用粒子优化生成多样危险驾驶场景，提升自动驾驶测试效果
From Particles to Perils: SVGD-Based Hazardous Scenario Generation for Autonomous Driving Systems Testing
- 结合随机初始化与粒子梯度优化，生成多样化初始条件
- 在CARLA上使安全违规率最高提升27.68%，场景多样性增9.6%
- 可直接接入现有测试框架，适合自动驾驶安全验证
自动驾驶系统（ADS）的仿真测试需揭示密集异构交通中的真实且多样的故障。现有基于搜索的种子生成方法（如遗传算法）在高维空间中易陷入局部模式，遗漏多种故障场景。本文提出PtoP框架，融合自适应随机种子生成与斯坦变分梯度下降（SVGD），生成多样且诱发故障的初始状态。SVGD通过吸引至高风险区域、粒子间相互排斥，实现风险导向但分布均匀的种子生成。PtoP可即插即用，增强现有在线测试方法（如强化学习测试器）的性能。在CARLA平台上对两个工业级ADS（Apollo、Autoware）及一个端到端系统进行评估，结果显示，相较于基线方法，PtoP将安全违规率提升最多27.68%，场景多样性提升9.6%，地图覆盖度提升16.78%。
原文摘要 · Abstract (English)
Simulation-based testing of autonomous driving systems (ADS) must uncover realistic and diverse failures in dense, heterogeneous traffic. However, existing search-based seeding methods (e.g., genetic algorithms) struggle in high-dimensional spaces, often collapsing to limited modes and missing many failure scenarios. We present PtoP, a framework that combines adaptive random seed generation with Stein Variational Gradient Descent (SVGD) to produce diverse, failure-inducing initial conditions. SVGD balances attraction toward high-risk regions and repulsion among particles, yielding risk-seeking yet well-distributed seeds across multiple failure modes. PtoP is plug-and-play and enhances existing online testing methods (e.g., reinforcement learning--based testers) by providing principled seeds. Evaluation in CARLA on two industry-grade ADS (Apollo, Autoware) and a native end-to-end system shows that PtoP improves safety violation rate (up to 27.68%), scenario diversity (9.6%), and map coverage (16.78%) over baselines.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。