Mobile Robot Path Optimization Technique Based on Reinforcement Learning Algorithm in Warehouse Environment

Abstract

This paper reports on the use of reinforcement learning technology for optimizing mobile robot paths in a warehouse environment with automated logistics. First, we compared the results of experiments conducted using two basic algorithms to identify the fundamentals required for planning the path of a mobile robot and utilizing reinforcement learning techniques for path optimization. The algorithms were tested using a path optimization simulation of a mobile robot in same experimental environment and conditions. Thereafter, we attempted to improve the previous experiment and conducted additional experiments to confirm the improvement. The experimental results helped us understand the characteristics and differences in the reinforcement learning algorithm. The findings of this study will facilitate our understanding of the basic concepts of reinforcement learning for further studies on more complex and realistic path optimization algorithm development.

Storehouse: a Reinforcement Learning Environment for Optimizing Warehouse Management

“…Thus, the GA defines the goal of the agent, and the RL algorithm manages the movements to achieve that goal. The results of this article provide insight that has been verified by other authors [LSK + 19], [YJL20], [LJ21]. These works compare RL algorithms such as Q-Learning [WD92] or DQN [MBM + 16] with classic path-finding algorithms, such as A* [HNR68], and conclude that RL algorithms achieve at least the same efficiency as A*, and enhance it in some cases, but all these works are limited to address the path-finding part of the warehouse management problem, without addressing the macro-planning problem using complex data analysis.…”

Speeding-Up Action Learning in a Social Robot With Dyna-Q+: A Bioinspired Probabilistic Model Approach

“…More recently, Hayamizu et al [48] have proven how a Guided Dyna-Q architecture (GDQ) allows a mobile robot to successfully navigate reducing the exploration speed of the environment. Similarly, Lee et al [49] highlight the benefits provided by a Dyna-Q architecture in comparison with classical Qlearning during path planning tasks for mobile robots. In their results, Lee at al.…”

A Comprehensive Study of Mobile Robot: History, Developments, Applications, and Future Research Perspectives

“…Problems associated with path planning include finding an appropriate path from the initial position to the final position. Reinforcement learning algorithms can play a vital role in the optimization of the path for a single-and multiple-agent environment to resolve more realistic and complex problems [108]. There are mainly two types of problems related to the mobile robot that are mostly famous among researchers; these are the single-and multi-objective path planning of mobile robots.…”