A novel double-hybrid learning method for modal frequency-based damage assessment of bridge structures under different environmental variation patterns

•Proposing a novel double-hybrid learning method for damage assessment under different environmental variation patterns.•Leveraging a self-adaptive neighbor searching algorithm for automatically determining the number of nearest neighbors.•Improving a hybrid clustering algorithm with a simple but effective approach to selecting the number of clusters.•Proposing a non-parametric hybrid anomaly detector based on local outlier factor.