Handling multiple intentions using action heuristics

The aim of this paper is to describe our new results that increase AgentSpeak(L)-based agent flexibility of decision making process. Namely we introduce several Algorithms and program constructions that allow us to implement non-trivial intention and substitution selection. Intention selection is based on a heuristics that estimates possible plan execution, based on current agent's state, and then we search for the best action consensus between particular intentions, which are then selected for execution. Substitution selection is based on searching biggest consensus of intention environments to achieve largest possible set of simultaneously covered actions. We also present a case study that shows benefits of our approach.