Ridesharing Path Computation Using Trajectory Block Overlap
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for finding ridesharing paths, such as the grid map method, often overlook potential candidates by only considering paths within the same block, even if adjacent blocks are close, leading to a lack of discovered ridesharing opportunities.
Innovation Solution
A computing method that determines a sequence of blocks for both target and candidate trajectories using GPS points, comparing these blocks to identify overlapping sections and calculate ridesharing paths, allowing for a more comprehensive search of potential ridesharing paths without relying on a road network.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If the grid map method is used to divide the map into blocks and only consider paths in the same block, then the computation is simplified and faster, but many potential ridesharing paths in adjacent blocks are overlooked
Solution Approach 1:
The patent divides the trajectory into multiple blocks based on spatial distance thresholds. Instead of using a fixed grid map, it segments the actual trajectory into blocks that represent meaningful segments of the ride path. This allows comparison between blocks of different trajectories to identify ridesharing opportunities while maintaining computational efficiency.
Solution Approach 2:
The patent introduces a temporal dimension to the block comparison by considering not only spatial overlap but also time overlap between trajectories. By comparing blocks in both space and time dimensions, it identifies ridesharing paths more accurately without requiring road network data, thus resolving the contradiction between simplified computation and comprehensive path discovery.
2Measurement precision
If road network data is used to calculate ridesharing paths, then the accuracy of path identification is improved, but the device complexity and data requirements increase
Solution Approach 1:
The patent extracts only the necessary information from GPS trajectory data - specifically the sequence of blocks formed by trajectory points - and performs comparison based on these extracted blocks. It deliberately excludes road network data and other complex external data sources, achieving accurate ridesharing path identification using only the extracted trajectory information, thus reducing device complexity while maintaining precision.
Solution Approach 2:
The method uses the trajectory data itself to define the blocks and perform comparisons, without requiring external road network data. The trajectory points self-organize into blocks that naturally represent the ride path segments, and the comparison is performed directly on these self-generated blocks, making the system self-sufficient and reducing external data requirements.
Data Source
AI summary
A computing method for ridesharing paths, a computing apparatus and a recording medium using the same are provided. This invention finds the ridesharing paths out from the target trajectory which have ridesharing relationship with a plurality of candidate trajectories. People on the ridesharing paths can share a ride on a part of or whole target trajectory. In the computing method, a trajectory is transformed into a sequence of blocks. After target trajectory and a plurality of the candidate trajectories are transformed, it is determined whether each of the blocks of the candidate trajectories overlaps with at least one of the blocks of the target trajectory. Then, ridesharing paths are calculated according the screened candidate trajectories. This method is many-to-one method which compares a plurality of the candidate trajectories to one target trajectory at one time.