Multi-agent dynamic systems have received growing attention in the last decade. The main objective of such systems is to optimize the use of group resources by benefiting from synergy, while using local information for control. Examples of such systems are groups of unmanned vehicles, automated highway systems, and complex distributed manufacturing processes. This is an emerging area of research with numerous applications across many disciplines. As is evident from the number of papers in recent control conferences, this is one of the main research topics in the field. One reason for the explosion of research in this area is the availability of inexpensive and advanced hardware technology (sensors and actuators) and ease of communicating between systems by efficient wireless technology. The main challenges associated with the analysis and development of control algorithms for multi-agent dynamic systems are: the computational complexity of solving large sized cooperative problems with coupled tasks, executing tasks under limited information, and obtaining scalable solutions.
In the preparation of the special issue, our goal was to solicit recent, and exploratory, research on this topic. We believe a subset of papers in this area reflective of current research in the field was assembled. This Special Issue, consisting of 17 papers (15 regular papers and 2 technical briefs), presents recent research that can be roughly divided into the following topic areas: (i) Task Assignment, (ii) Path Planning, (iii) Formation Control, and (iv) Cooperative Control.
In the first paper entitled “Recent Research in Cooperative Control of Multi-Vehicle Systems,” Murray presents a survey of the research field. It first describes driving applications of cooperative control; then relevant technology that has been developed over the past decade is presented; and finally some future research directions are outlined.
Task Assignment
Key issues pertaining to the problem of allocating a group of vehicles to tasks is addressed in this set of three articles. The paper “Autonomous Vehicle-Target Assignment: A Game Theoretical Formulation” by Arslan and Shamma introduces a game theoretical formulation of the vehicle-target assignment problem in which the vehicles are viewed as self-interested decision makers. Optimization is performed by each vehicle making individually rational decisions to optimize their own utility functions that are aligned with a global utility function. Finke and Passino present a mathematical model for the behavior study of a spatially distributed group of heterogeneous vehicles in the paper entitled “Stable Cooperative Vehicle Distributions.” Based on the mathematical model and underlying conditions for achieving a stable distribution, a cooperative control scheme for multi-vehicle surveillance is presented. In the paper entitled “UAV Team Decision and Control Using Efficient Collaborative Estimation” Shima et al. present a decision-estimation methodology allowing a team of agents cooperating under communication imperfections to perform multiple tasks on multiple ground targets. For the path optimization Dubins trajectories are used and efficient algorithms are presented for the estimation process computation and communication.
Path Planning
The selected papers in this topic area address the cooperative path planning for a group of agents. In the paper “Differential Geometric Path Planning of Multiple UAVs” by Shanmugavel et al. Pythagorean Hodograph curves are used to obtain smooth trajectories for multiple vehicles using continuous lateral control. It is shown that using this method simultaneous arrival on a target can be obtained while ensuring collision avoidance. Inspired by cyclic pursuit Pavone and Frazzoli present in the paper “Decentralized Policies for Geometric Pattern Formation and Path Coverage” a decentralized control policy for symmetric formations in multi-agent systems and study the formation stability. A strategy to make the agents totally anonymous and a potential application to coverage path planning are discussed. In the technical brief “Control of Multi-Agent Systems Using Linear Cyclic Pursuit With Heterogenous Controller Gains” Sinha and Ghose study the behavior of a group of autonomous mobile agents under cyclic pursuit. It is shown that by selecting suitable controller gains, collective behavior of agents can be controlled to obtain not only point convergence but also directed motion.
Formation Control
The subset of five papers under this title focuses on formation of a group of vehicles. Different control strategies which address specific formation problems are discussed. In the paper “A Composite Model for Vehicle Formation and Path Selection on a Cellular Structured Map” Zheng and Ozguner propose a new framework for multi-vehicle system modeling and control which emphasizes team behavior in a multi-level, multi-resolution manner. In the paper “Formation of a Group of Vehicles With Full Information Using Constraint Forces” Zou et al. present a formation control strategy to achieve, and maintain, a formation of a group of vehicles using the notion of constraint forces. In the paper “Passive Decomposition Approach to Formation and Maneuver Control of Multiple Agents with Inertias” Lee and Li present a formation control strategy via passive decomposition of a nonlinear group dynamics into two decoupled systems. By controlling these systems separately and individually it is shown that precise maneuver and formation controls can be achieved simultaneously without any crosstalk between them. In the paper “High-Order and Model Reference Consensus Algorithms in Cooperative Control of Multi-Vehicle Systems” Ren et al. study high-order consensus algorithms and present their effectiveness through simulations of a multi-vehicle cooperative control application which mimics the flocking behavior in birds. Control of deep-space aircraft formation flying using the virtual structure approach and the theta-D suboptimal control technique was investigated in the paper “Position and Attitude Control of Deep-Space Spacecraft Formation Flying via Virtual Structure and Theta-D Technique” by Xin et al.
Cooperative Control
The five papers under this topic investigate group behavior and cooperative strategies for different types of applications. In the paper “Cooperative Avoidance Control of Multi-Agent Systems” Stipanovic et al. propose a cooperative avoidance control law which guarantees collision avoidance in multi-agent systems; an optimization scheme that ensures avoidance laws are active only in the bounded sensing regions of each agent was given. Yao et al. in the technical brief entitled “Swarm Tracking Using Artificial Potentials and Sliding Mode Control” present a stable decentralized control strategy for multiple agents to capture a moving target. The paper “Phantom Track Generation Through Cooperative Control of Multiple ECAVS Based on Feasibility Analysis” by Maithripala et al. presents a framework to derive sufficient conditions for the existence of feasible solutions for an affine nonlinear control system comprising of a team of nonholonomic mobile agents having to satisfy actuator and inter-agent constraints. Yang et al. present in “Multi-UAV Cooperative Search Using an Opportunistic Learning Method” a framework for on-line planning and control of a group of UAVs for cooperative search. Decentralized coordination of two hydraulic actuators using reinforcement learning was investigated by Karpenko et al. in the paper “Decentralized Coordinated Motion Control of Two Hydraulic Actuators Handling a Common Object.”
Although this special issue does not exhaustively cover the ongoing research in multi-agent dynamic systems, we believe that it reflects a representative sample of recent research in this area by key researchers. The authors are affiliated with many professional societies, primarily from ASME, IEEE, and AIAA. We hope that this special issue will be beneficial to the readers of this journal as well as control researchers in this exciting field in various societies.
As guest editors, we would like to thank the authors for submitting their work, the reviewers for providing timely evaluation of papers, and the Editor Professor Suhada Jayasuriya and editorial staff of the journal for their vision and support.
