1. 研究目的与意义（文献综述包含参考文献）

Nanjing tech university Proposal for thesis Opening report student ID: L1501150135 Graduate Name: Hussan zakir Teacher: Qinwen laoshi research direction: Formation control of multi agent system without velocity measurement Essay topic: Research on Multi-Agent System College: School of Electrical Engineering and Control Science Admission time: 2015 september 14 Opening time: 2019 july 20 Background of multi agent system More than twenty years of research in the field of agents and multi-agent systems have offered remarkable results from both theoretical and practical point of view. However, after so many years of the research there is no general consensus on some basic notions such as ”what is an agent?” What is a multi-agent systems agreement on the terminology used, etc. Indeed, this fact makes confused those interested in applying agent based or multi agent base technology to solve practical problems. This short note is intended to serve as a ”gentle” introduction to the field of agents and multi-agent systems particularly for those interested in using these technologies in solving practical engineering problems. The main purpose of the note is to familiarize readers with basic terminology and definitions. This note will not enumerate (many) and elaborate different views on the subject but rather presents a view on the subject that (according to authors of the note) may help engineering people to get feeling on this important subject. Introduction Recently, natural evolution of Multi-Agent Systems (MAS) has leaded them to migrate from the research laboratories to the software industry. This migration is good news for the entire MAS community, but it also leads to new expectations and new questions about multi-agent system methodologies, tools, platforms, reuse, specification, and so on. This introduction of Software Engineering techniques into Multi-Agent Systems gains more and more interest in both the research area and the industry. The best example is the brunch of announcements about new multi-agent platforms, usually supposed to be the ultimate tool to build multi-agent systems. Since this kind of affirmation seems far-fetched compared to the difficulty of the problem, we need some elements to evaluate and to compare multi-agent platform we propose an analysis grid, built from criteria that evaluate each of the stages encountered during the creation of multi-agent systems. Of course, as in benchmarks, any criteria is relevant to a specific outside need, and a platform can only be compared relatively to another one because there is nothing like an absolute measurement scale for multi-agent platforms. Multi consistency agent as a basic control of the focal points, the system design for a simple control protocol, such as the agent between the system to each other, relying on the law of interaction, such that some of the final amount of the system reaches a consistent or synchronized. These quantities can be the formation between cars the final destination of bird migration the formation between drones, and so on Because of the wide range of fields involved in the consistency, the direction is wide which makes people more enthusiastic about research. It will not only raise awareness for the control and use of intelligent body, can promote the development of more practical applications. This short note is intended to serve as a ”gentle” introduction to the field of agents and multi-agent systems particularly for those interested in using these technologies in solving practical engineering problems. The main purpose of the note is to familiarize readers with basic terminology and definitions. This note will not enumerate (many) and elaborate different views on the subject but rather presents a view on the subject that (according to authors of the note) may help engineering people to get feeling on this important subject. Throughout of this note, when deemed necessary, simple examples are provided to illustrate different concepts. All the examples are based on the idea to use agent or multi-agent technology to design effective load shedding scheme. For readers interested for a deeper inside in the subject. Research status The research of multi agent system dated back to 1980.time by time there is the big reveal in the field of multi agent system which is well described by the Olfati-Saber, R.， Fax, J.A.， Murray, R.M. They described the frame work for consensus algorithms for multi agent system with directed information flow, robustness, time delay and performance guarantees the method they used to do this is matrix theory, algebraic graph theory and control theory and the end as the result they establish direct connection between spectral and structural properties of complex networks and the speed of information diffusion. A introduction was provided by him on nonlocal information flow that are considerable faster than distributed system with lattice-type nearest neighbor interactions. [1] The basic idea about distributed consensus in multi-vehicle cooperative control each vehicle update its information state based on information states of its local(possibly time-varying}neighbors in such a way that final information state of each vehicle converge to a common value the common theme was to coordinate the movement of multiple vehicles in a certain way to accomplish an object. The potential application for autonomous vehicle are usually in civil and commercial, military and mainland security. [2] Some dynamic neighbor-based rules, consisting of distributed controllers and observers for the autonomous agents, are developed to keep updating the information of the leader. With the help of an explicitly constructed common Lyapunov function it is proved that each agent can follow the active leader Distributed estimation via observers design for multi-agent coordination is an important topic in the study of multi-agent networks, with wide applications especially in sensor networks and robot networks, among many others. Yet, very few theoretic results have been obtained to date on distributed observers design and measurement-based dynamic neighbor-based control designs. A multi-agent network provides an excellent model for describing and analyzing complex interconnecting behaviors, with applications in many disciplines of physics, biology, and engineering [3] Hong Y, Hu J, Gao L With this background, we consider a consensus problem with an active leader with an under dynamics. Here, some variables (that is, the velocity and maybe the acceleration) of an active leader cannot be measured, and each agent only gets the measured information (that is, the position) of the leader once there is a connection between them. To track such a leader, a neighbor-based local controller together with a neighbor-based state-estimation rule is given for each autonomous agent. Then we prove that, with the proposed control scheme, each agent can follow the leader if the (acceleration) input of the active leader is known. If the information of the input a(t) can be used in local control design, we can prove that all the agents can follow the leader, though the leader keeps changing. If not, we can also get some estimation of the tracking error. [4] Their research based on the formation control of multiple nonholonomic mobile robots. They did the main researched on nonholonomic mobile robot, formation control and kinematics model. And the advantage of this model leads to the derivation of a controller free of possible singularity points that happen when using polar coordinate representations. And as a results simulation results are included in to demonstrate the feasibility of the presented model and controller. [5] A new distributed observer-type consensus protocol designed based only on the relative output measurements of neighboring agents is introduced. Compared with some existing observer-type protocols, the protocol presented in this paper without involving the relative states of observers. Finally, some numerical simulations are performed to show the effectiveness of the theoretical results. [6] Distributed reduced-order observer-based consensus protocols are proposed, based on the relative outputs of neighboring agents. A multi-step algorithm is presented to construct a reduced-order protocol, under which a continuous-time multi-agent system whose communication topology contains a directed spanning tree can reach consensus. This algorithm is further modified to achieve consensus with a prescribed convergence rate all they deal with the consensus, reduced-order observer and the convergences rate the consensus problems for multi-agent systems with continuous- and discrete-time linear dynamics and directed Cooperative control of a group of agents has received compelling attention from various scientific communities. A group of autonomous agents can coordinate with each other via communication or sensing networks to perform certain challenging tasks, which cannot be well accomplished by a single agent. Its potential applications include spacecraft formation flying, sensor networks, and cooperative surveillance. In the area of cooperative control of multi-agent systems, consensus is an important and fundamental problem, which is closely related to formation control and flocking problems. The main idea of consensus is to develop distributed control policies that enable a group of agents to reach an agreement on certain quantities of interest. [7-8] References [1] Olfati-Saber R, Fax A, Murray R M. Consensus and cooperation in networked multi-agent systems[J]. Proceedings of the IEEE, 2007, 95(1): 215-233. [2] Ren W, Beard R W. Distributed consensus in multi-vehicle cooperative control[M]. Springer-Verlag, London, 2008. [3] Hong Y, Chen G, Bushnell L. Distributed observers design for leader-following control of multi-agent networks[J]. Automatica, 2008, 44(3): 846-850. [4] Hong Y, Hu J, Gao L. Tracking control for multi-agent consensus with an active leader and variable topology[J]. Automatica, 2006, 42(7): 1177-1182. [5] Li Xiaohai, Xiao Jizhong, Tan Jindong. Modeling and Controller Design for Multiple Mobile Robots Formation Control[C]//Proceedings of IEEE International Conference on Robotics and Biomimetics. Shenyang, China: 2004: 838-843. [6] Zhao Y, Wen G, Duan Z, et al. A new observer‐type consensus protocol for linear multi‐agent dynamical systems[J]. Asian Journal of Control, 2013, 15(2): 571-582. [7] Li Z, Liu X, Lin P, et al. Consensus of linear multi-agent systems with reduced-order observer-based protocols[J]. Systems Control Letters, 2011, 60(7): 510-516. [8] Li Z, Liu X, Lin P, et al. Consensus of linear multi-agent systems with reduced-order observer-based protocols[J]. Systems Control Letters, 2011, 60(7): 510-516.

2. 研究的基本内容、问题解决措施及方案

Content of the research The consensus protocol of first-order multi-agent systems over fixed directed network is designed. Each agent only communicates with its neighboring agents at some specific conditions. under the Lyapunov function, we get the sufficient condition of agents achieving consensus. On the basis of the first order system we determined the second order multi-agent systems, and sufficient condition of consensus are obtained We study the multi agent systems consist of leader is the second order and follower is the first order. Assume that the velocity of the leader can not be measured, we design observe based formation control protocol to let the follower to keep doing the consensus with leader under the fixed topology. Under the basics of fixed topology we studied the multi-agent system with the switching topology. As the result we get the sufficient condition of consensus under the fixed and switching topology. From here we see the leader is the second order and follower is also the second order multi agent system. Under the assumption that the leader#8217;s velocity cannot be obtained, we came to observed the formation control and let the follower to do consensus with the leader by using the Lyapunov function. Under the basics of fixed topology we studied the multi-agent system with the switching topology, . As the result we get the sufficient condition of consensus under the fixed and switching topology Basic of the research Knowledge of the graph theory. Let G = {V,E,A} be a weighted directed graph of a multiagent system containing N orders, where V= {v_1, v_2, #8230; , v_n}is the node set, E #8838; {(v_i, v_j) ∶ vi, v#119895; ∈ V} is the edge set, and A = a_ij ∈ R^NNis the weighted adjacency matrix An edge in G is denoted by v_ij = (v_i, v_j)(i ≠ j). The adjacency elements in satisfy that a_ji ≠ 0 if and only if qi#119895; ∈ A, and a_ij = 0 otherwise. The in-degree of node vi is defined as degin(vi) =∑_(J=1)^N#9618;a_ij The agents can be divided into leaders and followers in the multiagent system.R^NN Lyapunov function Matlab