobjective optimization solution strategy for the precalciner is constructed.Through the analysis of production quality,pollutant emission,economy and safety in the operation process,performance index is quantified.The optimization objective function and constraint conditions are designed,and a multi-objective optimization problem that closely fits the actual production is constructed,that is,minimize the set value track deviations and minimize economical costs of the production process.The multi-objective evolutionary algorithm is used to solve the optimization problem,and the key variables of the production process are fitted based on the mathematical model of the decomposition furnace established in this paper,so as to reduce the calculation time of large-scale population optimization and quickly obtain the optimal operating parameters under the current working conditions.The algorithm can provide guidance for production control.Based on the above mathematical model and optimization suggestions,a predictive control method is introduced to realize multi-objective real-time optimal control of the precalciner.The mathematical model of the decomposition furnace established in this paper is linearized at the steady state point as a prediction model,and a multi-model switching strategy with a hysteresis factor is designed.The quality indicators,safety indicators and economic indicators of the precalciner operation process are introduced into the predictive control framework.By calculating and predicting the variation trends of controlled variables such as temperature and NOx concentration in the time domain,constructing the incremental constraints and amplitude constraints of coal injection and ammonia injection in the control process,and taking the ideal point of the multi-objective optimization problem described above as the tracking target,a multi-objective predictive controller is built.Through the analysis of multiple simulation cases,it is found that compared with the conventional predictive controller,this method can realize the coordinated optimal control of the amount of coal injection and ammonia injection,and effectively overcome the control difficulties caused by the nonlinearity and time delay of the complex system.The stability and economy of the control system can still be maintained under the conditions of set point change,external disturbance of flue gas and raw meal,and model mismatch.Finally,based on the App Designer platform,the decomposition furnace optimization combustion control software is developed.Combined with the previous research results,three functional modules of parameter prediction,multi-objective optimization and realtime control are embedded in the software framework,and a visual interactive interface is designed to realize the engineering application of the research content.