GPS病態定位方法之評估與設計

Abstract

一般GPS接收機定位時，需要至少能夠觀測到四顆衛星。但若環境高樓遮蔽，造成衛星訊號不佳的問題，使得最低需求無法滿足，則可能造成接收機無法使用的情況。 為解決此病態定位問題，本文使用多種方式，諸如高度固定輔助法、鐘差預測法與虛擬距離預估器等，解決病態定位下的問題。再進一步精準的定位，我們考慮最小平方法、卡爾曼濾波器、約束濾波器等，並用最好的方式求得定位解。 不同衛星數目的情況下，使用不同方式處理，有使用的時間限制與約束條件；因此除了整合各種演算法，成功完成病態定位外；再利用模擬器設計靜態與動態實驗，針對衛星不足情況發生時，評估在容忍誤差範圍內，各方法可使用的時間。

For normal positioning of using GPS receiver, it is required that there are at least four GPS satellites in view. However, due to the blockage of the tall buildings and receive unhealthy satellites signal, it is sometimes not able to reach that minimum requirement so that the operation of GPS receiver may be interrupted. In dealing with such ill-conditioned problems, different algorithm such as altitude-hold method, clock bias predictor and pseudorange predictor are used in this thesis. In order to attain more accurate positioning, we consider different methods to find the best positioning solution : LS(Least Squares), WLS(Weighted Least Squared), KF(Kalman Filter), RKF(Robust Kalman Fitler) and CF(Constraint Filtering). For different number of visible satellites, the applicability of different methods depends on the capability of the adopted method. In addition to solving the ill-conditioned problem, static and dynamic experiments were conducted by using the simulator SimGen to evaluate the duration of usage of various schemes for acceptable errors.