Learning and position estimation of a mobile robot in an indoor environment using FuzzyART neural network

The perception of the local environment is a crucial issue in mobile robot position estimation. Generating a control sequence for achieving a certain goal is also important. Our approach to solving the problem of environment learning and position estimation uses percept - action - percept graphs based on ultrasound sensor readings. We use a fuzzyART (fuzzy adaptive resonance theory) neural network for categorization of the position of the robot based on acquired and preprocessed sensor data (percept). Every position is associated with a fuzzyART category. The graph is then used in the process of localization by matching the current sensory reading category with an existing node category. Our approach allows the robot to generate a set of controls to reach a desired destination based on the previous knowledge about the environment. The approach was evaluated for its ability to recognize states which the robot already visited.