Ferromanganese Smelting Electric Arc Furnace Electrode Control System Simulation Studies

Electric arc furnaces converted electric power into heat energy, thus to melting metal. The main parameters of electric arc furnaces were transformer design capacity, long-term allow overload capacity, common secondary side voltage, the design of furnace parameters and design parameters. The smelting furnace was volatility, therefore in a constant drain on the operation of the core question was how to controlling three-phase electrodes, controlling arc length, and then control arc current.Electrode control system consisted of rectifier link, pressure regulating circuit, ac torque motors, arc links etc, because arc length and current was nonlinear, and the whole system was non-linear. In smelting process, electric arc furnaces operation parameters was changing with time, and the uncertainty of interference, and three current was coupling, electric arc furnace to run on a stable hard working nearby. Especially in smelting furnace melting period, the changes were very intense; the traditional PID control was hard to meet control requirements, the control law for a reasonable to solve this complex nonlinear time-varying object was one of the difficulties in the control theory. And in the actual cases, because of large current, high temperature, accurate measurement data was difficult to obtain, the accurate control for the system increased difficulty. The control research at electrode was:expert control and fuzzy control, neural network control and adaptive control, etc.Based on electric arc furnace of femn smelting process and electrical parts, analyzed the electrode control system of each section of model. discussed the nonlinear characteristics arc link, and arc current increased with the arc length. Compared some measuring method of arc current and arc length. Using BP neural network to nonlinear object approximation ability identified the model and inverse model of the arc link, and designed the adaptive inverse controller and generalized predictive controller. through without interference and interference simulation. The results showed that generalized predictive control had very good tracking ability and adaptive inverse control had better anti-interference ability.