Open Access

In recent years, the development of machine learning (ML) techniques has led to significant progress in the field of structural health monitoring with ultrasonic-guided waves. However, a number of challenges still need to be resolved for reliable operation in realistic settings. In this work, we consider the complex problem of experimental damage detection under varying temperature or load conditions where damage locations are not included in the training set. The ML techniques proposed here include supervised and unsupervised methods originally developed for image and time series classification combined with ensemble voting. A performance demonstration of the ML techniques is presented using benchmark datasets from the open-guided waves platform. The unsupervised approach is then applied to a new dataset from an experimental campaign carried out on a composite over-wrapped pressure vessel used for hydrogen storage with real defects. Results show that ensemble voting enables the effective combination of the predictions of multiple transducer pairs, even with a limited number of strong individual classifiers. When applied to unsupervised learning, this returns high accuracy also when real damage over the structure is considered.

Despite proven approaches available in the literature, structural health monitoring by ultrasonic guided waves under varying environmental and operational conditions is still challenging. The use of machine learning approaches is discussed in this work, considering the complex problem of experimental damage detection under varying load conditions in a composite overwrapped pressure vessel for hydrogen storage. Specifically, unsupervised methods originally developed for image and time series classification are combined with ensemble voting to conceive reliable damage detection technique. This enables the effective combination of the predictions of multiple transducer pairs, even with a limited number of strong individual classifiers. A performance demonstration of the technique is presented using a real damage scenario dataset.