Design and Implementation of Test Scenario Library Data System for Autonomous Vehicles

In recent years, with the development of autonomous driving technology, the unsafe, unreliable and unstable of autonomous driving systems have gradually been exposed. It is necessary to build test scenarios through a large amount of data in the real driving scene database to conduct comprehensive and strict tests and evaluation. There is a lack of an autonomous vehicle test scenario library data system that provides unified standard data. This paper designs and implements a test scenario library data system for autonomous vehicles. It explains the build steps of data system and analyses the methods of the implementation. It explains the build steps of data system and analyses the methods of the implementation. The data system contains real driving scene management module and test scenario management module. The system is based on B/S structure and decomposed into front-end and back-end. It is designed and implemented hierarchically that is composed application layer, data layer and support layer. The real driving scene management module is responsible for data collection, it could validate the data format, and then process, combine and store submitted data. The test scenario management module integrates the functions such as data filtering, data generation and scenario rebuilding. It extracts scenario element action libraries and provides algorithm-based test datasets. At present, the basic function of the data system has been developed. The test scenarios can be generated and export by using the system directly. It solves the problems in data processing and maintenance of massive autopilot data, and can give out standard data to construct virtual simulation cases. It would make the data support for building a suitable test environment of autonomous vehicles.