The 20th and 21st joint COTA International Conference of Transportation Professionals (CICTP 2020-21)

A lot of resources are required to construct autonomous vehicles (AVs) proving ground, so it is necessary to design proving ground with the goal of improving its testing ability. In this paper, a mathematical optimization model of road horizontal alignment design based on genetic algorithm (GA) for continuous loop area in AVs proving ground is proposed, the test case coverage is used as an index to measure the testing capability of it. This method takes continuous loop area boundary and test cases of autonomous vehicles as input, and generates road horizontal alignment design of continuous loop area automatically. Finally, the method is validated by an example of practical AVs proving ground. The result shows that compared with the design based on subjective experience, the test cases coverage of the design obtained by this method is increased by 16.2% and 10.4% under the two different test cases databases respectively.

CICTP