Three-way analysis of structural health monitoring data

Prada, Miguel A.; Toivola, Janne; Kullaa, Jyrki; Hollmén, Jaakko

doi:10.1016/j.neucom.2011.07.030

Cited by 29 publications

(17 citation statements)

References 22 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Tensor decomposition has been used in civil engineering [26,93] for detection of abnormal changes in the structure vibration response. Different sensors are employed in different parts of the structures and their vibration responses are measured during a period of time.…”

Section: Engineeringmentioning

confidence: 99%

“…Dimensionality reduction based Categorical target [26,93] Numerical target [30] Classification based Support tensor machines [89] Supervised tensorLearning [111] Tensor least square [112] Multilinear discriminant analysis [114] Factorization machines [115] Tensor subspace learning [116]…”

Section: Supervisedmentioning

confidence: 99%

“…In the first group of methods [26,93,109] , it is assumed that we have two sets, train and test, where train set contains normal samples. Tensor decomposition is applied on the normal tensor as a dimensionality reduction tool.…”

Section: Tensor Decomposition For Dimensionality Reductionmentioning

confidence: 99%

“…Therefore, the built model from train set is used to predict the label (normal or abnormal) of observation in the test factor matrix. For instance, in [26,93] , a PARAFAC decomposition with k number of components is applied on the space × time × frequency tensor corresponding to the normal samples and then the derived time factor matrix is trained via k-NN (features are the latent variables). The built model is then used for classification of time points in the arriving data.…”

Section: Tensor Decomposition For Dimensionality Reductionmentioning

confidence: 99%

See 3 more Smart Citations

Tensor-based anomaly detection: An interdisciplinary survey

Fanaee-T

Gama

2016

Knowledge-Based Systems

126

View full text Add to dashboard Cite

Traditional spectral-based methods such as PCA are popular for anomaly detection in a variety of problems and domains. However, if data includes tensor (multiway) structure (e.g. space-time-measurements), some meaningful anomalies may remain invisible with these methods. Although tensor-based anomaly detection (TAD) has been applied within a variety of disciplines over the last twenty years, it is not yet recognized as a formal category in anomaly detection. This survey aims to highlight the potential of tensor-based techniques as a novel approach for detection and identification of abnormalities and failures. We survey the interdisciplinary works in which TAD is reported and characterize the learning strategies, methods and applications; extract the important open issues in TAD and provide the corresponding existing solutions according to the state-of-the-art.

show abstract

Section: Engineeringmentioning

confidence: 99%

Section: Supervisedmentioning

confidence: 99%

Section: Tensor Decomposition For Dimensionality Reductionmentioning

confidence: 99%

Section: Tensor Decomposition For Dimensionality Reductionmentioning

confidence: 99%

See 2 more Smart Citations

Tensor-based anomaly detection: An interdisciplinary survey

Fanaee-T

Gama

2016

Knowledge-Based Systems

126

View full text Add to dashboard Cite

show abstract

“…Both DI expressions return values greater than zero, if any variation in the structural dynamic behavior occurs, and they will return "zero", if there is no damage in the structure. Another important approach was presented by Prada et al (2012) and it was defined as Change in Measured Parameters Metric (CMPM). This metric uses previously numerical or experimental modal characteristics of the undamaged structure.…”

Section: Case Study 2 -Damage Identification Without Pzt Transducermentioning

confidence: 99%

Experimental analyses of metal-composite bonded joints: damage identification

2014

View full text Add to dashboard Cite

The advent of composite co-cured and co-bonded integrated construction in aircraft structures has lead to the replacement of fastened joints with bonded joints between the skins and the stiffeners. Skin-stiffener debondings could occur due to impact or other operational reasons and it is usually internal failure. Damage identification of bonded components, which are often vital elements in many structures, is crucial for the prevention of failure of the entire structure. Thus, different researchers have investigated vibration-based methods as an alternative technique to be used in the structural health monitoring (SHM) systems. Hence, this work consists of investigating experimentally through the vibration-based method, the dynamic behavior changes in a bonded metal-composite structure by using piezoelectric transducer and accelerometers in order to monitory the damage. The damage is an artificial debonding in the joint, which was simulated by inserting Teflon™ tapes within the joint. In-situ inspection as ensured by accelerometer and piezoelectric transducers (PZT) bonded to the structure. Indeed, with a simple comparison of the frequency response functions is difficult to conclude if there is damage in the structure, unless a large damage is presented. However, by using damage metrics, it is possible to identify the damage with more accuracy. Thus, the experimental results obtained by the accelerometers were compared to the data provided by the smart composite sensors (PZT). Finally, it was discussed the advantages and limitations of the experimental analyses and the identification technique proposal.

show abstract