Active filter design and synthesis for hybrid neuro-fuzzy and robust PID controllers

To achieve robust optimal performance in the presence of control signal range limitation, the combination of controllers is suitable. Several objectives are considered in an actual system, such as stability and robust performance, optimal time characteristics, minimizing noise effects, and limiting control signal range. Neuro-fuzzy controllers are primarily used in the transient state to provide a fast dynamic response and minimize the energy of the impulse response. While the robust PID controllers are in a steady state to reduce disturbance and take care of the tracking, the robustness guarantees the system's stability. This paper's neuro-fuzzy and robust PID hybrid controllers are designed according to the optimal objectives. Then, considering that each controller has an optimal response in a range of frequencies, they are combined by active filters. Usually, each of the mentioned controllers has superiority in time and frequency response, the combination of controllers is of interest. Three approaches for the design of filters have been presented, and their efficiency against uncertainty and disturbance on aircraft pitch control has been examined. Finally, filter synthesis has been done. The simulation results show that the proposed controller has reduced overshoot than the robust PID controller. The neuro-fuzzy controller has a steady state error of 0.12 while the proposed controller's steady state error is zero. Also, the proposed controller is stable against disturbance and uncertainty and has better performance than robust PID controllers and neuro-fuzzy.