Experimental Investigation of Fully Informed Particle Swarm Optimization tuned Multi Loop L-PID and NL-PID Controllers for Gantry Crane System

Abstract This paper presents a novel stabilizing nonlinear PID controller with real time experimental validation on a gantry crane system. Our adaptive control technique describes the underactuated system as a nonlinear mapping of control signal with the hyperbolic tangent function. The proposed function is used to impose a constraint on the proportional, integral and derivative errors such that the system results an effective and robust performance. The main trait of the controller is to bound the high control error to result in a lesser correction value. An enhanced hybrid method based on the bioinspired meta-heuristic algorithm Particle Swarm Optimisation is utilised to tune the gain coefficients Kpδ, KIδ and KDδ. The Linear controller and the proposed controllers were compared based on the stabilizing action on the real time system. The real time results depict that the nonlinear PID performs better than the conventional PID.