Nowadays, the assisted driving function has become an important reference to evaluate the intelligence of a car. Major manufacturers, especially the new forces that build cars, are willing to flaunt the level of vehicle driving assistance when releasing new models. But at present, there is no unified evaluation standard and test method in the industry. Therefore, we regard fully automatic driving as a beautiful vision. From the user's point of view, we integrate the common assisted driving functions in the market into daily car scenes, experience their functional differences, and bring different feelings to daily use.

In the experience part of the advanced driver assistance function of vehicles, considering the different strategies of different manufacturers to achieve driver assistance, the hardware level is not taken as a reference, and only the actual performance is evaluated subjectively and objectively. In the comprehensive score, the functional experience of goods accounts for 75% and the subjective use experience accounts for 25%. The functional experience of the project is divided into basic assisted driving function, navigation assisted driving function and automatic parking function, while the subjective experience is divided into functional interactive experience and humanization degree of driving assistance function.

The video is loading. ...

Before we officially start this content, we want to reiterate that so far, any high-level driver assistance function is not automatic driving. When using these functions, we still need to pay attention to the road conditions at all times and be ready to take over the vehicle at any time. To quickly understand the full text, you can click on the video above to watch.

The model we welcome in this issue is Weilai ES6, and the system version carried by the vehicle during use is NIO OS 3.0.0. In the sensor part, Weilai ES6 is equipped with 9 cameras (including a three-eye camera), 5 millimeter-wave radars and 12 ultrasonic radars. The processor adopts Mobileye EyeQ4 chip, with 2.5TOPS computing power, and the high-precision map is provided by Baidu.

In the part of basic driving assistance function, there is no rigid limit on the scope of use of the experience item, which can be activated in all cases in theory. If the vehicle can pass or identify smoothly, it can greatly reduce our daily driving fatigue.

Weilai ES6 can quickly identify the most basic scenes, such as road signs, road markings, construction cones and so on. And some information can also be displayed in the HUD synchronously, thus reducing the time for eyes to leave the road and improving safety.

Especially when facing the road ahead, the cone bucket is used to guide the lane change. At present, Weilai ES6 does not support changing lanes with the cone barrel, but adopts the scheme of stopping before the cone barrel. We let the vehicles pass at 30 km/h, 40 km/h and 50 km/h respectively. Therefore, in most cases, vehicles can stop, and in a few cases, collisions will occur. Therefore, Weilai ES6 will eventually stop at this project.

In the adaptive cruise part, we should not only test the success rate of car following, but also take into account its acceleration performance and response speed during acceleration and deceleration, which determines the comfort and the possibility of congestion in the process. The performance of Weilai ES6 here can be said to be more handy, even the smacking of an old driver. In particular, Weilai ES6 will not directly copy the action of 1: 1 when the preceding vehicle just started to brake, but will leave a certain reaction time slightly, so that the passengers in the car can feel the whole process softer and more comfortable.

In terms of lane keeping function, Weilai ES6 can almost completely reassure you. Whether on a flat road with a good view or on a congested road with a view blocked by the preceding car, Weilai ES6 can stably stay in the middle of the lane without frequent direction correction.

After the lane keeping function, further is the ability to face the curve. In order to restore the daily road scene as much as possible, we designed four bends from easy to difficult, namely the entry-level 90-degree right bend; Experience the ability of the system to continuously change the direction of the vehicle; The experience sensor can sense the large curvature curve of vehicle handling ability in short distance. On this basis, the concept of uphill and downhill is introduced to try to detect the different performances of sensor angle and vehicle speed control.

Weilai ES6 performed equally well here. The upper four corners are not difficult for it, and the direction control of the corners is very soft. When it is possible that the vehicle cannot pass because of the too fast speed, the vehicle will slow down in time to ensure that it can pass.

In this project, we also designed an extra test link for all gifted students like Weilai ES6. We found the most difficult compound curve in Beijing, with greater curvature and more interference factors, so as to feel the control of vehicles facing extreme curves. The performance of Weilai ES6 didn't disappoint us either. In the process, the vehicle is very confident in the control of speed and direction, and almost no reverse correction is needed. The performance in the corner is orderly.

Another thing to note is that in the scene we designed, the traffic speed of all corners is the road speed limit. However, for vehicles supporting navigation-assisted driving such as Weilai ES6, the speed control will give priority to the system. If you can't pass or the speed is lower than the speed limit of the curve, we will manually set the speed as the road speed limit. In the actual experience, the speed of Weilai ES6 passing through all corners will be slightly higher than the speed limit of corners, and there is a certain risk of speeding, so we should pay more attention to it in actual driving.

Then there is the ability of the vehicle to change lanes autonomously. In the system update of NIO OS 3.0.0, Wei Lai particularly emphasized the optimization of lane change judgment logic. The same is true in actual experience. Toggle the turn light pole to trigger the parallel command, and the vehicle can be easily completed in most cases, greatly improving the overall success rate and fluency. If there is a high-speed vehicle coming behind the target lane, or a vehicle with the same lane change, Weilai ES6 can also go back to the original lane and wait.

In the scene where vehicles occupy the road in front of the same lane, we test the situation that the number of vehicles occupying the road in front is 65, 438+000%, 50% and 25% at the speeds of 30km/h, 40km/h and 50km/h respectively. If they all pass smoothly, we will further test the reaction of the vehicle we are driving after the vehicle suddenly evades in the following States.

Weilai ES6 can quickly find the front car to avoid collision in all the above scenarios, but the control strategy is relatively conservative, and it adopts the operation of rear parking. Even if only 25% of the front car body protrudes into the lane, it will not take the overtaking logic of turning and avoiding, and there is a risk of being chased by the rear car.

In the assisted driving function based on navigation, we mainly experience the scene of vehicles changing ramps independently, overtaking automatically and passing through intersections without road markings. If the vehicle has the above functions, when the advanced driving assistance is activated, the driver can basically avoid paying attention to navigation and other information after entering the closed loop or high speed.

The first thing Weilai ES6 should praise is that it uses Baidu's high-precision map. Compared with brands such as Tesla, which doesn't use high-precision maps, or Tucki, which uses high-precision maps in Gaode, there will be an extra five-ring main road in Beijing, which can greatly enhance its frequency of use in Beijing.

Compared with the basic assisted driving function, the most important improvement of high-speed navigation assisted driving is that vehicles can enter and leave the ramp independently according to high-precision maps. Here we will experience the scenes of vehicles entering and leaving two-way ramps, three-way ramps and one-lane and two-lane ramps respectively.

The performance of Weilai ES6 in this scene is not ideal. First of all, before entering the ramp, the deceleration is sudden, and there is a risk of being chased by the rear car. Secondly, when the two-lane ramp merges into the main road, if the vehicle happens to be in the inner lane of the two-lane ramp, it will not change lanes to the left and enter the main road at the first time after driving to the intersection of the ramp and the main road. Instead, in the process of gradually narrowing the ramp, it will change lanes to the outer lane first and then try to merge into the main road. In many experiences, I have also encountered the situation of ignoring the car coming from behind the ramp. This function is very risky as a whole.

Even better, when you need to drive into the ramp, if the vehicle thinks that it can't get to the outermost lane in front of the ramp, it will give priority to inform you through NOMI to avoid the embarrassment of making a wrong intersection on the expressway.

When it is necessary to automatically change lanes and overtake under navigation-assisted driving, Weilai ES6 has a better experience, and the decisive degree of lane change is similar to the feeling of controlling lane change with turn signals mentioned above. The system will also automatically determine the remaining mileage from the front exit. If it is less than a certain distance, it will not rush to overtake from the left, reducing the extra risks brought by the merger.

& ltimg class = " imgborder " src = " http://image . bitautooi