Stable control strategy for a second-order nonholonomic planar underactuated mechanical system

• Published in: International journal of systems science 2019-08, Vol.50 (11), p.2126-2141
• Main Authors: Lai, Xuzhi, Xiong, Peiyin, Wu, Min
• Format: Article
• Language: English
• Subjects: Algorithms; Angular velocity; Computer simulation; Control systems; Controllers; Convergence; Mechanical systems; nilpotent approximation; Particle swarm optimization; position control; PSO algorithm; Strategy; Underactuated mechanical system
• ISSN: 0020-7721; 1464-5319
• DOI: 10.1080/00207721.2019.1647304

This paper presents a stable control strategy for a planar four-link active-passive-active-active (APAA) underactuated mechanical system. The control objective is to move its end-point from any initial position to any target position. First, the controllers are designed to move the first link to the target angle which ensures the target position is within the reachable area of the planar three-link passive-active-active (PAA) system. Meantime, the system is reduced to a planar virtual Pendubot. Then, the periodic controllers are designed based on the nilpotent approximation model to make the first link return to the target angle and all angular velocities converge to zeroes, where the fuzzy modulating is added to adjust the convergence of angular velocities. Thus, the system is reduced to a planar virtual PAA system. Next, the control is divided into two stages, and based on the angle constraint relationships, the target angles of the planar virtual PAA system for the target position are obtained by using the online particle swarm optimisation algorithm. Each stage controllers are designed according to the target angles, so the end-point of the planar APAA system is controlled to the target position. Finally, the validity of the control strategy is demonstrated via simulations.