Research And Application Of Temperature Control In Ceramic Sintering Based On Predictive Control

China is the largest producer of ceramic in the world, but the craft and technology of production process still has a gap compared with foreign countries. Ceramic kiln plays a very important role in the process, and the level of automation in kiln determines the development of the ceramic industry. Therefore, the control technology of kiln will directly affect the quality of ceramic, and the key point is the kiln temperature control system.It is difficult to obtain precise mathematical model because of the ceramic production process has lagged, nonlinear, time-varying characteristics, and the traditional control method is also difficult to achieve good control effect. But there is a new type of control, the predictive control, due to the weak dependence on the model, easy to implement and to achieve good control effect, has been widely used in the field of industrial process control.In this dissertation, based on the characteristics of the kiln control technology, the predictive control is used to the kiln temperature control system.Ceramic kiln current status and problems of control technology are studied, as well as the development of the predictive control. Then the structure of the kiln with the process and the sintering system, the methods of furnace temperature measurement are researched, obtained the basic model of the temperature control system and determined the control program.Through studying the ceramic sintering temperature control system, designed the hardware of ceramics kiln temperature control system, including microprocessor module, the temperature acquisition module, the executive body module, driver module, and hardware anti-jamming. The software is designed by VC++6.0, including PC control, temperature setting, real-time temperature display, upper and lower computer communications, and algorithms called.According to pure time delay characteristics of the object, dynamic matrix control has good dynamic and static performance and strong robustness, which has provided the possibility for achieved precise control of ceramic sintering temperature. Artificial neural networks and support vector machine (least squares support vector machine and a based on optimization algorithm least squares support vector machine) are use to predict the model, which can be replaced step response of the actual production process and predicted results are compared. The simulation results showed that the system prediction model can be made more accurate by least squares support vector machines based on particle swarm optimization. Through the simulation research and analysis some parameters of dynamic matrix control of, it drew the law of control system parameters on the performance. Based on the object of study ceramic kiln temperature control system and its dynamic matrix predictive simulation, and compared with the traditional Smith, internal model control, the results prove the effectiveness of this method.