Device Status Evaluation Method Based on Deep Learning for PHM Scenarios

The emergence of fault prediction and health management (PHM) technology has proposed a new solution is suitable for implementing the functions improving intelligent control system. However, research application PHM model in system electronic equipment are few at present, there many problems that need to be solved urgently itself. In order solve such problems, this paper studies equipment-status-assessment method based on deep learning scenarios, conduct in-depth equipment. experimental results show change unimproved very subtle before performance point, while improvements increase value by about 10 times. Thus, improved amplifies changes early degradation slows down mutations late failure points. At same time, comparing health-index-evaluation indicators, it can concluded although monotonicity index low, its robustness correlation significantly improved. Additionally, close 1, making curve more line with traditional cognition convenient application. Therefore, an study methods assessment practical significance.