GNSS Correction Modeling With Gaussian Processes for Precise Positioning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current GNSS solutions face inaccuracies and computational challenges, particularly in Network RTK, where model parameter space increases non-linearly with network size, and Kalman filters struggle with time-variant parameter modeling.
Innovation Solution
Implementing Gaussian processes to update positioning corrections generation models, allowing for more accurate satellite positioning by explicitly modeling parameter time variance and reducing computational complexity.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional differential correction methods (e.g., RTK, PPP) are used to achieve high positioning precision, then positioning accuracy can reach centimeter or decimeter level, but the system complexity and computational resources required increase significantly
Solution Approach 1:
The patent uses pre-computed, time-limited correction messages that are generated offline and distributed to receivers. These correction messages have limited validity periods and are replaced regularly, allowing the system to achieve high precision without requiring complex real-time computational infrastructure at the receiver end.
Solution Approach 2:
The system performs correction computations in advance during offline processing, generating correction messages before they are needed. This preliminary action transfers the computational burden from the receiver to the reference station, allowing receivers to simply apply pre-computed corrections rather than perform complex real-time calculations.
2Measurement precision
If real-time kinematic (RTK) or precise point positioning (PPP) methods are employed to achieve centimeter-level accuracy, then positioning precision is improved, but the time required to achieve fixed solution and computational load increase
Solution Approach 1:
The system pre-computes correction messages offline that incorporate all necessary atmospheric and orbital corrections. When receivers apply these pre-prepared corrections, they achieve high precision positioning immediately without requiring the extended convergence time needed by traditional RTK or PPP methods to compute corrections in real-time.
Solution Approach 2:
The patent introduces a correction message as an intermediary that carries pre-computed correction data from reference stations to receivers. This intermediary enables receivers to obtain high-precision positioning results quickly by applying ready-made corrections rather than performing lengthy real-time computations.
3Reliability
If multiple reference stations are deployed to improve correction coverage and accuracy, then positioning reliability increases, but the infrastructure cost and system complexity increase
Solution Approach 1:
The patent creates a universal correction message format and distribution system that can serve multiple reference stations and various types of receivers (single-frequency, dual-frequency, multi-constellation) through a single infrastructure. This multi-functional approach allows the same correction infrastructure to support diverse positioning needs without requiring separate systems for each reference station or receiver type.
Data Source
Figure 1
Figure 2
Figure 3A~3B
AI summary
A system and method for generating a set of GNSS corrections using a GNSS corrections model comprising a Gaussian process.