A platform for research: civil engineering, architecture and urbanism
A platform for research: civil engineering, architecture and urbanism
Dam safety monitoring data anomaly recognition using multiple-point model with local outlier factor
Abstract Anomalous values commonly observed in monitoring data can severely impact dam structural health monitoring. Hence, it is crucial to establish reliable models for identifying various anomalies in dam monitoring systems. This paper develops a multi-point anomaly recognition model by coupling improved Local Outlier Factor (LOF) and mutual validations considering spatio-temporal correlation. The proposed model is evaluated using real-world displacement data from concrete and rockfill dams, and the results demonstrate its satisfactory accuracy and superior robustness across the whole dam monitoring section. This work provides a dependable database for dam safety assessment and a valuable reference for recognizing the cause of anomalies. Further research directions should be focused on the recognition of anomalies in more complex monitoring data and the processing of data after removing outliers.
Highlights A framework is developed to construct two-dimensional point sets as identification samples. The proposed multi-point model achieves improved accuracy and robustness of dam monitoring data anomaly recognition. Analysis for spatiotemporal correlation of outliers offers a novel perspective on the anomaly cause recognition. The identification results can provide a reliable database for further dam safety assessment.
Dam safety monitoring data anomaly recognition using multiple-point model with local outlier factor
Abstract Anomalous values commonly observed in monitoring data can severely impact dam structural health monitoring. Hence, it is crucial to establish reliable models for identifying various anomalies in dam monitoring systems. This paper develops a multi-point anomaly recognition model by coupling improved Local Outlier Factor (LOF) and mutual validations considering spatio-temporal correlation. The proposed model is evaluated using real-world displacement data from concrete and rockfill dams, and the results demonstrate its satisfactory accuracy and superior robustness across the whole dam monitoring section. This work provides a dependable database for dam safety assessment and a valuable reference for recognizing the cause of anomalies. Further research directions should be focused on the recognition of anomalies in more complex monitoring data and the processing of data after removing outliers.
Highlights A framework is developed to construct two-dimensional point sets as identification samples. The proposed multi-point model achieves improved accuracy and robustness of dam monitoring data anomaly recognition. Analysis for spatiotemporal correlation of outliers offers a novel perspective on the anomaly cause recognition. The identification results can provide a reliable database for further dam safety assessment.
Dam safety monitoring data anomaly recognition using multiple-point model with local outlier factor
Rong, Zhuo (author) / Pang, Rui (author) / Xu, Bin (author) / Zhou, Yang (author)
2024-01-16
Article (Journal)
Electronic Resource
English
| British Library Conference Proceedings | 2023
| Wiley | 2022
Local outlier factor-based fault detection and evaluation of photovoltaic system
| British Library Online Contents | 2018