The Control System Design Of Acetylene Hydrogenation Reactor Based On MVC~3

The production of ethylene is one of the most important links in the petroleum chemical industry. The purity of the ethylene product does not meet the requirements cannot be used in subsequent production. Especially in the presence of acetylene impurities will affected to each section of the reaction, which use ethylene as raw materials. However, the process of removing acetylene is very complicated. The traditional PID controller and the direct identification method are both difficult to achieve satisfied effects. Nowadays, an olefin plant of Sinopec is still adopting manual operation on the acetylene hydrogenation device, which has plenty of defects:in order to meet the outlet acetylene concentration is less than 5ppm; the reactor is often running in the condition of over hydrogenation in a long period. It will cause the waste of hydrogen, the reduction of activity and selectivity of catalyst and the yield of ethylene production. This paper focuses on solving these practical problems in industrial:Firstly, according to the reaction mechanism of acetylene hydrogenation process, and considering the energy balance, material balance and the kinetics equation, the model is linearized near the operation points; the temperature of each bed is regarded as the state. And the structure of the state space form is determined. The parameters of the model are calculated by the least square principle. The parameter matrixes are modified on-line and verified off-line. Besides, the on-line and real-time measurement of outlet acetylene concentration of the reactor, the activity and selectivity of catalyst are put into reality, which is in preparation for the design of control system.Secondly, the minimum covariance constraint control (MVC3) algorithm is adopted in the control part, which puts the input and output variance constraints of the system into account to minimize a new comprehensive index for the purpose. Besides, the mathematical tool LMI and Schur lemma are applied to obtain the optimal feedback control law. It has practical significance for industrial production process. Compared with the traditional manual operation, the advantages of this method are obvious. The facts have proved that this method has good control effects, and it can meet the needs of industrial production well.Finally, when the controlled object changes greatly and the model of the system cannot be reconstructed in time, the feedback control law based on the minimum covariance constraint control (MVC3) algorithm may no longer adapt to the process which has changed, and the control effects are not satisfied any more either. Therefore, the active disturbance rejection control (ADRC) technique is introduced in this paper. It does not need the exact model of the controlled object and it has strong robustness. When the MVC3 controller is not suitable for the current condition, switch to the ADRC controller as the main controller to guide the process.