.png){kind=icon}
.png){kind=icon}
Return
Random Forest-Based Optimization Algorithm for Shipborne GNSS Vector Tracking Loop
.png){kind=icon}
PRE
AI
DOI：10.1080/01490419.2025.2575974.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Global Navigation Satellite System (GNSS) serves as a pivotal technology for achieving global positioning. Within GNSS receivers, the Vector Tracking Loop (VTL) stands out as an advanced signal processing method, offering enhanced anti-interference capabilities and robustness in dynamic environments compared to the Scalar Tracking Loop (STL). However, traditional VTL face significant challenges under conditions of low signal-to-noise ratio (SNR), signal blockage, multipath effects, and non-line-of-sight scenarios. These challenges can compromise the stability and reliability of the system, leading to substantial positioning errors or navigation failures. To address this issue, this paper proposes an optimization method for shipborne GNSS VTL based on Random Forest (RF). Initially, satellite signal SNR, satellite elevation angle, and coordinate information are identified as key features. The pseudorange and pseudorange rate errors, derived from the VTL under favorable signal conditions, serve as output variables. By employing an improved ensemble bagging decision tree learning approach, predictive models for pseudorange and pseudorange rate errors are developed. A shipborne experiment found that in locations with signal blockage, the proposed RF-based VTL (RFVTL) optimization method enhanced horizontal positioning accuracy by 94% and 72% compared to traditional VTL and VTL with loop filters, respectively.
Keywords:
random forest
vector tracking loop
ship navigation system
Journal
M
IF:
1.4
Papers:
20
Citations:
0