ABSTRACT
Recent advances in sensing technologies make it feasible and practical to instrument sensors to high expenditure civil infrastructure such as bridges for safety monitoring during construction as well as for long-term condition assessment. This paper presents an investigation of the structural response acquired through the structural health monitoring system installed in the Seohae Bridge (Cable-stayed Bridge) in South Korea. Preliminary analysis of the measured data revealed that the responses of the bridge are mainly affected by temperature variation, thus influence of the temperature is closely investigated using ARX model, statistical time series analysis method. Fourteen channels of temperature data are utilized as input and selected static responses such as vertical displacement of main span, movement of expansion joint, or tilting angle of a pylon are utilized as output. Number of input channels and previous data length included in the model is varied to determine optimum structure of ARX model. Feasibility of ARX model for analysis of bridge monitoring data has been proved, and the underlying relationship between the temperature and the structural response could be explicitly modeled with this approach.
