Vehicle Positioning Using Error State Kalman Filter
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing vehicle positioning and orientation systems face challenges in achieving precise determination due to limitations in integrating inertial measurement data with odometer and satellite geolocation data, leading to inaccuracies and reduced precision.
Innovation Solution
A method utilizing a closed-loop integration scheme with an Error State Kalman Filter (ESKF) that combines inertial navigation data with odometer and satellite geolocation data, correcting position and orientation estimates by accounting for measurement biases and scale factor variations, and averaging angular speed measurements over intervals to improve accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If inertial measurement integration module constructs position and orientation from accelerometer and gyrometer measurements, then continuous position and orientation estimation is achieved, but measurement precision deteriorates due to accumulation of errors
Solution Approach 1:
The correction module uses feedback from multiple sensors (odometer, satellite geolocation unit) to continuously correct the position and orientation estimates generated by the inertial measurement integration module. This feedback loop compensates for error accumulation and maintains measurement precision while preserving continuous estimation capability.
Solution Approach 2:
The system merges data from multiple sensor sources (accelerometer, gyrometer, odometer, satellite geolocation unit) through a unified correction module that processes all measurements together. This combination allows the system to leverage the strengths of each sensor while compensating for their individual weaknesses, resolving the contradiction between continuous estimation and precision.
2Measurement precision
If correction module uses odometer and satellite geolocation measurements to correct position estimates, then measurement precision improves, but device complexity increases
Solution Approach 1:
The correction module is designed with multi-functionality, handling corrections from multiple sensor types (odometer, satellite geolocation unit) through a unified processing architecture. This universal approach allows the system to improve measurement precision using multiple sensors while managing complexity through a single integrated correction mechanism rather than separate processing paths.
3Measurement precision
If system integrates multiple sensor measurements, then measurement precision improves, but processing time increases
Solution Approach 1:
The system performs preliminary processing of sensor measurements by pre-calculating correction values and preparing correction data in advance. This preliminary action reduces the computational burden during real-time position estimation, allowing the system to integrate multiple sensor measurements for improved precision without significant increases in processing time.
Data Source
Figure 1~2
Figure 3
Figure 4~5
AI summary
This process includes: - for several instants tk between instants toj-1 and toj, the construction (84) of an estimate of the instantaneous value of the physical quantity from the measurements of an inertial navigation unit, then - the construction (106) of an estimate of the physical quantity for the instant toj by calculating the arithmetic mean of these instantaneous values constructed, then - the calculation (108) of a difference between a measurement of this physical quantity obtained from the measurement of an odometer and this estimate of the physical quantity at the instant toj, and - the correction (110), as a function of the difference calculated for the instant toj, of the estimated position and orientation of the vehicle, to obtain a corrected position and a corrected orientation.