A new master/slave model driven, and an optimization algorithm-based proportional–integral–derivative (PID) plus second-order derivative (PIDD2) controller is proposed in this work for a stable and efficient operation of an automatic voltage regulator (AVR) system. In this context, an ideal reference model of Bode is used as a master model. The new improved optimization algorithm is constructed via integrating the Lévy flight mechanism into Runge–Kutta optimizing algorithm. This algorithm optimally tunes the PIDD2 controller with the aid of a cost function known as integral of squared error. The latter control mechanism forms the slave model. As the PIDD2 controller and the intelligent tuning algorithm attempt to follow the response dictated by the ideal reference model of the master model, a significant improvement is achieved for the efficiency and the stability of the AVR system. The proposed master/slave driven, and intelligent optimization algorithm-based PIDD2 control approach presents more excellent transient response (steady state error, rise time, settling time, peak time, percent overshoot), frequency response (gain margin, phase margin and bandwidth), robustness and stability. Nonideal conditions such as measurement noise and the saturation at the input of the generator in the AVR are also considered to demonstrate the efficacy of the proposed method. Furthermore, the existing fifty-eight techniques in the literature are also used for performance comparison in order to present the more excellent efficiency of the proposed method from a wider perspective.