Heavy vehicle AEB braking method and system based on multi-source data fusion

Abstract

The invention discloses a heavy vehicle AEB braking method and system based on multi-source data fusion. The AEB braking method comprises the following steps that (1) a model is established; 2) scoring driving; 3) judging a driving style; 4) AEB strategy adjustment; 5) automatically matching a preset strategy; and 6) secondary adjustment of the AEB strategy. According to the invention, the problem of poor braking effect caused by insufficient adjustment of a single environment or vehicle factors is solved.

Description

Technical Field

[0001] The present invention relates to the technical field of vehicle automatic emergency braking, and in particular to a heavy vehicle AEB braking method and system based on multi-source data fusion. Background Art

[0002] Currently, existing vehicle AEB (Automatic Emergency Braking) systems on the market typically use fixed thresholds (such as fixed collision time TTC, fixed brake pedal amplitude) to trigger braking, leading to the following problems:

[0003] 1) Lack of personalized adaptation: Different drivers react differently to emergency braking, and fixed parameters cannot adapt to aggressive or conservative driving styles.

[0004] 2) Poor environmental adaptability: Heavy vehicles have long braking distances due to factors such as load and inertia. Fixed parameters in rainy, snowy weather or complex road conditions can easily lead to false triggering or insufficient braking.

[0005] 3) Insufficient data utilization: Existing technologies do not fully integrate driver behavior (such as attention and fatigue level) with environmental data (such as weather and road conditions) for dynamic adjustment.

[0006] With the rapid development of intelligent driving technology in recent years, multi-source data fusion technology has become key to improving the intelligence and adaptability of braking systems. However, there is still a lack of an efficient braking control solution that can integrate the environment, driver status, and vehicle conditions. Summary of the Invention

[0007] In view of the above-mentioned defects or deficiencies in the prior art, it is desired to provide a heavy-duty vehicle AEB braking method and system based on multi-source data fusion, which solves the problem of poor braking effect due to insufficient adjustment of a single environment or vehicle factors.

[0008] The present invention provides a heavy vehicle AEB braking method based on multi-source data fusion, comprising the following steps:

[0009] 1) Model building: Obtain driver identity information and build a driver information model;

[0010] 2) Driving score: Obtain driving data that meets the preset driving time and mileage, score the driver based on the driving data, and store the score in the corresponding driver information model;

[0011] 3) Driving style assessment: The driver's driving style is classified based on the score, and the driving style assessment results are stored in the corresponding driver information model;

[0012] 4) AEB strategy adjustment: The AEB strategy is adjusted based on the driver's driving style and stored in the corresponding driver information model;

[0013] 5) Automatically match preset strategies: When the driver enters the driving seat, the system identifies the driver's identity information and automatically matches the AEB strategy in the corresponding driver information model;

[0014] 6) Secondary adjustment of AEB strategy: When the driver is driving the vehicle, the AEB strategy is adjusted by adding information from external factors.

[0015] Furthermore, in step 1), the driver's facial information is obtained as identity information through the driver monitoring system.

[0016] Furthermore, in step 2), the preset driving time is not less than 20 hours, and the preset driving mileage is not less than 500 km.

[0017] Furthermore, in step 2), the driving data includes forward warning data, lane departure data, sudden acceleration data, and sudden deceleration data;

[0018] The forward warning data includes the number of forward collision warnings, pedestrian collision warnings, close vehicle distance warnings, and automatic emergency braking;

[0019] The lane departure data is the number of lane departure warnings;

[0020] The rapid acceleration data is the number of times that the average acceleration of the vehicle exceeds the acceleration threshold within the preset speed interval within the preset time within each preset speed interval;

[0021] The rapid deceleration data is the number of times, in each preset speed interval, the average deceleration of the vehicle within a preset time exceeds the deceleration threshold within the preset speed interval.

[0022] Furthermore, in step 2), the score = forward warning score + lane departure score + sudden acceleration score + sudden deceleration score;

[0023] Among them, the forward warning score = forward collision warning score + pedestrian collision warning score + close vehicle warning score + automatic emergency braking score;

[0024] The forward collision warning score is as follows: when the vehicle speed is ≤ 60 km / h, each forward collision warning is worth 0.3 points; when the vehicle speed is greater than 60 km / h, each forward collision warning is worth 0.5 points;

[0025] The pedestrian collision warning score is: 0.5 points for each pedestrian collision warning;

[0026] The scoring of the vehicle-to-vehicle distance warning is as follows: when the vehicle speed is ≤ 60 km / h, each vehicle-to-vehicle distance warning is scored as 0.5 points; when the vehicle speed is > 60 km / h, each vehicle-to-vehicle distance warning is scored as 0.3 points;

[0027] The automatic emergency braking score is: when the vehicle speed is ≤ 60 km / h, 1 point is awarded for each automatic emergency braking; when the vehicle speed is > 60 km / h, 1 point is awarded for each automatic emergency braking;

[0028] Lane departure scoring is: 1 point for each lane departure warning;

[0029] The rapid acceleration score is: 1 point for each occurrence of the average acceleration exceeding the acceleration threshold;

[0030] The score for sudden deceleration is: 1 point is awarded for each occurrence of the average deceleration exceeding the deceleration threshold.

[0031] Furthermore, in step 3), when the score is ≤5 points, the driver has a conservative driving style;

[0032] When 5 points < score ≤ 10 points, the driver has a moderate driving style;

[0033] When the score is >10, the driver has an aggressive driving style.

[0034] Furthermore, in step 4), when the driver has a moderate driving style, the original AEB strategy is maintained;

[0035] When the driver has a conservative driving style, the system increases the TTC time by 15-20% and reduces the pedal braking amplitude by 15-20% based on the existing AEB strategy.

[0036] When the driver has an aggressive driving style, the TTC time is reduced by 15-20% and the pedal braking amplitude is increased by 15-20% based on the original AEB strategy.

[0037] Furthermore, in step 6), the external factor information includes operating condition factors and driver factors; the operating condition factors include slippery road surface judgment and downhill road judgment; the driver factors include fatigue judgment and concentration judgment.

[0038] Furthermore, in step 6), adjusting the AEB strategy by superimposing external factor information includes:

[0039] 61) Slippery road detection: Identifies whether the current road is slippery and simultaneously queries weather data to determine whether it has rained or snowed within the past 2-4 hours or is currently raining or snowing. If so, increases the TTC time by 15-20% based on the current AEB strategy; otherwise, maintains the current AEB strategy.

[0040] 62) Road downhill judgment: The IMU data in the vehicle CAN signal is used to determine whether the vehicle is tilted > -5°. At the same time, the 3D map data shows whether there is a slope within 10m of the current positioning. If both are yes, the TTC time is increased by 15-20% based on the current AEB strategy; otherwise, the current AEB strategy is maintained.

[0041] 63) Fatigue assessment;

[0042] First, if the driver is judged to be moderately or severely fatigued, the TTC time will be increased by 5-10% based on the current AEB strategy;

[0043] Then, if the driver is judged to be slightly fatigued, the pedal pull-down amplitude is increased by 3-5% based on the current AEB strategy; if the driver is judged to be moderately fatigued, the pedal pull-down amplitude is increased by 10-15% based on the current AEB strategy; if the driver is judged to be severely fatigued, the pedal pull-down amplitude is increased by 20-25% based on the current AEB strategy;

[0044] 64) Attention judgment: If the current driver is judged to be inattentive, the TTC time will be increased by 5-10% based on the current AEB strategy.

[0045] In addition, the present invention also provides an AEB braking system that adopts the above-mentioned heavy vehicle AEB braking method based on multi-source data fusion.

[0046] Compared with the prior art, the present invention has the following beneficial effects:

[0047] (1) The present invention uses a DMS camera to identify the driver, establish a driver-specific braking data model, and integrate driver behavior, environmental, and vehicle data to provide a personalized braking strategy with comprehensive braking force judgment capabilities. Dynamically adjust braking force to avoid misoperation caused by errors in a single data source, thereby improving driving experience and safety.

[0048] (2) Integrate multi-dimensional inputs such as ADAS cameras, map data, weather data, and sensors to ensure the accuracy of data sources, collect vehicle and external environment information in real time, dynamically calculate the optimal braking force, and adapt to complex and changing driving environments.

[0049] (3) When the driver is extremely tired, the system can directly awaken the driver from fatigue both physically and psychologically by increasing the braking force, thus achieving a strong reminder effect. This solves the problem that the existing technology using sound and light reminders is difficult to solve the problem of driver fatigue in the long term.

[0050] (4) This application uses the vehicle's existing ADAS sensors, radars, DMS cameras, and body IMU data to score drivers. Without adding various hardware devices, it can assign corresponding AEB strategies to different driving styles, making it easier to promote and use.

[0051] It should be understood that the contents described in the summary of the invention are not intended to limit the key or important features of the embodiments of the present invention, nor are they intended to limit the scope of the present invention. Other features of the present invention will become readily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] Other features, objects and advantages of the present invention will become more apparent upon reading the detailed description of non-limiting embodiments made with reference to the following drawings:

[0053] Figure 1 This is a flow chart of the AEB braking method for heavy vehicles based on multi-source data fusion;

[0054] Figure 2 Flowchart for AEB strategy adjustment;

[0055] Figure 3 Flowchart for secondary adjustment of AEB strategy. DETAILED DESCRIPTION

[0056] The present invention will be further described in detail below with reference to the accompanying drawings and examples. It should be understood that the specific embodiments described herein are intended only to illustrate the relevant invention and are not intended to limit the invention. It should also be noted that, for ease of description, only portions relevant to the invention are shown in the accompanying drawings.

[0057] It should be noted that, in the absence of conflict, the embodiments and features of the embodiments of the present invention can be combined with each other. The present invention will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0058] Please refer to Figures 1 to 3 An embodiment of the present invention provides a heavy vehicle AEB braking method based on multi-source data fusion, characterized in that it includes the following steps:

[0059] 1) Model building: The driver’s facial information is obtained from the driver monitoring system as identity information, and a driver information model is built;

[0060] 2) Driving score: Obtain driving data of at least 20 hours of driving time and at least 500 km of driving mileage, score the driver based on the driving data, and store the score in the corresponding driver information model;

[0061] Among them, driving data includes forward warning data, lane departure data, sudden acceleration data and sudden deceleration data;

[0062] Forward warning data includes the number of forward collision warnings, pedestrian collision warnings, close vehicle warnings, and automatic emergency braking;

[0063] Lane departure data is the number of lane departure warnings;

[0064] The rapid acceleration data is the number of times the average acceleration of the vehicle exceeds the acceleration threshold within each preset speed range within a preset time period;

[0065] The rapid deceleration data is the number of times the average deceleration of the vehicle exceeds the deceleration threshold within the preset speed range within the preset time.

[0066] Score = forward warning score + lane departure score + sudden acceleration score + sudden deceleration score;

[0067] Among them, the forward warning score = forward collision warning score + pedestrian collision warning score + close vehicle warning score + automatic emergency braking score;

[0068] The forward collision warning score is as follows: when the vehicle speed is ≤ 60 km / h, each forward collision warning is worth 0.3 points; when the vehicle speed is greater than 60 km / h, each forward collision warning is worth 0.5 points;

[0069] The pedestrian collision warning score is: 0.5 points for each pedestrian collision warning;

[0070] The scoring of the vehicle-to-vehicle distance warning is as follows: when the vehicle speed is ≤ 60 km / h, each vehicle-to-vehicle distance warning is scored as 0.5 points; when the vehicle speed is > 60 km / h, each vehicle-to-vehicle distance warning is scored as 0.3 points;

[0071] The automatic emergency braking score is: when the vehicle speed is ≤ 60 km / h, 1 point is awarded for each automatic emergency braking; when the vehicle speed is > 60 km / h, 1 point is awarded for each automatic emergency braking;

[0072] The lane departure score is: 1 point for each lane departure warning;

[0073] The rapid acceleration score is: 1 point for each occurrence of the average acceleration exceeding the acceleration threshold;

[0074] The score for sudden deceleration is: 1 point for each occurrence of the average deceleration exceeding the deceleration threshold;

[0075] 3) Driving style assessment: The driver's driving style is classified based on the score, and the driving style assessment results are stored in the corresponding driver information model;

[0076] Among them, when the score is ≤5 points, the driver has a conservative driving style;

[0077] When 5 points < score ≤ 10 points, the driver has a moderate driving style;

[0078] When the score is >10, the driver has an aggressive driving style;

[0079] 4) AEB strategy adjustment: The AEB strategy is adjusted based on the driver's driving style and stored in the corresponding driver information model;

[0080] Among them, when the driver has a moderate driving style, the original AEB strategy is maintained;

[0081] When the driver has a conservative driving style, the TTC time is increased by 15-20% and the pedal braking amplitude is reduced by 15-20% based on the original AEB strategy;

[0082] When the driver has an aggressive driving style, the system reduces TTC time by 15-20% and increases the pedal braking amplitude by 15-20% based on the original AEB strategy.

[0083] 5) Automatically match preset strategies: When the driver enters the driving seat, the system identifies the driver's identity information and automatically matches the AEB strategy in the corresponding driver information model;

[0084] 6) Secondary AEB strategy adjustment: The driver adjusts the AEB strategy based on external factors while driving the vehicle;

[0085] Among them, external factor information includes working condition factors and driver factors; working condition factors include slippery road surface judgment and road downhill judgment; driver factors include fatigue judgment and concentration judgment;

[0086] Adjusting AEB strategies by adding external factors includes:

[0087] 61) Slippery road detection: Identifies whether the current road is slippery and simultaneously queries weather data to determine whether it has rained or snowed within the past 2-4 hours or is currently raining or snowing. If so, increases the TTC time by 15-20% based on the current AEB strategy; otherwise, maintains the current AEB strategy.

[0088] 62) Road downhill judgment: The IMU data in the vehicle CAN signal is used to determine whether the vehicle is tilted > -5°. At the same time, the 3D map data shows whether there is a slope within 10m of the current positioning. If both are yes, the TTC time is increased by 15-20% based on the current AEB strategy; otherwise, the current AEB strategy is maintained.

[0089] 63) Fatigue assessment;

[0090] First, if the driver is judged to be moderately or severely fatigued, the TTC time will be increased by 5-10% based on the current AEB strategy;

[0091] Then, if the driver is judged to be slightly fatigued, the pedal pull-down amplitude is increased by 3-5% based on the current AEB strategy; if the driver is judged to be moderately fatigued, the pedal pull-down amplitude is increased by 10-15% based on the current AEB strategy; if the driver is judged to be severely fatigued, the pedal pull-down amplitude is increased by 20-25% based on the current AEB strategy;

[0092] 64) Attention judgment: If the current driver is judged to be inattentive, the TTC time will be increased by 5-10% based on the current AEB strategy.

[0093] In this embodiment, in the rapid acceleration scoring, the average acceleration of the vehicle within 3 seconds is read through the IMU data of the vehicle body CAN signal. The acceleration thresholds for different speed ranges are different, as shown in Table 1. If it is judged as dangerous once, 1 point is scored.

[0094] Table 1 Acceleration thresholds in rapid acceleration scoring

[0095] Speed ​​range (km / h) <![CDATA[Safe acceleration (m / s 2 )]]> <![CDATA[Dangerous acceleration (m / s 2 )]]> (100,120] (0,0.9] (0.9,+∞) (80,100] (0,1.2] (1.2,+∞) (60,80] (0,1.3] (1.3,+∞) (40,60] (0,1.4] (1.4,+∞) (30,40] (0,1.5] (1.5,+∞) (0,30] (0,1.7] (1.7,+∞)

[0096] In the rapid deceleration scoring, the average deceleration within 3 seconds is read through the IMU data of the vehicle body CAN signal. The deceleration thresholds for different speed ranges are different. See Table 2 for details. Each dangerous deceleration is scored as 1 point.

[0097] Table 2 Deceleration thresholds in rapid deceleration scoring

[0098] Speed ​​range (km / h) <![CDATA[Safe acceleration (m / s 2 )]]> <![CDATA[Dangerous acceleration (m / s 2 )]]> (100,120] [-0.7,0) (-∞,-0.7) (80,100] [-0.9,0) (-∞,-0.9) (60,80] [-1.1,0) (-∞,-1.1) (40,60] [-1.2,0) (-∞,-1.2) (30,40] [-1.3,0) (-∞,-1.3) (0,30] [-1.5,0) (-∞,-1.5)

[0099] Aggressive driving style: Because this type of driver has an aggressive driving style, the driving characteristics reflected are that when encountering a risk of forward collision, they brake relatively late, but the force of the brakes will be deeper, the probability of point braking is low, and the probability of sudden braking is high. Therefore, the AEB strategy is adjusted to reduce the TTC time and increase the pedal braking amplitude. This will trigger AEB later, but the braking force after triggering will be greater, which is more in line with the driver's own style.

[0100] Conservative driving style: Because this type of driver drives conservatively, their driving characteristics are that when encountering a risk of forward collision, they brake early, but the force applied is not light and is gentle. The probability of point braking is high and the probability of sudden braking is low. Therefore, adjusting the AEB strategy to increase the TTC time and reduce the pedal braking amplitude will trigger AEB earlier, but the braking force after triggering will be gentle, similar to point braking to slow down, which is more in line with the driver's own style.

[0101] Moderate driving style: Because this type of driver's driving style is neither aggressive nor conservative, it is more in line with the original AEB strategy, so this style does not need to be modified.

[0102] Because heavy trucks often have multiple drivers rotating behind the wheel, the moment a driver enters the vehicle to begin driving, the DMS immediately identifies the driver and automatically selects a pre-set strategy based on historical driver profiles, choosing between "aggressive," "moderate," or "conservative." This differs from traditional methods, where drivers must manually adjust their strategy before each drive. This approach makes strategy adjustment seamless and automated. Furthermore, the adjusted driving process better aligns with the driver's preferences, minimizing driver interference and ensuring a more comfortable driving experience while ensuring safety.

[0103] like Figure 3 As shown, in the actual operating conditions of heavy-duty vehicles, they often drive in varying weather conditions and road conditions. Furthermore, heavy-duty vehicles are often driven for extended periods of time. Even with rotating drivers, they often work from dawn to dusk, spending the majority of their 24-hour shift in traffic. Therefore, traditional AEB systems fail to consider the actual driving scenarios of vehicles and fail to incorporate additional external factors to identify and adjust strategies to ensure AEB effectiveness. Therefore, this application adjusts the AEB braking strategy based on both operating conditions and driver factors.

[0104] Driver factors are divided into fatigue and concentration. This strategy adjustment serves two primary purposes. First, fatigue and inattention increase the risk of a collision. AEB triggering often requires a secondary takeover, necessitating earlier early warning intervention to give the driver more time to react. Both fatigue and inattention require increased TTC time to ensure the driver can more quickly and accurately take over and resume driving. However, traditional AEB does not use fatigue and inattention as a basis for adjusting the AEB braking strategy. Secondly, when the driver is fatigued, both the DMS warning and the forward collision warning use sound and light reminders. The driver has become accustomed to it and is not sensitive enough, so the reminders have little effect. Therefore, increasing the TTC time is to increase the number of reminders. At the same time, increasing the pedal pull-down amplitude according to different fatigue levels is to more effectively wake the driver from fatigue to a deeper level with stronger braking force and high-risk physical reflex after the AEB event occurs. Because as drivers, we all have a consensus that the most thorough way for us to wake up when we are fatigued driving is the risk of an emergency collision. At this time, a deeper emergency brake will immediately wake people up from their sleepiness, and they will not be tired again for a long time afterwards.

[0105] In fatigue assessment, if the driver blinks more than 20 times per minute, 0.5 points will be awarded;

[0106] If the driver closes his eyes for more than 3 seconds per minute, 0.5 points will be deducted; if the driver closes his eyes for more than 4 seconds, 1 point will be deducted; if the driver closes his eyes for more than 6 seconds, 1.5 points will be deducted;

[0107] If the driver yawns once per minute, 0.3 points will be deducted; if he yawns twice per minute, 0.8 points will be deducted; if he yawns three times per minute, 1.5 points will be deducted; if he yawns more than three times per minute, 2 points will be deducted;

[0108] If the driver's head posture deviates by more than 20 degrees for a period of more than 1 second or more than 1 second cumulatively within 1 minute, 1.5 points will be awarded; if it deviates for more than 2 seconds or more than 3 seconds cumulatively within 1 minute, 1 point will be awarded; if it deviates for more than 3 seconds or more than 5 seconds cumulatively within 1 minute, 2 points will be awarded;

[0109] A cumulative score of 3 points or more is considered mild fatigue; a cumulative score of 5 points or more is considered moderate fatigue; a cumulative score of 7 points or more is considered severe fatigue.

[0110] In the concentration judgment, if the driver is detected smoking, making a phone call, eating, or looking around once within 10 minutes, 0.5 points will be awarded for each time. If the cumulative score is ≥1 point, the driver will be judged as not concentrating.

[0111] This system uses a DMS camera to identify the driver and build a driver-specific braking data model. It then integrates driver behavior, environmental data, and vehicle data to provide a personalized braking strategy. This system provides comprehensive braking force determination capabilities. Dynamically adjusting braking force avoids misoperation caused by errors in a single data source, improving driving experience and safety.

[0112] It integrates multi-dimensional inputs such as ADAS cameras, map data, weather data, and sensors to ensure the accuracy of data sources, collect vehicle and external environment information in real time, dynamically calculate the optimal braking force, and adapt to complex and changing driving environments.

[0113] When the driver is extremely tired, the system enhances braking force, directly waking the driver from fatigue both physically and mentally, providing a strong reminder effect. This solves the problem that existing technologies using sound and light reminders are difficult to effectively solve the problem of driver fatigue.

[0114] In addition, an embodiment of the present invention further provides an AEB braking system that adopts the above-mentioned heavy vehicle AEB braking method based on multi-source data fusion.

[0115] In this embodiment, the driver can be scored using the vehicle's existing ADAS sensors, radar, DMS camera, and body IMU data. Without adding various hardware devices, corresponding AEB strategies can be assigned to different driving styles, making it easier to promote and use.

[0116] Example 1 (Dynamic Matching of AEB Strategy Based on Driving Style)

[0117] (1) Background

[0118] For a logistics company, heavy truck drivers A (aggressive) and B (conservative) took turns driving the same vehicle. Traditional AEB systems, due to fixed parameters, caused A to brake excessively, and the AEB system was too sensitive, impacting transportation and fuel efficiency. Meanwhile, B braked too infrequently, and the AEB system triggered too late, leading to frequent braking and significant energy consumption.

[0119] (2) Implementation steps

[0120] 1) Driving style judgment

[0121] Retrieve historical driving data (A: 8 FCW alarms in 500 kilometers, multiple sudden accelerations and decelerations, 500 kilometers score of 12 points (aggressive); B: 3 HMW alarms in 500 kilometers, mainly constant speed driving, 500 kilometers score of B is 1.5 points (conservative)).

[0122] 2) Strategy Adjustment

[0123] When the DMS camera recognizes that driver A is driving: the TTC time is shortened by 20% (for example, the original 2s trigger is adjusted to 1.6s), and the brake pedal amplitude is increased by 20%.

[0124] When the DMS camera recognizes that driver B is driving: the TTC threshold is extended by 20% (for example, the original 2s trigger is adjusted to 2.4s), and the brake pedal amplitude is reduced by 20%.

[0125] (3) Effect

[0126] Compared to before using the system, Driver A experienced a 0.4-second delay in AEB triggering at highway speeds, resulting in a 30% reduction in AEB brake activations and an 8% improvement in fuel efficiency. However, the increased pedal braking force during AEB triggering effectively prevented an accident each time. Driver B also experienced a 40% reduction in manual braking, significantly reducing physical effort and energy consumption.

[0127] Example 2 (Compounded Environmental Compensation for Rainy, Snowy Weather and Downhill Road Conditions)

[0128] (1) Background

[0129] A heavy truck fleet was driving downhill in rainy and snowy mountainous areas. The road was slippery and the slope was as high as -8°. Some vehicles in the fleet had the system deployed, while others did not. The fleet was driving in a front-and-rear following manner, which made rear-end collisions more likely.

[0130] (2) Implementation steps

[0131] 1) Working conditions

[0132] The ADAS camera identifies reflective features on the road (with a slippery confidence level >90%), and the real-time weather API returns "moderate snow," indicating the road is slippery. The IMU detects a vehicle tilt angle of -8°, and the 3D map shows a -7° slope on the current road section, indicating a downhill slope.

[0133] 2) Strategy Adjustment

[0134] Wet road compensation: TTC time is increased by 20% (superimposed compensation);

[0135] Downhill compensation: TTC time is increased by an additional 20%;

[0136] Final adjustment: TTC threshold increased from 2.0s to 2.88s (2×1.2×1.2).

[0137] (3) Effect

[0138] When the vehicle in front suddenly braked on the slope, the system triggered AEB 0.88 seconds earlier than vehicles without the system and stopped in time without an accident. However, another vehicle without the system triggered AEB braking on the same road section but ended up rear-ending it.

[0139] Example 3 (Dual Intervention of Fatigue Driving and Distracted State)

[0140] (1) Background

[0141] A driver with an aggressive driving style developed severe fatigue (frequent eye closure and head drooping) after driving continuously on the highway for 6 hours while distracted by operating a mobile phone.

[0142] (2) Implementation steps

[0143] 1) Driving style judgment

[0144] Retrieving historical driving data, analyzing the 500km data shows that the acceleration in the (60,80] speed range exceeded 1.3m / s. 2 The number of times was 3, which was counted as 3 points; at a speed greater than 60km / h, HMW and AEB occurred 5 and 4 times respectively, totaling 7.5 points, and the driver received a total of 10.5 points (aggressive type).

[0145] 2) Driver factors

[0146] Fatigue assessment: ① Blinking frequency 28 times / min for 2 minutes, 1 point; ② Eye closure duration 4.2 seconds / min for 5 minutes, 5 points; ③ Head deviation angle 25° for 3 seconds, 1 point; a total of 7 points (severe fatigue);

[0147] Concentration judgment: Operating the phone twice within 10 minutes, a total of 1 point (not focused).

[0148] 3) Strategy Adjustment

[0149] ① The preset style is aggressive: TTC time is reduced by 20%, and pedal braking amplitude is increased by 20%;

[0150] ② Fatigue compensation: TTC time increases by 10%, and pedal braking amplitude increases by 25%;

[0151] ③ Concentration compensation: additional 10% TTC time;

[0152] Final adjustment: TTC threshold dropped from 2s to 1.936s (2×0.8×1.1×1.1), and pedal braking amplitude increased by 45% (20%+25%).

[0153] (3) Effect

[0154] A broken-down vehicle appears in the middle of the road ahead of the vehicle, and AEB is triggered (1.936s threshold), performing 85% deep emergency braking (40% for traditional systems), generating a 1.2m / s 2 The deceleration wakes up the driver, who regains effective control within 0.8 seconds and immediately changes lanes to avoid a rear-end collision.

[0155] In summary, it can be seen that the adoption of the AEB braking solution of the present application greatly improves the safety of vehicle driving, is suitable for the driving styles of different drivers, provides a good driving experience, and has excellent promotion value.

[0156] Throughout this specification, terms such as "one embodiment" or "some embodiments" mean that the specific features, structures, materials, or characteristics described in conjunction with that embodiment or example are included in at least one embodiment or example of the present application. In this specification, schematic representations of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples.

[0157] The above are merely preferred embodiments of the present application and are not intended to limit the present application. Those skilled in the art will readily appreciate that various modifications and variations are possible. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of the present application shall be included within the scope of protection of the present application.

Claims

1. A heavy vehicle AEB braking method based on multi-source data fusion, characterized in that: The steps include: 1) Model building: Obtain driver identity information and build a driver information model; 2) Driving score: Obtain driving data that meets the preset driving time and mileage, score the driver based on the driving data, and store the score in the corresponding driver information model; 3) Driving style assessment: The driver's driving style is classified based on the score, and the driving style assessment results are stored in the corresponding driver information model; 4) AEB strategy adjustment: The AEB strategy is adjusted based on the driver's driving style and stored in the corresponding driver information model; 5) Automatically match preset strategies: When the driver enters the driving seat, the system identifies the driver's identity information and automatically matches the AEB strategy in the corresponding driver information model; 6) Secondary adjustment of AEB strategy: When the driver is driving the vehicle, the AEB strategy is adjusted by adding information from external factors.

2. The heavy vehicle AEB braking method based on multi-source data fusion according to claim 1, characterized in that: In step 1), the driver's facial information is obtained as identity information through the driver monitoring system.

3. The heavy vehicle AEB braking method based on multi-source data fusion according to claim 1, characterized in that: In step 2), the preset driving time is not less than 20 hours, and the preset driving mileage is not less than 500 km.

4. The heavy vehicle AEB braking method based on multi-source data fusion according to claim 1, characterized in that: In step 2), the driving data includes forward warning data, lane departure data, sudden acceleration data, and sudden deceleration data; The forward warning data includes the number of forward collision warnings, pedestrian collision warnings, close vehicle distance warnings, and automatic emergency braking; The lane departure data is the number of lane departure warnings; The rapid acceleration data is the number of times that the average acceleration of the vehicle exceeds the acceleration threshold within the preset speed interval within the preset time within each preset speed interval; The rapid deceleration data is the number of times, in each preset speed interval, the average deceleration of the vehicle within a preset time exceeds the deceleration threshold within the preset speed interval.

5. The heavy vehicle AEB braking method based on multi-source data fusion according to claim 4, characterized in that: In step 2), the score = forward warning score + lane departure score + sudden acceleration score + sudden deceleration score; Among them, the forward warning score = forward collision warning score + pedestrian collision warning score + close vehicle warning score + automatic emergency braking score; The forward collision warning score is as follows: when the vehicle speed is ≤ 60 km / h, each forward collision warning is worth 0.3 points; when the vehicle speed is greater than 60 km / h, each forward collision warning is worth 0.5 points; The pedestrian collision warning score is: 0.5 points for each pedestrian collision warning; The scoring of the vehicle-to-vehicle distance warning is as follows: when the vehicle speed is ≤ 60 km / h, each vehicle-to-vehicle distance warning is scored as 0.5 points; when the vehicle speed is > 60 km / h, each vehicle-to-vehicle distance warning is scored as 0.3 points; The automatic emergency braking score is: when the vehicle speed is ≤ 60 km / h, 1 point is awarded for each automatic emergency braking; when the vehicle speed is > 60 km / h, 1 point is awarded for each automatic emergency braking; The lane departure score is: 1 point for each lane departure warning; The rapid acceleration score is: 1 point for each occurrence of the average acceleration exceeding the acceleration threshold; The score for sudden deceleration is: 1 point is awarded for each occurrence of the average deceleration exceeding the deceleration threshold.

6. The heavy vehicle AEB braking method based on multi-source data fusion according to claim 5, characterized in that: In step 3), when the score is ≤5 points, the driver has a conservative driving style; When 5 points < score ≤ 10 points, the driver has a moderate driving style; When the score is >10, the driver has an aggressive driving style.

7. The heavy vehicle AEB braking method based on multi-source data fusion according to claim 6, characterized in that: In step 4), when the driver has a moderate driving style, the original AEB strategy is maintained; When the driver has a conservative driving style, the system increases the TTC time by 15-20% and reduces the pedal braking amplitude by 15-20% based on the existing AEB strategy. When the driver has an aggressive driving style, the TTC time is reduced by 15-20% and the pedal braking amplitude is increased by 15-20% based on the original AEB strategy.

8. The heavy vehicle AEB braking method based on multi-source data fusion according to claim 7, characterized in that: In step 6), the external factor information includes operating condition factors and driver factors; the operating condition factors include slippery road surface judgment and road downhill judgment; the driver factors include fatigue judgment and concentration judgment.

9. The heavy vehicle AEB braking method based on multi-source data fusion according to claim 8, characterized in that: In step 6), adjusting the AEB strategy by superimposing external factor information includes: 61) Slippery road detection: Identifies whether the current road is slippery and simultaneously queries weather data to determine whether it has rained or snowed within the past 2-4 hours or is currently raining or snowing. If so, increases the TTC time by 15-20% based on the current AEB strategy; otherwise, maintains the current AEB strategy. 62) Road downhill judgment: The IMU data in the vehicle CAN signal is used to determine whether the vehicle is tilted > -5°. At the same time, the 3D map data shows whether there is a slope within 10m of the current positioning. If both are yes, the TTC time is increased by 15-20% based on the current AEB strategy; otherwise, the current AEB strategy is maintained. 63) Fatigue assessment; First, if the driver is judged to be moderately or severely fatigued, the TTC time will be increased by 5-10% based on the current AEB strategy; Then, if the driver is judged to be slightly fatigued, the pedal pull-down amplitude is increased by 3-5% based on the current AEB strategy; if the driver is judged to be moderately fatigued, the pedal pull-down amplitude is increased by 10-15% based on the current AEB strategy; if the driver is judged to be severely fatigued, the pedal pull-down amplitude is increased by 20-25% based on the current AEB strategy; 64) Attention judgment: If the current driver is judged to be inattentive, the TTC time will be increased by 5-10% based on the current AEB strategy.

10. An AEB braking system using the heavy vehicle AEB braking method based on multi-source data fusion according to any one of claims 1 to 9.

Images