Multi-dimensional comprehensive evaluation method and device for automatic driving automobile

Abstract

The invention relates to a multi-dimensional comprehensive evaluation method and device for an automatic driving automobile. The method specifically comprises the following steps: screening a test scene set in an automatic driving ODD range; calculating to obtain a test scene environment complexity score and a test scene task complexity score based on the tested scene features; calculating to obtain a vehicle driving autonomy evaluation result, a vehicle traffic coordination evaluation result and a vehicle learning evolutionary evaluation result based on the scene test result; and finally, incombination with results output by the test condition evaluation module and the test result evaluation module, calculating to obtain an 'integrated' intelligent comprehensive performance result of theautomatic driving automobile. The method can comprehensively reflect the advantages and disadvantages of different-dimension intelligent performance of the automatic driving vehicle in different complex environments, solves a problem of multi-dimension evaluation of the automatic driving vehicle under the condition that the height of a man-vehicle environment task complex system is uncertain, andprovides a support for the development of the technology and industry of the automatic driving vehicle.

Description

technical field [0001] The invention relates to the technical field of automatic driving vehicles, in particular to a multi-dimensional comprehensive evaluation method and device for automatic driving vehicles. Background technique [0002] With the development of smart car technology, the penetration rate of smart cars on real roads is gradually increasing. At present, the automatic driving system is still challenged by many factors, so testing and evaluation technology is a key link to support the development of automatic driving technology. [0003] However, the evaluation object of autonomous driving tests has changed from the traditional binary independent system of human and vehicle to a strongly coupled system of human-vehicle-environment-task. The test scenarios and tasks are difficult to exhaust, and the evaluation dimensions are complicated. At present, a complete autonomous driving test and evaluation system has not been established worldwide, and its core techno...

AI Technical Summary

Problems solved by technology

[0003] However, the evaluation object of autonomous driving test has changed from the traditional binary independent system of human and vehicle to the strong coupling system of human-vehicle-environment-task. The test scenarios and tasks are difficult to exhaust, and the evaluation dimensions are complicated.

At present, a complete autonomous driving test evaluation system has not been established worldwide, and its core technology needs to be broken through

Images