High precision experimentally validated adaptive neuro fuzzy inference system controller for DC motor drive system.

Area of Science:

• Electrical Engineering
• Control Systems
• Artificial Intelligence

Background:

• DC motor drives commonly use Proportional-Integral (PI) controllers for speed and torque regulation.
• Tuning PI controllers presents challenges, often resulting in significant overshoot and long settling times.
• Eliminating overshoot without compromising speed in DC motor control is a persistent issue.

Purpose of the Study:

• To present an experimental implementation of Adaptive Neuro-Fuzzy Inference System (ANFIS) controllers for high-precision DC motor drive control.
• To eliminate overshoot and simultaneously improve settling time in DC motor speed and torque control.
• To compare the performance of ANFIS controllers against traditionally tuned PI controllers.

Main Methods:

• Utilized a 2-quadrant DC motor drive experimental setup with a dSPACE DS1104 R&D controller board.
• Employed the Bode Plot method for optimal tuning of Proportional-Integral (PI) controller gains (Kp and Ki).
• Designed, implemented, and validated Adaptive Neuro-Fuzzy Inference System (ANFIS) controllers through both simulation and experimental testing.

Main Results:

• ANFIS controllers successfully eliminated overshoot, achieving 0% in experimental results.
• ANFIS controllers significantly improved settling time, reducing it to 0.18 seconds.
• Experimental and simulation results demonstrated that ANFIS controllers outperform tuned PI controllers.

Conclusions:

• Adaptive Neuro-Fuzzy Inference System (ANFIS) controllers offer a superior solution for DC motor drive control compared to PI controllers.
• ANFIS-based control effectively addresses the limitations of PI controllers, namely overshoot and settling time.
• The proposed ANFIS controllers provide high-precision control for DC motor drives, validated through rigorous simulation and experimentation.