A Method for Learning a Petri Net Model Based on Region Theory

2020 
The deployment of robots in real life applications is growing. For better control and analysis of robots, modeling and learning are hot topics in the field. This paper proposes a Petri net model learning method based on region theory, which can learn some information from the limited attempts of robots and generate an accurate Petri net model accordingly. In order to demonstrate the effectiveness of the method, we programmed and simulated the scene in which the robot solved the problem in the building block world, and the Petri net model generation algorithm in that scene. The results of experiments conducted on building block world show that our algorithm is feasible and effective.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    5
    References
    0
    Citations
    NaN
    KQI
    []