Development of learning control algorithms for mechatronics systems

Abstract

In this research work, several iterative learning control (ILC)algorithms have been developed for mechatronics systems. These ILCalgorithms are classified into three types of schemes: previouscycle learning scheme (PCL), current cycle learning scheme (CCL)and previous and current cycle learning scheme (PCCL). Thesethree learning algorithms have been analyzed systematically.Their convergence properties have been discussed correspondingly.To verify the analytical results, several learning schemes aredeveloped for certain practical systems, such as hard disk driveservo, ball-and-beam test bed and a model of a microrobot.Extensive simulations and experimental results show that thelearning schemes developed achieve good performance.