The Design And Implementation Of Adaptive MPC Control System For Electromagnetic Stirrers

In modern metallurgical enterprise, Electromagnetic stirring is is one important aspect to obtain high-quality materials. The control precision of EMS has a direct impact on mixing effects.However, electromagnetic stirring devices, which are made by hand, are difficult to have the same electrical characteristics, and the system parameters between products are also changed greatly, let alone inherent delay and inertia. All this create a barrier for the system to rapidly achieve setting current and frequency.The open-loop PID algorithm is hard to attained general use, without mentioning time-consuming and labor-intensive cased by finding parameters of the algorithm. It can be said that this algorithm is one source of the inefficient, during actual operation. Therefore, there is need for a new algorithm which can not only accurately identify the system, but also precise control system, while the adaptive model predictive algorithm be able to meet this demand.Base on large amounts of data statistics, it shows some characteristics of system which can be used in control. Using the characteristics, Self-adapting MPC has been implemented and tested under different parameters. When internal model is close to actual ones, it works good. The system integration component can enable two-way communicattion between electromagnetic stirring system and WinCC. By re-use and expansion of MICROWIRE serial communication protocol, a control framework for three-phase mixing system has been implemented and tested.A preliminary analysis been carried out for the three -phase mixing system. Base on TMS320F2812, a new pulse-width modulation has been implemented and optimized, which can replace the old chip (SA866).