Behavior-Based Control with Learning from Demonstration for Path Following Applied to Mobile Robots Soccer

Abstract

Learning from Demonstration (LfD) enables robots to learn new tasks through an imitation process. This methodology can be combined with Behavior-Based Control (BBC), which facilitates the abstraction of behaviors for the robot by dividing them into micro-behaviors. These can be implemented using Artificial Neural Networks trained by bio-inspired algorithms. In this work, a novel methodology based on the LfD and BBC methods is proposed for controlling soccer robots, specifically for the Very Small Size Soccer (VSSS) category, which demands quick and dynamic responses. Simulations using CoppeliaSim were conducted to demonstrate how the proposed methodology can be used for path planning and path-following problems in the IEEE VSSS category.