Research On The Method And System Of Electrode Control For AC Ladle Furnace

The method of Ladle Furnace is one mode of dominating refine. But, the electrode's control system of Ladle Furnace has the characters of non-linearity, multi-variable and strong couple etc. At present, traditional PID is still adopted in the electrode's control system, the effect isn't well. It results in the imbalance of three-phase electrode power, the increase of power consumption and the reduction of productivity. the capability of the electrode's control system influence efficiency of Ladle Furnace. So, It's important to develop automatic adjuster with high performance in order to increase the quality of production and recede consume of electric energy.The paper is based on the electrode's control system of the Ladle Furnace of No.2 Steeling Plant of Tangshan Iron and Steel. According to the control demands on the electrode's control system of Ladle Furnace, the paper does research on control policy of the electrode's control system, confirms control policy of invariable impedance.In this paper ,the model reference self-adaptive control is adopted. Adopting to the theory of piece wise-linear, mathematics model is constituted and model reference self-adaptive algorithm is derived. Furthermore, the central process of the control is designed. The results of the experiments show that the model reference self-adaptive control can overcome the influence caused by uncertainties of the plant and has very strong self-adaptive and is robust.Although fuzzy logic system can express knowledge easily and has strong fuzzy logic inference ability, it is difficult in self-learning. Artificial neural networks have particular superiority in self- learning and function approximation ability. Based on the integration of fuzzy logic system and neural networks the thesis constructs a fuzzy neural network controller, which is used in the electrode's self-learning adaptive control system. Because the fuzzy logic system is realized in neural networks, the fuzzy neural networks controller's self-learning ability is improved greatly. It is applied to the control of industrial environment. Therefore, it is adaptive and robust and the arc direction also can ensure the stability through the closed loop control system.