Route Calculation System Using Driver Preference Weighting
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing route calculation systems fail to provide recommended routes that reflect driver's route selection know-how for roads with limited or no probe data, leading to unsatisfactory route guidance.
Innovation Solution
A route calculation system and method that uses weighting parameters calculated from probe data to adjust road link costs, ensuring route recommendations consider driver preferences even on roads with incomplete data by classifying routes based on probability and attributes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If route search uses only probe data from roads with sufficient data, then route recommendations reflect driver know-how accurately, but roads with limited or no probe data cannot provide recommended routes
Solution Approach 1:
The system pre-calculates weighting parameters for road links using probe data during periods when data is available, storing these parameters for later use. This preliminary action enables the system to provide route recommendations on roads even when real-time probe data is insufficient or unavailable, as the pre-computed weighting parameters can be applied to maintain route recommendation accuracy across all roads including those with limited data coverage
Solution Approach 2:
The system introduces weighting parameters as an intermediary element that bridges the gap between roads with sufficient probe data and roads with limited or no data. These weighting parameters, derived from probe data analysis, serve as a mediator that allows the route search algorithm to extend driver know-how based recommendations to roads lacking direct probe data coverage, thus improving both accuracy and coverage simultaneously
2Reliability
If route search prioritizes roads with multiple probe data records, then route recommendations reflect driver preferences accurately, but roads with poor or no probe data are excluded from recommendations
Solution Approach 1:
The system changes the parameter representation by introducing weighting parameters that quantify driver preferences and route characteristics. Instead of requiring multiple probe data records for each road, the system uses these derived parameters to represent road link characteristics, allowing reliable route recommendations even for roads with limited data. The weighting parameters transform the availability issue into a parameter estimation problem that can be solved with fewer data points
Solution Approach 2:
The system creates a simplified representation (copy) of the complex probe data through weighting parameters. Rather than requiring the full probe data records for route calculation, the system uses these compact parameter copies that capture the essential driver preference information. This copying approach maintains route selection reliability while reducing the data requirements, enabling recommendations on roads with poor probe data coverage
3Productivity
If conventional route search minimizes time and distance, then efficient routes are provided, but routes may include narrow roads or roads with poor visibility that compromise comfort and safety
Solution Approach 1:
The system changes the cost parameters used in route search by introducing weighting parameters that incorporate comfort and safety considerations. Instead of using only time and distance as cost metrics, the weighting parameters modify the link costs to reflect driver preferences and route quality factors such as road width, visibility, and other comfort-related attributes. This parameter change enables the route search to balance efficiency with comfort and safety, providing routes that satisfy multiple criteria simultaneously
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A route calculation system includes: a route selection probability calculation unit that calculates a route selection probability of each route; a weighting parameter acquisition unit that acquires at least two kinds of weighting parameters representing degrees of influences of at least a route traveling distance and a route traveling time on the route selection probability; a weighting parameter selection unit that selects at least two kinds of weighting parameters corresponding to a specified set of a starting point and ending point combination related to a combination of a departure point and a destination; and a recommended route calculation unit that calculates a recommended route based on at least two kinds of link costs including a link traveling distance and a link traveling time and based on the at least two kinds of weighting parameters corresponding to the specified set of the starting point and ending point combination.