Msaewe, Hussein Alwan Mahdi
(2018)
Assessment of the accuracy and the contribution of multi-GNSS in structural monitoring.
PhD thesis, University of Nottingham.
Abstract
Structural health monitoring requires precise techniques with high accuracy, due to the relatively small deformations that can be found in many structures. It has been proved that GPS is capable of detecting the characteristics of the response of flexible structures.
Although GPS can be applied for displacement monitoring, some existing constraints may limit the accuracy of the monitoring application. Some sources of these limitations are the multipath error, random noise, cycle slips and the geometry of the satellite constellation. Studies have been conducted to overcome these limitations using a combination of the GPS and other sensors: geodetic (e.g. Robotic Total Stations) or non-geodetic (e.g. accelerometer). Although these studies succeeded generally to reduce the effects of these limitations, they are still restricted on specific cases of monitoring, due to difficulties of using the accelerometer for monitoring static or quasi-static displacements and RTS due its limited range.
The introduction of the Global Navigation Satellite Systems (GNSSs), apart from GPS, such as GLONASS, BeiDou, and Galileo provides an alternative solution by combined GPS data with additional observations from other GNSS constellations to overcome GPS-only limitations. Therefore, the current study aims to develop a multi-GNSS method for structural health monitoring. The integration of different constellations contributes to improving the accuracy and availability of the solution, which allows the structural response and its characteristics effectively. The positioning performance in this study is investigated and achieved in three conducted stages:
Firstly, a series of GNSS zero baseline measurements are conducted simultaneously for 12 consecutive days in the UK and China sites to investigate the noise level of different GNSS solutions. The main aim is to investigate the correlation between the geometry of satellite constellation and the performance of the GPS-only and multi-GNSS solution. It is proved that for periods of a poor GPS-only solution, due to the satellite constellation or problematic satellites, the multi-GNSS solution leads to more accurate and of higher availability solution.
Secondly, short baselines of GNSS measurements of static (1 Hz) and kinematic (10 Hz and 20 Hz) observations are collected to evaluate the precision and accuracy of different GNSS solutions. Following the same analysis approach and using the 3D best fitting of non-linear least squares adjustment models, it is proved that the multi-GNSS solution is significantly improved in comparison with the GPS-only solution for periods of weak geometry or problematic satellites. This can be attributed to the geometry improvement of the combined solution, which can reach 30% for GDOP values.
Finally, the GNSS measurements are investigated at the Severn Bridge and the Forth Road Bridge in the UK for real structural monitoring of long-span suspension bridges. The analysis was based on an assessment of workflow methodology applied in this study. The noise level of GNSS data was assessed by zero baseline at the base station of the Severn Bridge. It was approved that the combined GPS/GLONASS solution reduced the noise level and led to more accurate results with less discontinuous time intervals. Regarding the spectral analysis, the GPS-only solution led to a higher noise level of the spectrum and less easy to be detected peaks for some intervals relatively to the spectra of the multi-GNSS solution.
It was inferred from the results presented in the current study that the noise errors and discontinuity problems, as well as other limitations of the GPS-only solution, were significantly reduced and improved by combined GNSS solution. These findings are promising for many real structural monitoring applications.
Actions (Archive Staff Only)
 |
Edit View |
Tools
Tools
![[img]](/style/images/fileicons/application_pdf.png)