■Relying on a 300-acre enclosed test field in Zengcheng, Guangzhou (including a high-speed loop road (with a total length of 2800 meters), a dynamic square (with a diameter of 230 meters)), a straight performance road (with a total length of 1450 meters and a width of ≥ 50 meters), and a 645-acre bay area intelligent connection test field in Pingshan, Shenzhen (including a high speed testing area (including characteristic scenes such as inbound and outbound ramps, ETC toll stations, tunnels, emergency lanes, etc.) The urban road testing area (including characteristic scenes such as adjustable signal light intersections, intersections, roundabouts, bus stops, and outdoor parking lots), ADAS dedicated testing area (with adjustable street light illumination, including characteristic scenes such as roadside grass, roadside gravel, and roadside lane lines), flexible testing area (with a dynamic square diameter of 260 meters), slope testing area, etc.), integrates the leading domestic and foreign OXFORD high-precision positioning and orientation system, ABD driving robot system, etc. The 4A series high simulation (dummy fake car) soft target object and ABD series soft target mobile transportation platform, as well as other autonomous driving testing systems, have the ability to simulate typical traffic flow such as vehicle to vehicle, vehicle to two wheeled vehicle, vehicle to pedestrian, and vehicle to road markings in conventional and extreme weather environments. They can quickly and risk controllably carry out research and development testing of active safety, assisted driving, and autonomous driving functions.
■The intelligent connected closed field testing and evaluation platform can cover national standards, operational standards, European standards, ISO and other automotive active safety, assisted driving, autonomous driving standards and regulations, mainstream national (regional) new car evaluation regulations NCAP research and development testing, and enterprise standard research and development testing.
■The scope of evaluation standards/regulations is as follows:
National standards: GB/T 30677, GB/T 20608, GB/T 39901, GB/T 39323, GB/T 39265, GB/T 33577, GB/T 26773, GB/T 38186, GB/T 41796, GB/T 41630, GB/T 41798;
Line labels: JT/T 883, JT/T 1242, JT/T 1358, JT/T 1447, JT/T 1446, JT/T 1428;
European standards: UN R48, UN R79, UN R130, UN R131, UN R140, UN R151, UN R152, UN R157, UN R158, UN R159, (EU) 1341, (EU) 1958, (EU) 646;
ISO: 15622, 17387, 11270, 17361, 16787;
NCAP: C-NCAP, EuroNCAP, ASEAN-NCAP;
Other: C-ICAP, etc.
■In terms of virtual simulation testing, an intelligent car hardware in the loop simulation testing system is integrated based on the NI real-time cabinet, which includes a 4-target radar echo simulator, a camera video black box, a 6-degree of freedom driving simulator, and supporting scene construction software such as Carmaker/Matlab Simulink, forming a coverage of autonomous driving perception layer, decision layer, and execution layer The autonomous driving testing capability includes software/hardware/driver in the loop at the level of human-computer interaction. At the same time, combined with on-site testing and public road testing, the collected scene data, human-computer interaction data, and driving behavior data from actual driving are clustered, analyzed, and reorganized into high mileage software/hardware/driver in the loop testing scenarios and test cases, Form an integrated development and testing platform for autonomous driving simulation testing software, hardware, and scene library.
■In terms of public road testing and evaluation, in accordance with the national and local government's policy guidelines for intelligent connected vehicle public road testing, we have deeply explored the front-end research and development verification needs of enterprises, and built an intelligent connected vehicle public road evaluation system based on a fusion perception scheme of LiDAR+camera+high-precision positioning. With the use of high bandwidth and high-capacity data acquisition equipment, we collect real-time driving scene information on public roads, Evaluate the reliability and comfort of intelligent connected vehicle functions based on the subjective experience of professional testers. At the same time, advanced technologies such as big data analysis and 4G/5G communication have been applied to build a cloud data management system that integrates testing process supervision, data processing, and online analysis. It can support intelligent connected vehicle public road testing for various testing tasks such as single vehicle, multi vehicle, and queue.
■In terms of complex meteorological simulation testing, we have completed the construction of the first self-driving closed testing platform in China that simultaneously has complex lighting, rain, fog quantification, and combination simulation. It is 200 meters long, 10 meters wide, and has a maximum speed of 100 km/h. It can achieve individual simulation and arrangement combination simulation of 0.1lux-2500lux light intensity, 5mm/h-100mm/h rainfall intensity, and 15-100m fog visibility, quantitatively reproducing light switching and daytime rainfall Typical meteorological scenarios such as nighttime rainfall, daytime fog, and nighttime fog, combined with advanced testing equipment such as navigable soft targets, autonomous driving robots, centimeter level indoor and outdoor positioning systems, accurately reproduce traffic scenes such as vehicle to vehicle, vehicle to two wheeled vehicle, vehicle to pedestrian, and vehicle to traffic signs and markings in complex meteorological environments. It can provide software for perception, decision-making, and control of autonomous driving Hardware development provides basic conditions for on-site verification.
