Wear particle analysis based on self-organizing clusters

The focus of this paper is integration of system process information obtained through an image-processing system with an evolving knowledge database to improve the accuracy and predictability of wear particle analysis. We describe the process of deriving such parameters from images of wear particles using an image-processing system. The objective is to classify this wear particle information for possible cluster analysis via self-organizing maps. This can be achieved using relationship measurements among corresponding attributes of various measurements for wear particle analysis. As a result, it helps in predicting wear failure modes in engines and other machinery. Finally, simulations are performed to estimate the efficiency of the proposed system along with visualization techniques that help the viewer in understanding and utilizing these relationships that enable accurate diagnostics and provide the in-depth data needed to support results.