Comparative Analysis of A* Algorithm and Improved Q Learning Algorithm for an Optimal Path Planning of Autonomous Mobile Robot

Path planning is a critical task in robotics, where efficient navigation through static grid-based environments is essential. This study compares two popular approaches-Qlearning and the <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\mathrm{A}^{*}$</tex> algorithm-by evaluating their path quality, learning speed, and computational efficiency. <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$A^{*}$</tex>, a heuristic search algorithm, performs well in structured grids but struggles with large or complex environments due to high computational demands. In contrast, Q-learning, a reinforcement learning method, adapts to various grid patterns through experience, enabling efficient navigation with lower resource usage. Simulation results show that Q-learning converges faster and performs better in complex grids, while <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$A^{*}$</tex> remains effective in simple, well-structured layouts but lacks adaptability. This study offers valuable insights into the strengths and limitations of both methods, contributing to advancements in path planning for robotics and autonomous systems.