本論文提出一種新的M-estimation方式調控的穩健無跡卡爾曼濾波器(RUKF),用於在具有各種非視距(NLOS)出現概率的環境中進行移動定位。由於合併了提出的測量驗證,因此提出的RUKF可以顯著降低NLOS測量雜訊所帶來的不利影響。此外,將基於M-estimation自適應調控測量雜訊協方差的測量雜訊協方差調控方法合併到濾波器中,以增強追蹤性能。通過電腦模擬證明了所提出的RUKF的有效性。
This paper proposes a new M-estimation based robust unscented Kalman filter (RUKF) for mobile positioning in environments with various probabilistic non-line-of-sight(NLOS) occurrences. Due to the incorporation of the proposed measurement validation, the proposed RUKF can significantly mitigate the adverse effect of NLOS measurement noises. Furthermore, owing to the incorporation of M-estimation methodology based adaptive measurement noise covariances, the RUKF enhances tracking performance. Simulations demonstrate the effectiveness of the proposed RUKF.