Route Search Using Probe Segments for Driver Know-how
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing route search methods using probe car data often fail to optimize routes holistically, as they divide running history into link units, losing continuous information and not always presenting information about directly connecting routes, leading to locally optimized but discontinuous route suggestions.
Innovation Solution
A route searching system that uses probe car data to identify main branch nodes and divide data into probe segments, allowing for the generation of derived routes that reflect the running history of multiple contiguous links, ensuring continuity and user preference by evaluating and selecting routes based on frequency and ease of driving.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If probe car data is divided into link units for route searching, then running frequency per link can be obtained, but continuous information straddling multiple links is lost and route continuity is not considered
Solution Approach 1:
The patent segments probe car data not by arbitrary link units but by meaningful running patterns identified through clustering. Continuous running information is preserved by grouping links into pattern-based segments that maintain the temporal and spatial continuity of actual driving behavior, resolving the contradiction between precise frequency measurement and information loss.
Solution Approach 2:
The patent introduces running patterns as an intermediary layer between raw probe car data and route searching. These patterns act as mediators that aggregate continuous running information across multiple links while preserving the essential characteristics needed for frequency analysis, thus preventing information loss during the segmentation process.
2Productivity
If route search optimizes individual links based on high running frequency, then local link optimization is achieved, but overall route continuity and holistic optimization are not considered
Solution Approach 1:
The patent performs preliminary clustering of probe car data to identify running patterns before the actual route searching process. This preliminary action organizes continuous running information into reusable patterns that can be applied during route search, ensuring both local link optimization and overall route continuity are considered simultaneously.
Solution Approach 2:
The patent merges individual link optimizations by combining them into coherent routes using identified running patterns. Instead of treating links in isolation, the system combines link-level frequency information with pattern-level continuity requirements, achieving holistic route optimization that maintains both efficiency and reliability.
3Device complexity
If probe car data is processed per link, then data processing complexity is reduced, but information about routes directly connecting departure to destination is lost
Solution Approach 1:
The patent applies segmentation based on actual running patterns rather than fixed link boundaries. This pattern-based segmentation naturally preserves route connection information by grouping links that are actually traversed together in real driving, maintaining route integrity while keeping processing complexity manageable through pattern reuse.
Solution Approach 2:
The patent creates copies of identified running patterns and applies them to multiple route searches. Instead of reprocessing raw data for each route query, the system copies and reuses pattern information, significantly reducing processing complexity while preserving complete route connection information across different departure-destination pairs.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
It is an object of the present invention to provide a route searching which can reflect know-how such as easiness of running included in the probe car data in an entire route from departure place to destination. A center device detects main branch nodes from probe car data received from an in-vehicle terminal device by means of a main branch node detecting section (105). A probe car data dividing section (107) divides the probe car data into probe segments by the main branch nodes. A route dividing section (111) divides an initial route generated by an initial route generating section based on specified departure place and destination, and a derived route generating section (112) substitutes probe segments for the divided routes so as to generate derived routes. The route selecting section (113) scores the derived routes and select one so as to provide a recommended route in which know-how such as easiness of running included in the probe car data is reflected.