Distributed Tracking Control of Uncertain Multiple Manipulators Under Switching Topologies Using Neural Networks
The distributed tracking control of a group of manipulators under switching directed topologies is studied. Each manipulator is modeled by the Euler-Lagrange dynamics which includes uncertainties and external disturbances. The proposed controller has the
- PDF / 274,415 Bytes
- 9 Pages / 439.37 x 666.142 pts Page_size
- 62 Downloads / 208 Views
State Key Laboratory of Management and Control for Complex Systems, Institute of Automation, Chinese Academy of Sciences, Beijing 100190, China [email protected] 2 Department of Petroleum Engineering, Harbin Institute of Petroleum, Harbin 150028, China 3 Department of Computing, Bournemouth University, Poole BH12 5BB, UK 4 School of Statistics and Mathematics, Central University of Finance and Economics, Beijing 100081, China
Abstract. The distributed tracking control of a group of manipulators under switching directed topologies is studied. Each manipulator is modeled by the Euler-Lagrange dynamics which includes uncertainties and external disturbances. The proposed controller has the neural network approximation unit for compensating uncertainties and the robust term for counteracting external disturbances. It can be proved that when the communication topology switches among a set of graphes which have a spanning tree and have no loop structure, the final tracking error can be reduced as small as possible. Keywords: Leader-following problem Neural networks · Switching topologies
1
·
Manipulator
·
Uncertainty
·
Introduction
The distributed coordination control of a group of manipulators has been studied extensively due to its potential applications in the industrial assembly and manufacturing. In the literature, many research papers have been published to investigate this topic with two control objectives: the leaderless consensus problem [1–9] and the leader-following/tracking problem [10–17]. For the leaderless consensus problem, all manipulators are required to achieve a consensus on certain value in the joint space [2–8] or the task space [1,9]. However, the leaderless consensus can only be regarded as the “self-organization behavior”. To make the whole group have some predesigned global motions, the leader-following problem should be considered, which is relatively more difficult to be solved compared to the consensus problem. And the leader-following problem becomes more challenging if the leader’s state is set as a time-varying trajectory rather than a fixed value. c Springer International Publishing Switzerland 2016 L. Cheng et al. (Eds.): ISNN 2016, LNCS 9719, pp. 233–241, 2016. DOI: 10.1007/978-3-319-40663-3 27
234
L. Cheng et al.
From the communication constraint aspect, most results in the literature assume that the communication topologies among manipulators are timeinvariant. Because of the link failure and the packet drop, this assumption cannot always hold in practice. To address this issue, few attempts have been made to study the coordination control of multiple-manipulator systems under switching topologies. In [7], the consensus of a group of manipulators with actuator faults has been considered and the communication topology can switch among a family of connected graphs. In [5,9], the consensus problem of networked manipulators is investigated and it is proved that the consensus can be achieved in the joint space [5] and in the task space [9] if the switching topologies are joint
Data Loading...
