GNSS Post-Processing Without Kalman Filtering for Epoch Accuracy
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing GNSS post-processing techniques rely on Kalman filtering, which assumes normal distribution and independence of errors, leading to inaccurate and optimistic position solutions when the system model is complex or inconsistent, and direct combination of forward and backward solutions ignores consistency and ambiguity distinctions, degrading accuracy.
Innovation Solution
Dynamic application of multiple GNSS post-processing methods to each epoch, determining consistency between epochs and combining solutions based on shared ambiguity values to generate a more accurate combined position solution.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If Kalman filtering is used to process forward and backward GNSS data, then the position solution can be generated efficiently, but the accuracy deteriorates when the system model is overly complex or inconsistent with actual state variables
Solution Approach 1:
The patent segments the processing into multiple independent post-processing methods (e.g., precise point positioning, differential GNSS, carrier phase ambiguity resolution) rather than relying on a single Kalman filter. Each method processes the GNSS data independently, allowing selection of the most appropriate method for each epoch based on data quality and consistency, thereby maintaining accuracy without sacrificing efficiency.
Solution Approach 2:
The patent dynamically selects and switches between different post-processing methods for each epoch based on the consistency check results and data quality metrics. This dynamic adaptation allows the system to optimize accuracy for each specific epoch while maintaining overall processing efficiency, rather than using a static Kalman filter approach.
2Quantity of substance
If forward and backward solutions are directly combined, then the data from two correlated solutions can be combined to enhance raw data output, but the accuracy deteriorates due to ignoring solution consistency and ambiguity distinctions
Solution Approach 1:
The patent implements a feedback mechanism through consistency checks that evaluate the quality and consistency of forward and backward solutions before combination. The system uses ambiguity distinctions and solution consistency metrics to determine whether and how to combine solutions, providing feedback that prevents inaccurate combinations and maintains high accuracy in the final position solution.
Data Source
AI summary
A post-processed GNSS solution from raw GNSS data acquired from a vehicle and optionally other sources is dynamically calculated at the epoch level without using a Kalman filter. Instead, a plurality of GNSS processing methods is applied to each epoch in either the forward or backward direction and the most accurate solution for each epoch of the forward generated solution and backward generated solution is combined. The combined post-processed GNSS solution is determined based on which of the GNSS processing methods were used to generate an epoch of the forward solution and a corresponding epoch of the backward solution.