At present, the heat brought by the topic of new infrastructure continues. How can autonomous driving catch up on this express train, what core technologies are involved in autonomous driving, and how can the CAN smart cloud be used to improve the safety performance of autonomous driving? This article will give a brief introduction to this.
Focusing on the deep transformation of the automotive industry, we can regard the self-driving car as a “smart wheeled robot with human eyes” that can identify surrounding vehicles, faults, pedestrians, etc., and make precise actions. It is not difficult to know that autonomous driving relies on artificial intelligence AI, sensors, big data, 5G and other technical fields. Fortunately, autonomous driving stands at the intersection of these technologies.
1.1 How does autonomous driving enter the “fast lane”?
The key to why autonomous driving will explode with the energy brought by the new infrastructure is determined by the core technology of autonomous driving. Autonomous driving actually contains three questions: First, where am I? Second, where am I going? Third, how do I get there? Only when these three problems are solved perfectly can it be regarded as true autonomous driving. The technical modules included in it are shown in Figure 1.
Figure 1 Block diagram of autonomous driving technology
l Sensor technology: camera, lidar, millimeter-level radar and other sensor technologies are integrated to identify the external environment;
l High-precision positioning: use GPS, GNSS, IMU and other technologies to achieve centimeter-level positioning;
l AI artificial intelligence: the key technology, which makes full use of the obtained information for analysis and decides how the vehicle should drive. The core task of the algorithm is shown in Figure 2;
l V2X communication security: including car and user information, user authentication and data encryption, relying on 5G’s large broadband, low-latency information exchange;
l HIM human-computer interaction: The human-computer interaction interface is used to manually handle situations that autonomous driving robots cannot handle.
Figure 2 Core algorithm tasks of autonomous driving AI
1.2 ZWS-CAN Smart Cloud Assists Autonomous Driving
As shown in Figure 2 above, after automatic driving passes a series of core algorithm tasks, it will eventually control the work of the vehicle ECU through the CAN-bus bus, so the quality of CAN communication has a huge impact on the performance of the autonomous vehicle. Feedback helps optimize AI algorithms. So, in the autonomous driving road test stage, how can the technicians remotely analyze the vehicle operation in real time?
Vehicle CAN-bus data logging CANDTU series
When an autonomous vehicle encounters a problem during the road test, it is difficult for the tester to analyze the situation at the scene or to judge whether the vehicle is running well. Under normal circumstances, only by analyzing the CAN messages of the vehicle can we understand the speed and pressure of the vehicle, and then further feedback and optimize the task of the AI algorithm.
ZLG Zhiyuan Electronics has launched a CAN network bus “black box”, which we call CANDTU. Testers can use CANDTU to record CAN message data in the road test phase, so as to conduct overall fault diagnosis of the vehicle.
CANDTU product performance is as follows:
l Integrate 2 or 4 independent CAN channels conforming to ISO11898 standard;
l The standard storage medium is 32G high-speed SD card, which supports long-time recording, conditional recording, pre-triggered recording and other recording modes, and can store large data;
l Support ASC, CSV and other record data storage format conversion, which is convenient for later software analysis;
l Pass strict anti-vibration and anti-shock tests to meet the needs of industrial users;
l With 2-channel DI record and 2-channel DO alarm output;
l Support GPS positioning, real-time upload to the cloud through 4G communication, and real-time viewing of vehicle positioning, instruments, oil temperature and oil pressure through mobile phones and other terminals, as shown in Figure 3.
Figure 3 Application of CANDTU
ZWS-CAN Smart Cloud
In the road test phase of autonomous vehicles, if the testers can remotely monitor the vehicle information in real time and grasp the information of the vehicle at the first time, it will be of great help to improve the performance of the vehicle. ZLG Zhiyuan Electronics provides the ZWS-CAN smart cloud solution, which connects to the cloud server through the 4G communication of CANDTU series products, echoes the CAN message data to the scene, and realizes remote monitoring of autonomous vehicles and fault diagnosis. So, what special services and functions can ZWS-CAN smart cloud provide for autonomous vehicles? A brief introduction follows.
1. Cloud curve, visual analysis of CAN messages
ZWS-CAN smart cloud can realize the visual analysis of DBC, combined with a variety of graphic controls, timely visual Display of CAN (FD) data, and intuitive analysis of vehicle operation. As shown in Figure 4, it can display signal values and track signals. .
Figure 4 Visual analysis of CAN messages
2. Support vehicle UDS diagnosis
Users can directly perform standard UDS diagnosis on the vehicle through the ZWS-CAN cloud server to monitor the equipment, as shown in Figure 5.
Figure 5 UDS diagnostic function
3. Support Beidou/GPS positioning
As shown in Figure 6, logging into the ZWS-CAN cloud server can perform map visualization positioning, display the vehicle running speed in real time, and record and store the vehicle running track. The positioning accuracy is about 2m, and the vehicle fault analysis and positioning has a high reliability, and the AI algorithm is optimized according to the vehicle operating conditions.
Figure 6 GPS positioning map
4. Custom Test Script
ZWS-CAN cloud server supports custom test scripts. Through the editor and actuator functions, as shown in Figure 7, actions such as sending, waiting, verifying, verifying response, etc. can be performed to realize automatic testing of equipment and facilitate remote testing by users. vehicle performance.
Figure 7 Custom test script