Intersection-Specific Driver Sensitivity Determination

The patent segments the entire recorded vehicle velocity information into intersection-specific data groups. Instead of analyzing all velocity data uniformly, the system divides the data by intersection, allowing separate determination of standard driving operation levels for each intersection. This segmentation enables accurate comparison of driving operations at the same intersection while eliminating the diluting effect of variations between different intersections, thereby resolving the contradiction between simple processing and accurate determination.

The patent applies local quality by determining standard driving operation levels separately for each intersection rather than using a unified standard for all intersections. The base station identifies intersections where the standard driving operation levels match between comparison targets, ensuring that the determination is based on locally consistent driving patterns at the same intersection. This local quality approach maintains sufficient data volume while improving prediction accuracy for specific intersection scenarios.

The patent uses copying by creating a reference standard driving operation level for each intersection based on aggregated data from multiple vehicles. Instead of directly comparing individual velocity data that varies with visibility and traffic conditions, the system creates a standardized reference copy that represents the typical driving operation at each intersection. This copying approach eliminates the harmful variations caused by environmental factors while preserving the essential driving patterns needed for accurate unexpectedness prediction sensitivity determination.