DATA-DRIVEN ITERATIVE LEARNING CONTROL OF ANTI-WINDUP PROPORTIONAL-INTEGRAL FUZZY CONTROLLERS FOR TOWER CRANE SYSTEMS

This paper suggests the data-driven Iterative Learning Control (ILC) cost-effective anti-windup Proportional-Integral (PI) fuzzy control for tower crane systems in terms of a novel direct data-driven fuzzy control approach. The presentation is focused on payload position control. The PI fuzzy controller structure includes a back-calculation and tracking anti-windup mechanism to avoid integrator windup and compensate for the process’s dead zone and saturation nonlinearity. The lifted form representation specific to ILC is employed using fuzzy basis functions to model the fuzzy control system. The direct data-driven fuzzy control approach is based on experiments conducted on the closed-loop fuzzy control system to compute the Markov coefficients. The anti-windup PI fuzzy controller’s parameters are tuned optimally to fit the Markov coefficients in an appropriately defined optimization problem. This problem is solved using the hybrid Particle Filter-Particle Swarm Optimization algorithm, which has been improved by adding an information feedback model. Experiments and comparisons validate the suggested direct data-driven fuzzy control approach and highlight performance enhancement.