Accelerated Evaluation of Automated Vehicles Safety in Lane-Change Scenarios Based on Importance Sampling Techniques

Abstract

Automated vehicles (AVs) must be thoroughly evaluated before their release and deployment. A widely used evaluation approach is the Naturalistic-Field Operational Test (N-FOT), which tests prototype vehicles directly on the public roads. Due to the low exposure to safety-critical scenarios, N-FOTs are time consuming and expensive to conduct. In this paper, we propose an accelerated evaluation approach for AVs. The results can be used to generate motions of the other primary vehicles to accelerate the verification of AVs in simulations and controlled experiments. Frontal collision due to unsafe cut-ins is the target crash type of this paper. Human-controlled vehicles making unsafe lane changes are modeled as the primary disturbance to AVs based on data collected by the University of Michigan Safety Pilot Model Deployment Program. The cut-in scenarios are generated based on skewed statistics of collected human driver behaviors, which generate risky testing scenarios while preserving the statistical information so that the safety benefits of AVs in nonaccelerated cases can be accurately estimated. The cross-entropy method is used to recursively search for the optimal skewing parameters. The frequencies of the occurrences of conflicts, crashes, and injuries are estimated for a modeled AV, and the achieved accelerated rate is around 2000 to 20 000. In other words, in the accelerated simulations, driving for 1000 miles will expose the AV with challenging scenarios that will take about 2 to 20 million miles of real-world driving to encounter. This technique thus has the potential to greatly reduce the development and validation time for AVs.

Accelerated Evaluation of Automated Vehicles in Car-Following Maneuvers

“…In other words, the proposed accelerated evaluation method generates high risk maneuvers, some of which could be considered in future government certification processes. The crash and conflict events are binary events with indication function defined in (32). When a crash occurs, injury may or may not occur.…”

“…In our previous work [32], [33], the AE approach is applied to study AV behavior in lane change scenarios. In this paper, we extend the AE concept to the car-following scenarios, which is a more challenging problem, as in the lane change scenarios, the randomness of the motion of the lane-changing vehicle is modeled as a set of distributions but sampled only once at the lane change moment.…”

Modeling and simulations on automated vehicles to alleviate traffic congestion

“…Without considering the braking effect [5], the relationship between the acceleration and throttle opening angle is obtained:…”

“…In order to obtain the corresponding macroscopic characteristics of traffic flow, all vehicles in each section of the road are collected and processed according to Eq. (5). Take the first section as an example:…”

Evaluation of a semi-autonomous lane departure correction system using naturalistic driving data

“…Such models may also allow deeper insight into the physiological and cognitive behaviors of human drivers so that present or future driver-automation interfaces can be optimized. Recently, simulation-based evaluation methods have become popular [3]- [8] in evaluating the safety of automated vehicles and the performance of advanced driver assistance systems. In this paper, we propose a simulationbased method to evaluate the lane departure correction (LDC) system.…”