TWI831488B - Advanced emergency braking system for vehicle driving assistance and control method thereof - Google Patents

Abstract

An advanced emergency braking system for vehicle driving assistance includes an in-vehicle monitoring system, a vehicle host, an out-of-vehicle monitoring system, and an artificial intelligence (AI) module. The in-vehicle monitoring system is used to monitor a driver to generate a feature information of the driver’s attention. The vehicle host is used to provide a vehicle operation information. The out-of-vehicle monitoring system is used to monitor the driving environment outside of the vehicle to generate driving environmental information. The AI module afterward can determine a braking assist strategy based on the feature information of the driver’s attention, the vehicle operation information, and the vehicle environmental information.

Description

Advanced emergency braking system and control method applied in vehicle driving assistance

The present invention relates to a vehicle driving assistance system, and in particular to an advanced emergency braking system (Advanced Emergency Braking System) and its control method.

The Society of Automotive Engineers (SAE) divides autonomous driving levels into six levels, of which level 0 is no automation, level 1 is driving assistance, level 2 is partial automation, level 3 is conditional automation, and level 4 is high Automatic, level 5 is fully automatic. Autonomous driving levels 0 to 2 require humans to monitor the driving environment, while autonomous driving levels 3 to 5 require system monitoring. Currently, the Advanced Driver Assistance System (ADAS) belongs to Level 2~3 of the autonomous driving level. The advanced emergency braking system is one of the advanced driver assistance systems. The advanced emergency braking system can automatically detect potential collision risks ahead. When it is determined that the collision risk reaches a preset value, it can remind the driver to slow down through sounds, lights or vibrations. When the advanced emergency braking system has continued to warn for a preset time and the driver does not respond, the advanced emergency braking system will automatically perform the braking function to reduce the vehicle's speed and avoid a collision.

However, the driving behaviors of different drivers are obviously different, and even the driving behaviors of the same driver at different times and different models will be different. The brake assist strategy of traditional advanced emergency braking systems cannot be adjusted in real time based on the driver's driving behavior, the driver's status or the conditions of the surrounding environment. Therefore, when the driver's driving behavior, the driver's state or the surrounding environment changes, the original brake assist strategy may not be applicable. For example, when the driver is in low spirits or distracted, the driver's braking response to changes in vehicle conditions ahead may become slow. If the original braking assistance strategy is still used, the risk of collision will be greatly increased.

The purpose of the present invention is to provide an advanced emergency braking system that self-adapts to the driver's state and a control method thereof.

According to the present invention, an advanced emergency braking system used in vehicle driving assistance includes an in-vehicle monitoring system, a vehicle host computer, an external vehicle monitoring system and an artificial intelligence module. The in-car monitoring system is used to monitor a driver to generate driver focus characteristic information. The vehicle host computer is used to provide vehicle operation information. The vehicle exterior monitoring system is used to monitor the vehicle's external environment to generate vehicle environment information. The artificial intelligence module determines a braking assistance strategy based on the driver's focus characteristic information, the vehicle operation information, and the vehicle environment information.

According to the present invention, a control method for an advanced emergency braking system used in vehicle driving assistance includes: monitoring the driver to generate driver concentration characteristic information; providing vehicle operation information through a vehicle host computer; monitoring the vehicle's The external environment is used to generate a vehicle environment information; and the driver's focus characteristic information, the vehicle operation information and the vehicle environment information are input into an artificial intelligence module to determine a brake assist strategy.

The advanced emergency braking system and its control method of the present invention can determine the driver's current driving behavior, the driver's status or the condition of the surrounding environment through the driver's concentration characteristic information, the vehicle operation information and the vehicle environment information. to adjust the brake assist strategy to help improve safety.

Figure 1 shows an advanced emergency braking system 10 of the present invention applied to vehicle driving assistance. In FIG. 1 , the advanced emergency braking system 10 includes an in-vehicle monitoring system 11 , a vehicle host 12 , an exterior monitoring system 13 , a playback device 14 and an artificial intelligence (AI) module 15 . The in-vehicle monitoring system 11 is used to monitor a driver to generate driver concentration characteristic information A. In one embodiment, the in-car monitoring system 11 includes a camera 111 and an AI chip 112. The camera 111 is used to photograph the driver to obtain an image or video. The AI chip 112 can determine the driver based on the image or video. behavior and/or state to generate driver concentration characteristic information A. For example, the AI chip 112 can determine based on the image or video whether the driver is using a mobile phone, whether the driver is smoking, whether the driver is drinking, the way the driver is holding the steering wheel, whether the driver is paying attention to the road ahead, and / Or whether the driver is mentally disturbed, etc., and driver concentration characteristic information A is generated based on the judgment results. . In one embodiment, the AI chip 112 can be replaced by an image processing chip.

The vehicle host computer 12 can control the operation of various devices in the vehicle, such as controlling the front wheels of the vehicle according to the rotation direction and angle of the steering wheel, or controlling the speed of the vehicle according to the accelerator state. The vehicle host computer 12 can provide a vehicle operation information B according to the driver's operation. Vehicle operation information B includes but is not limited to at least one of accelerator information, brake information, steering wheel angle information, and vehicle speed information.

The vehicle exterior monitoring system 13 is used to monitor the vehicle's external environment to generate vehicle environment information C. The external environment of the vehicle includes but is not limited to lane offset, front and rear distance, and/or the speed limit of the road. In one embodiment, the vehicle exterior monitoring system 13 includes a monitoring device 131 and an AI chip 132 . The monitoring device 131 includes but is not limited to radar, cameras and/or global positioning system (GPS). The radar can be used to detect the distance between the vehicle and the vehicle in front and/or behind. The camera can be used to obtain images of the outside of the vehicle. GPS You can locate the road you are currently on and its speed limit. The AI chip 132 determines the distance between the front and rear vehicles based on the radar detection data, determines the lane offset amount based on the image obtained by the camera, and determines the speeding frequency based on the speed limit of the road where it is located, and generates vehicle environment information C based on the determination results. In another embodiment, the vehicle exterior monitoring system 13 may also use an image processing chip (not shown) to process the images obtained by the camera to determine the lane deviation amount.

The artificial intelligence module 15 is connected to the in-vehicle monitoring system 11 , the vehicle host computer 12 and the out-of-vehicle monitoring system 13 . The artificial intelligence module 15 has a trained inference model (not shown in the figure), and can analyze the driver's focus characteristic information A, vehicle operation information B and vehicle environment information C based on the parameters F obtained through training. , to determine a brake assist strategy D based on the driver's operating habits. In one embodiment, the artificial intelligence module 15 can preset multiple brake assist strategies and select a more appropriate brake assist strategy from the multiple brake assist strategies according to the driver's operating habits. The driver's operating habits include but are not limited to the driver's control ability and concentration. In one embodiment, the inference model has an automatic supervision edge learning function and can therefore be connected to an automotive edge computing platform to improve the reliability and trust of autonomously adapting to driver behavior. The automatic supervised edge learning function means that the artificial intelligence module 15 can connect to the cloud server through the Internet. The cloud server can collect more vehicle conditions and different driver data through the Internet to obtain a more accurate artificial intelligence model. Through the Internet connection, the artificial intelligence module 15 can update the parameters in the artificial intelligence module 15 through the cloud server to provide a more appropriate braking assistance strategy. In one embodiment, the artificial intelligence module 15 includes, but is not limited to, machine learning algorithms such as support vector machines (SVM), deep neural networks (DNN), or decision trees.

The brake assist strategy D determined by the artificial intelligence module 15 is received and executed by the vehicle host computer 12 . In one embodiment, the brake assist strategy D includes warning the user, and the vehicle host 12 sends a warning signal E to a playback device 14 (such as a display or speaker) according to the brake assist strategy D determined by the artificial intelligence module 15 to generate A warning sound, text and/or image to remind the driver. In one embodiment, the brake assist strategy D further includes active braking. When the vehicle host 12 sends the warning signal E and the driver still does not respond after a preset time, the vehicle host 12 controls the braking system to perform a braking operation.

The advanced emergency braking system 10 of the present invention can further reduce the risk of collision by selecting an appropriate braking assistance strategy based on the driver's status, the vehicle's status, and the vehicle's surrounding environment through the artificial intelligence module 15 .

Figure 2 shows an embodiment of a training system for obtaining training parameters F. In the training system 20 of Figure 2, an in-vehicle monitoring system 21, a vehicle host computer 22 and an external vehicle monitoring system 23 respectively generate driver concentration characteristic information A', vehicle operation information B' and vehicle environment information C'. The in-vehicle monitoring system 21, the vehicle host computer 22, and the vehicle exterior monitoring system 23 may refer to the vehicle interior monitoring system 11, the vehicle host computer 12, and the vehicle exterior monitoring system 13 in Fig. 1 . Multiple pieces of driver focus characteristic information A', vehicle operation information B' and vehicle environment information C' are stored in a knowledge base 24. The training model (Training model) 25 has the same architecture as the inference model of the artificial intelligence module 15 in Figure 1 . Using multiple pieces of driver focus characteristic information A', vehicle operation information B' and vehicle environment information C' in the knowledge base 24 to feed the training model 25, the training model can learn to identify the driver's operation habits to obtain training parameters. F. In one embodiment, the training model 25 classifies the driver's operating habits into conservation type, regular type and adventure type. For example, drivers who often drive faster than the road speed limit, often apply sudden braking, often lose concentration, often fail to maintain a safe distance from the vehicle in front, and often deviate from the center of the lane will be classified as risk-taking. Drivers whose vehicle speed is always lower than the road speed limit, do not slam on the brakes, stay focused, always maintain a safe distance from the vehicle in front, and never deviate from the center of the lane will be classified as conservative. Drivers in between are classified as ordinary.

Figure 3 shows the control method of the advanced emergency braking system 10 of the present invention. Referring to FIGS. 1 and 3 , as shown in steps S10 , S11 and S12 , the in-vehicle monitoring system 11 , the vehicle host computer 12 and the out-of-vehicle monitoring system 13 respectively generate driver concentration characteristic information A, vehicle operation information B and vehicle environment information C. Give 15 to the AI module. Then, the artificial intelligence module 15 can determine a braking assistance strategy D based on the driver's concentration characteristic information A, the vehicle operation information B and the vehicle environment information C, as shown in step S13. Then, the vehicle host computer 12 executes the brake assist strategy D determined in step S13, that is, step S14.

In the embodiment of FIG. 3 , step S13 of determining the brake assist strategy includes steps S131 and S132. In step S131, the artificial intelligence module 15 determines the driver's operating habits based on the driver's concentration characteristic information A, vehicle operation information B, and vehicle environment information C. After determining the driver's operating habits, the artificial intelligence module 15 determines the brake assist strategy D based on the operating habits, as shown in step S132. In one embodiment, the advanced emergency braking system 10 can preset multiple brake assist strategies. For example, three brake assist strategies can be set: conservation mode, regular mode and adventure mode, respectively corresponding to the target. Classified into conservative, normal and adventurous drivers. The collision time (time-to-collision; TTC) of the conservative mode brake assist strategy can be set to 1 second (s) and the maximum deceleration can be set to 4m/s ² , and the TTC of the normal mode brake assist strategy can be set to 1.5s and the maximum deceleration can be set to 6m/s ² . The TTC of the adventure mode brake assist strategy can be set to 2s and the maximum deceleration can be set to 8m/s ² . Although three types of drivers and three brake assist strategies are used as examples for explanation, the invention is not limited thereto. In the above embodiment, a driver's operating habit corresponds to a brake assist strategy. Therefore, in different embodiments, the artificial intelligence module 15 can also directly determine the braking assistance strategy based on the driver's focus characteristic information A, vehicle operation information B and vehicle environment information C, without first judging the driver's operating habits. , and then determine the braking assistance strategy corresponding to the driver's operating habits.

In the embodiment of FIG. 3 , step S14 is for the vehicle host computer 12 to execute the brake assist strategy D determined in step S14 , which includes steps S141 and S142 . In step S141, the vehicle host computer 12 calculates a collision risk value. The collision risk value represents the probability of collision with the vehicle in front. Next, in step S142, when the vehicle host computer 12 determines that the collision risk value reaches the threshold of the current brake assist strategy, the warning signal E is sent. In one embodiment of step 142, the vehicle host computer 12 calculates the collision risk value based on the "distance to the vehicle in front" and "current vehicle speed" in the vehicle environment information C. The vehicle environment information C may be provided to the vehicle host 12 by the artificial intelligence module 15 . In another embodiment, the vehicle host 12 is further connected to the vehicle exterior monitoring system 13 , and the vehicle exterior monitoring system 13 directly transmits the vehicle environment information C to the vehicle host 12 .

In order to make it easier to understand the present invention, specific examples are used to illustrate the present invention. In one embodiment, when the traffic flow ahead is sparse, the advanced emergency braking system 10 of the present invention recognizes that the driver is distracted, such as talking on a cell phone, drinking a drink, lighting a cigarette, etc. However, the driver can still maintain the vehicle's speed by controlling the steering wheel. Driving on the center line of the lane, keeping a safe distance from the vehicle in front, and the vehicle speed does not change drastically. Although the driver's concentration level is low at this time, he still has good control over the vehicle conditions. Therefore, the artificial intelligence model of the advanced emergency braking system 10 of the present invention Group 15 infers a brake assist strategy that can be maintained in the normal mode (ie, the default mode). In another embodiment, if the driver also performs distracted actions, but in a crowded traffic environment, the driver's ability to control the vehicle condition is reduced, so the artificial intelligence module 15 infers the risk of collision due to the deterioration of the driving condition. Therefore, the brake assist strategy of adventure mode will be selected to provide more stringent braking warning conditions, an earlier time point for active braking intervention (such as setting TTC=2s), and a deceleration value that strengthens the braking action (such as the maximum deceleration value). is 8m/s ² ). Adventure mode's brake assist strategy includes warning the user in a more prominent way (such as a loud emergency sound) and allowing the car to automatically brake faster to avoid a collision. In another embodiment, if the driver's concentration and control ability are both high, the artificial intelligence module 15 will infer that the driver is a conservative type, and will select a conservative mode brake assist strategy. At this time, the brake assist control The settings will be looser, that is, the braking warning conditions will be relaxed, the time point for active braking intervention will be shortened (such as setting TTC=1s), and the deceleration value of active braking will be reduced (such as the maximum deceleration value being 4m/s ² ). The conservative mode brake assist strategy includes warning the driver in a gentler manner and allowing the car to automatically brake more gently, giving the driver a more comfortable feeling.

While the vehicle is moving, the aforementioned control method can be continuously performed to switch to a suitable braking assistance strategy at any time based on the real-time environment and driver status.

The above are only embodiments of the present invention, and do not limit the present invention in any form. Although the present invention has been disclosed in the embodiments above, they are not used to limit the present invention. Anyone with ordinary knowledge in the technical field, Without departing from the scope of the technical solution of the present invention, the technical content disclosed above can be used to make some changes or modifications to equivalent embodiments with equivalent changes. Any simple modifications, equivalent changes and modifications made to the above embodiments still fall within the scope of the technical solution of the present invention.

10:Advanced emergency braking system 11: In-car surveillance system 111:Camera 112:AI chip 12:Car host 13:Outdoor surveillance system 131:Monitoring device 132:AI chip 14:Playback device 15:Artificial intelligence module 20:Training system 21:In-car surveillance system 22:Car host 23:Outdoor surveillance system 24:Knowledge base 25:Training module A: Driver focus characteristic information A’: Driver focus characteristic information B: Vehicle operation information B’: Vehicle operation information C: Vehicle environment information C’: Vehicle environment information D: Brake assist strategy E: warning signal F: training parameters

Figure 1 shows an advanced emergency braking system applied to vehicle driving assistance according to the present invention. Figure 2 shows an embodiment of the training system. Figure 3 shows the control method of the advanced emergency braking system of the present invention.

10:Advanced emergency braking system

11: In-car surveillance system

111:Camera

112:AI chip

12:Car host

13:Outdoor surveillance system

131:Monitoring device

132:AI chip

14:Playback device

15:Artificial intelligence module

A: Driver focus characteristic information

B: Vehicle operation information

C: Vehicle environment information

D: Brake assist strategy

E: warning signal

Claims (10)

An advanced emergency braking system used in vehicle driving assistance, including: an in-vehicle monitoring system for monitoring a driver to generate driver focus characteristic information; a vehicle host computer for providing vehicle operation information; An off-vehicle monitoring system for monitoring the external environment of the vehicle to generate vehicle environment information; and an artificial intelligence module coupled to the in-vehicle monitoring system, the vehicle host computer and the off-vehicle monitoring system, according to the driver The focus characteristic information, the vehicle operation information and the vehicle environment information determine a brake assist strategy; wherein, the artificial intelligence module determines the driver's operating habits from the driver's focus characteristic information, the vehicle operation information and the vehicle environment information. , to determine the brake assist strategy. The advanced emergency braking system as described in claim 1, wherein the in-vehicle monitoring system includes: a camera for photographing the driver to obtain an image; and an artificial intelligence (AI) chip connected to the camera, through the The image generates information on the driver's concentration characteristics. The advanced emergency braking system as described in claim 1, wherein the vehicle exterior monitoring system includes: a monitoring device for obtaining external environment information of the vehicle; and an artificial intelligence (AI) chip connected to the monitoring device, according to the The external environment information generates the vehicle environment information. The advanced emergency braking system as described in claim 3, wherein the monitoring device includes At least one of a camera, a radar and a global positioning system. The advanced emergency braking system as described in claim 1, wherein the artificial intelligence module has an automatic supervised edge learning function. The advanced emergency braking system as described in claim 1, wherein the vehicle operation information includes at least one of accelerator information, brake information, steering wheel angle information and vehicle speed information. A control method for an advanced emergency braking system applied in vehicle driving assistance, including: A1. Monitoring the driver to generate driver focus characteristic information; A2. Providing vehicle operation information through a vehicle host; A3. Monitoring the The external environment of the vehicle to generate a vehicle environment information; and A4. Input the driver's focus characteristic information, the vehicle operation information and the vehicle environment information into an artificial intelligence module to determine a brake assist strategy; wherein, the step A4 includes The driver's operating habits are determined based on the driver's concentration characteristic information, the vehicle operating information, and the vehicle environment information, and the brake assist strategy is determined based on the operating habits. The control method of the advanced emergency braking system as described in claim 7, wherein the step A1 includes: photographing the driver to obtain an image; and generating the driver's focus characteristic information through the image. As for the control method of the advanced emergency braking system described in claim 7, step A3 includes: obtaining the external environment information of the vehicle; and generating the vehicle environment information based on the external environment information. The control method of the advanced emergency braking system as described in claim 7, wherein the artificial intelligence module has an automatic supervised edge learning function.