Research and Application of Hausdorff Metric Based i-v Vague Clustering Method

While vague clustering, as generalization of fuzzy clustering, has been attracting more and more attentions, its generalization, i-v Vague clustering, is still ignored by researchers. In order to achieve the clustering of i-v Vague set, a system clustering method is proposed, which is based on similarity generated by Hausdorff metric. The method firstly utilizes Hausdorff metric to generate similarity between two i-v Vague values, and then establish similarity matrix. By setting a threshold, i-v Vague clustering is finally achieved through direct clustering. The feasibility of the method is validated through analyzing of mechanical properties degeneration data about some material caused by fatigue damage. Furthermore, the potential application of the method to state identification is also discussed.