Adaptive scheduling of predictive maintenance interval via the integration of multi-source information

As a pivotal technology in digital asset management, intelligent inspection and replacement play a crucial role in reducing operational costs, ensuring reliability, and enhancing the overall efficiency of various industrial plants. Traditionally, inspection and replacement intervals have been set based on a static reference cycle, which can often lead to either excessive or insufficient resource allocation. To overcome this issue, we propose an adaptive, predictive interval optimization approach that leverages multi-source health information. Specifically, we develop a data fusion method to track the dynamic degradation trend of assets, employing a Kalman Filter to facilitate the real-time updating of degradation rates and the distribution of remaining useful life. Subsequently, a dynamic interval management strategy is introduced, predicated on real-time state conditions. This strategy informs the optimization of a cost model to determine the optimal time for replacement. A case study is included to demonstrate the practicality and effectiveness of the proposed scheduling model.