AGV multi-target task scheduling method and system based on deep reinforcement learning

• Main Authors: WU QINGYAO, QIN ZHUORUI, WU XIAOQIAN, ZHENG YIMIN
• Format: Patent
• Language: Chinese; English
• Published: 08.09.2023
• Subjects: CALCULATING; COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS; COMPUTING; COUNTING; DATA PROCESSING SYSTEMS OR METHODS, SPECIALLY ADAPTED FORADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL, SUPERVISORYOR FORECASTING PURPOSES; PHYSICS; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE,COMMERCIAL, FINANCIAL, MANAGERIAL, SUPERVISORY OR FORECASTINGPURPOSES, NOT OTHERWISE PROVIDED FOR

The invention discloses an AGV multi-target task scheduling method based on deep reinforcement learning. The AGV multi-target task scheduling method comprises the steps of obtaining environment real-time state information and making a decision through a deep Q network to obtain an optimal scheduling sequence; constructing a deep reinforcement learning task scheduling model; training a task scheduling model by taking the processed real-time environment state information, actions and rewards as samples; and performing deployment prediction on the trained task scheduling model. According to the method, AGV tasks in a specific scene are dynamically scheduled in real time by using a deep reinforcement learning method, a Transform module is introduced to enable a model to pay attention to global task information, features and importance of different tasks can be learned, and general knowledge between the tasks can be learned at the same time; a reward function of reinforcement learning is improved, AGV collision wa