从实验室到街道：解决加速自动车辆测试的挑战.pdf

jenk

254

9页

0次

2021-02-22

40墨值下载

May 2017 1

From the Lab to the Street:

Solving the Challenge of Accelerating

Automated Vehicle Testing

EXECUTIVE SUMMARY

As automated vehicles and their technology become more advanced and technically

sophisticated, evaluation procedures that can measure the safety and reliability of these

new driverless cars must develop far beyond existing safety tests. To get an accurate

assessment in ﬁeld tests, such cars would have to be driven millions or even billions of

miles to arrive at an acceptable level of certainty – a time-consuming process that would

cost tens of millions of dollars.

Instead, researchers afﬁliated with the University of Michigan’s Mcity connected

and automated vehicle center have developed an accelerated evaluation process that

eliminates the many miles of uneventful driving activity to ﬁlter out only the potentially

dangerous driving situations where an automated vehicle needs to respond, creating a

faster, less expensive testing program. This approach can reduce the amount of testing

needed by a factor of 300 to 100,000 so that an automated vehicle driven for 1,000 test

miles can yield the equivalent of 300,000 to 100 million miles of real-world driving.

While more research and development needs to be done to perfect this technique, the

accelerated evaluation procedure offers a ground-breaking solution for safe and efﬁcient

Contents

1 Executive Summary

2 The Problem

3 Approach

8 Conclusions

9 Resources

DING ZHAO, PhD

Assistant Research Scientist

Mechanical Engineering

University of Michigan

HUEI PENG, PhD

Director, Mcity

Roger L. McCarthy Professor

of Mechanical Engineering

University of Michigan

AUTOMATED VEHICLES | SIMULATION AND TESTING

From the Lab to the Street: Solving the Challenge of Accelerating Automated Vehicle Testing May 2017 2

testing that is crucial to deploying automated vehicles.

THE PROBLEM

The dawning of driverless vehicles presents a number of challenges for automakers,

regulators and city planners, from the design of software and hardware in the vehicles,

to redesigning the road infrastructure, to clarifying the challenging legal issues about

potential liability in an accident.

But before consumers will embrace automated vehicles – especially cars with no driver

controls at all – the people who will buy and ride in these “cars of the future” will need

to be assured that the vehicles are reliable and safe.

Safety testing in today’s cars and trucks is a well-deﬁned, standardized effort: For

crashworthiness, get your test vehicle, install the crash-test dummies and sensors, put it

on a test sled, roll the video cameras and see what happens when the car hits the wall.

For rollover vulnerability, conduct a few well deﬁned steering maneuvers, and compute

a rollover “score” using results of the vehicle-in-motion test and taking into account

the vehicle’s shape and weight distribution. The results are easily measured and can

be repeated in a way that assures car buyers, government regulators, and insurance

companies.

The crashworthiness test measures the outcome of a single event: What happens when

a car crashes at a particular speed in a particular way and how badly are the occupants

hurt? The rollover tests rate the propensity for a tip-over. But gauging with any kind of

certainty how an automated vehicle will react is vastly more difﬁcult than looking to see

whether the crash-test dummy’s arm got broken. Test methods for traditionally driven

cars are something like having a doctor take a patient’s blood pressure or heart rate, while

testing for automated vehicles is more like giving someone an IQ test. The variables of

trafﬁc and road conditions, weather, time of day, and the unpredictable actions of other

drivers and vehicles present a constantly changing tangle of variables that an automated

car will need to recognize and process to make the right, safe choice.

In fact, even the question of how to design such tests is much more complicated. Instead

of, “What happens in a crash?” the tests for automated vehicles must measure how

effectively these cars can keep one from happening.