VR-Based Learning Platform for the Application of BPP Classification in 5G Learning Factory
摘要
In recent years, fuel cells have been widely used because of their efficiency, cleanliness and economy. The structural and physical parameters of the battery stack change depending on the applied assembly loads, so the bipolar plates (BPP) and membrane electrodes (MEA) require high surface quality and positioning accuracy, which is one of the prerequisites for the proper operation of the stack. This article proposes a VR-based BPP classification and clustering learning platform on the 5G cloud server. The platform realizes virtual and real data sharing and continuously updates algorithm parameters through the 5G connection. In virtual learning platform, students can learn all the image information and label bipolar plates with less warping and more similarity based on the convolutional neural network and clustering algorithms. Learning Factory can make intelligent algorithms learning with the help of VR devices come true. In other manufacturing scenarios, students can train machine models and tune relevant parameters to achieve the classification and clustering of other products.