Time series data play a crucial role in Prognostics and Health Management by enabling the continuous monitoring and predictive maintenance of critical assets across various industries. Industrial systems, such as energy infrastructure, transportation networks, and manufacturing processes, generate vast amounts of time-dependent data through sensors that capture parameters like temperature, vibration, acoustic signals, and electrical performance.
However, challenges such as noisy signals, varying operating conditions, and the need for long-term forecasting demand advanced analytical approaches. Our lab focuses on developing robust machine learning and deep learning techniques to address these challenges and extract actionable insights from time series data.
Our research in time series analysis for PHM can be grouped into three main categories:
-
Time Series Denoising and Feature Extraction
-
Time Series Forecasting and Predictive Modeling
-
Condition Monitoring and Industrial Applications
