Study on prognosis and health management for the liquid tank

• Published in: 2017 Prognostics and System Health Management Conference (PHM-Harbin) pp. 1 - 7
• Main Authors: Qing-zhong Hu, Fu-lei Chu
• Format: Conference Proceeding
• Language: English
• Published: IEEE 01.07.2017
• Subjects: Data models; Hidden Markov models; HMM; Markov processes; PHM; Predictive models; Reliability; The Space launch Site
• DOI: 10.1109/PHM.2017.8079217

For a new generation of spaceflight launch site online air supply mode, new health status of the equipment is difficult to effectively evaluate the single sample problem, put forward a kind of equipment based on hidden markov state of health management and prediction method; The method use the history of equipment monitoring data hid-den markov health status evaluation model was constructed, based on different observation sequence and the pre-diction accuracy of the simulation under different observation times, determine the optimal model parameters, and the real-time monitoring data generation into the model, according to the model, the calculation results from the minimum to judge the health status of the equipment. Finally, the current data and historical data fitting, estimate the system safe and reliable life, through the actual test, this method is effective to solve the single sample equipment health assessment.