High-Resolution Route Aggregation for Efficient Map Matching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Mapping service providers face significant challenges in reducing computation while maintaining navigation accuracy and efficiency, particularly in processing high volumes of vehicle sensor data for traffic anomaly detection.
Innovation Solution
The method involves receiving a sequence of location data points, generating an initial link set of traveled link options, determining probable matching road links, computing probabilities for pairs of traveled and matching road links, and aggregating these pairs to reduce the number of link options, thereby reducing computation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If probability-based route-building methods (e.g., Hidden Markov Model) are used to accurately map-match all vehicle location data, then navigation accuracy is maintained, but computational resources and processing time increase significantly
Solution Approach 1:
The patent segments the vehicle trajectory into multiple path segments between consecutive location points. For each segment, multiple candidate road links are identified and evaluated independently using probability calculations. This segmentation allows the system to process complex routes by breaking them into manageable segments, maintaining accuracy while reducing overall computational burden through localized analysis.
Solution Approach 2:
The patent applies partial action by calculating probabilities only for a limited set of candidate road links that are spatially and temporally relevant to each trajectory point, rather than evaluating all possible road links in the network. The system generates candidate links within a predefined distance threshold and time window, performing probability-based matching only on this reduced subset, thus significantly reducing computational resources while preserving navigation accuracy.
2Loss of information
If all vehicle trajectory points are processed through map-matching algorithms, then complete route information is obtained, but processing time increases
Solution Approach 1:
The patent performs preliminary actions by pre-calculating and storing road link attributes, spatial relationships, and temporal information before actual map-matching occurs. Candidate road links are pre-identified based on spatial proximity and temporal plausibility, and their attributes are pre-computed. This preliminary preparation enables faster real-time processing by eliminating redundant calculations during the actual route-building phase.
Solution Approach 2:
The patent applies local quality by adjusting processing intensity based on local trajectory characteristics. For trajectory points in areas with high road density or ambiguous geometry, the system generates more candidate links and performs more thorough probability evaluations. For points in areas with clear, unambiguous road geometry, the system reduces candidate link generation and processing effort. This adaptive local quality approach ensures complete route information is captured where needed while reducing processing time in simpler areas.
3Measurement precision
If high-density sampling of vehicle location data is performed to improve route accuracy, then navigation precision increases, but data processing complexity increases
Solution Approach 1:
The patent implements dynamics by making the candidate link generation and probability calculation processes adaptive to local conditions. The system dynamically adjusts the number of candidate links generated, the distance thresholds for candidate selection, and the probability evaluation depth based on local road network complexity, trajectory quality, and geometric characteristics. This dynamic adaptation maintains high route accuracy in complex areas while reducing processing complexity in simpler areas.
Solution Approach 2:
The patent applies parameter changes by modifying processing parameters such as candidate link distance thresholds, time window durations, and probability calculation precision based on local trajectory density and road network characteristics. In areas with high trajectory density, the system uses tighter thresholds and more precise calculations. In areas with lower density, it relaxes parameters to reduce computational complexity. This parameter adaptation maintains route accuracy while managing data processing complexity.
Data Source
AI summary
An approach is provided for aggregating an incident route based on high-resolution sampling. The approach involves receiving location data points representing a path traveled. The approach also involves generating an initial link set comprising a first predetermined number of traveled link options based on an initial location data point. The approach further involves determining a second predetermined number of probable matching road links of the geographic database for a next location data point. The approach further involves computing probabilities for respective pairs of the traveled link options and the probable matching road links. Each probability indicates a pair of a likelihood that a traveled route option and a probable matching road link are connected. The approach further involves aggregating the respective pairs to reduce a number of the respective pairs into the first or another predetermined number of traveled link options of a new link set based on the probabilities.