Simulation of robot task allocation for warehouse automation

• Main Author: Han, Jiayu
• Other Authors: Hu Guoqiang
• Format: Thesis-Master by Coursework
• Language: English
• Published: Nanyang Technological University 2024
• Subjects: Engineering; Multi-robot system; Task allocation; Path planning; Action-bid method; Floyd alogorithm
• Online Access: https://hdl.handle.net/10356/179033
• Institution: Nanyang Technological University

Muti-robot systems are now widely used in warehouses to transport goods, but task allocation for robots is still a topic worth studying. In this dissertation, research will be introducing in this field. First of all, the goal of the dissertation will be confirmed to be achieved is to make the muti-robot system in warehouse complete the work in the shortest time through task allocation, and there will be no robot collision and other problems. For this purpose, some path planning methods will be used to help the task allocation system achieve optimization. In the experiment, the shortest path of robots by using action-bid task allocation model will be calculated finally and combined with Floyd algorithm. By assigning different weights, 100 robots could complete all tasks within 200s. Simultaneously, to prevent robot collisions, adjustments will be made to bid parameters based on the density distribution of robots within the system's workspace. Additionally, new motion modes will be planned for predicting robot collisions. Other diverse methods were also explored during the study, and detailed descriptions of these methods and the experimental process will be provided. The studies have been conducted solely through simulation, without considering additional factors, to ensure the visualization of experimental outcomes that can be applicable across various scenarios. The proportion of bids can be adjusted to accommodate diverse requirements and environments.