Self-adaptive traffic incident detection method based on small step adjustment

Abstract

The invention discloses a self-adaptive traffic incident detection method based on small step length adjustment. The method comprises the following steps: Step 1, dynamically determining a detection frequency; Step 2, calculating traffic parameters and traffic parameter statistics; Step 3, calculating traffic parameter characteristics; Step 4, carrying out threshold comparison, and outputting a traffic incident detection result; and Step 5, calculating a historical false drop rate, and carrying out small-step adjustment on the traffic parameter characteristic threshold according to the historical false drop rate. According to the method, the detection frequency is dynamically determined, ETC portal data is acquired, then traffic parameter characteristics including the road section average speed variation and the vehicle speed dispersion variation are calculated according to the acquired data, meanwhile, the traffic parameter characteristics are compared with threshold values of the traffic parameter characteristics, a traffic event detection result is determined, finally the historical false detection rate is calculated, an updating formula of a traffic parameter characteristic threshold is determined according to a historical false detection rate, so as to carry out small-step adjustment on the threshold, thereby realizing adaptive real-time efficient detection of the traffic event.

Description

technical field [0001] The invention relates to the technical field of road segment traffic event detection, in particular to an adaptive traffic event detection method based on small step size adjustment. Background technique [0002] Traffic incidents refer to traffic incidents that occur non-periodically and reduce the traffic capacity of a certain section of road, such as traffic accidents, breakdown parking, scattered goods, road maintenance, vehicles going against the road, and traffic jams. The number of motor vehicles in my country is large, the road network environment is complex, and the frequency of traffic accidents remains high, which not only seriously endangers the travel safety of travelers, but also causes great losses to the social economy. Therefore, in order to prevent and reduce the occurrence of traffic incidents, solve the traffic delays caused by traffic incidents, and avoid secondary damage caused by traffic incidents, it is necessary to detect traff...

AI Technical Summary

Problems solved by technology

[0005] Finally, the traditional traffic event detection algorithm based on the threshold method uses a fixed threshold, but as time changes, the traffic patterns of different time periods may be different, and the fixed threshold cannot accurately reflect the road section of each time period. Time conditions, so the threshold needs to be adaptively adjusted to improve the accuracy of traffic incident detection

Images