A Scenario-Based Development Framework for Autonomous Driving
This work addresses the need for systematic testing in autonomous driving, but it is incremental as it builds on existing research by summarizing and organizing concepts.
The paper tackles the challenge of testing and developing autonomous vehicles by proposing a scenario-based framework, which includes definitions, ontology elements, data sources, processing methods, and a V-Model, along with automated methods for random and dangerous scenario generation.
This article summarizes the research progress of scenario-based testing and development technology for autonomous vehicles. We systematically analyzed previous research works and proposed the definition of scenario, the elements of the scenario ontology, the data source of the scenario, the processing method of the scenario data, and scenario-based V-Model. Moreover, we summarized the automated test scenario construction method by random scenario generation and dangerous scenario generation.