You are here: Home :: Area NEM :: Geoscience :: Geodesy

M. Heinert, W. Niemeier

From fully automated observations to a neural network model inference: The Bridge "Fallersleben Gate" in Brunswick, Germany 1999–2006

Keywords: structural monitoring, continuous measurements, time series analysis, Kalman filtering, system analysis

Since the beginning of June 1999 the Institute of Geodesy and Photogrammetry is monitoring the displacements of a small, but unstable bridge within the city of Brunswick. As basic sensor for automatic and continuous monitoring an automated total station is installed below the bridge, which takes measurements to 180 reflecting targets on the structure three times per day. This paper begins with the outline of the basic concepts for the instrument setup, the longterm control of its stability and possible influencing factors. Then the processing of the time series of raw data is discussed, emphasing the filtering concepts to account for the detection of outliers, due to target corrosion or distortion. Finally the bridge is considered as a closed system and all the available information is used to generate a non-parametric model to describe its behaviour with tools from system analysis.

Journal of Applied Geodesy, Walter de Gruyter

Print ISSN: 1862-9016
Volume: 1, 09/2007
Pages: 71 - 80

Show full article (external site)

Show all available items of this journal

 

Area NEM