- 學位論文
衛星導航系統即時精確定位法研究
A Study on the Real-Time Positioning Algorithm with GNSS
摘要
在2013年，本研究團隊利用GPS單頻接收器所得到的載波相位觀測量與都普勒頻移觀測量，配合GPS之觀測模型並利用LAMBDA(Least-Square AmBiguity Decorrelation Adjustment)法解算二次差分周波未定值並與卡爾曼濾波器整合實現RTK，是為本研究中所述之第一型RTK [27] 。 由於第一型RTK演算法定位精度以及準確度仍有需加強的空間，為處理精度不足之問題，本研究改變第一型RTK演算法之階層架構，並將二次差分周波未定值加入卡爾曼濾波器的狀態向量中，是為第二型RTK。 此外，第二型RTK雖然精度較第一型RTK略有提升，但仍無法解決定位偏差之問題。本研究遂基於第二型RTK之數學模型發展模糊邏輯輔助適應性卡爾曼濾波器，藉由模糊適應性卡爾曼濾波器之參數來解決此項問題。實驗結果顯示確實能在維持較高精度的狀況下，解決準確度不足的問題。在衛星接收良好的狀況之下，第二型RTK可得到較佳的定位精度。此外，我們亦探討模糊邏輯輔助適應性卡爾曼濾波器應用於RTK演算法以增加定位準確度之可行性，並且實驗證實確實能夠有效增加定位準確度。
並列摘要
In 2013, an RTK(real-time kinematic) algorithm has been developed by using differential measurement data from two GPS single frequency receivers. The integer ambiguities for double-difference observables are resolved by applying LAMBDA (Least-Square AmBiguity Decorrelation Adjustment) method. That algorithm shall be referred as Type I RTK in this research [27]. Due to the fact that the positioning accuracy and precision of Type I RTK algorithm are not satisfied, this research adopts another hierarchy which includes double-differenced integer ambiguity in the state vector in applying the Kalman filter. Such algorithm is referred to as Type II RTK in this research. However, although the positioning precision of Type II RTK has been enhanced in comparing with Type I RTK, the positioning accuracy is still not improved. To solve this problem, this research develops an adaptive Kalman filter by fuzzy logic based on the model of Type II RTK. The experimental results show that fuzzy logic RTK can improve the positioning accuracy with the same level of precision as that of Type II RTK. It is then observed that, in normal condition, type II RTK gives rise to more precise positioning result that that from using Type I RTK, while fuzzy adaptive RTK is more accurate that Type II RTK.