Modelo de razonamiento basado en creencias, deseos e intenciones para la toma de decisiones en un algoritmo de planificación de trayectorias

Abstract

This work focused on improving the reactive technique known as Artificial Potential Fields, widely used in navigation trajectory planning, because this method is analytical, simple and easy to implement. In more complex environments, this method does not provide an optimal solution as it can lead the robot to become trapped in local minima, preventing it from reaching the target. To address this problem, a 2D navigation strategy was developed that integrates the decision making model based on practical reasoning for BDI (Beliefs, Desires and Intentions) intelligent agents. This methodology allows the robot to avoid local minima by optimizing repulsive potential fields. An approximate analysis is performed for the deterministic version of the selected model, evaluating both qualitative and quantitative results.