Simulation-Based Evaluations for Maze Navigation and Wall-following Behavior of the Robot using Reinforcement Learning Algorithms