Integrating Learning into a BDI Agent for Environments with Changing Dynamics
Dhirendra Singh, Sebastian Sardina, Lin Padgham and Geoff James
In this paper we present a framework for integrating learning into the popular and robust BDI agent programming paradigm. We summarise previous work which has integrated decision trees into the context condition used for plan selection, and then develop a confidence measure which allows the agent to adjust its reliance on the decision tree dynamically, facilitating both initial learning and re-learning. We evaluate this with an example of an embedded controller for energy management.