Research on Targeting Technology in Disassembly of Retired Power Battery Packs

Along with the process of industrialization, the new energy vehicle industry has been rapid development, but also brought a series of problems, a large number of EVB (Electric Vehicle Battery) to reach the end of its useful life is about to be recycled or end-of-life treatment, through the disassembly of decommissioned power battery packs recycling can reduce the heavy metals in the battery packs of the pollution of the environment, and at the same time, bring great Economic Benefits. This paper combines deep learning to disassemble decommissioned power packs for targeting research, disassembly first need to identify the fasteners on the power battery pack, this paper takes the typical fasteners hexagonal bolts as the parts to be detected, acquires the hexagonal bolt image through the cell phone terminal, completes the processing of the image using data enhancement technology and completes the establishment of the datasets, and at the same time trains the detection model based on EDGE IMPULSE deep learning platform, and analyzes the performance of the model. model, and experimental analysis of the performance of the model, and finally deploy the model to the cell phone to complete real-time detection. Compared with the existing datasets, the model trained from the self-built datasets has higher accuracy and better performance, which improves the disassembly efficiency and reduces the labor cost.