Design and Optimization of a Multi-AMR Scheduling System Based on QT

In the context of box storage environments, traditional scheduling systems have proposed solutions based on conventional path-planning algorithms to address the scheduling problem of autonomous mobile robots. However, these solutions suffer from inefficiencies and high error rates. To solve these problems, this paper proposes and implements a software design scheme for a scheduling system based on the deep reinforcement learning algorithm, utilizing the QT platform. In the context of box storage environments, this paper proposes a software architecture based on a layered structure and implements it in the QT development environment. Firstly, the overall framework of the scheduling system is introduced. Subsequently, the principles and advantages of the deep reinforcement learning algorithm adopted by the scheduling system, as well as the specific structure and functionality of the scheduling system, are elaborated. Finally, through practical testing, it is concluded that the proposed design scheme is feasible.