2018/03/30 by Majid Mazouchi, Mazouchi, Majid, Mohammad Bagher Naghibi Sistani +7
Computer Science · #Adaptive Dynamic Programming Control #Distributed Control Multi-Agent Systems #FOS: Computer and information sciences #FOS: Electrical engineering #Machine Learning (cs.LG) #Multiagent Systems (cs.MA) #Reinforcement Learning in Robotics #Systems and Control (eess.SY) #electronic engineering #information engineering
paper · pdf · doi:10.48550/arxiv.1803.11411
openalex publication_date 2018/03/30 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28
This paper develops an optimal relative output-feedback based solution to the\ncontainment control problem of linear heterogeneous multi-agent systems. A\ndistributed optimal control protocol is presented for the followers to not only\nassure that their outputs fall into the convex hull of the leaders' output\n(i.e., the desired or safe region), but also optimizes their transient\nperformance. The proposed optimal control solution is composed of a feedback\npart, depending of the followers' state, and a feed-forward part, depending on\nthe convex hull of the leaders' state. To comply with most real-world\napplications, the feedback and feed-forward states are assumed to be\nunavailable and are estimated using two distributed observers. That is, since\nthe followers cannot directly sense their absolute states, a distributed\nobserver is designed that uses only relative output measurements with respect\nto their neighbors (measured for example by using range sensors in robotic) and\nthe information which is broadcasted by their neighbors to estimate their\nstates. Moreover, another adaptive distributed observer is designed that uses\nexchange of information between followers over a communication network to\nestimate the convex hull of the leaders' state. The proposed observer relaxes\nthe restrictive requirement of knowing the complete knowledge of the leaders'\ndynamics by all followers. An off-policy reinforcement learning algorithm on an\nactor-critic structure is next developed to solve the optimal containment\ncontrol problem online, using relative output measurements and without\nrequirement of knowing the leaders' dynamics by all followers. Finally, the\ntheoretical results are verified by numerical simulations.\n