Intersection Travel Path Data Generation Using Trajectory Fitting
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for generating travel path data inside intersections are inaccurate and unrealistic, as they rely solely on GPS positions or unrealistic assumptions about vehicle trajectories, making it difficult to create reliable data for automated driving, especially on local roads and intersections.
Innovation Solution
A travel path data generation apparatus that fits an estimated vehicle trajectory into lane network data using both absolute GPS trajectories and sensor-derived trajectories, optimizing the connection between travel path data and lane network data to generate accurate travel paths inside intersections.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If GPS positions alone are used to generate travel path data, then the data generation process is simple, but the accuracy and reliability of the travel path data inside intersections deteriorates
Solution Approach 1:
The patent combines multiple data sources including GPS positions, sensor-derived trajectories (accelerometers, gyroscopes, magnetometers), and map data to generate travel path data. This merging of multiple measurement systems compensates for the weaknesses of individual sources, particularly GPS signal degradation in intersection areas, thereby improving overall accuracy while maintaining a manageable processing framework.
Solution Approach 2:
The travel path data is constructed as a composite by integrating information from heterogeneous sources: satellite-based GPS, inertial sensors, magnetic field sensors, and pre-existing map data. This composite approach creates a more robust and accurate representation of vehicle trajectories by leveraging the complementary strengths of each data source.
2Measurement precision
If sensor-derived trajectories are used to improve accuracy, then the measurement precision improves, but the device complexity increases
Solution Approach 1:
The patent introduces map data and trajectory fitting algorithms as intermediaries between raw sensor data and final travel path data. These intermediaries process and reconcile the complex sensor inputs (multiple accelerometers, gyroscopes, magnetometers) with GPS positions and lane network data, reducing the direct complexity burden while maintaining high measurement precision through sophisticated data fusion.
Solution Approach 2:
The system replaces complex mechanical measurement systems (dedicated vehicles with expensive sensors) with a software-based data fusion approach using standard sensors in automated driving vehicles. The complexity is shifted from physical hardware to computational algorithms for trajectory estimation and data fusion, achieving comparable or superior accuracy without dedicated measurement vehicles.
3Measurement precision
If dedicated vehicles with expensive sensors are used, then the measurement precision improves, but the productivity and cost-effectiveness deteriorates
Solution Approach 1:
The patent enables automated driving vehicles to generate their own high-precision travel path data using their existing sensors (GPS, accelerometers, gyroscopes, magnetometers) without requiring external dedicated measurement vehicles. This self-service capability allows regular fleet vehicles to contribute to map data improvement, dramatically increasing productivity and eliminating the need for expensive specialized equipment and manual operations.
Solution Approach 2:
The system transforms automated driving vehicles from单纯的 driving machines into multi-functional platforms that simultaneously perform automated driving and high-precision mapping functions. By utilizing the same vehicle sensors for both navigation and map generation, the system achieves dual purposes, improving productivity while reducing the need for specialized dedicated vehicles and expensive equipment.
4Ease of manufacture
If unrealistic assumptions about vehicle trajectories are made, then the data generation process is simplified, but the reliability of travel path data deteriorates
Solution Approach 1:
The patent employs feedback mechanisms where estimated trajectories are continuously refined by comparing sensor-derived positions with GPS positions and map data. The system uses trajectory fitting algorithms that adjust estimated paths based on actual vehicle behavior patterns and lane network constraints, ensuring realistic and reliable travel path data without requiring complex manual interventions or unrealistic assumptions.
Data Source
AI summary
A travel path data generation apparatus for generating travel path data inside an intersections for automated driving includes a travel path data generator that generates the data of the travel path in such a way that, using an absolute trajectory of actual traveling of a vehicle inside the intersection, the travel path data generator fits an estimated trajectory of the actual traveling of the vehicle inside the intersection into data of lane network connected to the intersection.