Research On Modeling And Control Of Varnish Reaction Kettle System

In the modern home decoration, varnish has been widely accepted by people and become a kind of the mainstream paint, due to its advantages of excellent luster, fast film forming and good waterproof. Its demand is growing by the day. To meet the market requirement of small batch and multitudinous variety, varnish is mainly produced intermittently and the batch reactor is used as the main equipment. Varnish production is a complicated physical and chemical reaction process and has the characteristics of large delay and serious nonlinearity, which makes it difficult to control the temperature of reaction kettle, while the temperature of reaction kettle directly affects the quality of the termination product.The article firstly depicts the selection of the varnish production system and introduces the structure and function of each equipment of the reaction kettle system in the form of subsystem. Then it describes the varnish production process, including feeding, reaction kettle reacting, condensation, water diversion, olefin kettle reacting and the final discharge slag formation of finished products. A control system is designed to regulate parameters in the process after determining the process. According to the characteristics of varnish producing and combined with the modern industrial control technology, we design a control system which takes IPC as the upper operating level and DCS as the basic control level to control the temperature of the reaction kettle. We take the AI-T8000 distributed control system as the control station and take the Wizcon Supervisor software as the operating station for monitoring.In order to get a better control of the temperature in the reaction kettle, the article establishes a mathematical model about the temperature. While the traditional mechanism modeling can well describe the internal laws of things, it can hardly describe the varnish reacting precisely for the reaction kettle system has the characteristics of large delay, serious nonlinearity and so on. However, the statistical modeling can achieve the purpose of establishing an object model without a deep understanding of the system’s internal mechanism, using the mathematical processing methods. The article establishes the reaction kettle systems’ RBF neural network model with the orthogonal least squares method by the production data collected through the field. This model has the advantages of simple structure, concise training and can approximate nonlinear functions with arbitrary precision.Finally, the article puts forward a control strategy of the temperature in the reaction kettle. Combining the advantages of the superior dynamic performance of fuzzy control with the steady precision of traditional PID control, we build a fuzzy-PID controller to control the temperature of the reaction kettle, taking the fuzzy controller when the error, which is the difference of the set value and the feedback value, is large and taking the traditional PID controller when the error is small. The simulation results show that the fuzzy-PID controller has a better control performance than the traditional PID controller.