This thesis presents and evaluates a decision making framework that enables autonomous agents to dynamically select the mechanism they employ in order to coordinate their inter-related activities. Adopting this framework means coordination mechanisms move from the realm of something that is imposed upon the system at design time, to something that the agents select at run-time in order to fit their prevailing circumstances and their current coordination needs. Using this framework, agents make informed choices about when and how to coordinate, when to respond to requests for coordination and when it is profitable to drop contracts in order to exploit better opportunities. The framework's efficacy is investigated in a range of coordination situations; such as scenarios in which agents have varying dispositions to work cooperatively, in which agents make decisions based on alternative assumptions about their environment, and in which the task of predicting other agents' behaviour varies in difficulty. In all of these cases, a systematic empirical analysis is undertaken to evaluate the effect of such situations on the agent's decision making mechanisms and their effectivene.
展开▼
