Modeling And Control Methods For The Alumina Evaporation Process

Alumina is the main raw material for the production of electrolysis aluminium. The special property alumina is used widely in the fileds of electron, petroleum, chemistry, pharmacy, material, paper making, military affairs, and spaceflight, etc. Thus, the alumina takes an important role in national economy. Evaporation process is the important part in the production of alumina. The objective of the evaporation process is to remove the redundant water of the spent caustic liquor, so that the caustic concentration of the liquor meets the production demands to make the liquor be recycled. Whether the production quality of the evaporation process satisfies the production demands or not will directly affect the other working procedures, such as mill grinding process, leaching process and eventually affect the quality the alumina. Thus, the research on the modeling and control methods for the alumina evaporation process are of the great significance.The evaporation process is composed of the six-effect countercurrent evaporation system, the four-effect flash evaporation system and the forced-circulation evaporation process. The forced-circulation system and the six-effect countercurrent evaporation system have complex dynamics, such as strong nonlinearity, strong coupling. The existing control method often results in large fluctuations of the level and the slow tracking of the product density with respect to the setpoint. Therefore, the existing control strategy can not satisfy the production demand and the production quality, even leading to the system instability.This dissertation is based on the project "The700thousand tons bayer alumina production integrated automatic system of Zhengzhou alumina factory". In order to ensure the stability of the system, improve the control effect and the control performance and reduce the energy consuming, modeling and control methods for the evaporation process of Zhengzhou alumina factory, a classical process industry, is presented in this dissertation. The main works are as follows:(1) The dynamic modeling of the forced-circulation system is developed based on the mass and energy balances principles. The inputs of dynamic model are discharge flow and the steam heat flow. The outputs of the dynamic model are density and the level of the evaporator. The validation of model can be verified using the actual data.(2)The adaptive decoupling switching control strategy based on generalized predictive control is proposed for the forced-circulation system characterized by strong nonlinearities, coupling and uncertainties. The control strategy is composed of a linear adaptive decoupling controller, a neural-network-based nonlinear adaptive decoupling controller and a switching mechanism. The linear adaptive decoupling controller is designed to ensure the boundedness of the input and output signals of the closed-loop system. The neural-network-based nonlinear adaptive decoupling controller is developed to improve the transient performance of the closed-loop system. The stability and convergence of the closed-loop system are achieved simultaneously by using multi-model-based switching mechanism.(3)By taking the actuating valves and the forced-circulation evaporation system as a augmented plant, the developed adaptive decoupling switching control strategy solve the nonlinear, strong coupling and uncertain problems for the forced-circulation process. Simulation results show that the proposed method not only can effectively decouple the dual two loops effectively for the forced-circulation evaporation system and improve the evaporation efficiency, but also has strong robustness in case of large disturbance and uncertainties.(4)The dynamic model of the six-effect countercurrent evaporator process is presented. The inputs of the dynamic model are discharge flow, steam flow. The outputs of the dynamic model are discharge density, temperature of the second evaporator and the level of the evaporator. In order to cope with the strong coupling and nonlinearities, the input-output linearization decoupling technology that has globally linearizing control (GLC) structure is adopted. PI control law applies to the linearized system. Simulation results show that input-output linearization control strategy eliminates the coupling of each loop and can track the set point rapidly and efficiently. At the same time, it can meet the aim of saving energy and reducing energy consumption.