<![CDATA[This study compares two PID controller tuning methods, particle swarm optimization (PSO) and Cohen-Coon, employed for speed control of an omnidirectional Mecanum-wheel electric wheelchair. Mecanum wheels improve maneuverability on powered mobility platforms; yet, controlling these systems is difficult due to nonlinearities and directional coupling effects. This work investigates the effectiveness of PSO as a sophisticated alternative to traditional PID tuning methods, effectively tackling this issue. This study evaluates P, PI, PD, and PID controllers tuned by both Cohen-Coon and PSO methods, applied to a DC motor system simulating real-world wheelchair actuation. Step response-based system identification models the motor using MATLAB/Simulink. Simulations of a 12V DC motor are examined using controlled-step time-domain inputs. Every controller configuration is subjected to evaluation for overshoot, root mean square error (RMSE), rise time, and settling time. The PSO-tuned PID controller exhibited enhanced performance, characterized by a rise time of 2.06 s, a settling time of 2.37 s, an overshoot of 0.78%, and an RMSE of 4.59, far surpassing the Cohen-Coon variant, which had a settling time of 6.12 s and an overshoot of 20.14%. The results indicate that PSO enhances both transient and steady-state performance in intricate and disturbance-sensitive systems, including Mecanum wheelchairs. Despite PSO's increased computing complexity during offline tuning and the necessity for meticulous parameter selection, its advantages can be precomputed and effectively utilized in real-time embedded systems. This study highlights the importance of safety, dependability, and responsiveness, illustrating that PSO is a scalable and efficient method for improving assistive robotic systems.]]>
Copyrights © 2025
