Autonomous Vehicle Localization Correction via Predictive Feedback
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Autonomous driving vehicles face inaccuracies in position determination due to measurement errors from sensors like IMUs and GPS, affecting motion planning and control.
Innovation Solution
A dynamic model comprising a localization predictive model and a localization correction model is trained using driving statistics data to calibrate the predicted position of the vehicle, integrating sensor data to improve accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If sensor output (IMU/GPS) is used to determine vehicle position, then localization can be achieved, but measurement errors cause mismatch between sensor output and actual position
Solution Approach 1:
The system uses a feedback mechanism where the predicted position from the dynamic model is compared with actual sensor measurements, and the localization correction model adjusts the predicted position based on this feedback to compensate for sensor errors
Solution Approach 2:
A localization correction model acts as an intermediary between the dynamic model and sensor output, correcting the predicted position by learning the systematic errors from historical driving data and applying corrections to improve accuracy
2Measurement precision
If a dynamic model with localization correction is implemented, then position accuracy is improved, but system complexity increases
Solution Approach 1:
The localization correction model is trained offline using historical driving data before deployment, performing the complex learning task in advance rather than during real-time operation, thus reducing online computational complexity
Solution Approach 2:
The system uses a predictive model that copies and simulates vehicle dynamics based on historical patterns, allowing the correction model to learn from simulated trajectories without requiring complex real-time sensor fusion
Data Source
AI summary
In one embodiment, a set of parameters representing a first state of an autonomous driving vehicle (ADV) to be simulated and a set of control commands to be issued at a first point in time. In response, a localization predictive model is applied to the set of parameters to determine a first position (e.g., x, y) of the ADV. A localization correction model is applied to the set of parameters to determine a set of localization correction factors (e.g., Δx, Δy). The correction factors may represent the errors between the predicted position of the ADV by the localization predictive model and the ground truth measured by sensors of the vehicle. Based on the first position of the ADV and the correction factors, a second position of the ADV is determined as the simulated position of the ADV.