Host Vehicle Position Estimation with Dynamic Weighting
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing host vehicle position estimation methods using target objects on the road face instability due to factors like scraped marking lines or erroneous detection, leading to unsteady position estimates.
Innovation Solution
A host vehicle position estimation device that incorporates a target object recognition unit, observation position estimation unit, prediction position calculation unit, and unsteadiness determination unit, which uses a Kalman filter to dynamically weight prediction positions based on vehicle speed and yaw rate to stabilize position estimates, and includes an azimuth angle estimation for improved accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If host vehicle position estimation uses only target object recognition results, then the position estimation can be obtained from external sensor detection, but unsteadiness occurs due to target object state changes such as scraped marking lines or erroneous detection
Solution Approach 1:
The patent combines target object recognition results with prediction position information derived from vehicle speed and acceleration data. By merging these two independent information sources, the system achieves more reliable position estimation that maintains stability even when target object recognition becomes unstable due to scraped marking lines or detection errors.
Solution Approach 2:
The system uses unsteadiness determination to monitor the stability of position estimation results. When unsteadiness is detected in target object recognition, the system provides feedback to adjust the weighting toward prediction position information, thereby maintaining reliable position estimation under varying conditions.
2Device complexity
If fixed ratio weighting is used for prediction position and observation position, then the estimation process is simple, but it cannot suppress unsteadiness when target object detection becomes unreliable
Solution Approach 1:
The patent implements dynamic weighting adjustment based on the unsteadiness of position estimation results. Instead of using fixed ratio weighting, the system adaptively changes the weighting between prediction position and observation position according to real-time detection reliability, thereby suppressing unsteadiness while maintaining manageable system complexity through automated adjustment.
Solution Approach 2:
The system changes the weighting parameter dynamically based on unsteadiness determination. When target object detection becomes unreliable, the weighting parameter is adjusted to favor prediction position information, allowing the system to maintain stable position estimation without requiring complex manual intervention or system redesign.
3Reliability
If more weighting is given to prediction position when unsteadiness is detected, then position estimation stability improves, but the response to actual position changes may be delayed
Solution Approach 1:
The system uses continuous unsteadiness determination as feedback to dynamically adjust weighting. When unsteadiness is detected, weighting shifts toward prediction position for stability; when unsteadiness subsides, weighting automatically increases for observation position to improve responsiveness. This feedback mechanism resolves the trade-off between stability and responsiveness.
Solution Approach 2:
The weighting between prediction and observation positions is made dynamic rather than fixed. The system adapts the weighting in real-time based on detection quality, allowing it to prioritize stability when needed and responsiveness when target object detection is reliable, thereby resolving the contradiction between these two opposing requirements.
Data Source
AI summary
A host vehicle position estimation device includes an observation position estimation unit configured to estimate an observation position of the vehicle based on a result of recognition of the target object performed, a prediction position calculation unit configured to calculate a prediction position of the vehicle from a result of estimation of the host vehicle position in the past based on a result of measurement performed by an internal sensor, a host vehicle position estimation unit configured to estimate the host vehicle position based on the observation position and the prediction position. The host vehicle position estimation unit is configured to give more weighting to the prediction position in the estimation of the host vehicle position such that the host vehicle position is estimated to be close to the prediction position if it is determined that a result of estimation of the host vehicle position is unsteady.