OPTIMUM SHORT PATH FINDER FOR ROBOT USING Q-LEARNING

Abstract

Programming robots is a useful tedious task, so there is growing interest in building robots which can learn by themselves. This paper describes the Reinforcement Learning and teaching approach like Queue Learning (Q-Learning) to be implemented for robotics technology environment navigation and exploration. Q – Learning algorithm is one of the widely used online learning methods in robotics; it is simple, efficient, and not need to complex process as in adaptive system. The aim of this work is to empower the agent to learn a certain goal directed navigation strategy and to generate a shortest path in static environment which contain static obstacles; it uses one of the important intelligent search methods the “heuristic”. It makes a necessary modification for the search algorithm to suit the way of solving the problem. In our approach of learning from demonstration, the robot learns a reward function from the demonstration and a task model from repeated attempts (trials) to perform the task. A simplified reinforcement learning algorithm based on one-step Q-Learning that is optimized in speed and memory consumption is proposed and implemented in Visual Basic language (VB). The robot can be built using stepper motors and

any available microcontroller like 89c52 with its driver circuit to utilize of their matching.