Infrared detector fault classification and prediction technology based on sensitive parameter learning

Infrared detector is an important device with a wide range of applications. Based on the fault sensitive parameter data of infrared detectors, this paper studies the fault classification and fault prediction model of infrared detectors by using machine learning methods such as neural network BPNN and long and short term memory network LSTM. Through the establishment and verification analysis of the fault classification model, it provides a model reference and basis for the multi-type fault diagnosis of infrared detectors. Through the establishment and analysis of the fault prediction model, it provides a modeling method for the lifetime prediction of infrared detectors. The application of infrared detector fault classification and prediction technology can improve the reliability of infrared detector products.