A computer simulator for steel plant electrical arc furnace regulator

The function of the simulator is to imitate the behavior of the regulator loop, which is the main component of the Electrical Arc Furnace (EAF) control systems. In the past, the use of artificial intelligence methods, and in particular, the Adaptive Neuro Fuzzy Inference System (ANFIS) were successfully applied in the modeling and control of the EAF components individually. This research expands the use of ANFIS in building the full closed loop computer simulator for the three-phase regulator loop. THe ANFIS models inuts and outpus selected for this project were tried for the first time in this research. The simulator components were trained and verified by the use of plant recorded data in the open loop mode. The response of the closed loop simulator was tuned to follow the behavior of the plant EAF. Therefore the simulator works independent of the plant data or operation commands. The developed simulator, then, was used to measure the results of applying new controls in EAF such as fuzzy controllers, without disturbing the actual plant process.