Analysis of key technologies of intelligent vehicle software

Figure 8 Overall view of the Internet of Vehicles

The figure shows the main participants in the Internet of Vehicles business, including the communication terminal T-Box on the car, related controllers and car computers, as well as the backend TSP (Internet of Vehicles Service Integrator), mobile APP, web page, content provider and service provider, as well as OBU, RSU and base station that communicate with the road side. Among them, T-Box is the only bridge on the car that communicates with the outside world, which not only realizes the in-car networking but also realizes the communication outside the car; TSP is the integration part of the backend, which usually provides basic capability management (such as account and authentication) and connects with various terminals to realize service integration and information transmission; mobile APP, web page and car computer are direct user touch points, undertake the task of interacting with users, and are the embodiment of various services; content and service providers are the source of most data, and they are professional providers in the Internet segment.

The in-vehicle communication module T-BOX architecture usually contains dual-channel high-speed CAN transceivers, 4G/5G/V2X modules and high-performance microprocessor chips capable of real-time processing. It is mainly responsible for in-vehicle and out-of-vehicle communication services. The application scenarios and collaborative service types of the C-V2X technology of the Internet of Vehicles are gradually enriched, and the evolution path of technology and applications is also developing from node processing to more complex applications.

Figure 9 Neusoft vehicle communication module T-Box architecture diagram

6. Analysis of key technologies for high-precision maps

High-precision map refers to a navigation map with high precision, high freshness and high richness, with absolute and relative accuracy at the decimeter level. It is referred to as HD Map (High Definition Map) or HAD Map (Highly Automated Driving Map). High-precision map contains rich information, including road information such as road type, curvature, lane line position, as well as environmental object information such as roadside infrastructure, obstacles, traffic signs, etc., and also includes real-time dynamic information such as traffic flow and traffic lights. Different map information has different application scenarios and requirements for real-time performance. By hierarchical processing of information, the management and collection efficiency of maps can be effectively improved, as well as their wide application.

Compared with traditional on-board electronic maps, high-precision maps are more detailed and have richer dynamic elements. In addition, the volume of on-board maps is limited by the storage capacity of the embedded system. At present, the storage density of high-precision maps (centimeter level) used for autonomous driving is very high, and the overall capacity has far exceeded the storage capacity of the current mainstream controller solutions, so it is necessary to use cloud storage and cloud distribution to achieve it. In addition, the update frequency of traditional navigation electronic maps is static data (usually updated quarterly or monthly) and quasi-static data (daily). High-precision maps have higher requirements for the real-time nature of data, and the update frequency is usually quasi-dynamic data (updated every minute) and real-time dynamic data (updated every second or millisecond).

Figure 11 Schematic diagram of the basic form of high-precision maps for urban scenes