Investigation on data-driven life prediction methods

The theory of life prediction is the key foundation of the safety service of electromechanical device, and it is also an important aspect of modern mechanical design and manufacturing. Life prediction for major mechanical equipment has great significance for economic development and defense construction. This paper aims at the investigation of characteristics and applications of electromechanical device life prediction techniques. It summarizes the current progress of life prediction based on data-driven methods. The challenges of electromechanical device life prediction include theoretic models, experimental validation, data accumulation. It provides a possible direction for further research on life forecasting method.