Testing method and device for driverless automobiles

Abstract

The invention discloses a testing method and device for driverless automobiles. The method in one embodiment comprises steps of: selecting testing scenario parameter combinations for testing the performance of the driverless automobile from preset testing scenario parameter combinations; determining, on the basis of preset values and preset weighted value corresponding to various types of testing scenario parameters in the testing scenario parameter combinations, the complexity parameter of a testing scenario corresponding to each testing scenario parameter combination; aggregating the complexity parameters with values corresponding to the same preset value section; generating indicative information indicating whether the driverless automobile passes the testing scenario corresponding to the complexity parameter on the basis of a testing result whether the driverless automobile passes the testing scenario corresponding to the complexity parameter. The testing method and device quantify the complexities of the testing scenarios and may further determine various testing scenarios corresponding to the same difficulty so as to increase the testing accuracy of the driverless automobiles.

Description

technical field [0001] The present application relates to the field of artificial intelligence, specifically to the field of unmanned vehicles, and in particular to a testing method and device applied to unmanned vehicles. Background technique [0002] Vehicle performance testing is an important means to improve vehicle safety. In the test of the performance of the vehicle, different test scenarios need to be set, and then, the performance of the vehicle can be judged according to whether the vehicle can pass the test of test scenarios of different complexity. At present, the complexity of the test scene is usually manually determined, and the performance of the vehicle is tested under different complexity. [0003] However, there is a lack of quantification of the complexity of the test scenarios when testing vehicles in the above manner. On the one hand, there will be a problem of distortion of test results caused by manual judgment of the complexity of the test scenario...

AI Technical Summary

Problems solved by technology

[0003] However, there is a lack of quantification of the complexity of the test scenarios when testing vehicles in the manner described above

On the one hand, there will be the problem of distortion of test results caused by manual judgment of the complexity of the test scene

On the other hand, it is impossible to classify the test difficulty based on the complexity of the test scene, and it is difficult to accurately test the performance of the vehicle under different difficulties

Images