A Remote Bridge Health Monitoring System Using Computational Simulation and Single Frequency GPS Data

The IESSG at the University of Nottingham has been involved with the use of GPS for deformation monitoring for over a decade. The work started off by using static GPS to monitor long term deformations of structures such as reservoirs. More recently, kinematic GPS has been used as a basis in order to measure the deflections of structures in real time. This work has been carried out on a number of structures, mainly bridges; including the Humber Bridge. Previous research work with collaborators from Brunel University showed that it is indeed possible to use kinematic GPS to measure the deflections of structures and compare the results very favourably with Finite Element Models (FEMs).

The University of Nottingham in collaboration with Cranfield University was recently awarded a research grant by the Engineering and Physical Sciences Research Council’s Structural Integrity Programme. The title of this three year project is "A Remote Bridge Health Monitoring System Using Computational Simulation and GPS Sensor Data". The overall aim of the project is the creation of a system that employs computation simulation with GPS sensors to monitor the health of bridges without the need for onsite inspection.

The use of dual frequency code/carrier phase GPS data has been shown to be a feasible tool for bridge deflection monitoring. This is due to the quick On The Fly (OTF) integer ambiguity resolution that is possible through using the dual frequency data in order to create the wide lane observable. The applications of OTF dual frequency data are numerous, whereby OTF integer ambiguities are resolved in seconds. Single frequency code/carrier phase GPS receivers are less expensive than dual frequency receivers; however, the time taken to resolve integer ambiguities in a traditional OTF manner is much longer than that of dual frequency. This is due to the wide lane observable not being available. It is possible to accelerate the single frequency OTF search through various techniques. One such technique is to assume that the bridge does not in fact move too much, as it is in a semi-static situation. Therefore the OTF search can be based around a fairly well-known coordinate.

A variety of trials have been conducted as part of this work in order to develop the FEM and validate it. These trials include deploying the GPS receivers, accelerometers and pseudolites upon the Wilford Bridge in Nottingham and gathering data in order to ascertain the characteristics of the structure. The data is then meticulously processed and cleaned using a variety of adaptive filtering techniques developed for this project. During some of the trials, a group of people marched over the bridge, hence introducing some form of motion on the suspension bridge. The data is then used to validate the FEM of the bridge. Once the FEM is thought to be an accurate representation, it can be used to compare the subsequent GPS data with. Any discrepancies with the comparison will probably be due to the bridge not behaving as expected and prompt closer inspection.

The following paper details the precision obtainable using single frequency receivers and accelerometers, as well as the techniques used to accelerate the OTF search in the context of measuring the deflections, long term deformations and vibrations of a bridge. The application of the adaptive filtering is also discussed, illustrating how such filtering can be used to mitigate multipath on a daily basis. In addition, the development and comparison of the FEM is also discussed with a particular reference to the Wilford Bridge in Nottingham.