Multi-agent system motion planning under temporal logic specifications and control barrier function
- PDF / 467,963 Bytes
- 10 Pages / 595.28 x 841.89 pts (A4) Page_size
- 87 Downloads / 159 Views
Control Theory and Technology http://link.springer.com/journal/11768
Multi-agent system motion planning under temporal logic specifications and control barrier function Xinyuan HUANG, Li LI† , Jie CHEN College of Electronics and Information Engineering, Tongji University, Shanghai 201804, China; Shanghai Research Institute for Intelligent Autonomous Systems, Shanghai 201210, China Received 20 May 2020; revised 26 July 2020; accepted 29 July 2020
Abstract In this paper, we provide a novel scheme to solve the motion planning problem of multi-agent systems under high-level task specifications. First, linear temporal logic is applied to express the global task specification. Then an efficient and decentralized algorithm is proposed to decompose it into local tasks. Moreover, we use control barrier function to synthesize the local controller for each agent under the linear temporal logic motion plan with safety constraint. Finally, simulation results show the effectiveness and efficiency of our proposed scheme. Keywords: Temporal logic, multi-agent system, formal methods, control barrier function DOI https://doi.org/10.1007/s11768-020-0110-6
1
Introduction
A multi-agent system (MAS) is a collection of independent agents with the goal to achieve global or local (individual) tasks. Collaborative control deals with achieving global tasks, such as consensus [1], formation control [2], connectivity maintenance [3] and collision avoidance [4]. However, multi-agent systems still
face many practical problems with more complex objectives such as periodic, sequential, or reactive tasks, which require new methods of control and motion plans. These complex objectives can be typically formulated in formal languages based on temporal logic [5]. Therefore, ideas from the formal methods may be used, where the tasks are written as temporal logic formulas to describe sequential events, and the system dy-
† Corresponding
author. E-mail: [email protected]. Tel.: +86-21-69583907. This work was partially supported by the National Natural Science Foundation of China (No. 51475334), the National Key Research and Development Program of Science and Technology of China (No.2018YFB1305304) and the Shanghai Science and Technology Pilot Project (No. 19511132100). © 2020 South China University of Technology, Academy of Mathematics and Systems Science, CAS and Springer-Verlag GmbH Germany, part of Springer Nature
270
X. Huang et al. / Control Theory Tech, Vol. 18, No. 3, pp. 269–278, August 2020
namics are abstracted as a transition system [6, 7]. This transition system is then used to obtain a discrete plan, which is subsequently executed by continuoustime control laws. However, these abstraction-based approaches are subject to large computational burdens due to the curse of dimensionality, especially for a multiagent system. There are some literatures to reduce the computing complexity. Lindemann and Dimarogonas [8] provided a framework for controlling multi-agent systems under local signal temporal logic tasks, which improved the computa
Data Loading...
