自动驾驶汽车传感器综述

（ An overview of sensors in Autonomous Vehicles ）

QNS2202-0020 2022 年 07 月 19 《五道口电力电子观察》

日
来源 : https://www.sciencedirect.com/science/article/pii/S1877050921025540
作者： Henry Alexander Ignatious* , Hesham-El-Sayed,
Manzoor Khan|November 1-4, 2021

自动驾驶汽车传感器综述

自动驾驶是一项快速发展的技术，也是一个争议的来源。人们相信，自动驾驶汽车将通过提高道路安全、
降低基础设施费用和改善儿童、老人和残疾人的机动性，提供一个更美好的未来。另一方面，许多人担
心汽车黑客的发生率、致命车祸的可能性以及驾驶相关职业的损失。毫无疑问，对于许多人来说，自动
驾驶是一项复杂而有问题的技术。为了更好地理解自动驾驶汽车的安全性，有必要首先了解它们的功能，
以及自动驾驶汽车使用何种传感器来确定它们应该行驶的位置，并识别道路上的物体，以避免发生车祸。
传感器收集的数据表现出异构和多模态特征，这些特征进一步融合到框架有效的决策规则中。因此，传
感器在自主车辆的决策活动中发挥着重要作用。为了获取更多与传感器相关的信息，本文根据 AV 传感
器的强制属性分析和总结了不同类型的 AV 传感器。该分析有助于读者了解传感器对 AVs 决策的贡献，
并总结了不同传感器收集的数据类型。总结的推论将使 AVs 领域的大多数新兴研究人员和学生大开眼界，
根据他们的研究需要选择合适的传感器。该研究还简要介绍了市场上主要供应商生产的不同类别传感器
的规格。

文章链接 :https://www.sciencedirect.com/science/article/pii/S1877050921025540
Source : https://www.sciencedirect.com/science/article/pii/S1877050921025540
Author: Henry Alexander Ignatious*, Hesham-El-Sayed,
Manzoor Khan|November 1-4, 2021

An overview of sensors in Autonomous Vehicles

Autonomous driving is a rapidly developing technology that is also a source of debate. People believe that
autonomous vehicles will provide a better future by increasing road safety, lowering infrastructure expenses,
and improving mobility for children, the old, and the disabled. On the other hand, many individuals are
concerned about incidences of automotive hacking, the likelihood of fatal crashes, and the loss of driving-
related professions. Autonomous driving is, without a question, a complex and problematic technology for
many people. To better comprehend how safe self-driving cars are, it's necessary to first understand how they
function, as well as what kind of sensors autonomous vehicles use to determine where they should travel and
recognize things on the road in order to avoid automobile accidents. Data collected by the sensors exhibit
heterogeneous and multimodal characteristics, which are further fused to frame effective decision rules. Thus
sensors play a major role in decision making activity of Autonomous Vehicles (AVs). In order to acquire more
information related to the sensors, this paper analyses and summarizes different types of AV sensors based on
their mandatory attributes. This analysis helps the readers to understand the contribution of the sensors
towards decision-making in AVs and also summarizes the data types collected by different sensors. The
summarized inferences will be an eye-opener to most of the budding researchers and students in the field of
AVs to select the appropriate sensor based on their needs for their research. The study also gives brief
information regarding the specifications of different categories of sensors manufactured by leading vendors in
the market.
 自动驾驶汽车传感器综述
1 、引言 1. Introduction
 A self-driving vehicle is one that can sense its surroundings and operate
 自动驾驶车辆是一种能够感知周围环境并在无需人工干预的
without the need for human intervention.
情况下运行的车辆。  At no point is a human passenger required to assume control of the car, nor is
 任何时候都不需要人类乘客来控制汽车，也不需要人类乘客 a human passenger required to be present in the car at all.
出现在汽车中。  Self-driving cars are progressively gaining market penetration. While there
 自动驾驶汽车正逐步获得市场渗透。 2019 年，全球约有 were about 31 million machines with some level of automation in operation
3100 万台具有一定自动化程度的机器在运行，但预计到 around the world in 2019, that number is predicted to rise to 54 million by
2024.
2024 年，这一数字将增至 5400 万台。  As a result, the worldwide autonomous vehicle market is expected to expand.
 因此，全球自主汽车市场有望扩大。  Rapid advancements in electronics, information, and communications
 电子、信息和通信技术的快速发展（导致计算机、传感器和 technology (leading to downsizing and improved computer, sensor, and
网络性能的缩小和改进）催生了各种自主车辆（ AV ）技术。 networking performance) have spawned various autonomous vehicle (AV)
 在现实场景中，大多数自动驾驶系统都面临类似的障碍和限 technologies.
制，例如在恶劣天气下安全驾驶和导航，以及与行人和其他  In real-world scenarios, most autonomous driving (AD) systems face similar
obstacles and limits, such as safe driving and navigating in inclement weather,
车辆的安全交互。
and safe interactions with pedestrians and other vehicles.
 恶劣的天气条件，如眩光、雪、雾、雨、霾和雾，会对基于  Harsh weather conditions, such as glare, snow, mist, rain, haze, and fog, can
感知的传感器的感知和导航性能产生重大影响。 have a substantial impact on the perception and navigation performance of
 由于其他车辆的不可预测情况和行为，这些困难对于道路自 perception-based sensors.
动驾驶系统来说变得更加复杂。  Because of the unpredictable situations and behaviors of other cars, these
 因此，为了限制碰撞危险， AVs 必须具有能够识别所有位置 difficulties become more complex for on road AVs.
 As a result, in order to limit collision hazards, AVs must have a thorough
未来运动的全面预测模块。
prediction module that can identify all position future motions.

文章链接 : https://www.sciencedirect.com/science/article/pii/S1877050921025540
 自动驾驶汽车传感器综述
 自动驾驶车辆是一种能够感知周围环境并在无需人工干预的  Sensors are devices that convert sensed events or changes in the environment
情况下运行的车辆。 into a numerical measurement that may then be processed.
 Sensors are divided into two categories based on their operational principle.
 任何时候都不需要人类乘客来控制汽车，也不需要人类乘客
Internal state sensors, also known as proprioceptive sensors, record the
出现在汽车中。 dynamical state of a dynamic system and detect internal data such as force,
 自动驾驶汽车正逐步获得市场渗透。 2019 年，全球约有 angular rate, wheel load, battery voltage, and so on.
3100 万台具有一定自动化程度的机器在运行，但预计到  Inertial measurement units (IMU), encoders, inertial sensors (gyroscopes and
2024 年，这一数字将增至 5400 万台。 magnetometers), and location sensors (Global Navigation Satellite System
 因此，全球自主汽车市场有望扩大。 (GNSS) receivers) are examples of proprioceptive sensors.
 Exteroceptive sensors, or external state sensors, on the other hand, perceive
 电子、信息和通信技术的快速发展（导致计算机、传感器和
and gather information from the system's environment, such as distance
网络性能的缩小和改进）催生了各种自主车辆（ AV ）技术。 measurements or light intensity.
 在现实场景中，大多数自动驾驶系统都面临类似的障碍和限  Exteroceptive sensors include cameras, radio detection and range (Radar),
制，例如在恶劣天气下安全驾驶和导航，以及与行人和其他 light detection and range (LiDAR), and ultrasonic sensors
车辆的安全交互。
 恶劣的天气条件，如眩光、雪、雾、雨、霾和雾，会对基于
感知的传感器的感知和导航性能产生重大影响。
 由于其他车辆的不可预测情况和行为，这些困难对于道路自
动驾驶系统来说变得更加复杂。
 因此，为了限制碰撞危险， AVs 必须具有能够识别所有位置
未来运动的全面预测模块。

文章链接 : https://www.sciencedirect.com/science/article/pii/S1877050921025540
 自动驾驶汽车传感器综述

文章链接 : https://www.sciencedirect.com/science/article/pii/S1877050921025540
 自动驾驶汽车传感器综述
相机的传感器 Camera sensors
 Autonomous cars often have video cameras and sensors in order to see and
自动驾驶汽车通常装有摄像机和传感器，以便像人类司 interpret the objects in the road just like human drivers do with their eyes. By
机的眼睛一样观察和解读道路上的物体。通过在每个角 equipping cars with these cameras at every angle, the vehicles are capable of
度安装这些摄像头，车辆能够保持外部环境的 360° 视角， maintaining a 360° view of their external environment, thereby providing a
broader picture of the traffic conditions around them.
从而提供周围交通状况的更广泛的图片。  Today, 3D cameras are available and utilized for displaying highly detailed
今天， 3D 摄像机可用来显示高度详细和逼真的图像。 and realistic images. These image sensors automatically detect objects,
这些图像传感器会自动检测目标，对目标进行分类，并 classify them, and determine the distances between them and the vehicle. For
example, the cameras can easily identify other cars, pedestrians, cyclists,
确定目标与车辆之间的距离。例如，摄像头可以很容易 traffic signs and signals, road markings, bridges, and guardrails.
地识别其他车辆、行人、自行车、交通标志和信号、道
路标记、桥梁和护栏。

文章链接 : https://www.sciencedirect.com/science/article/pii/S1877050921025540
 自动驾驶汽车传感器综述
雷达传感器 Radar sensors
 Radar (Radio Detection and Ranging) sensors make up a crucial contribution
雷达 ( 无线电探测和测距 ) 传感器对自动驾驶的整体功 to the overall function of autonomous driving: they send out radio waves that
能做出了至关重要的贡献 : 它们发出无线电波来探测物 detect objects and gauge their distance and speed in relation to the vehicle in
体，并实时测量物体与车辆的距离和速度。 real time.
 Both short- and long-range radar sensors are usually deployed all around the
短期和长期雷达传感器通常部署在汽车周围，每个都有 car and each one has their different functions. While short range (24 GHz)
其不同的功能。短距离 (24ghz) 雷达应用可以实现盲点 radar applications enable blind spot monitoring, the ideal lane-keeping
监控、理想的车道保持辅助和停车辅助，而远程 assistance, and parking aids, the roles of the long range (77 GHz) radar
sensors include automatic distance control and brake assistance. Unlike
(77ghz) 雷达传感器的作用包括自动距离控制和制动辅 camera sensors, radar systems typically have no trouble at all when
助。与相机传感器不同，雷达系统在雾或雨中识别物体 identifying objects during fog or rain.
时通常没有任何问题。

文章链接 : https://www.sciencedirect.com/science/article/pii/S1877050921025540
 自动驾驶汽车传感器综述
激光雷达传感器 Lidar sensors
 Lidar (Light Detection and Ranging) sensors work similar to radar systems,
激光雷达 ( 光探测和测距 ) 传感器的工作原理与雷达系 with the only difference being that they use lasers instead of radio waves.
统类似，唯一的区别是它们使用激光而不是无线电波。 Apart from measuring the distances to various objects on the road, lidar
除了测量到道路上各种物体的距离，激光雷达还可以创 allows creating 3D images of the detected objects and mapping the
surroundings. Moreover, lidar can be configured to create a full 360-degree
建被探测物体的 3D 图像，并测绘周围环境。此外，激 map around the vehicle rather than relying on a narrow field of view. These
光雷达可以配置为围绕车辆创建完整的 360 度地图，而 two advantages make autonomous vehicle manufacturers such as Google,
不是依赖于狭窄的视野。这两大优势使得谷 Uber, and Toyota choose lidar systems.
 Moreover, lidar can be configured to create a full 360° map around the
歌、 Uber 、丰田等自动驾驶汽车厂商纷纷选择激光雷 vehicle rather than simply relying on a narrow field of view. These two
达系统。 advantages have led autonomous vehicle manufacturers such as Google, Uber,
此外，激光雷达可以配置为围绕车辆创建一个完整的 and Toyota to choose lidar systems for their vehicles.
360° 地图，而不是简单地依赖于狭窄的视野。这两大优
势使得谷歌、 Uber 和丰田等自动驾驶汽车制造商纷纷
选择激光雷达系统。

文章链接 : https://www.sciencedirect.com/science/article/pii/S1877050921025540
 备注

 备注 1
相机传感器还远远不够完美。雨、雾、雪等恶劣的天气条件会使摄像机无法清楚地看到道路上的障碍物，从而增加事
故发生的可能性。
 备注 2
如今汽车上使用的汽车雷达传感器只能正确识别 90% 到 95% 的行人，这不足以确保道路安全。
 备注 3
激光雷达传感器比自动驾驶汽车使用的雷达传感器昂贵得多。另一个问题是，雪或雾有时会阻碍激光雷达传感器，对
它们探测道路上物体的能力产生负面影响。

页码
 评论

 评论 1

页码
联系方式

 北京乾勤科技发展有限公司

电话： 010 － 82642324

传真： 010 － 82641640
地址： 北京市海淀区智造大街 D 栋 106
邮编： 100190
信箱： service@qianqin.com.cn
网址： www.qianqin.com.cn

页码 10
技术服务

智能车 QQ 群 B 站——智能车开发 淘宝—— Aurix 开发工具 公众号

11
IGBT/MOSFET 双脉冲测试平台（ DUT ）

页码
电机负载模拟器 (QSVPWMBench)

 北京乾勤科技的 QSVPWMBench 电机负载模拟测试系统，

由 SVPWM 矢量信号发生器（用于模拟向三相全桥功率
主电路提供 PWM 驱动信号的 DSP 或单片机）、高压可
调直流电源、三相负载电感（用于模拟电机负载的）、
水冷机组、传感器组件和上位机软件，以及被测组件组
成。被测组件为三相全桥，根据不同的被测 IGBT/SiC
MOSFET ，提供对应的测试工装。设备采用电感负载，
为无源型负载模拟器。

 系统可测试被测组件在特定工况（开关频率，基波频率，
电流等）和特定冷却系统参数（水温，流量）之下的温
升，进行功率模块的电流能力验证，明确模块的工作条
件，为模块在实际系统中的使用提供依据。同时，也可
以用于直流母线电容的纹波电流能力和温升验证。

页码