An anomaly detection method based on multi-feature fusion and t-distributed stochastic neighbor embedding (t-SNE) is proposed to identify the abnormal state of wind turbine towers based on monitored structural response signals. The method involves evaluating the time domain
frequency domain
and time-frequency domain statistical indicators of the monitored structure response signals to extract high-dimensional multi-feature vectors. The t-SNE algorithm is used for dimensionality reduction and fusion to obtain a visual representation of the data in a low-dimensional space. The -means clustering algorithm is employed to analyze the data state
and a quantitative analysis of the anomaly index is then constructed to realize structural anomaly detection. The proposed method has been successfully applied to engineering applications of wind turbine towers during typhoon and earthquake periods
demonstrating its effectiveness in detecting abnormal structural responses caused by environmental changes.
CASTELLANI F, GARIBALDI L, DAGA A P, et al.Diagnosis of faulty wind turbine bearings using tower vibration measurements[J]. Energies, 2020, 13(6): 1474.
ANOWAR F, SADAOUI S, SELIM B.Conceptual and empirical comparison of dimensionality reduction algorithms (PCA, KPCA, LDA, MDS, SVD, LLE, ISOMAP, LE, ICA, t-SNE)[J]. Computer science review, 2021, 40: 100378.
IKOTUN A M, EZUGWU A E, ABUALIGAH L, et al.K-means clustering algorithms: a comprehensive review, variants analysis, and advances in the era of big data[J]. Information sciences, 2023, 622: 178-210.