Vehicle rear early warning method and system based on ADAS

Abstract

The invention relates to the technical field of vehicle early warning, in particular to an ADAS-based vehicle rear early warning method, which comprises the steps of collecting rear peripheral data corresponding to a target vehicle in real time, and obtaining a vehicle real-time state corresponding to the target vehicle and a road speed limit dynamic state corresponding to a current road; judging whether the rear vehicle distance is smaller than a safety distance threshold based on the rear peripheral data, and determining a first early warning level; judging whether the speed difference meets a preset speed difference threshold value or not, and if yes, determining a second early warning level; inputting the real-time state of the vehicle into a pre-trained anomaly recognition model for recognition, and obtaining a corresponding third early warning level; and determining a corresponding target early warning level based on the first early warning level, the first weight data, the second early warning level, the second weight data, the third early warning level and the third weight data, determining a corresponding target early warning coping strategy, and performing execution based on the target early warning coping strategy. According to the invention, the early warning accuracy of the rear vehicle can be improved.

Description

Technical Field

[0001] The present application relates to the technical field of vehicle warning, and in particular to a vehicle rear warning method and system based on ADAS. Background Art

[0002] As the number of cars increases, road traffic safety has attracted much attention. ADAS technology has emerged and is widely used. Its purpose is to assist drivers and actively intervene through the collaborative work of multiple components to reduce the incidence of traffic accidents. Vehicle rear warning, as an important function of ADAS, is of great significance to driving safety. It is difficult for drivers to fully understand the situation behind the vehicle while driving, especially in scenarios such as changing lanes and reversing. Potential dangers behind the vehicle can easily cause serious accidents.

[0003] Although there are basic rear warning functions such as reversing radar and reversing image, they have obvious limitations. Each warning system is independent of each other, lacks data fusion and coordination, and cannot comprehensively assess danger. During data collection and processing, sensor data is not synchronized, affecting the accuracy of warning.

[0004] From the above, we can see that how to improve the accuracy of warning for rear vehicles remains to be solved. Summary of the invention

[0005] In order to improve the accuracy of warning for rear vehicles, the present application provides a vehicle rear warning method and system based on ADAS.

[0006] In the first aspect, the present application provides a vehicle rear warning method based on ADAS, which adopts the following technical solution:

[0007] A vehicle rear warning method based on ADAS includes: real-time collection of rear peripheral data corresponding to a target vehicle, obtaining the real-time vehicle state corresponding to the target vehicle and the road speed limit dynamics corresponding to the current road, wherein the rear peripheral data includes distance, speed, and driving trajectory; determining a corresponding safety distance threshold based on the real-time vehicle state and the road speed limit dynamics, judging whether the rear vehicle distance is less than the safety distance threshold based on the rear peripheral data, and determining a first warning level; determining speed difference data corresponding to the target vehicle and the rear vehicle, judging whether the speed difference meets a preset speed difference threshold, and if so, determining a corresponding second warning level; inputting the real-time vehicle state into a pre-trained abnormal recognition model for recognition, and obtaining a corresponding third warning level; retrieving first weight data corresponding to the first warning level, retrieving second weight data corresponding to the second warning level, and retrieving third weight data corresponding to the third warning level, determining a corresponding target warning level based on the first warning level, the first weight data, the second warning level, the second weight data, the third warning level, and the third weight data, determining a corresponding target warning response strategy based on the target warning level, and executing based on the target warning response strategy.

[0008] By adopting the above technical solution, through real-time collection of surrounding data behind the target vehicle, the real-time status of the vehicle and the dynamic road speed limit, multi-dimensional information is integrated to determine different warning levels and assign corresponding weights, so as to determine the target warning level and response strategy. This combination of multi-source data fusion and multi-dimensional warning comprehensively considers various vehicle driving conditions, realizes more accurate warning of rear vehicles, and effectively improves the accuracy and reliability of the warning.

[0009] Optionally, the method further includes: in the process of dynamically collecting the rear surrounding data, the real-time status of the vehicle and the road speed limit, using a time synchronization technology based on a high-precision clock chip.

[0010] By adopting the above technical solution, time synchronization technology based on high-precision clock chips is used in the process of dynamically collecting rear surrounding data, vehicle real-time status and road speed limits, which can ensure the temporal consistency of data from different sources and avoid data deviation and confusion caused by differences in collection time. This will provide an accurate and reliable data basis for subsequent multi-dimensional warning analysis, warning decision fusion, and feedback optimization, thereby ensuring the stable and efficient operation of the entire vehicle rear comprehensive warning system from the source, improving the accuracy and reliability of the warning, and enhancing driving safety.

[0011] Optionally, in the first warning level, the method also includes: dividing the first warning level into three levels, when the first warning level is a level one warning, the indicator light flashes at a frequency of once per second, and the prompt sound is a soft prompt sound with a frequency of 500 Hz and a duration of 0.5 seconds; when the first warning level is a level two warning, the indicator light flashes at a frequency of 3 times per second, and the prompt sound becomes a medium-intensity prompt sound with a frequency of 1000 Hz and a duration of 1 second; when the first warning level is a level three warning, the indicator light flashes rapidly at a high frequency of 5 times per second, and simultaneously emits a sharp alarm sound with a frequency of 1500 Hz and a duration of 1.5 seconds.

[0012] By adopting the above technical solution, the first warning level is divided into three levels. Through the differentiated settings of the indicator light flashing frequency and the prompt audio frequency and duration at different levels, it can provide drivers with clearer, more intuitive and gradient danger level prompts. This refined warning method allows drivers to quickly and accurately judge the urgency of the distance danger of the rear vehicle based on different warning performances, so as to take corresponding measures in time, effectively improving the warning effect based on distance judgment and enhancing driving safety.

[0013] Optionally, in the second warning level, the method also includes: obtaining the traffic flow conditions of the current road of the target vehicle, and determining a corresponding real-time congestion index based on the traffic flow conditions; when it is determined based on the traffic flow conditions that the traffic flow is large and the real-time congestion index exceeds a preset congestion index threshold, the speed difference threshold is lowered.

[0014] By adopting the above technical solution, by obtaining the traffic flow status of the target vehicle's current road and determining the real-time congestion index, the speed difference threshold is lowered when the traffic flow is large and the real-time congestion index exceeds the preset threshold. This measure can flexibly adjust the warning trigger conditions according to the complexity of the actual traffic environment. Early warnings are issued in complex traffic scenarios, making the warning system more suitable for changing road conditions, effectively improving the timeliness and accuracy of the speed difference warning for the rear vehicle, and better ensuring the driving safety of vehicles under complex road conditions.

[0015] Optionally, the weight data includes first weight data, second weight data and third weight data, and the method also includes: judging whether it is currently in normal weather conditions during the day, if so, the first weight data accounts for 40%, the second weight data accounts for 30%, and the third weight data accounts for 30%; judging whether it is currently driving at night, if so, the third weight data is increased to 40%, the first weight data is adjusted to 35%, and the second weight data is 25%; judging whether it is currently in bad weather, if so, the second weight data is increased to 40%, the first weight data is 30%, and the third weight data is 30%.

[0016] By adopting the above technical solution, by dynamically adjusting the first, second, and third weight data according to different driving environments (normal weather during the day, driving at night, and bad weather), the warning decision can more reasonably integrate multi-dimensional warning results, allowing the system to accurately judge the danger level in different scenarios, and then determine a more suitable final warning level and response strategy, thereby enhancing the adaptability and accuracy of the warning system under complex and changeable driving conditions.

[0017] Optionally, in the process of executing the target warning response strategy, the method also includes: obtaining driver operation data and vehicle operation data corresponding to the driver when executing the warning; determining the average response time corresponding to the driver based on the driver operation data and vehicle operation data, calling a pre-set response time threshold, and judging whether the average response time is greater than the response time threshold; if so, the trigger timing corresponding to the target warning level is advanced, and the warning prompt method is adjusted.

[0018] By adopting the above technical solution, when executing the target warning response strategy, the average response time is determined by obtaining the driver's operation data and vehicle operation data, and compared with the preset response time threshold. When the average response time is greater than the threshold, the target warning level trigger timing is advanced and the prompt method is adjusted. This can dynamically optimize the warning mechanism based on the driver's actual reaction situation, ensure that the warning system is in line with the driver's reaction characteristics, improve the driver's response efficiency to the warning, and further enhance the effectiveness and safety of the vehicle's rear warning.

[0019] In the second aspect, the present application provides a vehicle rear warning system based on ADAS, which adopts the following technical solution:

[0020] A vehicle rear warning system based on ADAS, comprising:

[0021] The rear surrounding data acquisition module is used to collect the rear surrounding data corresponding to the target vehicle in real time, obtain the real-time vehicle status corresponding to the target vehicle and the road speed limit dynamics corresponding to the current road, wherein the rear surrounding data includes distance, speed, and driving trajectory;

[0022] The early warning judgment module dynamically determines the corresponding safety distance threshold based on the real-time state of the vehicle and the road speed limit, determines whether the distance to the rear vehicle is less than the safety distance threshold based on the rear surrounding data, and determines the first early warning level; determines the speed difference data corresponding to the target vehicle and the rear vehicle, and determines whether the speed difference meets the preset speed difference threshold. If so, determines the corresponding second early warning level; inputs the real-time state of the vehicle into a pre-trained abnormal recognition model for recognition, and obtains the corresponding third early warning level;

[0023] The target warning level determination module retrieves the first weight data corresponding to the first warning level, retrieves the second weight data corresponding to the second warning level, retrieves the third weight data corresponding to the third warning level, and determines the corresponding target warning level based on the first warning level, the first weight data, the second warning level, the second weight data, the third warning level and the third weight data, determines the corresponding target warning response strategy based on the target warning level, and executes it based on the target warning response strategy.

[0024] In a third aspect, the present application provides a vehicle rear warning system based on ADAS, which adopts the following technical solution:

[0025] A vehicle rear warning method based on ADAS includes a processor, wherein the processor runs a program of any one of the above-mentioned vehicle rear warning methods based on ADAS.

[0026] In a fourth aspect, the present application provides a storage medium, which adopts the following technical solution:

[0027] A storage medium stores a program of any one of the above-mentioned ADAS-based vehicle rear warning methods.

[0028] In summary, the present application includes at least one of the following beneficial technical effects:

[0029] First, by combining multi-source data fusion with multi-dimensional warning, real-time data of the surrounding area behind the vehicle, the real-time status of the vehicle and the dynamic speed limit of the road are collected to provide comprehensive information support for the warning. Different warning levels are determined based on these data, and the target warning level and response strategy are obtained after assigning weights, comprehensively considering the driving conditions of the vehicle, and improving the accuracy of the warning for the rear vehicle from the overall architecture level.

[0030] Secondly, in the processing of warning details, the time synchronization technology of high-precision clock chips is used to ensure data consistency and provide a reliable basis for subsequent analysis; the first warning level is refined and graded, and different levels are equipped with different prompts to allow drivers to clearly perceive the degree of danger; the speed difference threshold is flexibly adjusted according to traffic flow and congestion index to warn of complex road conditions in advance; the weight is dynamically adjusted according to the driving environment to accurately judge the danger level. These measures improve the accuracy of warnings from each warning link.

[0031] Finally, when executing the warning response strategy, the average response time is determined based on the driver's operation and vehicle operation data. After comparing with the threshold, the warning triggering timing and prompt method are dynamically optimized to make the warning system fit the driver's reaction characteristics, further enhance the effectiveness and accuracy of the warning, and comprehensively ensure the safety of the rear vehicle. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] Figure 1The figure is a flow chart of a vehicle rear warning method based on ADAS according to an exemplary embodiment.

[0033] Figure 2 It is a structural block diagram of a vehicle rear warning system based on ADAS according to an exemplary embodiment. DETAILED DESCRIPTION

[0034] Embodiments of the present application are described in detail below, examples of which are shown in the accompanying drawings.

[0035] In the description of this specification, the description with reference to the terms "certain embodiments", "one embodiment", "some embodiments", "illustrative embodiments", "examples", "specific examples", or "some examples" means that the specific features, structures, materials, or characteristics described in conjunction with the embodiments or examples are included in at least one embodiment or example of the present application. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials, or characteristics described may be combined in any one or more embodiments or examples in a suitable manner.

[0036] The present application embodiment discloses a vehicle rear warning method based on ADAS, referring to Figure 1 ,include:

[0037] S100, collecting rear surrounding data corresponding to the target vehicle in real time, obtaining the real-time vehicle status corresponding to the target vehicle and the road speed limit dynamics corresponding to the current road.

[0038] Millimeter-wave radar: Millimeter-wave radar works by emitting millimeter-wave signals and receiving reflected waves. It continuously emits millimeter waves at certain time intervals. When the signal encounters a vehicle or other object behind, part of the signal will be reflected back and received by the radar. By measuring the time difference between the transmitted and received signals and the frequency change of the signal (Doppler effect), the distance and relative speed between the rear object and the target vehicle can be accurately calculated. For example, on a highway, the millimeter-wave radar can monitor the position and speed changes of the rear vehicle in real time within a range of several hundred meters, and can perform multiple measurement updates per second to ensure the timeliness and accuracy of the data.

[0039] Camera: The camera is installed at the rear of the vehicle and captures images or videos at a fixed frame rate (such as 30 frames per second). The captured images are analyzed through image recognition and computer vision technology. The outline, position and driving trajectory of the rear vehicle can be identified. For example, the feature point matching algorithm is used to track the position changes of the rear vehicle in continuous frame images to determine its driving direction and trajectory. If the rear vehicle suddenly changes lanes, the camera can capture this change in time and record its trajectory information.

[0040] Among them, it can also be combined with sensors such as laser radar (LiDAR) to further supplement the rear surrounding data. LiDAR creates a three-dimensional point cloud map of the rear environment by emitting laser beams and measuring the time of reflected light. It can more accurately depict the shape and position of objects behind, providing richer information for subsequent early warning analysis.

[0041] The speed sensor is installed on the vehicle's transmission system or wheels to determine the vehicle's speed by measuring the rotation speed of the wheels or the movement of the transmission components. The speed sensor transmits real-time speed data to the vehicle's electronic control unit (ECU) and can further provide it to the early warning system. For example, when the vehicle accelerates or decelerates, the speed sensor can reflect the change in speed in real time.

[0042] The steering sensor is used to measure the steering angle and steering speed of the vehicle. When the driver turns the steering wheel, the steering sensor transmits steering information to the system, which is critical for judging the change in the vehicle's driving direction. For example, when the vehicle turns, the rear surrounding data can more accurately assess the potential danger of the rear vehicle.

[0043] The navigation system of the vehicle stores detailed map data, including speed limit information of different roads. The warning system can obtain the speed limit value of the road where the vehicle is currently located through the interface with the navigation system. When the vehicle travels to different speed limit sections, the system can automatically update the speed limit information.

[0044] It should be noted that the present application includes V2X communication. Through the communication technology between vehicles and infrastructure (V2I) or vehicles and vehicles (V2V), vehicles can obtain the latest speed limit dynamics of the road in real time. For example, the traffic management department can send temporary speed limit adjustment information to surrounding vehicles through roadside communication equipment. After the vehicle receives this information, the early warning system can update the road speed limit data in time to adapt to different traffic conditions.

[0045] S110, dynamically determine the corresponding safety distance threshold based on the real-time status of the vehicle and the road speed limit, determine whether the distance to the rear vehicle is less than the safety distance threshold based on the rear surrounding data, and determine the first warning level; determine the speed difference data corresponding to the target vehicle and the rear vehicle, and determine whether the speed difference meets the preset speed difference threshold. If so, determine the corresponding second warning level; input the real-time status of the vehicle into a pre-trained abnormality recognition model for identification, and obtain the corresponding third warning level.

[0046] Among them, for the first warning data, the vehicle's speed is one of the important factors affecting the safety distance. Generally speaking, the faster the speed, the greater the safety distance requirement. The warning system can calculate the safety distance threshold based on a pre-set algorithm, combined with the current vehicle speed and road speed limit dynamics. For example, on a highway, when a vehicle is traveling at a speed of 100km / h, the safety distance threshold may be set to 100 meters; on urban roads, the speed is lower, and the safety distance threshold is correspondingly reduced.

[0047] In addition, different road conditions will also affect the setting of the safety distance. For example, on a slippery road, the vehicle's braking distance will increase, so the safety distance threshold needs to be increased accordingly. The early warning system can detect weather conditions through sensors (such as rain sensors) and dynamically adjust the safety distance threshold based on the road type (such as highways, urban roads, etc.).

[0048] First warning level confirmed

[0049] The distance between the rear vehicle and the target vehicle collected in real time is compared with the calculated safety distance threshold. If the distance of the rear vehicle is less than the safety distance threshold, the system will trigger the early warning mechanism. According to the degree to which the distance of the rear vehicle is less than the safety distance threshold, the first early warning level is divided into different levels. For example, it can be divided into a first-level early warning, a second-level early warning, and a third-level early warning. When the distance of the rear vehicle approaches the safety distance threshold, a first-level early warning may be triggered. At this time, the system can remind the driver by flashing the indicator light on the dashboard; when the distance is further reduced, a second-level early warning is triggered, and an early warning method such as a prompt tone may be added; when the distance is very close to the target vehicle, a third-level early warning is triggered, and a strong alarm may be sounded and emergency measures such as automatic braking may be taken.

[0050] It should be pointed out here that in the first warning level, the system calculates the safety distance threshold through a preset algorithm based on the real-time status of the target vehicle (such as speed, acceleration, etc.) and the speed limit dynamics of the current road. For example, the faster the speed and the more complex the road conditions, the greater the safety distance threshold. The distance data between the rear vehicle and the target vehicle collected in real time will be continuously compared with the safety distance threshold. When the distance of the rear vehicle is less than the safety distance threshold, the first warning level determination process is triggered. According to the degree to which the distance of the rear vehicle is less than the safety distance threshold, the first warning level is divided into three levels. For example, when the distance is between 80% and 100% of the safety distance threshold, it is determined as a first-level warning; when it is between 50% and 80%, it is determined as a second-level warning; when it is less than 50%, it is determined as a third-level warning.

[0051] Therefore, the method also includes:

[0052] S1111, the first warning level is divided into three levels. When the first warning level is level one, the indicator light flashes once per second, and the prompt sound is a soft prompt sound with a frequency of 500 Hz and a duration of 0.5 seconds.

[0053] Once the system determines that the current warning is level one, it will send a command to the indicator light control module on the dashboard. According to the command, the indicator light control module controls the indicator light to flash at a frequency of once per second. This frequency is relatively low, which can attract the driver's attention without disturbing the driver too much. At the same time, the system will send a command to the audio system to play a prompt sound. After receiving the command, the audio system generates a soft prompt sound with a frequency of 500Hz and a duration of 0.5 seconds, and plays it through the speakers in the car. This gentle prompt sound can remind the driver that the distance of the vehicle behind is close to the safety threshold without causing excessive fright.

[0054] S1112: When the first warning level is the second warning level, the indicator light flashes three times per second, and the warning sound becomes a medium-intensity warning sound with a frequency of 1000 Hz and a duration of 1 second.

[0055] Among them, when the system detects that the distance to the rear vehicle is further reduced and it is determined to be a level 2 warning, the indicator light control module will receive a new instruction to increase the flashing frequency of the indicator light to 3 times per second. A higher flashing frequency can attract the driver's attention more strongly. At the same time, the audio system will also receive an instruction to adjust the prompt sound, increasing the frequency of the prompt sound to 1000Hz and the duration to 1 second. The increase in frequency and duration makes the intensity of the prompt sound moderate, allowing the driver to clearly understand that the degree of danger has increased.

[0056] S1113: When the first warning level is level three, the indicator light flashes rapidly at a high frequency of 5 times per second, and emits a sharp alarm sound with a frequency of 1500 Hz and a duration of 1.5 seconds.

[0057] If the distance to the vehicle behind continues to decrease and reaches the judgment standard of the third-level warning, the indicator light control module will make the indicator light flash quickly at a high frequency of 5 times per second. This high-frequency flashing can give the driver a strong visual warning. At the same time, the audio system will generate and play a sharp alarm with a frequency of 1500Hz and a duration of 1.5 seconds. The sharp alarm sound can arouse the driver's high alertness in hearing, reminding the driver that the distance to the vehicle behind is already very dangerous and immediate measures need to be taken.

[0058] The warning level classification and prompt method scientifically calculates the safety distance threshold, continuously monitors the distance to the rear vehicle and provides graded warnings. It uses the changes in the indicator flashing frequency and the prompt sound parameters to feedback the driver the degree of danger of the rear vehicle approaching in a progressive and intuitive way. At the same time, the system's continuous monitoring and dynamic adjustment mechanism ensures that the warning information can adapt to the distance changes in a timely manner, helping the driver to detect danger in a timely manner and take countermeasures, effectively improving driving safety.

[0059] In addition, the speed information of the target vehicle and the rear vehicle has been obtained through sensors such as millimeter wave radar. The early warning system can calculate the speed difference between the two in real time. For example, if the speed of the target vehicle is 80km / h and the speed of the rear vehicle is 100km / h, the speed difference is 20km / h. The system pre-sets speed difference thresholds for different situations. These thresholds are determined based on factors such as road type and traffic flow. For example, on highways, the speed difference threshold may be set to 20km / h; on urban roads, due to the low speed, the speed difference threshold may be set to 10km / h. When the calculated speed difference meets the preset speed difference threshold, it means that the speed change of the rear vehicle relative to the target vehicle may bring danger, and the system will determine the corresponding second warning level.

[0060] Level division: Similar to the first warning level, the second warning level can also be divided according to the size of the speed difference. The larger the speed difference, the higher the warning level. For example, when the speed difference just reaches the threshold, it may trigger a first-level warning; when the speed difference is large, it triggers a second-level warning; when the speed difference is very large, it triggers a third-level warning. Different warning levels correspond to different warning methods and response measures.

[0061] It should be noted here that in the second warning level, the methods also include:

[0062] S1121, obtaining the traffic flow condition of the current road of the target vehicle, and determining the corresponding real-time congestion index based on the traffic flow condition.

[0063] Among them, the vehicle uses V2X technology to communicate with surrounding vehicles and traffic facilities. By communicating with surrounding vehicles, the vehicle can know the location and speed of nearby vehicles, thereby estimating the density of vehicles; by communicating with roadside facilities, such as smart traffic lights, it can obtain information such as the length of the vehicle queue at the intersection to determine the traffic flow. Then, the vehicle's own millimeter-wave radar, camera, etc. can also help. The radar can detect the number and distance of vehicles in front and behind, and the camera can identify the distribution of vehicles and comprehensively judge the degree of traffic congestion.

[0064] The vehicle density, speed, queue length and other data obtained by V2X and on-board sensors are aggregated together and processed using a pre-set congestion index calculation model. The congestion index calculation model takes into account multiple factors and weights that affect congestion, and calculates the real-time congestion index of the current road.

[0065] S1122: When it is determined based on the traffic flow condition that the traffic flow is heavy and the real-time congestion index exceeds a preset congestion index threshold, the speed difference threshold is lowered.

[0066] Among them, the system will judge the acquired traffic flow conditions and the calculated real-time congestion index. First, it is determined whether the traffic flow is large, which can be determined by the set vehicle density threshold or other relevant indicators. For example, if the vehicle density exceeds the threshold of 50 vehicles per kilometer, the traffic flow is considered to be large. Then, the real-time congestion index is compared with the pre-set congestion index threshold. This congestion index threshold is set based on a large amount of traffic data and actual experience, and represents the critical state of traffic congestion. For example, when the congestion index exceeds 70 (out of 100), it is considered that the traffic congestion has reached the level where the warning parameters need to be adjusted.

[0067] Then, when both conditions of heavy traffic and real-time congestion index exceeding the preset threshold are met, the system will lower the speed difference threshold. The speed difference threshold is a key parameter used to determine whether the speed difference between the rear vehicle and the target vehicle is dangerous. Lowering the speed difference threshold means that in the case of traffic congestion, the system will be more sensitive to the potential dangers caused by the speed change of the rear vehicle. For example, under normal circumstances, the speed difference threshold may be set to 20km / h, and when the above conditions are met, the system will reduce it to 15km / h. In this way, when the speed difference between the rear vehicle and the target vehicle reaches 15km / h, the second warning level will be triggered, instead of waiting until the speed difference reaches 20km / h, so as to issue a warning in advance and improve driving safety.

[0068] Finally, the anomaly recognition model is usually built based on machine learning or deep learning algorithms. During the training phase, a large amount of vehicle driving data is used as training samples, including data on normal driving states and various abnormal driving states. For example, data on the state of the vehicle in normal driving, sudden braking, sudden acceleration, serpentine driving, etc. is collected. By extracting features from these data and training the model, the model can learn the characteristic patterns of normal driving states and abnormal driving states.

[0069] The real-time status data of the target vehicle (such as speed, acceleration, steering angle, etc.) is input into the pre-trained anomaly recognition model. The anomaly recognition model will analyze and judge the input data and compare it with the learned normal and abnormal patterns. If the input data deviates from the normal pattern, the model will identify the abnormal situation and output the corresponding degree of abnormality.

[0070] According to the degree of abnormality output by the abnormality recognition model, the third warning level is divided into different levels. For example, when the degree of abnormality is relatively mild, a first-level warning may be triggered; when the degree of abnormality is moderate, a second-level warning may be triggered; when the degree of abnormality is severe, a third-level warning may be triggered. Different warning levels correspond to different warning methods and response measures to remind the driver to pay attention to the abnormal state of the vehicle.

[0071] S120, call the first weight data corresponding to the first warning level, call the second weight data corresponding to the second warning level, call the third weight data corresponding to the third warning level, determine the corresponding target warning level based on the first warning level, the first weight data, the second warning level, the second weight data, the third warning level and the third weight data, determine the corresponding target warning response strategy based on the target warning level, and execute it based on the target warning response strategy.

[0072] The system has pre-set corresponding weight data for the first warning level, the second warning level, and the third warning level. These weight data are determined based on the importance of different warning factors in actual risk assessment. For example, in some cases, the distance factor may be more important, and the first weight data corresponding to the first warning level may be higher; while in other cases, the speed difference factor or the abnormal state factor may be more critical, and the corresponding weight data will be increased. According to the current situation, the warning system retrieves the first weight data corresponding to the first warning level, the second weight data corresponding to the second warning level, and the third weight data corresponding to the third warning level from the database.

[0073] Specifically, the weight data includes first weight data, second weight data and third weight data, and the method also includes: judging whether it is currently in normal weather conditions during the day, if so, the first weight data accounts for 40%, the second weight data accounts for 30%, and the third weight data accounts for 30%; judging whether it is currently driving at night, if so, the third weight data is increased to 40%, the first weight data is adjusted to 35%, and the second weight data is 25%; judging whether it is currently in bad weather, if so, the second weight data is increased to 40%, the first weight data is 30%, and the third weight data is 30%.

[0074] Different driving scenarios (normal weather during the day, night, and bad weather) have different influencing factors on the danger behind the vehicle. This method dynamically adjusts the weights of each warning dimension according to the characteristics of these scenarios, so that the warning system can better adapt to different environmental conditions. For example, in normal weather during the day, the field of vision and road conditions are relatively good, and the various warning factors are relatively balanced; when driving at night, the field of vision is limited, and the abnormal behavior of the rear vehicle is more difficult for the driver to detect, so the weight of the abnormal behavior warning (third warning level) is increased; in bad weather, the road conditions are complex and vehicle speed control is more critical, so the weight of the speed difference warning (second warning level) is increased.

[0075] By reasonably allocating the weights of each warning level in different scenarios, the warning system can more accurately and comprehensively assess the danger behind. The adjustment of weights reflects the changes in the importance of each warning factor in different scenarios, avoiding the warning deviation that may be caused by using fixed weights in all cases. For example, after increasing the weight of abnormal behavior warning at night, the system can more keenly capture the abnormal trajectory and posture of the rear vehicle, and issue a timely warning that is more in line with the actual degree of danger.

[0076] Then, the first warning level, the first weight data, the second warning level, the second weight data, the third warning level and the third weight data are weighted. For example, a weighted summation method can be used to multiply each warning level by the corresponding weight, and then add the results to obtain a comprehensive warning value. According to the calculated comprehensive warning value, the target warning level is divided into different levels, for example, it can be divided into three levels: low, medium and high. The higher the comprehensive warning value, the higher the target warning level.

[0077] It should be pointed out here that the system has pre-established a warning response strategy library, which stores various response strategies corresponding to different target warning levels. According to the determined target warning level, the corresponding target warning response strategy is selected from the strategy library. For example, when the target warning level is low, the driver may only be reminded through the indicator light on the dashboard and a slight prompt sound; when the target warning level is medium, the intensity of the prompt sound may be increased, and detailed warning information may be displayed on the display; when the target warning level is high, braking measures may be automatically taken, and a strong alarm sound may be issued.

[0078] The early warning system sends the selected target early warning response strategy to the corresponding actuators, such as the instrument panel, audio system, braking system, etc. The actuators operate according to the strategy requirements, promptly convey danger information to the driver or take automatic safety measures.

[0079] Therefore, in the process of executing the target-based early warning response strategy, the method also includes:

[0080] S131, collecting driver operation data and vehicle operation data corresponding to the driver when obtaining execution warning.

[0081] Among them, various sensors inside the vehicle will continuously monitor the driver's operating behavior. For example, sensors installed on the brake pedal, accelerator pedal and steering wheel can respectively record the driver's actions of braking, stepping on the accelerator and turning the steering wheel; when the early warning system issues an early warning, these sensors will begin to record data such as the specific time point and operation range from the driver receiving the early warning to making the corresponding operation.

[0082] The vehicle's electronic control unit (ECU) will collect various operating parameters of the vehicle in real time, such as speed, acceleration, driving direction, etc. During the warning execution, these data will be recorded synchronously for subsequent analysis of the vehicle's status changes before and after the warning.

[0083] The collected driver operation data and vehicle operation data will be temporarily stored in the vehicle's data storage module. In order to facilitate subsequent processing, these data will be organized in a certain format, such as using timestamps as indexes to associate different types of data. Then, the data will be transmitted to the data analysis module of the early warning system through the vehicle's internal communication bus (such as CAN bus).

[0084] S132, determining an average response time corresponding to the driver based on the driver operation data and the vehicle operation data, retrieving a preset response time threshold, and determining whether the average response time is greater than the response time threshold.

[0085] After receiving the data, the analysis module will analyze the driver's operation data, and calculate the driver's response time for each warning by comparing the time when the warning is issued and the time when the driver first makes an operation. For example, if the warning is issued at t1 and the driver brakes at t2, then the response time is t2-t1.

[0086] The system will perform statistical analysis on the response time during multiple warnings and calculate the average response time. Generally speaking, the moving average method is used, that is, only the response time data in the most recent period is considered, which can reflect the driver's current response status more promptly. The warning system pre-stores a response time threshold, which is set based on a large amount of experimental data and actual driving experience, and represents the response speed that the driver should achieve under normal circumstances. The data analysis module will retrieve this preset response time threshold from the system's database. Compare the calculated average response time with the retrieved response time threshold. If the average response time is greater than the response time threshold, it means that the driver's response speed is slow and may not be able to respond to the dangerous situation prompted by the warning in time.

[0087] S133: If yes, the triggering time corresponding to the target warning level is advanced, and the warning prompt mode is adjusted.

[0088] When it is determined that the average driver response time is greater than the response time threshold, the warning system will adjust the trigger conditions corresponding to the target warning level. For example, the warning was originally triggered when the rear vehicle reached a certain distance from the target vehicle, but now the distance threshold will be appropriately increased so that the warning is issued earlier. This can give the driver more reaction time and improve the ability to deal with danger.

[0089] In addition to triggering the warning in advance, the system will also adjust the warning prompt method to enhance the reminder effect on the driver. For example, increase the volume of the prompt sound, change the frequency of the prompt sound to make it sharper and more obvious; or increase the flashing intensity and frequency of the indicator light to make it easier for the driver to notice the warning information. In addition, some additional prompt methods may be added, such as seat vibration, etc., to ensure that the driver can perceive the warning in time through multiple sensory stimulations.

[0090] In the embodiment of the present application, a time synchronization technology based on a high-precision clock chip is used in the process of dynamically collecting rear surrounding data, real-time vehicle status, and road speed limit.

[0091] Specifically, a high-precision clock chip is selected and installed in the vehicle data acquisition control unit. It can provide an accurate time signal. When the vehicle starts, the standard time is obtained from the vehicle clock system or external time server to initialize the chip and set its working parameters. When each sensor (millimeter wave radar, camera, etc.) collects data, it works according to the unified time signal of the chip. After collecting the data, a timestamp that accurately records the collection time is immediately added. The timestamped data is transmitted to the central processing unit through the internal network of the vehicle. After receiving it, the central processing unit first verifies the timestamp information. If there is an abnormality, it will handle the error. The central processing unit processes the data using a time calibration algorithm. The algorithm takes into account the sensor clock characteristics and transmission delays, etc., and fine-tunes the timestamp to unify all data to the vehicle system clock reference. The calibrated data can be accurately integrated and analyzed, allowing the vehicle rear warning system to combine the data of multiple sensors at the same time to accurately judge the rear situation and issue timely warnings.

[0092] The present application embodiment discloses a vehicle rear warning system based on ADAS, referring to Figure 2 , the system includes but is not limited to:

[0093] The rear surrounding data acquisition module 200 is used to collect the rear surrounding data corresponding to the target vehicle in real time, and obtain the real-time vehicle status corresponding to the target vehicle and the road speed limit dynamics corresponding to the current road, wherein the rear surrounding data includes distance, speed, and driving trajectory;

[0094] The warning judgment module 210 dynamically determines the corresponding safety distance threshold based on the real-time state of the vehicle and the road speed limit, determines whether the distance to the rear vehicle is less than the safety distance threshold based on the rear surrounding data, and determines the first warning level; determines the speed difference data corresponding to the target vehicle and the rear vehicle, and determines whether the speed difference meets the preset speed difference threshold. If so, determines the corresponding second warning level; inputs the real-time state of the vehicle into the pre-trained abnormal recognition model for recognition, and obtains the corresponding third warning level;

[0095] The target warning level determination module 220 calls the first weight data corresponding to the first warning level, calls the second weight data corresponding to the second warning level, calls the third weight data corresponding to the third warning level, and determines the corresponding target warning level based on the first warning level, the first weight data, the second warning level, the second weight data, the third warning level and the third weight data, determines the corresponding target warning response strategy based on the target warning level, and executes it based on the target warning response strategy.

[0096] An embodiment of the present application further discloses an ADAS-based vehicle rear warning system, comprising a processor, wherein a program of any one of the above-mentioned ADAS-based vehicle rear warning methods is running on the processor.

[0097] An embodiment of the present application further discloses a storage medium storing a program of any one of the above-mentioned ADAS-based vehicle rear warning methods.

[0098] Although the embodiments of the present application have been shown and described above, it can be understood that the above embodiments are exemplary and cannot be understood as limitations on the present application. Ordinary technicians in this field can change, modify, replace and modify the above embodiments within the scope of the present application.

Claims

1. A vehicle rear warning method based on ADAS, characterized in that: include: Collect the rear surrounding data corresponding to the target vehicle in real time, obtain the real-time status of the target vehicle and the speed limit dynamics of the current road, where the rear surrounding data includes distance, speed, and driving trajectory; Dynamically determine a corresponding safety distance threshold based on the real-time state of the vehicle and the road speed limit, determine whether the distance of the rear vehicle is less than the safety distance threshold based on the rear surrounding data, and determine a first warning level; determine speed difference data corresponding to the target vehicle and the rear vehicle, and determine whether the speed difference meets a preset speed difference threshold, and if so, determine a corresponding second warning level; input the real-time state of the vehicle into a pre-trained abnormality recognition model for recognition, and obtain a corresponding third warning level; Retrieve the first weight data corresponding to the first warning level, retrieve the second weight data corresponding to the second warning level, retrieve the third weight data corresponding to the third warning level, determine the corresponding target warning level based on the first warning level, the first weight data, the second warning level, the second weight data, the third warning level and the third weight data, determine the corresponding target warning response strategy based on the target warning level, and execute based on the target warning response strategy.

2. The vehicle rear warning method based on ADAS according to claim 1, characterized in that: The method also includes: In the process of dynamically collecting the rear surrounding data, the real-time status of the vehicle and the road speed limit, a time synchronization technology based on a high-precision clock chip is used.

3. The vehicle rear warning method based on ADAS according to claim 1, characterized in that: In the first warning level, the method further comprises: The first warning level is divided into three levels. When the first warning level is level 1, the indicator light flashes once per second, and the prompt sound is a gentle prompt sound with a frequency of 500 Hz and a duration of 0.5 seconds; When the first warning level is level 2, the indicator light flashes 3 times per second, and the warning tone changes to a medium-intensity tone with a frequency of 1000 Hz and a duration of 1 second. When the first warning level is level three, the indicator light flashes rapidly at a high frequency of 5 times per second, and at the same time emits a sharp alarm sound with a frequency of 1500Hz and a duration of 1.5 seconds.

4. The vehicle rear warning method based on ADAS according to claim 1, characterized in that: In the second warning level, the method further comprises: Obtaining the traffic flow condition of the current road of the target vehicle, and determining a corresponding real-time congestion index based on the traffic flow condition; When it is determined based on the traffic flow condition that the traffic flow is heavy and the real-time congestion index exceeds a preset congestion index threshold, the speed difference threshold is lowered.

5. The vehicle rear warning method based on ADAS according to claim 1, characterized in that: The weight data includes first weight data, second weight data and third weight data, and the method further includes: Determine whether it is normal weather conditions during the day. If so, the first weighted data accounts for 40%, the second weighted data accounts for 30%, and the third weighted data accounts for 30%; When judging whether the vehicle is currently driving at night, if so, the third weight data is increased to 40%, the first weight data is adjusted to 35%, and the second weight data is adjusted to 25%; Determine whether the current weather is bad. If so, the second weight data is increased to 40%, the first weight data is 30%, and the third weight data is 30%.

6. The vehicle rear warning method based on ADAS according to claim 1, characterized in that: In the process of executing the target early warning response strategy, the method further includes: Collect the driver's corresponding driver operation data and vehicle operation data when obtaining execution warning; Determine an average response time corresponding to the driver based on the driver operation data and the vehicle operation data, retrieve a preset response time threshold, and determine whether the average response time is greater than the response time threshold; If so, the trigger timing corresponding to the target warning level is advanced, and the warning prompt method is adjusted.

7. A vehicle rear warning system based on ADAS, characterized in that: include: The rear surrounding data acquisition module is used to collect the rear surrounding data corresponding to the target vehicle in real time, obtain the real-time vehicle status corresponding to the target vehicle and the road speed limit dynamics corresponding to the current road, wherein the rear surrounding data includes distance, speed, and driving trajectory; The early warning judgment module dynamically determines the corresponding safety distance threshold based on the real-time state of the vehicle and the road speed limit, determines whether the distance to the rear vehicle is less than the safety distance threshold based on the rear surrounding data, and determines the first early warning level; determines the speed difference data corresponding to the target vehicle and the rear vehicle, and determines whether the speed difference meets the preset speed difference threshold. If so, determines the corresponding second early warning level; inputs the real-time state of the vehicle into a pre-trained abnormal recognition model for recognition, and obtains the corresponding third early warning level; The target warning level determination module retrieves the first weight data corresponding to the first warning level, retrieves the second weight data corresponding to the second warning level, retrieves the third weight data corresponding to the third warning level, and determines the corresponding target warning level based on the first warning level, the first weight data, the second warning level, the second weight data, the third warning level and the third weight data, determines the corresponding target warning response strategy based on the target warning level, and executes it based on the target warning response strategy.

8. A vehicle rear warning system based on ADAS, characterized in that: The invention comprises a processor, wherein a program of the vehicle rear warning method based on ADAS as described in any one of claims 1 to 6 is run in the processor.

9. A storage medium, characterized in that: A program for the vehicle rear warning method based on ADAS as described in any one of claims 1 to 6 is stored.

Images