Study On Fuzzy Iterative Learning Control For Electro-slag Furnace

After years of practice and development, electro-slag re-melting furnace steel-making method has got significant progress. Its producing process has also evolved variedly. After remelting, the steel composition has greatly improved the quality of the ingot greatly improved, so the ESR is widely used in industrial production, especially the in the production of special steel. However, the control the ESR is so strict that we must use the appropriate control method to design optimal control systems to ensure the stability and reliability of the melting process.This article first briefly introduces the historical development of electroslag remelting, process principle, the control situation and the hardware structure of electroslag furnace. Then an iterative learning control method for normal melting process has been proposed. And iterative learning controllers have been designed respectively using open, closed-loop PID-type iterative learning algorithm. The entire control system has been studied through the simulation.The initial basic speed is calculated by the model based on the melting theory.For the robustness of the system is not strong enough, we proposed to optimize the PID parameters in the iterative learning law using the fuzzy control theory. A fuzzy controller has been designed using the systematic error and error change rate as input and the variation of the PID parameters as output. Also we gived the simulation analysis. Eventually from using the fuzzy iterative learning algorithm, the actual output melting rate of system is well tracking the given melting rate curve. At the same time the robustness of the system has been ensured when a large disturbance is added to eh system or when the parameters of system has been changed. Thus the remelting process could be more stable,and the quality of the remelting metal could be more improved.This paper study the application of new control methods based on the National Science and Technology major projects which is called "high-end CNC machine tools and basic manufacturing equipment". Speaking from the process, melting speed control has certain advantages than the original current control. From the simulation we can result that the effect of iterative learning control is significant, and the melting process become more stable. Especially after joining the fuzzy control, the robustness of the system has been greatly improved. All of the control algorithms have achieved with satisfactory results.