A lot of time and resources are required to construct autonomous vehicle (AV) proving ground, so it is necessary to design the proving ground with the design goal of improving the testing ability of proving ground. In the paper, a mathematical optimization model of road planning based on genetic algorithm for AV proving ground continuous loop area is proposed, the coverage rate of test cases is used as an index to measure the testing capability of continuous loop area of AV proving grounds. This method takes the continuous loop area boundary and the requirements of test cases as input, and generates the road planning scheme of continuous loop area automatically, so that the continuous loop area can cover the test cases requirements to the largest extent in any given limited space, and avoids the process of continuously adjustment relying on subjective experience. Finally, the method is validated by an example of real AV proving ground. The results show that compared with the planning based on expert experience, the test cases coverage rate of the planning obtained by this method is increased by 16.2% and 10.4% respectively under the two test cases requirement libraries.

CICTP