Method for quantitative evaluation of traffic complexity on the highway

As a traditional industry, the automobile industry nowadays is facing a lot of new challenges, which appear with the development of telecommunication technology, global climate change, change of consumers’ demand and attitudes, etc. New development trends caused by the influence of these factors are summarized with an acronym ‘C.A.S.E.’, indicating Connectivity, Autonomous, Sharing and Electrification. Among them the trend Autonomous Driving has earned the most attraction. Before a vehicle with automated functions can be launched in the market, large number of tests need to be executed to ensure the performance of these functions in the real-world traffic. There are several principles to generate scenarios for testing, for instance, based on comprehension and based on real-world traffic data. The metric developed in this work is based on the real-world data. However, since the scenarios from real-world traffic can be countless, it is necessary to find out the most representative test cases to make the testing process more effective, efficient and affordable. This work develops a method, which evaluates the traffic situation for automated vehicles on the highway quantitatively and the quantified result will be used as a measurement for degree of complexity of the scenario from the perspective of ego-vehicle. The larger the result is, the more complex is the evaluated scenario. Thus, scenarios with high degrees of complexity can be extracted and used as test scenarios.