Adversarial Driving Behavior Generation Incorporating Human Risk Cognition for Autonomous Vehicle Evaluation
This work addresses the need for robust testing of autonomous vehicles, particularly for safety-critical scenarios, though it appears incremental by extending existing methods to incorporate human risk factors.
The paper tackles the problem of evaluating autonomous vehicles by generating adversarial driving behaviors to expose risky events, using a reinforcement learning approach combined with human risk cognition modeling, and demonstrates effectiveness in a cut-in scenario on a Hardware-in-the-Loop platform.
Autonomous vehicle (AV) evaluation has been the subject of increased interest in recent years both in industry and in academia. This paper focuses on the development of a novel framework for generating adversarial driving behavior of background vehicle interfering against the AV to expose effective and rational risky events. Specifically, the adversarial behavior is learned by a reinforcement learning (RL) approach incorporated with the cumulative prospect theory (CPT) which allows representation of human risk cognition. Then, the extended version of deep deterministic policy gradient (DDPG) technique is proposed for training the adversarial policy while ensuring training stability as the CPT action-value function is leveraged. A comparative case study regarding the cut-in scenario is conducted on a high fidelity Hardware-in-the-Loop (HiL) platform and the results demonstrate the adversarial effectiveness to infer the weakness of the tested AV.