Improved particle swarm optimization for fractional order PID control design in robotic manipulator system: A performance analysis

This research seeks to promote the field via the design and implementation optimized robotic manipulator control systems, recognizing control techniques' vital role in current engineering applications. This study introduces an improved particle swarm optimization (IPSO) technique that maximizes the efficiency of a Fractional Order Proportional-Integral-Derivative (FOPID) controller by optimally adjusting FOPID gains in robotic manipulator systems. The controller has undergone refinement and enhancement using state-of-the-art particle swarm optimization (PSO) techniques incorporating a cost function and a representative bio-inspired algorithm. The IPSO algorithm enhances global search efficiency by preventing premature convergence and local minima trapping through chaos-based initialization and adaptive mutation strategies. The performance of IPSO-tuned FOPID controllers is benchmarked against conventional PSO-tuned FOPID controllers using various objective functions. The stabilizing fractional order PID controllers demonstrated a higher stability margin than traditional PID controllers. Numerical simulations support the developed strategy by analyzing the step and sinusoidal responses of the closed-loop system within the stability region. The results indicate that IPSO outperforms PSO with improvements of approximately 50% for 10 iterations, about 12% for 50 iterations, and around 20% for 100 iterations across ITSE, ITAE, and ITAE metrics, respectively. Furthermore the statistical analysis based on Wilcoxon sign rank test proof that the IPSO algorithm significantly improves convergence speed, controller accuracy, and overall performance, thereby enhancing the effectiveness of the IPSO technique such as in case of 10 iteration the confidence intervals do not include zeros, which indicates that IPSO outperformed the traditional POS in all scenarios.