فهرست نشریات علمی دسترسی آزاد ایران

This paper presents an adaptive inverse dynamic control (AIDC) designed as a fault-tolerant controller for nonlinear models of coreless induction furnace systems. Unlike other research focused on intelligent identification of nonlinear systems, this methodology developed an inverse intelligent process as a universal controller applicable to various nonlinear systems. In induction furnace operations, accurately tracking the target reference temperature is critical for maintaining the crystalline properties of metals; for instance, to produce iron ferrite with 12.22 wt% carbon, the temperature must be held at 912°C. The proposed AIDC approach features an online inverse model identifier, updated using the back-propagation (BP) algorithm. This involves three techniques: 1) multilayer perceptron (MLP), 2) adaptive neuro-fuzzy inference system (ANFIS), and 3) neural networks, all used to identify the system's inverse dynamics as a nonlinear controller. Key benefits of the AIDC include the convergence of faulty states to nominal conditions, robust system design, and reduced impact of faults on system performance. Simulation results demonstrate the effectiveness of this approach.