The Reserch Of Improving CPE Control System Based On Fuzzy Neural Network

Chlorinated polyethylene referred to CPE, a new polymer composite materials, has excellent weather resistance, aging resistance, electrical insulation, and other fine features. From the mid-1990s, China began to vigorously promote plastic building materials, restricting the use of wood, so that the construction industry to the rapid development of plastic and wood, plastic and steel. In recent years, with the rapid development of domestic manufacturing and construction, the demand for the chemical building materials is increasing, CPE as a new type of elastic material, due to the simple raw materials, lower production costs, as well as plastic doors and windows and other petrochemical products superior more and more people know, the domestic CPE products as its production and application of new materials is a rising trend. CPE’s production process mainly chloride polymerization is carried out in a batch reactor, the reactor temperature control in the production process of the chlorinated polyethylene exists a large hysteresis, large inertia characteristics, the common cascade control is difficult to get the ideal control effect, and prone to relatively large overshoot to adjust long time, robustness, and even affect the quality of the product. The conventional PID control cannot meet the control requirements, for which we propose the use of advanced control algorithms to give an effective transformation.In this paper, a chemical plant with an annual output of10,000tons of CPE polymerization reactor control system for the study of the background, after in-depth analysis of the polymerization reaction temperature’s characteristics and summed the difficulties of CPE polymerization and control, proposing the advanced control method based on fuzzy neural network control. The fuzzy logic system is easy to understand. The neural network has a robust adaptive learning ability. Make maximum use of fuzzy control theory and neural network learning algorithm based, targeted to the temperature of the control system, to design a controller using fuzzy neural network learning algorithm, to complete the advanced control algorithms for the CPE control system improvements.In order to be able to intuitively vivid performance of the CPE production processes in the chemical industry, this article designs a CPE production configuration with the SUPCON JX-300XP DCS distributed control system and SCKey configuration software which has reliable operation, real-time monitoring and dynamic characteristics. Using the SCKey configuration software to solve the recipe management, batch control and kettle temperature control of the polymerization reactor. In order to verify the availability of the algorithm, this article uses the Matlab simulation with powerful computing capabilities to simulate the algorithm, and the simulation shows the improvement effect, the fuzzy neural network algorithm is applied to the actual production to provide an effective way.