Petri Net Based Cyber-physical System for Factory Autonomous Mobile Robot Control Applications

• Main Authors: Yi-Jung Liu, 劉益榮
• Other Authors: Chung-Hsien Kuo
• Format: Others
• Language: zh-TW
• Published: 2018
• Online Access: http://ndltd.ncl.edu.tw/handle/udc9p7

碩士 === 國立臺灣科技大學 === 電機工程系 === 106 === This thesis presents a colored Petri net (CPN) based cyber-physical system (CPS) study for modeling and simulating the machining and material handling processes in an automated factory. An automated guided vehicle (AGV) was introduced to be a physical part of the CPS system for executing material transport task commands generated from the CPN cyber part. Hence, the CPN production model is responsible for dealing with the operation of a virtual factory (VF), production order handling and dispatching, and AGV movement command generation because of the CPN’s capabilities on dealing with concurrent and asynchronous events in systems. Moreover, a cloud-based web system was produced to perform the monitoring of the system operation, including production order, production status and AGV status. All the proposed subsystems were integrated via network and database. A production order is delivered to CPN VF model, and the CPN VF model is used for production and AGV dispatching. The AGV dispatching command is delivered to the physical AGV via Wi-Fi. The AGV will communicated with virtual machine model that was done with a tablet for exchange production information. In this manner, all the cyber CPN VF model, tablet virtual machines and a physical AGV were all properly integrated. The proposed system was worked with either simulation or physical operation mode for the AGV. No physical AGV was required for the simulation mode, and the simulation mode can evaluate the system performance in advance. The physical operation mode can be used to evaluate the physical AGV performance according to the VF operations. Finally, a VF with 4 tablet virtual machining stations and an AGV was set in our laboratory for experiments. The experiment results showed that the proposed CPN-based cyber-physical system is feasible for evaluating the AGV performance without the needs of real machines.