Design and Optimization of Controller-Based Approach for Magnetic-Field Driven Robotic Arm Joints and End-Effector

Abstract

Magnetic Levitation (Maglev) is a technique that involves suspending an object using a magnetic field. This study presents a novel approach for robotic arm joints and end effectors by utilizing the functioning prototype of the Maglev system due to their similar functionality. The proposed approach utilizes a fractional-order enhanced model reference adaptive controller (FOEMRAC) in conjunction with the Coyote optimization algorithm (COA) to control the stability of levitating magnetic objects. The FOEMRAC system employs a modified MIT rule as its adaptation mechanism. The simulation is performed using the Quanser Maglev system, and a comparison is done with other state-of-the-art techniques such as linear quadratic regulator (LQR), particle swarm optimization-LQR (PSO-LQR), LQR + proportional integral (LQR + PI), LQR + proportional integral derivative (LQR + PID), proportional integral voltage + PI (PIV + PI), enhanced model reference adaptive controller (EMRAC), FOEMRAC, and PSO-FOEMRAC, respectively. The robustness of the controllers is assessed using various integral error criteria, such as integral absolute error (IAE), integral square error (ISE), and integral time absolute error (ITAE), respectively. Additionally, rise time, settling time, overshoot, and undershoot have been employed for comparison purposes with load disturbance and parametric uncertainties. The results are also validated on real-time hardware, demonstrating the superior performance of COA-FOEMRAC as compared to various controllers. Thus, it can be effectively employed to improve the functionality of the magnetic joints and magnetic end effectors in real-time applications. A video demonstrating the functioning of the Maglev system is available at this link: https://drive.google.com/file/d/1FrD1YKqRXSTTe44S2-ap126KljiVmOEU/view?usp=drivesdk.