Multisensorial Data Fusion and Correlation for Dams Structural Health Monitoring (Multi Scale, Uncertainty-Sensitive Heterogeneous Data Integration)

The paper submit an innovative approach to dam surveillance question by means of multi-level heterogeneous sensory data fusion, in order to increase the aggregation efficiency of data obtained by different sensors, which combines fuzzy set theory with other multi-sensor data fusion techniques. The sensor information fusion model was designed, using partial coupled data gathered by various sensors. Behind the analysis and process of the multi level fusion model, the precise appraisal of the secure status of the dam environment can be achieved. Lastly, the approach, that could actually enrich the consistency of the secure situation of the dam, is validated through a case analysis. A rule-based expert sub-system for fusing the raw data acquired from sensors with complementary accuracy and bandwidth attributes is proposed. Furthermore, for particularly precise purposes it is proposed a high accuracy fuzzy predictor sub-system. The major benefits of the projected multi-modal sensor fusion method are illustrated by simulation of a dam surveillance sub-system employing the fusion components for behavior assessment.