The study of data densification for GPS kinematic observables

• Main Authors: Lee Hsin Yu, 李信佑
• Other Authors: Hwang Lih Shinn
• Format: Others
• Language: zh-TW
• Published: 2008
• Online Access: http://ndltd.ncl.edu.tw/handle/68741052511288512796

碩士 === 國防大學中正理工學院 === 軍事工程研究所 === 96 === If the GPS observables of base station were collected from a public service for post-processing type of kinematic GPS, an inconsistent GPS sampling rate, e.g. 30 second and 1 second interval adopted by the base station and rover, respectively, might occur and lead to a non-simultaneous observation. A mathematical algorithm is expected to work for data desification to enhance the available data availability and reduce the cost by not taking any supplementary observation. This study selected a curve fitting model as the data interpolation technique for the desification of phase observable applying to GPS kinematic positioning. During the tests, different epoch numbers of the original data and different orders of polynomial were used to find out the proper models to provide the optimal positioning data. It was found from the observation made at a short and medium baseline that an operation model based on every 4 epochs of data and a 3rd order of polynomial was able to effectively increase the data density up to 1 second interval for the stations with an original observation of every 30 seconds. The internal and external accuracy of 1~2 cm and 6~7 cm were obtained in horizontal and vertical component, respectively, by using such Interpolated one second data sets. Compared with the result using different sampling rates of original data and different density of Interpolated data, the accuracy difference can be maintained to be less than 0.5 cm. It proves that such an interpolateon algorithm is suitable for the short range of kinematic GPS.