Emergence of Global Speed Patterns in a Traffic Scenario

摘要

We investigate different analysis methods for traffic data. The measure for emergence can be used to identify global dependencies in data sets. The measure for target orientation can be used to identify dangerous situations in traffic. We apply these measures in a use case on a data set of the M30 highway in Madrid. The evaluation shows that the measures can be used to predict or to identify abnormal events like accidents in traffic by an evaluation of velocity data or flow data measured by detectors at the road. Such events leads to a decrease of the measures of emergence and target orientation.