Detection of Abnormal Traffic Flow with the Hurst Parameter for Traffic Management

In order to further reduce emissions of particulate matter, nitrogen oxides and carbon dioxide into the atmosphere from road transport, the causes of congestion on city streets should be eliminated. The movement of transport in urban conditions is quite cyclical, however, in the conditions of each city it has its own specifics. The use of adaptive traffic light control systems within the framework of intelligent transport systems (ITS) contributes to solving this problem. However, there is a need not only to directly respond to traffic jams and reduce the speed of vehicles, but also to predict these situations in advance. The article shows an example of using the normalized range method (Hurst exponent) for predictive detection of changes in traffic flow indicators on a city highway for subsequent correction of traffic lights.