A Comparative Analysis of Displacement Detection Methods Using Locata

Alarm algorithms for displacement detection methods are one of the key components of an automatic displacement monitoring system used for monitoring any large structure (such as dams, bridges etc.). Methods such as Shewhart algorithm, moving average algorithm weighted moving average algorithm (WMA), exponentially weighted moving average algorithm (EWMA), cumulative sum algorithm (CUSUM) are used extensively to detect displacement, or in other words shifts in the mean and/or process variance. These methods and conventional least squares adjustment analysis are used extensively in GNSS displacement monitoring applications. Locata Corporation‟s positioning technology “Locata” provides centimetre-level accurate position solution with millimetre-level precision. This technology can complement (or augment) or even replace GNSS in environments with bad satellite geometric distribution or poor received satellite signal levels. In this paper the Shewhart and CUSUM algorithms are implemented and compared with conventional least squares adjustment. Two experiments were conducted at the University of New South Wales to analyse Locata‟s performance. Simulated movement was introduced for algorithm testing. After generating the coordinate solution, the three displacement monitoring algorithms were tested. In addition the prerequisite requirements of these methods were also analysed. Prerequisite includes checking autocorrelation and checking residuals for zero mean normal distribution. An autoregressive model was used to model autocorrelation. Results show that CUSUM and conventional displacement test performs better than the Shewhart algorithm.