Waste Incineration Boiler Drum Level Control Model Predictive Control Technology Research

This paper has discussed the popular method of control and deeply analyzed the dynamic and static characteristics of the water level based on analyzing the main factors influencing the water level of drum.Solid waste boiler system is a complicated multi-variable system with nonlinear characteristic, and the parameters of this system change according to the condition of operation. In general the capability of solid waste boiler is small and the demand of control is easy to meet, so the satisfying effect will be got using normal control instruments.The test and control lab in thermal engineering graduate school in Zhejiang University has developed a set of DCS demonstrating control system of the solid waste plant. After putting this system in practice, the solid waste plant has gotten better effects. Based on this system this thesis brings forward applying the technology of forecast control into fired solid waste boiler control system, which makes the original system obtain another two functions of model identification and forecast control. The technology of predict control develops following with computer, communication and the theory of control.This paper begins with the balance theory of heat and substance of water in boiler' s bubble, and educes the universal mechanism controlling model to describe the water level of boiler, and studies the character of controlling model of the water level of solid waste boiler. In the following chapters the opened loop model of the water level of No 1 solid waste boiler in Qiaosi has been set up with the method of system identification and data of opened loop running. The closed loop model of the object also has been identified with the data of the closed loop. Looking from the statistical results of predictive error curves, the closed loop system benefits to set up the model of water level of the bubble. The mean of predictive error e(k)isE = 0.00082 , variance is 2 =1.863. The value is close to a misrelated random yawp serial, because the fluctuation of the identification data is much smaller in closed loop.The model based on the closed loop and the opened loop has been simulated respectively. The identification model can comparatively truly reflect localeobjects. In addition the simulation plane of this model has been set up, and the result of simulation is according to the real curve of water level. Under the appointed closed loop capability index, the parameters of simulating control based on identification model are Kp=2.38, Tt =30 and Td=0.Chapter four has studied the application of the technology of predictive self-adaptive control on controlling water lever of solid waste boiler. With the condition of having opened loop identified result as real model and having closed loop result as predictive model, the controlling effect has been analyzed through the simulation. When offset is 60 mm, the overtaken mete is -30mm, and the adjusting time is about 200 seconds.