A Versatile Approach to Evaluating and Testing Automated Vehicles based on Kernel Methods

TL;DR

This paper introduces a kernel-based statistical learning method to efficiently evaluate automated vehicles by identifying rare failure scenarios, reducing testing time and improving safety validation.

Contribution

It proposes a versatile kernel method to construct sampling distributions for accelerated evaluation, adaptable to various critical event sets in automated vehicle testing.

Findings

Robustly identifies rare failure scenarios in automated vehicles

Significantly reduces evaluation time compared to traditional methods

Uses Gaussian properties to tailor sampling distributions

Abstract

Evaluation and validation of complicated control systems are crucial to guarantee usability and safety. Usually, failure happens in some very rarely encountered situations, but once triggered, the consequence is disastrous. Accelerated Evaluation is a methodology that efficiently tests those rarely-occurring yet critical failures via smartly-sampled test cases. The distribution used in sampling is pivotal to the performance of the method, but building a suitable distribution requires case-by-case analysis. This paper proposes a versatile approach for constructing sampling distribution using kernel method. The approach uses statistical learning tools to approximate the critical event sets and constructs distributions based on the unique properties of Gaussian distributions. We applied the method to evaluate the automated vehicles. Numerical experiments show proposed approach can robustly identify the rare failures and significantly reduce the evaluation time.