Forward collision early warning device and method based on risk assessment and electronic equipment

Abstract

The invention discloses a forward collision early warning device and method based on risk assessment, electronic equipment and a storage medium, and relates to the field of driving prediction, and the device comprises a sensing unit which is used for obtaining the relative position data of a detected object, and carrying out the modeling according to the relative position data of the recognized detected object; detecting a driving state of a driver; detecting the motion state of the vehicle in real time; the processing unit is used for processing the data of the sensing unit in real time, performing target classification, running a dynamic risk assessment algorithm and outputting an early warning level; the prompt unit is used for displaying prompt information on the instrument panel display module and the vehicle-mounted terminal; by means of the scheme, the problems that when a static vehicle or a low-speed obstacle is detected by an existing system, due to the conservative property of TTC calculation, the time window for triggering early warning is later, collision is difficult to avoid when the vehicle is close to the static vehicle, and potential safety hazards exist are solved.

Description

Technical Field

[0001] The present application relates to the field of driving prediction, and in particular to a forward collision warning device, method, electronic device, and storage medium based on risk assessment. Background Art

[0002] Currently, forward collision warning (FCW) and automatic emergency braking (AEB) technologies in the industry primarily rely on "time to collision (TTC)" or "enhanced time to collision (ETTC)" as trigger thresholds. Warnings or braking are initiated only when the system detects a forward obstacle with a TTC lower than the set value. However, this logic has the following shortcomings:

[0003] Special characteristics of commercial vehicles: Commercial vehicles (such as trucks and buses) have large mass and high inertia, resulting in long emergency braking distances, especially on slippery roads. Furthermore, if the driver is distracted and fails to respond to regular warnings in a timely manner, the system may not be able to reach sufficient distance to actually brake. When responding to stationary / low-speed obstacle risks, existing systems trigger warnings late due to the conservative nature of TTC calculations when detecting stationary vehicles or low-speed obstacles. This makes it difficult to avoid collisions when approaching stationary vehicles, posing a safety hazard. Summary of the Invention

[0004] In view of this, the purpose of the present invention is to provide a forward collision warning device, method, electronic device and storage medium based on risk assessment, so as to achieve early risk warning for stationary / low-speed targets, reserve more sufficient reaction time for the driver, and reduce the risk of collision.

[0005] The present invention provides the following solutions:

[0006] According to one aspect of the present invention, a forward collision warning device based on risk assessment is provided, comprising: a sensing unit for acquiring relative position data of a detection object and modeling the relative position data of the detection object; detecting the driving state of the driver; and detecting the motion state of the vehicle in real time;

[0007] The processing unit is used to process the perception unit data in real time and perform target classification, run the dynamic risk assessment algorithm, and output the warning level;

[0008] The prompt unit is used to display prompt information on the instrument panel display module and the vehicle terminal; and judge the driver's attention status.

[0009] Further, including:

[0010] The perception unit includes: millimeter-wave radar, visual camera, lidar, driver status detection module, and vehicle speed and acceleration detection module;

[0011] Millimeter-wave radar is used to obtain the distance, speed, and angle information of an object by emitting electromagnetic waves in the millimeter wave band and measuring the time and frequency changes of the reflected waves.

[0012] A visual camera, configured to detect objects using a target detection neural network model, classify the detected objects, and determine their distance from the ego vehicle;

[0013] LiDAR, used to detect the target position and shape of the object and measure its speed, and model it in complex environments;

[0014] The driver status detection module is used to detect the driver's characteristic information through the camera, calculate the driver's driving status, and send the driver's driving status to the CAN bus;

[0015] The vehicle speed and acceleration detection module is used to detect the vehicle's speed (Vs), acceleration (as) and other motion states in real time through pre-installed wheel speed sensors and acceleration sensors, and transmit the data to the processing unit.

[0016] Further, including:

[0017] The driver status detection module uses a camera to monitor the driver's eye shape and head rotation angle to calculate whether the driver is fatigued or distracted, and sends the driver's status to the CAN bus.

[0018] Further, including:

[0019] The processing unit includes:

[0020] The risk level S is quantified by integrating multiple factors including the vehicle's speed, the target vehicle's speed, the time to collision (TTC), and the driver's distraction. The formula is:

[0021]

[0022] in:

[0023] ( W1, W2, W3 ) are weight coefficients;

[0024] Vs is the speed of the ego vehicle, and Vt is the speed of the target vehicle;

[0025] D is the driver distraction index, ranging from 0 to 1, determined by the fatigue monitoring system;

[0026] Among them, TTC is the time to collision, and its calculation formula is:

[0027]

[0028] in: is the acceleration of the target vehicle (m / is the acceleration (m / s2) of the ego vehicle; is the distance (m) between the ego vehicle and the target vehicle.

[0029] Further comprising:

[0030] obtaining a preset risk level threshold value:

[0031] when S≥ : triggering a first level of early warning; S1 is a first level of early warning threshold value;

[0032] when S≥ : triggering a second level of early warning; S2 is a second level of early warning threshold value.

[0033] Further comprising:

[0034] The processing unit adopts Kalman filtering algorithm to fuse the relative position data of the detected object in the sensing unit, so as to eliminate the measurement error of the multi-source sensor data.

[0035] Further comprising:

[0036] The instrument panel display module of the prompting unit supports hierarchical color warning and text prompt.

[0037] According to the two aspects of the present application, a forward collision warning method based on risk assessment is provided, comprising the following steps:

[0038] The position, speed, acceleration and other information of the target vehicle in front are detected by the sensing unit. The Kalman filtering method is adopted to fuse the sensing results of the multi-source sensors through the computing platform, and the final fused result is output;

[0039] The real-time speed and real-time acceleration of the ego vehicle are detected by the wheel speed sensor and the acceleration sensor equipped on the ego vehicle, and the detection results are input to the processing unit through the CAN bus;

[0040] The real-time state of the driver is detected by the driver state detection module, and the detection result is input to the processing unit through the CAN bus.

[0041] The dynamic risk assessment algorithm is run in the processing unit, and the risk level S is calculated based on the speed, acceleration, distance between vehicles and driver state parameters of the ego vehicle and the target vehicle.

[0042] The risk level S is compared with the preset risk threshold value;

[0043] If the risk assessment result S≥ , the first level of early warning is executed through the prompting unit;

[0044] ​​If the risk assessment result is S≥ , then a secondary warning is executed through the prompt unit. According to three aspects of the present invention, an electronic device is provided, comprising: a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory communicate with each other via the communication bus;

[0045] A computer program is stored in the memory. When the computer program is executed by the processor, the processor is caused to execute steps of a forward collision warning method based on risk assessment.

[0046] According to four aspects of the present invention, a computer-readable storage medium is provided, which stores a computer program executable by an electronic device. When the computer program runs on the electronic device, the electronic device executes the steps of a forward collision warning method based on risk assessment.

[0047] Through the above solution, the following beneficial technical effects are achieved:

[0048] This application uses an early warning mechanism to trigger early warning for stationary / low-speed targets in commercial vehicle scenarios, significantly extending the driver's reaction time.

[0049] This application uses multi-sensor fusion, combining radar, vision and lidar (optional), to improve the reliability of forward collision warning risk assessment in complex weather conditions (such as rain and fog).

[0050] This application uses driver status linkage to dynamically adjust warning strategies by detecting distracted behavior to adapt to the special needs of commercial vehicle driving.

[0051] This application uses graded warnings to avoid the driver's trust crisis caused by collision warnings in dangerous scenarios that are difficult to experience. Through the risk assessment model, the driver is reminded of the presence of low-speed vehicles ahead with a significant sense of experience, while ensuring the effectiveness of key scenarios.

[0052] This application uses millimeter-wave radar, visual camera (including target detection algorithm) and optional lidar and computing platform to detect the displacement, speed, acceleration and other parameters of the vehicle in front in real time; and adopts multi-sensor data fusion technology to improve the accuracy of target detection and distance monitoring range. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] Figure 1 1 is a structural diagram of a forward collision warning device based on risk assessment provided by one or more embodiments of the present invention.

[0054] Figure 2 This is a flowchart of a forward collision warning method based on risk assessment provided by one or more embodiments of the present invention.

[0055] Figure 3 It is a structural diagram of a computing platform of a specific embodiment of the present invention.

[0056] Figure 4 It is a schematic diagram of a fusion target according to a specific embodiment of the present invention.

[0057] Figure 5 It is a risk assessment flow chart of a specific embodiment of the present invention.

[0058] Figure 6 This is a structural block diagram of an electronic device according to a forward collision warning method based on risk assessment provided by one or more embodiments of the present invention. DETAILED DESCRIPTION

[0059] The technical solution of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the embodiments described are only some embodiments of the present invention, not all embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0060] Figure 1 1 is a structural diagram of a forward collision warning device based on risk assessment provided by one or more embodiments of the present invention.

[0061] like Figure 1 The apparatus shown comprises:

[0062] The sensing unit is used to obtain the relative position data of the detected object, build a model based on the relative position data of the detected object, detect the driver's driving state, and detect the movement state of the vehicle in real time;

[0063] The processing unit is used to process the perception unit data in real time and perform target classification, run the dynamic risk assessment algorithm, and output the warning level;

[0064] The prompt unit is used to display prompt information on the instrument panel display module and the vehicle terminal; and judge the driver's attention status.

[0065] Specifically, it solves the problems of insufficient detection accuracy and low data reliability of traditional single sensors (such as only visual cameras or millimeter-wave radars) in complex environments (such as bad weather, lighting changes, and obstacle obstructions), and multi-sensor data is prone to conflicts due to measurement errors and is difficult to effectively integrate.

[0066] The perception unit integrates multiple sensors such as millimeter-wave radar (excellent at ranging and speed measurement in adverse weather conditions), visual cameras (excellent at target classification), and lidar (high-precision modeling range), and combines it with the Kalman filter algorithm of the processing unit to eliminate measurement errors, thereby solving the detection limitations of a single sensor in complex environments. It enables accurate acquisition of data such as object distance, speed, shape, and type, and is adaptable to complex scenarios such as rainy days, strong light, and occlusion.

[0067] Among them, existing technologies have low accuracy in identifying dangerous conditions such as driver fatigue and distraction (such as judging only by a single feature), and are unable to capture key features such as the driver's eye shape and head rotation in real time, resulting in difficulty in early warning of accident risks caused by human factors.

[0068] The driver status detection module monitors eye morphology (such as blinking frequency) and head rotation angle to accurately determine fatigue or distraction status and transmits the information to the system in real time, solving the problem of "delayed discovery" of accident hazards caused by human factors and providing key human factor input for risk assessment.

[0069] Traditional risk assessment relies solely on a single parameter (such as vehicle-to-vehicle distance) and fails to comprehensively consider dynamic factors such as vehicle speed, relative acceleration, and driver status. This leads to inaccurate risk judgment (for example, the risk difference between high-speed and low-speed driving at the same distance is ignored) and a lack of quantitative assessment standards.

[0070] The processing unit's dynamic risk assessment algorithm integrates multiple parameters, including vehicle speed (Vs), target vehicle speed (Vt), time to collision (TTC), and driver distraction index (D), to quantify the risk level (S) through a formula, overcoming the one-sidedness of traditional single-factor assessments. Specifically, TTC incorporates relative acceleration and vehicle-to-vehicle distance calculations to better reflect the dynamic risk changes in real-world driving scenarios, making risk assessment more scientific.

[0071] Existing warning systems are often simple (e.g., just an audible alarm) with no clear risk level. This prevents drivers from quickly assessing the severity of an incident, leading to accidents caused by delayed response or misjudgment. Furthermore, risk warnings lack clear thresholds, leading to false and missed alerts.

[0072] The prompt unit uses "graded color warning + text prompt" combined with clear thresholds ( =0.5 is level one, =0.8 for Level 2), so that warning information can intuitively distinguish the degree of urgency: Level 1 warning indicates potential risks, and Level 2 warning indicates high risks, helping drivers quickly judge and take corresponding measures (such as slowing down and avoiding) to reduce reaction delays.

[0073] Further, including:

[0074] The perception unit includes: millimeter-wave radar, visual camera, lidar, driver status detection module, and vehicle speed and acceleration detection module;

[0075] Millimeter-wave radar is used to obtain the distance, speed, and angle information of an object by emitting electromagnetic waves in the millimeter wave band and measuring the time and frequency changes of the reflected waves.

[0076] A visual camera, configured to detect objects using a target detection neural network model, classify the detected objects, and determine their distance from the ego vehicle;

[0077] LiDAR, used to detect the target position and shape of the object and measure its speed, and model it in complex environments;

[0078] The driver status detection module is used to detect the driver's characteristic information through the camera, calculate the driver's driving status, and send the driver's driving status to the CAN bus;

[0079] The vehicle speed and acceleration detection module is used to detect the vehicle's speed (Vs), acceleration (as) and other motion states in real time through pre-installed wheel speed sensors and acceleration sensors, and transmit the data to the processing unit.

[0080] Further, including:

[0081] The driver status detection module uses a camera to monitor the driver's eye shape and head rotation angle to calculate whether the driver is fatigued or distracted, and sends the driver's status to the CAN bus.

[0082] Further, including:

[0083] The processing unit includes:

[0084] The risk level S is quantified by integrating multiple factors including the vehicle's speed, the target vehicle's speed, the time to collision (TTC), and the driver's distraction. The formula is:

[0085] ;

[0086] in:

[0087] ( W1, W2, W3 ) are weight coefficients;

[0088] Vs is the speed of the ego vehicle, and Vt is the speed of the target vehicle;

[0089] D is the driver distraction index, ranging from 0 to 1, determined by the fatigue monitoring system;

[0090] Among them, TTC is the time to collision, and its calculation formula is:

[0091]

[0092] in: is the acceleration of the target vehicle (m / ); is the acceleration of the vehicle (m / ); is the vehicle distance (m).

[0093] Further, including:

[0094] Risk level threshold:

[0095] ( S≥ ): triggering a level 1 warning; is the first-level warning threshold (e.g. S1=0.5);

[0096] (S≥ ): trigger the second level warning; It is the second-level warning threshold (such as S2=0.8).

[0097] Further, including:

[0098] The processing unit adopts the Kalman filter algorithm to perform fusion processing on the relative position data of the detected object in the perception unit to eliminate the measurement error of the multi-source sensor data.

[0099] Further, including:

[0100] The instrument panel display module of the prompt unit supports graded color warnings and text prompts.

[0101] Furthermore, it includes: the millimeter wave radar operates in the 77GHz frequency band, has the ability to resist electromagnetic interference, can detect at least 16 targets in front at the same time, and can distinguish between stationary and moving targets.

[0102] Furthermore, it includes: the visual camera uses a 12-megapixel high-definition lens, is equipped with an infrared fill light module, supports target detection in night and low-light environments, and the image acquisition frame rate is not less than 30fps.

[0103] Furthermore, it includes: the point cloud density of the laser radar is not less than 200 points / ㎡, the detection distance range is 0.5-200 meters, the angular resolution is ≤0.1°, and it can maintain an effective detection rate of more than 80% in complex weather conditions such as rain, snow, and heavy fog.

[0104] Furthermore, the weight coefficients (W1, W2, W3) have a value range of W1∈[0.2,0.4], W2∈[0.4,0.6], W3∈[0.1,0.3], and satisfy W1+W2+W3=1. They can be dynamically adjusted according to the vehicle speed. When the vehicle speed is ≥80km / h, the weight of W2 increases by 5%-10%.

[0105] Furthermore, it includes: the dashboard display of the first-level warning is a green light flashing twice per second, and the text prompt "There is a vehicle within 500 meters ahead" is displayed simultaneously; the dashboard display of the second-level warning is a red light flashing 4 times per second, and the text prompt "There is a low-speed vehicle within 200 meters ahead, please slow down" is displayed simultaneously, and the voice reminder volume of the vehicle terminal is 30% higher than the normal playback volume, and is repeated twice.

[0106] Furthermore, it also includes: when using the Kalman filter method to fuse multi-source sensor data, data is updated every 10ms, and abnormal data is eliminated using the 3σ criterion during the fusion process to ensure the stability of data fusion.

[0107] Furthermore, it also includes: the value of the first-level warning threshold S1 is 0.5±0.05, the value of the second-level warning threshold S2 is 0.8±0.05, and can be customized within the range of ±0.1 through the user setting interface of the vehicle terminal.

[0108] Furthermore, when the driver state detection module calculates the driver distraction index D, if the driver does not look at the road ahead for 3 consecutive seconds or turns his head more than 45 degrees, the D value increases by 0.2; if the driver blinks less than 5 times per minute, it is determined to be a fatigue state, and the D value is directly taken as 0.8.

[0109] Figure 2 This is a flow chart of a forward collision warning method based on risk assessment provided by one or more embodiments of the present invention. Figure 2 As shown, the following steps are included:

[0110] In step S1, the sensing unit detects the position, speed, acceleration, and other information of the target vehicle ahead. Using the Kalman filter method, the computing platform fuses the sensing results of multiple sensors and outputs the final fused result.

[0111] In step S2, the real-time speed and acceleration of the vehicle are detected by the wheel speed sensor and acceleration sensor equipped on the vehicle, and the detection results are input to the processing unit through the CAN bus.

[0112] Step S3, detecting the real-time status of the driver through the driver status detection module, and inputting the detection result to the processing unit through the CAN bus;

[0113] Step S4: Running a dynamic risk assessment algorithm in the processing unit to calculate the risk level based on the speed, acceleration, and vehicle distance of the ego vehicle and the target vehicle, as well as the driver's state parameters, to generate a risk level S;

[0114] Step S5, comparing the risk level S with a preset risk threshold;

[0115] If the risk assessment result S≥ , then the prompt unit will execute the first-level warning;

[0116] If the risk assessment result S≥ , then the second-level warning is executed through the prompt unit.

[0117] Specifically, multi-source sensor fusion (combined with Kalman filtering) effectively eliminates single-sensor measurement errors, making the target vehicle's position, velocity, acceleration, and other data more stable and reliable. Whether in sunny or rainy conditions, or in complex road conditions (such as curves and obstructions), the system outputs consistent and accurate perception results, laying a high-quality data foundation for subsequent risk assessments.

[0118] Dedicated sensors are used to directly detect the dynamic parameters of the ego vehicle, and the high real-time performance of the CAN bus (millisecond-level transmission) is used to ensure that the data is delivered to the processing unit without delay. Step S3 simultaneously captures the driver's status (such as distraction and fatigue) and transmits it to the system, realizing real-time linkage of the three-factor data of "ego vehicle-target vehicle-driver", avoiding risk misjudgment due to missing or delayed data.

[0119] The dynamic risk assessment algorithm integrates the speed, acceleration, and inter-vehicle distance of the ego and target vehicles, as well as driver status parameters, to convert risk into a quantifiable level (S) through a formula, overcoming the limitations of traditional "qualitative judgment." For example, at the same inter-vehicle distance, the risk level is precisely differentiated between high- and low-speed driving, and between distracted and focused drivers, making the assessment more accurate and accurate to the actual level of danger.

[0120] In another embodiment,

[0121] Through multi-source data fusion and real-time synchronization, by setting:

[0122] Roadside Perception Layer: Using V2X roadside units (RSUs), devices like lidar, millimeter-wave radar, and cameras collect real-time information on traffic events (such as accidents and construction), road conditions (such as slippery and icy), and signal phases. For example, Huawei's road network digitalization service covers 306 V2X events, including traffic accidents and severe weather conditions, and delivers these information to vehicles in real time via the 5G network.

[0123] On-board terminal layer: The PCC system receives V2X data through the onboard communication unit (OBU) and integrates it with high-precision maps and vehicle dynamics models. For example, Hirain's OBU supports dual-interface communication, PC5 and Uu, and can receive real-time target detection results from the roadside perception unit (FPU), extending the vehicle's perception range to 500 meters.

[0124] Cloud Platform Layer: Data preprocessing is performed through edge computing (MEC) nodes, dynamically matching V2X events with PCC preview paths. After MEC processing of roadside equipment data, end-to-end latency can be compressed to less than 50ms, meeting the requirements of emergency braking coordination.

[0125] The C-V2X communication architecture utilizes the 3GPP-defined dual modes of PC5 (direct communication) and Uu (cellular communication). The PC5 interface is used for short-range vehicle-to-vehicle and vehicle-to-road interaction (latency <10ms), while the Uu interface enables wide-area data transmission (coverage radius >10km) via the 5G network. For example, Tunnel Group's expressway V2X application uses the PC5 interface to transmit real-time information about abnormally stopped vehicles, providing a 1km advance warning to vehicles behind.

[0126] Dynamic traffic incident response

[0127] Construction area warning: When the roadside camera identifies construction ahead, the RSU broadcasts a BSM message through the PC5 interface. The PCC system adjusts the vehicle speed 2 kilometers in advance (for example, from 90km / h to 60km / h) and automatically switches to a low gear to use engine braking.

[0128] Emergency vehicle priority: By obtaining the ambulance's location and driving path through V2I communication, the PCC system proactively plans an avoidance route, such as guiding vehicles to change lanes in advance on congested roads to ensure a clear rescue channel.

[0129] When the roadside millimeter-wave radar detects an oncoming vehicle, it sends a warning message via V2X. The PCC system, combined with the curve curvature data of the high-precision map, triggers a steering wheel vibration prompt 150 meters in advance and automatically limits the vehicle speed to a safe cornering speed (such as 40km / h).

[0130] By integrating meteorological sensor data and real-time weather warnings from V2X, the PCC system dynamically adjusts the braking deceleration threshold (for example, from -8m / s² to -6m / s²) and enhances the detection capability of stationary vehicles through lidar (extending the distance to 250 meters).

[0131] Multi-vehicle data sharing is achieved through V2V communication. When the leading vehicle suddenly brakes, the PCC system of the following vehicle can receive the braking signal 0.8 seconds in advance and trigger automatic deceleration, shortening the following distance of the formation to within 5 meters and reducing wind resistance by 15%-20%.

[0132] In high-density traffic scenarios, the V2X system avoids communication conflicts through resource reservation coordination algorithms (such as the CRR protocol), ensuring that the status information transmission delay between platoon vehicles is stable within 50ms.

[0133] Figure 3 It is a structural diagram of a computing platform of a specific embodiment of the present invention.

[0134] like Figure 3 As shown,

[0135] A forward collision warning device based on risk assessment and system components are provided, including:

[0136] 1. Perception Unit

[0137] Millimeter-wave radar (which transmits millimeter-wave electromagnetic waves and measures the time and frequency changes of reflected waves to obtain the distance, speed, and angle information of objects. It is less affected by weather and has good penetration in rain, snow, fog, and dust);

[0138] Vision camera (including a target detection neural network model that can automatically detect objects, classify them, and determine their distance, supporting day and night operation);

[0139] Optional LiDAR (a technology that uses lasers to detect target position and shape and measure speed, for high-precision modeling in complex environments).

[0140] The driver status detection module uses a camera to monitor the driver's eye shape and head rotation angle to calculate whether the driver is in a state of fatigue or distraction, and sends the driver's status to the CAN bus.

[0141] The vehicle speed and acceleration detection module detects the vehicle's motion status in real time, including its speed and acceleration, through pre-installed wheel speed sensors and acceleration sensors.

[0142] 2. Processing Unit

[0143] Computing platform (processes multi-sensor data in real time and performs target classification, runs dynamic risk assessment algorithms, and outputs warning levels; integrates with vehicle networks (such as the CAN bus) to control warning devices and braking systems)

[0144] 3. Prompt unit

[0145] Instrument panel display module: supports graded color warnings and text prompts;

[0146] Vehicle terminal: Voice prompts such as "low-speed vehicle ahead" and other warning messages;

[0147] Driver status monitoring module: monitors driving behavior through cameras and determines the driver's attention status.

[0148] Figure 4 It is a schematic diagram of a fusion target according to a specific embodiment of the present invention.

[0149] Through millimeter-wave radar, visual camera (including target detection algorithm) and optional lidar and computing platform, the displacement, speed, acceleration and other parameters of the vehicle in front are detected in real time; multi-sensor data fusion technology is used to improve the accuracy of target detection and distance monitoring range.

[0150] Figure 5 It is a risk assessment flow chart of a specific embodiment of the present invention.

[0151] like Figure 5 The process shown:

[0152] Millimeter-wave radar, visual cameras, and optional lidar sensors detect the position, speed, acceleration, and other information of the target vehicle ahead. Using Kalman filtering, the computing platform fuses the perception results of multiple sensors and outputs the final fused result, improving the accuracy of target perception.

[0153] The real-time speed and acceleration of the vehicle are detected by the wheel speed sensor and acceleration sensor equipped on the vehicle, and the detection results are input to the computing platform module through the CAN bus.

[0154] The driver's real-time status is detected by the driver fatigue detection module, and the detection results are input to the computing platform module through the CAN bus.

[0155] The dynamic risk assessment model is deployed in the computing platform. The risk assessment model calculates the risk level based on the speed, acceleration, vehicle distance and driver status parameters of the ego vehicle and the target vehicle.

[0156] Compare the calculated risk level S with the preset risk threshold. When the risk assessment result S≥ When the vehicle is approaching, the instrument panel light (such as flashing green) will prompt "vehicle ahead" and provide a first-level warning;

[0157] When the risk assessment result S≥ When the system is in operation, the instrument light prompts and the vehicle terminal adds a voice reminder to remind the driver that there is a "low-speed vehicle" ahead, issuing a secondary warning.

[0158] Figure 6 This is a structural block diagram of an electronic device according to a forward collision warning method based on risk assessment provided by one or more embodiments of the present invention.

[0159] like Figure 6 As shown, the present application provides an electronic device, comprising: a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory communicate with each other via the communication bus;

[0160] A computer program is stored in the memory. When the computer program is executed by the processor, the processor is caused to perform steps of a forward collision warning method based on risk assessment.

[0161] The present application also provides a computer-readable storage medium storing a computer program executable by an electronic device. When the computer program runs on the electronic device, the electronic device executes the steps of a forward collision warning method based on risk assessment.

[0162] For simplicity of description, the method embodiments are described as a series of actions. However, those skilled in the art should be aware that the embodiments of the present invention are not limited by the order of the actions described, because certain steps can be performed in other orders or simultaneously according to the embodiments of the present invention. Secondly, those skilled in the art should also be aware that the embodiments described in this specification are all preferred embodiments, and the actions involved are not necessarily required by the embodiments of the present invention.

[0163] Through the description of the above embodiments, it can be seen that those skilled in the art can clearly understand that the present application can be implemented by means of software plus the necessary general hardware platform. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a storage medium such as ROM / RAM, a magnetic disk, an optical disk, etc., and includes a number of instructions for enabling a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in various embodiments of the present application or certain parts of the embodiments.

[0164] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the above embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A forward collision warning device based on risk assessment, characterized in that: include: A sensing unit is used to obtain relative position data of the detected object and to build a model based on the relative position data of the detected object; Detecting the driver's driving status; Real-time detection of the vehicle's motion status; The processing unit is used to process the perception unit data in real time and perform target classification, run the dynamic risk assessment algorithm, and output the warning level; The prompt unit is used to display prompt information on the instrument panel display module and the vehicle terminal; and judge the driver's attention status.

2. The forward collision warning device based on risk assessment according to claim 1, characterized in that: The perception unit includes: millimeter wave radar, visual camera, laser radar, driver status detection module and vehicle speed and acceleration detection module; Millimeter-wave radar is used to obtain the distance, speed, and angle information of an object by emitting electromagnetic waves in the millimeter wave band and measuring the time and frequency changes of the reflected waves. A visual camera, configured to detect objects using a target detection neural network model, classify the detected objects, and determine their distance from the ego vehicle; LiDAR, used to detect the target position and shape of the object and measure its speed, and model it in complex environments; The driver status detection module is used to detect the driver's characteristic information through the camera, calculate the driver's driving status, and send the driver's driving status to the CAN bus; The vehicle speed and acceleration detection module is used to detect the vehicle's speed (Vs), acceleration (as) and other motion states in real time through pre-installed wheel speed sensors and acceleration sensors, and transmit the data to the processing unit.

3. The forward collision warning device based on risk assessment according to claim 2, characterized in that: The driver status detection module monitors the driver's eye shape and the driver's head rotation angle through a camera to calculate whether the driver is in a fatigue or distracted state, and sends the driver's status to the CAN bus.

4. The forward collision warning device based on risk assessment according to claim 1, characterized in that: The processing unit includes: The risk level S is quantified by integrating multiple factors including the vehicle's speed, the target vehicle's speed, the time to collision (TTC), and the driver's distraction. The formula is: ； in: ( W1, W2, W3 ) are weight coefficients; Vs is the speed of the ego vehicle, and Vt is the speed of the target vehicle; D is the driver distraction index, ranging from 0 to 1, determined by the fatigue monitoring system; Among them, TTC is the time to collision, and its calculation formula is: ； in: is the acceleration of the target vehicle (m / ); is the acceleration of the vehicle (m / ); is the vehicle distance (m).

5. The forward collision warning device based on risk assessment according to claim 4, characterized in that: Obtain the preset risk level threshold; When S≥ : Triggering the first level warning; S1 is the first level warning threshold; When S≥ : Triggering the second level warning; S2 is the second-level warning threshold.

6. The forward collision warning device based on risk assessment according to claim 5, characterized in that: The processing unit uses a Kalman filter algorithm to perform fusion processing on the relative position data of the detected object in the perception unit to eliminate measurement errors of multi-source sensor data.

7. A forward collision warning device based on risk assessment according to claim 1, characterized in that: The instrument panel display module of the prompt unit supports graded color warnings and text prompts.

8. A forward collision warning method based on risk assessment, characterized in that: The sensing unit detects the position, speed, acceleration, and other information of the target vehicle in front. Using the Kalman filter method, the computing platform fuses the perception results of multiple sensors and outputs the final fusion result. The real-time speed and acceleration of the vehicle are detected by the wheel speed sensor and acceleration sensor equipped on the vehicle, and the detection results are input to the processing unit through the CAN bus; The driver's real-time status is detected by the driver status detection module, and the detection result is input to the processing unit through the CAN bus; The dynamic risk assessment algorithm is run in the processing unit to calculate the risk level based on the speed, acceleration, and distance between the ego vehicle and the target vehicle, as well as the driver's state parameters, to generate a risk level S. Compare the risk level S with the preset risk threshold; If the risk assessment result S≥ , then the prompt unit will execute the first-level warning; If the risk assessment result is S≥ , then the second-level warning is executed through the prompt unit.

9. An electronic device comprising: A processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory communicate with each other via the communication bus; A computer program is stored in the memory. When the computer program is executed by the processor, the processor executes the steps of a forward collision warning method based on risk assessment as claimed in claim 8.

10. A computer-readable storage medium storing a computer program executable by an electronic device, wherein when the computer program is executed on the electronic device, the electronic device executes the steps of a forward collision warning method based on risk assessment as claimed in claim 8.

Images