Crack Monitoring for Hot-Spot Areas Under Time-Varying Load Condition Based on FCM Clustering Algorithm

Crack monitoring for damage-prone structural area, named hot-spot, is one of the most interesting concerned technologies for aircraft structural health monitoring. This paper proposes an efficient technique based on Fuzzy c-means (FCM) clustering algorithm to monitor the crack growth under time-varying load condition. The proposed method first extracts damage indexes from the collected signals to construct the baseline model as well as the online FCM model. Migration index of the online model is calculated by Kullback-Liebler Divergence (KLD) and Bhattacharyya Distance (BD), respectively. By comparing with the Gaussian mixture model (GMM), experimental results show that the proposed FCM algorithm has less computation time and its migration index curve is smoother when the crack size is small. Furthermore, KLD achieves a higher sensitivity than BD as crack grows.