Hidden Markov Model Position Correction for GPS Accuracy
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing satellite navigation systems for movable objects, such as vehicles, have limited accuracy and are not suitable for real-time position correction, especially in urban environments where line-of-sight to satellites is obstructed, leading to inaccurate GPS positioning.
Innovation Solution
A method using a hidden Markov model (HMM) with the Viterbi algorithm to correct position estimates by determining the most likely road segment a vehicle is on, based on a sequence of position estimates, and transmitting the corrected position in real-time, while filtering out inaccurate estimates and incrementally updating the HMM instance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If satellite navigation system is used for position estimation, then position estimation is available in real-time, but accuracy is limited
Solution Approach 1:
The patent introduces an intermediary system consisting of multiple reference receivers at known positions and a central server that processes carrier phase measurements. This intermediary infrastructure mediates between the satellite signals and the mobile receiver, providing differential corrections that significantly improve positioning accuracy while maintaining real-time capability
Solution Approach 2:
The system implements feedback by continuously monitoring carrier phase measurements from multiple reference receivers, computing differential corrections at the central server, and applying these corrections to the mobile receiver's position estimates. This closed-loop feedback mechanism maintains high accuracy in real-time operation
2Measurement precision
If traditional GPS positioning is used, then positioning is simple to implement, but accuracy deteriorates in urban environments with obstructed satellite signals
Solution Approach 1:
The patent deploys reference receivers at known positions throughout the service area as intermediaries. These reference receivers capture satellite signals at multiple locations, enabling the system to compute and correct for local environmental effects such as multipath and signal blockage, thereby improving accuracy in urban environments without requiring complex modifications to mobile devices
Solution Approach 2:
The system merges data from multiple reference receivers with the mobile receiver's measurements through differential processing. By combining observations from multiple geographically distributed reference stations, the system achieves robust positioning performance in challenging urban environments while keeping individual receiver hardware relatively simple
3Measurement precision
If high-accuracy position correction is implemented, then positioning accuracy is improved, but processing time and computational complexity increase
Solution Approach 1:
The system performs preliminary actions by continuously collecting and storing carrier phase measurements from multiple reference receivers at known positions. This pre-collected data is ready for rapid differential processing when a mobile receiver requests position correction, enabling high-accuracy results to be computed quickly without requiring time-consuming real-time calculations
Solution Approach 2:
The positioning system is segmented into independent functional components: reference receivers collecting data independently, a central server performing differential calculations, and mobile receivers applying corrections. This segmentation allows parallel processing of measurements from multiple reference stations, reducing overall computation time while maintaining high accuracy
Data Source
AI summary
Aspects concern a method for correcting position estimates of a movable object. According to various embodiments, the method comprises establishing (1001) a hidden Markov model HMM, instance for a movable object and, for positioning times of a sequence of positioning times, receiving (1002) a position estimate from a positioning device of the movable object for a respective positioning time, determining (1003) a set of candidate path segments for the positioning time, determining (1004) likelihoods for the candidate path segments to correspond to the position estimate by application of the Viterbi algorithm to the HMM instance, expanding (1005) the HMM instance by the determined likelihoods for the candidate path segments for the positioning time and determining (1006) a corrected position estimate from a candidate path segment of the set of candidate path segments with the highest likelihood.