Distributed control of multi-agent systems using extremum seeking
MetadataShow full item record
A model free control technique (extremum seeking) is employed to address problems of large-scale systems involving multi-agents in real-time. This thesis focuses on the use of extremum seeking control in a distributed, coordinated and a cooperative fashion to solve distributed control and optimization problems. First, the problem of maximizing the power produced in a wind farm is considered. To tackle this problem, a distributed time-varying extremum seeking control (TVESC) technique is employed to overcome the need to provide accurate models of aerodynamic wake interactions among the wind turbines. Solutions in continuous-time and discrete-time are presented. Secondly, precise knowledge of the structure of network connectivity has been utilized in solving cooperative optimization problems of multi-agent systems to achieve global objectives. In this thesis, a distributed proportional-integral extremum seeking control technique is designed to tackle such problems over unknown networks.