A Traffic Crash Warning Model for BOT E-Tolling Operations Based on Predictions Using a Data Association Framework

As a result of the increasing use artificial intelligence technology in transportation, numerous real-time crash prediction techniques have been developed. In context highway traffic management, machine learning models and classifiers are used to analyze electronic toll collection (ETC) vehicle detector (VD) data predict occurrences. However, accidents influenced by multiple factors, such as speed differences, density, weather conditions, direct associations may not exist between sensor incidents. Therefore, integration association methods must be examine ETC VD through flow theories, extract key from datasets facilitate model training. this study, method framework combined with deep was proposed construct warning for national highways Taiwan. The results revealed accuracy 94%, indicating that had low error rate suitable accidents. Overall, study provides referential Freeway Bureau Taiwan conduct comprehensive assessments develop strategies prevention.