Intelligent Driving | Key Parameters and Design Points of LiDAR

Publisher:RadiantExplorerLatest update time:2024-08-18 Reading articles on mobile phones Scan QR code
Read articles on your mobile phone anytime, anywhere

In the field of high-speed autonomous driving, L3-level systems require accurate and reliable sensors to ensure safe and efficient operation on highways. As one of the key sensors, the performance of LiDAR directly affects the vehicle's environmental perception capabilities.


picture


● Safety is the primary consideration in LiDAR design, especially in highway driving environments. It must be able to accurately detect moving objects, obstacles that cannot be crossed, and obstacles that cannot be lowered. The detection requirements determine the resolution, detection range, and frame rate of LiDAR. For high-speed vehicles, high resolution and long-distance detection capabilities are particularly important, especially in the center of the field of view.


● The field of view (FOV) and region of interest (ROI) of the LiDAR are key parameters in the design. The central field of view has higher requirements because it is the main detection area in the direction of the vehicle's advance. By dynamically adjusting the laser energy to focus on the region of interest, the detection range and resolution of a specific area can be improved. This design ensures higher detection performance where it is needed, thereby improving driving safety.


The installation location of the LiDAR is also very important. Common installation locations include the front grille, roof, behind the windshield, and near the headlights. Each location has its own unique advantages and disadvantages, affecting factors such as the LiDAR's field of view, size, environmental adaptability, and cleaning system design.


01


Demand analysis of LiDAR

● Horizontal field of view

picture


The horizontal field of view (HFOV) determines the horizontal angle range that can be captured in each frame. For highway driving, a wide HFOV is necessary to help the vehicle predict and track potential collision objects. Especially in "cut-in scenarios", when an overtaking vehicle suddenly enters the self-lane, the LiDAR needs to have a wide enough field of view to detect such situations.

HFOV design considerations mainly include:

◎ Installation location: LiDAR installed on the front grille of the vehicle requires a larger HFOV because this location detects the intruding vehicle later. In contrast, when installed on the roof, the LiDAR can detect the intruding vehicle earlier, so a smaller HFOV is required.

◎ Vehicle width: A vehicle with a width of approximately 1.9 meters requires a HFOV of at least 120° to ensure that the cut-in scene can be detected.

Speed ​​and reaction time: At a speed of 130 km/h, LiDAR needs to operate at at least 20 frames per second (FPS) to ensure that it can accurately detect the oncoming vehicle before the vehicle reacts.

picture



The horizontal field of view ROI (ROI HFOV) must take into account changes caused by factors such as road slope changes, vehicle acceleration and deceleration, wind force, and sensor direction calibration errors. These changes must be covered by the fixed field of view within the ROI. In order to meet the needs of basic scene understanding, it is also required to scan multiple lanes and different types of curves.

In addition, an additional field of view is required as a buffer to allow for factory assembly tolerances. The ROI is approximately 30° (including tolerances). Knowing the horizontal field of view (HFOV) and the minimum detection distance of 100 meters, we can infer that the vehicle is traveling at 130 km/h (80 mph).

● Vertical field of view

The vertical field of view (VFOV) defines the range of the LiDAR in the vertical direction. This is critical for detecting objects on the road surface and obstacles ahead. In highway driving, the LiDAR needs to be able to detect small objects within a range of 3 meters to ensure that the vehicle does not collide with undetected obstacles.

picture


VFOV design considerations mainly include:

◎ Mounting height: Different mounting locations (such as front grille, roof, windshield or headlight) have different VFOV requirements. When mounted on the roof, the LiDAR requires a smaller VFOV because it is easier to detect protruding objects, while a front grille installation requires a larger VFOV.

◎ Detection distance: For a mounting height of 50 cm, the VFOV should be between 25° and 40° to ensure that loads at a height of 1.5 m and drivable obstacles such as low bridges can be detected.

picture


● Resolution

picture


Resolution determines the details that LiDAR can accurately capture. High resolution provides more accurate object detection and classification capabilities, which is the key to achieving safe driving. In L3 level autonomous driving, special attention needs to be paid to horizontal and vertical resolution to ensure accurate detection of distant and close objects. Resolution design considerations.

◎ Horizontal resolution (HRes): In order to detect pedestrians at a distance of 150 meters, a horizontal resolution of at least 0.07° is required. This ensures that the width of a pedestrian can be detected by 2 to 3 pixels at a distance of 150 meters, thereby achieving correct identification and classification with a high probability.

◎ Vertical Resolution (VRes): At a distance of 100 meters, LiDAR requires a vertical resolution of 0.05° to ensure that objects 14 cm high can be detected. This is critical to avoid collisions due to undetected obstacles when the vehicle is traveling at high speeds.

picture


02

Other parameter measurement range

● Measuring range

picture


The measurement range of LiDAR determines the maximum distance it can detect. In highway driving, a long detection range can provide longer reaction time, allowing the vehicle to drive more smoothly. A comfortable driving experience requires LiDAR to detect vulnerable road users (VRUs), such as pedestrians and other vehicles, at a distance of at least 200 meters.

Measurement range design considerations:

◎ Environmental conditions: LiDAR performance is affected by ambient light, weather conditions, and sensor aging. Therefore, a detection distance of more than 200 meters is desirable to ensure effective detection under various conditions.

◎ Reflectivity: The measurement range of LiDAR will be different for objects with different reflectivity. Within a distance of 150 meters, a black metal car needs a reflectivity of 10% to be detected, while a reflectivity of 1% is required within a distance of 50 meters.

● Frame rate

The frame rate defines how many times a LiDAR can fully refresh all pixels per second. A higher frame rate can provide faster reaction time and motion detection capabilities. In L3 level autonomous driving, 20 frames per second (FPS) is the ideal frame rate to ensure accurate detection and classification of objects when the vehicle is moving quickly.

Frame rate design considerations:

◎ Reaction time: A higher frame rate can reduce reaction time and thus reduce braking distance. For example, a frame rate of 20FPS can refresh all pixels within 0.35 seconds, providing faster reaction capabilities.

◎ Detection accuracy: A higher frame rate can improve detection accuracy and reliability, ensuring that the vehicle can accurately detect all objects in the surrounding environment when driving at high speeds.

picture


Comprehensive considerations mainly include the need to comprehensively consider the relationship between various performance parameters when designing an L3-level LiDAR system. For example, a higher frame rate may reduce resolution and field of view, so it is necessary to weigh and optimize various parameters.

Reflectivity data is generated by measuring the amount of light collected by each pixel. By normalizing this value with the measurement distance, the reflectivity of the object can be extracted. This information helps identify lane markings and improve the classification of vehicles and pedestrians.

When evaluating LiDAR, it is important to consider both specifications and parameters, as improving a single parameter may negatively impact other parameters. For example, increasing the frame rate may reduce vertical field of view and resolution, and all parameters must be considered when designing a LiDAR system to ensure optimal performance without compromising any critical performance.

picture


summary

Designing an L3 LiDAR system for highway driving requires comprehensive consideration of multiple key performance parameters, including horizontal field of view, vertical field of view, resolution, measurement range, and frame rate. Each parameter is critical to the overall performance of the LiDAR and the vehicle's ability to drive safely.

In practical applications, various parameters need to be weighed and optimized to ensure that the LiDAR system can work stably and reliably under various environmental conditions, thereby providing a safe and comfortable driving experience.


Reference address:Intelligent Driving | Key Parameters and Design Points of LiDAR

Previous article:RoboSense's revenue doubled in the first half of the year, and total sales of LiDAR exceeded 580,000 units
Next article:AK protocol + high cost performance + vibration suppression + ASIL-B! Saizhuo Electronics launches new wheel speed sensor chip SC9684

Latest Automotive Electronics Articles
Change More Related Popular Components

EEWorld
subscription
account

EEWorld
service
account

Automotive
development
circle

About Us Customer Service Contact Information Datasheet Sitemap LatestNews


Room 1530, 15th Floor, Building B, No.18 Zhongguancun Street, Haidian District, Beijing, Postal Code: 100190 China Telephone: 008610 8235 0740

Copyright © 2005-2024 EEWORLD.com.cn, Inc. All rights reserved 京ICP证060456号 京ICP备10001474号-1 电信业务审批[2006]字第258号函 京公网安备 11010802033920号