A fuzzy perceptual model for map building and navigation of mobile robots

In the area of the intelligent mobile robots the hybrid reactive-deliberative architectures for navigation are aimed at an efficient integration of reactive and deliberative skills. Reactive skills allow the robot a fast reaction to unexpected events whereas deliberative skills permit the generation of plans to carry out tasks. In these systems a high level of accuracy in the modeling of the environment is not usually necessary but some representation structure is needed to generate safe paths for the robot navigation. Thus, the perception of the environment and the construction of useful models are two main problems to deal with. In this work, we propose a fuzzy perceptual model and a map building process which would allow us to build a world topological map giving us the possibility for reasoning and planning about the robot motion in the world. The perceptual model deals with the uncertainty and vagueness underlying the sensor data, it carries out the data fusion from different sensors and it allows us to establish various levels of interpretation in the sensor data. The topological map is used to generate high-level abstraction paths and then the navigation is carried out using our own hybrid architecture and taking into account the perceptual model to represent the robot's beliefs about the world. Experiments in simulation and in real office-like environments are shown for validating the proposal.