At present, most of the autonomous driving models mass-produced in China are in the L2 stage. In the L2 stage, sensors are used to determine the surrounding driving environment, perceive the road, pedestrians, and road conditions, and perform warning or braking actions based on the perceived information to assist drivers in safe driving. Therefore, radar sensors are the most core component of this autonomous driving stage. Radar sensors detect the relative position, velocity, and angle of objects through TOF or FMCW technology, and with the assistance of algorithms and decision-making platforms, achieve more than ten functions such as AEB emergency automatic braking, FCW forward collision warning, BSW blind spot monitoring, etc.
1. AEB (Automatic Emergency Braking) function
Definition: AEB is an important function of ADAS system, which is divided into three levels:
1）FCW， When there is a potential collision risk, the AEB system will promptly notify the driver through visual, auditory, and/or tactile alarms (such as seat vibrations);
2) Level 2 alarm, if the driver is unable to respond, an automatic speed limit prompt will appear;
3) Automatically intervene in braking, and if necessary, AEB automatically controls the braking system to complete the braking operation, thereby reducing or avoiding collision injuries. This function can be activated within a range of 7-180km.
According to the Insurance Institute for Highway Safety, the AEB system can reduce the probability of rear end collisions by 56%
Category: (Difficulty from low to high)
According to the type of accident, E-NCAP categorizes AEB systems into three types:
1. Urban specific AEB system - often occurs at intersections with traffic congestion, slow speeds (<20km/h), and low collision severity;
2. Expressway specific AEB system - commonly found in high-speed (50-80km/h) driving scenarios where drivers are fatigued;
3. Pedestrian protection dedicated AEB system - detects pedestrians and bicycles, requires predicting their calculation path, and the algorithm is complex.
The first two are more suitable for using millimeter wave radar, while the third requires image capture based on the first two, so a camera needs to be added.
Working principle: AEB can be achieved through technologies such as visual sensors, millimeter wave radar, and LiDAR. Radar sensors or camera sensors measure the distance to the preceding vehicle or obstacle, and the data analysis module compares the measured distance with the safe distance and warning distance. When the distance is too small, the AEB system will issue a collision warning. If the driver fails to brake in a timely manner, the AEB system will issue a braking request to the braking system, achieving automatic braking.
Millimeter wave radar: It has the advantage of working 24/7, but there are horizontal radiation blind spots.
Lidar: In addition to its long detection range, it can also measure lateral position, but it is susceptible to weather conditions.
Visual sensors: have obvious cost advantages, but recognition accuracy still needs to be improved, and are greatly affected by weather conditions.
Cost: The single camera version of AEB costs about 600 yuan, the millimeter wave radar version costs about 1500 yuan, and the laser radar version costs over 10000 yuan.
Considering the overall performance and cost, the optimal sensor configuration for the AEB system is a combination of "millimeter wave radar+visual sensor".