Silicon Single Crystal Diameter Control Systembased On Dual Predictive PI

Single crystal silicon is the basic raw material in the microelectronics industry and solar cell production primarily by Czochralski method. In the production of silicon in the process, the flow of argon and the crucible rotation cause fluctuations of thermal field gradient in the furnace. Inertial process objects, large delay, nonlinear and multi-interference is the difficulty of control. Surface temperature and the pulling speed is not only related to the diameter of the silicon rod, but also affects the structure of the atomic arrangement of single crystal silicon. Surface temperature is higher, the smaller the diameter, slow pull star, the bigger the diameter, to coordinate the surface temperature and the casting speed is the key of control of single crystal silicon production furnace.This paper analyzes the basic structure and principle of dual control system, analyze its performance and compare the effect of control of the predicted PI controller and Smith Predictor Control for a delay system. Based on Kalman filter, a solution that accurately measures temperature of surface and thermal field was proposed. On the one hand, we used dual control algorithm to coordinate the casting speed and the temperature of liquid surface. On the other hand, two cascade systems was used for quickly suppressing thermal field temperature and the temperature of the liquid surface disturbance. In addition, we used predictive PI algorithm to solve the problem that the object is large inertia and delay. Then the dual predictive PI control algorithm to control the diameter of single-silicon was put forward. This paper derived the deputy controller object model of dual control system, provides a theoretical for deputy controller selection and parameter tuningIn this paper, the speed and robustness of double prediction PI control system was confirmed. Pull Speed-Diameter as the main control loop to respond quickly for a variety of disturbances within the control system. Temperature-Diameter-Pull Speed as a sub-control loop that try to make the pull speed in the ideal state, steady and rapid control system to overcome the disturbance of thermal field, and ensure product quality and production efficiency.OPTO22 PAC as a controller hardware and OPTO22 PAC Project as the software platform in this paper. Then real-time monitoring software was developed. Based on the PAC controller hardware, OPTO22 OPC servers and Simulink the data exchange interface that provides rapidly and economic debugging was built. Online real-time simulation results show that the double predictive PI control system can be fully coordinate the pull speed and surface temperature and quickly and smoothly response to disturbance, verified the practicality of double predictive PI control system.