Several Distributed Consensus Problems Of Multi-Agent Cooperative Systems With Different Aspects

The concept of consensus problem in multi-agent coordination possesses an important theoretical significance and practical application. This dissertation includes several consen--sus problems related to multi-agent coordination which are solved with the appropriate methods.Two types of coordination are proposed to address multi-agent communication issues. One is coordination via variable elimination, in which agents are eliminated one after another by a sequential maximization procedure. The other is coordination by message passing, in which neighborhood agents in the graph send messages to each other repeatedly, where message is a local payoff function for the receiving agent and messages will exchanged under the condition that they converge to some specific point, or some external signal terminate the process. Finally, the algorithms are proposed to obtain the best optimum response function for both types of coordination approaches.The ‘formation’ is a key issue in a multi-agents consistency system which is solved here by proposing the vehicle formation tracking controller on the basis of cascaded system theory. A novel consensus controller based vehicle formation is proposed. And simulations of various formations of vehicles are examined in MATLAB.The input and communication delays generally appeared to be the main issues in the way of solving multi-agent coordination problems to achieve the consensus. Therefore this particular research has been focused to address the problems of delays and a state predictor feedback approach has been developed to handle the same. The aforesaid problem might have been addressed in multiple solutions but this study has been focused to introduce a state buffer between the communication links and subsequently to obtain an improved and faster consensus.Moreover two distributive adaptive protocols have been proposed to address any distance errors existing among the agent’s relative output information. One of the said adaptive protocol which assigns a time varying coupling weight to each edge in the communication graph, whereas the other adaptive protocol which uses a time varying coupling weight for each node. Both the protocols are designed to ensure that consensus is achieved without using any global information. Several examples illustrate the above sufficient conditions effectively.