Hierarchical ensemble-based data fusion for structural health monitoring

Abstract

In structural health monitoring, damage detection results always have uncertainty becauseof three factors: measurement noise, modeling error and environment changes. Data fusioncan lead to the improved accuracy of a classification decision as compared to a decisionbased on any individual data source alone. Ensemble approaches constitute a relativelynew breed of algorithms used for data fusion. In this paper, we introduced ahierarchical ensemble scheme to the data fusion field. The hierarchical ensemblescheme was based on the Dempster–Shafer (DS) theory and the Rotation Forest(RF) method, it was called a hierarchical ensemble because the RF method itselfwas an ensemble method. The DS theory was used to combine the output of RFbased on different data sources. The validation accuracy of the RF model wasconsidered in the improvement of the performance of the hierarchical ensemble. Healthmonitoring of a small-scale two-story frame structure with different damages subjectto shaking table tests was used as an example to validate the efficiency of theproposed scheme. The experimental results indicated that the proposed scheme willimprove the identification accuracy and increase the reliability of identification.