Adaptive Change Detection for Long-Term Machinery Monitoring Using Incremental Sliding-Window

Abstract

Detection of structural changes from an operational process is a major goal in machine condition monitoring. Existing methods for this purpose are mainly based on retrospective analysis, resulting in a large detection delay that limits their usages in real applications. This paper presents a new adaptive real-time change detection algorithm, an extension of the recent research by combining with an incremental sliding-window strategy, to handle the multi-change detection in long-term monitoring of machine operations. In particular, in the framework, Hilbert space embedding of distribution is used to map the original data into the Re-producing Kernel Hilbert Space (RKHS) for change detection; then, a new adaptive threshold strategy can be developed when making change decision, in which a global factor (used to control the coarse-to-fine level of detection) is introduced to replace the fixed value of threshold. Through experiments on a range of real testing data which was collected from an experimental rotating machinery system, the excellent detection performances of the algorithm for engineering applications were demonstrated. Compared with state-ofthe-art methods, the proposed algorithm can be more suitable for long-term machinery condition monitoring without any manual re-calibration, thus is promising in modern industries.

Adaptive Threshold for Anomaly Detection in ATM Radar Data Streams

“…An interesting insight is brought by combining the using of sliding windows and probabilistic tools. T. Wang et al [8] present a martingale-based method to learn the regularity of the observed data in a sliding window of variable size and to identify the shifts in the data stream. The threshold is computed according to a global factor that ascertains the confidence level of the detection.…”

Graph-based change detection for condition monitoring of industrial machinery: an enhanced framework for non-stationary condition signals

“…Obviously, a smaller value of error indicates a better segmentation result, and vice versa. Additional discussions of estimation of λ can be referred to our recent article [38].…”

“…Comparison of detection results by two methods.In detection, we first performed a prior estimation to confirm the value of λ, and then used this value for generating the final result. Additional discussions of estimation of λ can be referred to our recent article[38].…”