Automatic identification algorithm for freeway bottleneck

A bottleneck is defined as a point upstream of which exists a queue and downstream of which exists freely flowing traffic. Bottleneck automatic identification technology is based on loop detector data using identification algorithm to identify bottleneck locations and their activation and deactivation times. This research is an improvement to previous speed-based bottleneck recognition methods, and various parameter determinations of standards and algorithm evaluation standards were discussed. The proposed algorithm was applied to detect the bottlenecks in Shanghai inner ring viaduct Dabaishu-Guangzhong Road section, and verified by cumulative curve method. It was found out that the occupancy threshold value based algorithm had a high precision. The proposed real-time bottleneck detection method is useful for transportation authorities to identify bottlenecks as soon as possible and track their impact over time.