Memory-based robust adaptive control of a variable length stepping nanomanipulator

摘要

This paper presents the modeling and memory-based robust adaptive control of a variable length stepping nanomanipulator. A three degree of freedom (3DOF) nanomanipulator with revolute revolute prismatic (RRP) actuator structure, namely here MM3A, is utilized for a variety of nanomanipulation tasks. Unlike widely used Cartesian-structure nanomanipulators, the MM3A is equipped with revolute-piezoelectric actuators which result in outstanding performance for controlling the nanomanipulator's tip alignment during the nanomanipulation process. However, the RRP structure of the nanomanipulator introduces complicity in kinematic and dynamic equations of the system which needs to be addressed in order to control the nanomanipulation process. Dissimilar to the ordinary piezoelectric actuators which provide only a couple of micrometers working range, the piezoelectric actuators utilized in MM3A, namely Nanomotors~R, provide wide range of action (120° in revolute actuators and 12mm in prismatic actuator) with sub-nano scale precision (0.1 urad in revolute actuators and 0.25 nm in prismatic actuator). This wide range of action combined with sub-nano scale precision is achieved using a special stick/slip moving principle of the Nanomotors~R. However, such stick/slip motion results in stepping movement of the MM3A. Hence, due to the RRP structure and stepping movement principle of the MM3A nanomanipulator, controller design for the nanomanipulation process is not a trivial task. In this paper, a novel memory-based robust adaptive controller is proposed to overcome these shortfalls. Following the development of the memory-based robust adaptive controller, numerical simulations of the proposed controller are preformed to demonstrate the positioning performance capability of the controller in nanomanipulation tasks.