In the wave of vigorous development of the digital economy
Intelligent networked vehicles have become an important direction of innovation in the automotive industry
In recent years, CRRC Electric will
"Smart car", "smart road", "all-round cloud"
The three are closely integrated
The application scenarios of intelligent networked vehicles
From a beautiful vision to a living reality
Since the end of 2018, on the open road of the smart bus demonstration line in Xiangjiang New Area, Changsha, the CRRC electric 12-meter autonomous bus has been tested and operated so far; As a new achievement of CRRC to create a new business card of "high-speed rail on rubber wheels", the driverless microcirculation bus small V has been operated in Chengdu, Chongqing, Yibin, Shanghai and other places.
Through the extensive use of "cloud brains" and algorithms, intelligent networked vehicles are becoming more and more "human". When to cross the traffic light? Where can I park my car? For intelligent networked vehicles, these are behavioral norms that need to be constantly "learned".
Get on the 5.3-meter-long CRRC electric driverless micro-circulation bus small V and drive on the smart bus demonstration line in Xiangjiang New Area, Changsha. The vehicle is equipped with many equipment such as lidar and ultrasonic radar, which can carry out decision-making control through perception fusion, data fusion and positioning, automatic steering, lane change and other highly automated driving behaviors to achieve convenient connection and accurate stopping.
Therefore, in the case of unmanned operation, the micro-circulation bus automatically adjusts the steering angle, recognizes the surrounding situation, and runs normally on the road.
The small V is the latest intelligent networked vehicle that can achieve L4 unmanned driving, which can complete high-level autonomous driving in a relatively fixed driving area, and meet the operational needs of various park scenarios such as industrial parks, parks, schools, and airports, without any manual operation.
When passing through an intersection, the speed is significantly slower than other vehicles on the road. "There were a lot of passengers who reacted to the problem." According to CRRC electric technology experts, the speed reduction is to ensure the safety of driving on public roads.
In fact, perception and recognition is still a bottleneck that needs to be solved urgently for intelligent networked vehicles. Vehicles driving on the road, will encounter a variety of complex road conditions, whether the intelligent driving system can make effective judgments like humans, the first hurdle to pass is identification.
For example, when driving behind a sprinkler truck, the intelligent networked car cannot accurately determine whether the water mist sprayed by the sprinkler is a soft object or a hard object, and whether it can pass through.
In order to solve these identification problems, for more than 5 years of test operation, the CRRC electric technical team has been exploring various driving scenarios, and then the methods of identification, judgment and decision-making, through software and algorithms, form new behavioral norms, "implant" the brain of intelligent networked vehicles, continuously update and "domesticate" vehicles, and can be closer to human driving behavior.
At the beginning, CRRC Electric's intelligent driving new energy buses will slam on the brakes when passing through traffic lights, regardless of whether there are pedestrians and vehicles next to them, and the ride will not feel good. "At that time, there was no sensing capability on the side of the vehicle, and the sensing capability at the rear was only 20 to 30 meters away." According to experts, through the improvement of software and hardware and algorithm capabilities in recent years, the intelligent driving bus currently operating online has a sensible range of 40 meters on the left and right sides, and a sensible range of 50 meters to 80 meters at the rear, and the perceptual ability has also been improved from the original digital information to image information.
Another example is the parking at the pit stop. If the platform is crowded by social vehicles, resulting in the vehicle being unable to drive to the specified position, the "brainless" intelligent networked car will "refuse" to open the door to pick up and drop off passengers. Through these years of continuous "domestication", now, it has learned to be flexible, and will choose the right position to park and pick up and drop off passengers before and after the social vehicle.
As the first enterprise in China to build an integrated vehicle-road-cloud management platform, CRRC leads the industry in the number of open road test licenses for electric intelligent networked buses. Up to now, CRRC Electric has more than 63,000 new energy vehicles operating around the world, of which the cumulative mileage of unmanned driving exceeds 200,000 kilometers, and a large amount of intelligent perception information data is generated every day. The overall data volume has reached 2PB, which is equivalent to the storage capacity of 8.39 million mobile phones.
In the future, CRRC's electric technology team will continue to use data and algorithms to enable intelligent networked vehicles to complete more behavioral norms and empower more urban public transport intelligent service upgrades.
Manuscript material: Gao Xiaoyan, Wang Dan
Editor: Tian Jianxiao
Proofreading: Ouyang Pengfei Tang Rong
Review: Xie Yanbing
(This article is provided by the company)