期刊名称：Balkan Journal of Electrical & Computer Engineering
印刷版ISSN：2147-284X
出版年度：2019
卷号：7
期号：3
页码：235-244
DOI：10.17694/bajece.532746
语种：English
出版社：Kirklareli University
摘要：In mobile robotics, navigation is considered asone of the most primary tasks, which becomes more challenging during localnavigation when the environment is unknown. Therefore, the robot has to exploreutilizing the sensory information. Reinforcement learning (RL), abiologically-inspired learning paradigm, has caught the attention of many as ithas the capability to learn autonomously in an unknown environment. However,the randomized behavior of exploration, common in RL, increases computationtime and cost, hence making it less appealing for real-world scenarios. Thispaper proposes an informed-biased softmax regression (iBSR) learning processthat introduce a heuristic-based cost function to ensure faster convergence.Here, the action-selection is not considered as a random process, rather, isbased on the maximum probability function calculated using softmax regression.Through experimental simulation scenario for navigation, the strength of theproposed approach is tested and, for comparison and analysis purposes, the iBSRlearning process is evaluated against two benchmark algorithms.
