Driving status recognition methods, devices, electronic equipment and vehicles

Abstract

This invention provides a driving state recognition method, device, electronic device, and vehicle. The method includes: acquiring different driving parameters of a target vehicle within a preset distance range from the current vehicle using different types of sensors; determining initial judgment values ​​for the driving state of the target vehicle by different types of sensors based on the different driving parameters; acquiring the current weather type and the real-time confidence levels of different types of sensors; determining the target weather environment attenuation factor value corresponding to the current weather type; determining the allocation weights corresponding to different types of sensors based on the real-time confidence level and the target weather environment attenuation factor value; determining whether the target vehicle is in an intelligent driving state based on the initial judgment values ​​and allocation weights of different types of sensors; and if it is in an intelligent driving state, sending a warning reminder to reduce safety hazards.

Description

Technical Field

[0001] This invention relates to the field of vehicle technology, and in particular to a driving state recognition method, device, electronic device, and vehicle. Background Technology

[0002] Currently, most vehicles rely on Vehicle to Everything (V2X) wireless communication technology to determine the driving status of other vehicles, and then use this information to remind the driver, reducing vehicle safety hazards.

[0003] However, because V2X terminals are not yet mandatory standard equipment on vehicles, the proportion of vehicles with actual communication capabilities on the road is low. Therefore, vehicles without V2X terminals cannot communicate with those equipped with them. This means vehicles without V2X terminals cannot obtain the driving status of other vehicles, and vehicles equipped with V2X terminals are completely unaware of the presence of vehicles without them when they enter the detection range. Consequently, in such situations, vehicles cannot alert drivers, creating safety hazards. Furthermore, existing V2X broadcast protocols typically transmit only basic fields such as location, speed, and heading regarding driving status. This shallow data, in adverse weather conditions or when sensors are interfered with, may lead to misjudgments of other vehicles' true driving status, resulting in delayed or false warnings, further exacerbating safety risks. Summary of the Invention

[0004] In view of this, the present invention aims to propose a driving state recognition method, device, electronic device, and vehicle to solve the problem that vehicles not equipped with V2X terminals cannot communicate with vehicles equipped with V2X terminals, which brings certain safety hazards. Furthermore, the V2X broadcast protocol may misjudge the true driving state of other vehicles in adverse weather conditions or when sensors are interfered with, leading to delayed or false warnings. The specific technical solution is as follows:

[0005] According to a first aspect of the present invention, a driving state recognition method is provided, the method comprising:

[0006] Different driving parameters of target vehicles within a preset distance range are acquired using different types of sensors;

[0007] Different types of sensors are used to determine the initial values ​​for the target vehicle's driving state based on different driving parameters.

[0008] Obtain the current weather type and the real-time confidence level of the sensors for different types;

[0009] Determine the value of the target weather environment attenuation factor corresponding to the current weather type;

[0010] The allocation weights for different types of sensors are determined by the real-time confidence level and the target weather environment attenuation factor.

[0011] By assigning initial values ​​and weights to different types of sensors, it is determined whether the target vehicle is in an intelligent driving state.

[0012] If the vehicle is in intelligent driving mode, a warning reminder will be sent.

[0013] Optionally, the different types of sensors include cameras, millimeter-wave radar, and lidar, and the driving parameters include the lateral offset distance between the center point of the target vehicle and the center line of the lane where the target vehicle is located, the longitudinal speed of the vehicle, and the lateral displacement.

[0014] The acquisition of different driving parameters of a target vehicle within a preset distance range from the current vehicle using different types of sensors includes:

[0015] The camera is used to obtain the lateral offset distance between the center point of the target vehicle and the center line of the lane in which the target vehicle is located;

[0016] The longitudinal velocity of the target vehicle is obtained using millimeter-wave radar;

[0017] The lateral displacement of the target vehicle is obtained using lidar.

[0018] Optionally, the step of obtaining the lateral offset distance between the center point of the target vehicle and the lane centerline of the lane where the target vehicle is located via the camera includes:

[0019] The camera identifies the actual lane the vehicle is currently in and the lateral center point of that lane.

[0020] Get the center point of the current vehicle;

[0021] A coordinate system is established based on the current vehicle's center point and the lateral center point, wherein the X-axis of the coordinate system is determined based on the current vehicle's center point, and the Y-axis of the coordinate system is determined based on the lateral center point.

[0022] The target vehicle's position coordinates in the coordinate system are obtained using a camera;

[0023] Substitute the location coordinates into the preset lane centerline equation to determine the lateral offset distance between the center point of the target vehicle and the lane centerline of the lane where the target vehicle is located. Different lanes correspond to different lane centerline equations.

[0024] Optionally, the initial determination of the target vehicle's driving state by the millimeter-wave radar includes a first initial determination value and a second initial determination value.

[0025] The initial determination of the target vehicle's driving state by different types of sensors based on different driving parameters includes:

[0026] If the lateral offset distance is less than or equal to a preset distance threshold within a first preset time period, then the initial determination of the camera's driving state of the target vehicle is assigned to the target calibration value.

[0027] If it is determined that the target vehicle has not entered the new road, and if the change in the longitudinal speed of the vehicle is less than the first speed change threshold, then the first initial determination of the driving state of the target vehicle by the millimeter-wave radar is assigned as the target calibration value.

[0028] If the target vehicle enters a new road or interacts with other traffic participants, and the change in the vehicle's longitudinal speed is greater than or equal to a second speed change threshold within a second preset time period, then the second initial determination of the target vehicle's driving state by the millimeter-wave radar is assigned as the target calibration value.

[0029] Obtain the standard deviation of the lateral displacement within a third preset time period;

[0030] If the standard deviation is less than the standard deviation threshold, then the initial determination of the driving state of the target vehicle by the lidar is assigned to the target calibration value.

[0031] Optionally, before obtaining the real-time confidence levels of the sensors for the current weather type and different types, the method further includes:

[0032] Acquire image frames captured by the camera, millimeter-wave radar data, and lidar data;

[0033] Obtain the average brightness, image sharpness, and historical trajectory consistency of the image frame;

[0034] Obtain the signal-to-noise ratio and velocity-time consistency of the millimeter-wave radar data;

[0035] Obtain the point cloud average intensity, reflection intensity consistency, and geometric fitting residual inverse index of the lidar data;

[0036] The average brightness of the image, the image sharpness, and the consistency of the historical trajectory are substituted into the first preset linear relationship algorithm to generate the real-time confidence of the camera;

[0037] The signal-to-noise ratio and velocity-time consistency are substituted into the second preset linear relationship algorithm to generate the real-time confidence level of the millimeter-wave radar;

[0038] The average intensity of the point cloud, the consistency of reflection intensity, and the inverse index of geometric fitting residual are substituted into the third preset linear relationship algorithm to generate the real-time confidence level of the lidar.

[0039] Optionally, the weighting of the different types of sensors is obtained by the following formula:

[0040]

[0041] in, This refers to the assigned weight of the i-th sensor. This refers to the real-time confidence level of the i-th sensor. This refers to the weather and environmental attenuation factor value, where n is the total number of sensors. This refers to the real-time confidence level of the k-th sensor.

[0042] Optionally, the step of determining whether the target vehicle is in an intelligent driving state by initially assigning values ​​and weights to different types of sensors further includes:

[0043] The driving state determination value is obtained by weighting and summing the initial judgment values ​​of different types of sensors according to the weighting assignment.

[0044] If the driving state determination value is greater than the preset value, the target vehicle is determined to be in intelligent driving state;

[0045] If the driving status determination value is less than or equal to the preset value, the target vehicle is determined not to be in intelligent driving mode.

[0046] According to a second aspect of the present invention, a driving state recognition device is provided, the device comprising:

[0047] The first acquisition module is used to acquire different driving parameters of target vehicles that are within a preset distance range from the current vehicle through different types of sensors;

[0048] The first determining module is used to determine the initial judgment values ​​of different types of sensors on the driving state of the target vehicle based on different driving parameters;

[0049] The second acquisition module is used to acquire the current weather type and the real-time confidence level of the sensors for different types;

[0050] The second determining module is used to determine the value of the target weather environment attenuation factor corresponding to the current weather type;

[0051] The third determining module is used to determine the allocation weights corresponding to different types of sensors by using the real-time confidence level and the target weather environment attenuation factor value;

[0052] The judgment module is used to determine whether the target vehicle is in an intelligent driving state by assigning initial judgment values ​​and weights to different types of sensors.

[0053] The warning module is used to send warning reminders if the vehicle is in intelligent driving mode.

[0054] According to another aspect of the present invention, an electronic device is also provided, comprising:

[0055] processor;

[0056] Memory used to store the processor's executable instructions;

[0057] The processor is configured to execute the instructions to implement the driving state recognition method as described above.

[0058] According to another aspect of the present invention, a vehicle is also provided, including the above-described driving state recognition device.

[0059] The driving state recognition method provided by this invention acquires different driving parameters of a target vehicle within a preset distance range using different types of sensors. This facilitates a comprehensive and accurate reflection of the driving state based on these parameters. The method uses different driving parameters to determine the initial judgment values ​​of different types of sensors for the target vehicle's driving state, reducing the error of single-sensor judgments. It also acquires the real-time confidence levels of the current weather type and different types of sensors, using these real-time confidence levels to evaluate sensor reliability and avoid reliance on failed data. Furthermore, by acquiring the current weather type, an appropriate weather environment attenuation factor value can be set, thus determining the target weather environment attenuation factor value corresponding to the current weather type. By considering the impact of weather on sensor accuracy, the subsequent judgment results are more accurate. By using real-time confidence levels and target weather environment attenuation factors, the assigned weights for different types of sensors are determined. These weights, calculated based on real-time confidence levels and target weather environment attenuation factors, can automatically reduce the weights of low-confidence sensors in adverse weather conditions, improving system robustness and the accuracy of subsequent judgments. Through initial judgment assignments and weight allocations for different types of sensors, the system determines whether the target vehicle is in intelligent driving mode. If it is, a warning is sent. By comprehensively judging the driving status of the target vehicle using different types of sensors, "information blind spots" can be avoided. Furthermore, in adverse weather conditions or when sensors are interfered with, the system combines the recognition parameters of multiple types of sensors for judgment, improving system robustness and the accuracy of the judgment results, avoiding misjudgments of the true driving status of other vehicles, and sending warnings reasonably and promptly to reduce safety hazards.

[0060] The above description is merely an overview of the technical solution of the present invention. In order to better understand the technical means of the present invention and to implement it in accordance with the contents of the specification, and to make the above and other objects, features and advantages of the present invention more apparent and understandable, specific embodiments of the present invention are described below. Attached Figure Description

[0061] Various other advantages and benefits will become apparent to those skilled in the art upon reading the following detailed description of preferred embodiments. The accompanying drawings are for illustrative purposes only and are not intended to limit the invention. Furthermore, the same reference numerals denote the same parts throughout the drawings. In the drawings:

[0062] Figure 1 This is a flowchart of the steps of a driving state recognition method provided by the present invention;

[0063] Figure 2 yes Figure 1 The flowchart shown is a step 101 of a driving state recognition method provided by the present invention;

[0064] Figure 3 yes Figure 1 The diagram shown is a schematic representation of the distribution of the current vehicle and the target vehicle in the coordinate system in a driving state recognition method provided by the present invention.

[0065] Figure 4 yes Figure 1 The flowchart shown is a step 102 of a driving state recognition method provided by the present invention;

[0066] Figure 5 yes Figure 1 The flowchart shown is a step 106 of a driving state recognition method provided by the present invention;

[0067] Figure 6 This is a schematic diagram of the structure of a driving state recognition device provided by the present invention;

[0068] Figure 7 This is a schematic diagram of the structure of an electronic device provided by the present invention. Detailed Implementation

[0069] To make the objectives, technical solutions, and advantages of the present invention clearer, the various embodiments of the present invention will be described in detail below with reference to the accompanying drawings. However, those skilled in the art will understand that many technical details are presented in the various embodiments of the present invention to facilitate a better understanding of the invention. However, the technical solutions claimed in the present invention can be implemented even without these technical details and with various variations and modifications based on the following embodiments. The division of the various embodiments below is for ease of description and should not constitute any limitation on the specific implementation of the present invention. The various embodiments can be combined with and referenced by each other without contradiction.

[0070] Currently, the driving status of other vehicles is typically obtained through Vehicle-to-Everything (V2X) wireless communication technology broadcasting. However, V2X is not mandatory for all vehicles on the road. Therefore, vehicles equipped with V2X terminals cannot obtain the driving status of other vehicles without them, and consequently cannot provide alerts to their own drivers based on that status, posing a safety hazard. Furthermore, the V2X broadcast protocol may misjudge the true driving status of other vehicles in inclement weather or when sensors are interfered with, leading to delayed or false warnings. Based on these problems, this invention proposes a driving status recognition method. (Refer to...) Figure 1 The diagram illustrates a flowchart of a driving state recognition method provided by the present invention, the method comprising:

[0071] Step 101: Acquire different driving parameters of a target vehicle that is within a preset distance range from the current vehicle using different types of sensors.

[0072] The present invention utilizes various types of sensors, including cameras, millimeter-wave radar, and lidar. Driving parameters acquired through these sensors include the lateral offset distance between the target vehicle's center point and the lane centerline of its lane, the vehicle's longitudinal speed, and its lateral displacement. The target vehicle is another vehicle within a preset distance range from the current vehicle. This preset distance range can be set according to actual conditions, such as 100 meters, 150 meters, etc., and is not specifically limited herein. Since the present invention acquires different data through different types of sensors, step 101, as... Figure 2 As shown:

[0073] Step 1011: Obtain the lateral offset distance between the center point of the target vehicle and the center line of the lane where the target vehicle is located using a camera.

[0074] Step 1012: Obtain the longitudinal speed of the target vehicle using millimeter-wave radar.

[0075] Step 1013: Obtain the lateral displacement of the target vehicle using lidar.

[0076] Among these methods, cameras can compensate for the shortcomings of radar sensors in geometric feature recognition, thus enabling high-precision identification of lane lines and vehicle positions. Millimeter-wave radar has strong anti-interference capabilities (such as in rain, snow, fog, and haze), compensating for the delay or error of cameras in speed measurement and ensuring dynamic response in high-speed scenarios. Therefore, millimeter-wave radar can directly measure the radial velocity of target vehicles with high accuracy and good real-time performance. LiDAR provides high-resolution spatial measurement through point cloud data, compensating for the shortcomings of cameras in complex scenarios (such as obstruction and nighttime), and providing more stable lateral motion data. Therefore, LiDAR is used to accurately capture the lateral movement trajectory of vehicles. In summary, the above steps reduce the risk of single sensor failure and improve system robustness by acquiring data from multiple types of sensors.

[0077] In obtaining the lateral offset distance between the center point of a target vehicle and the center line of the lane in which the target vehicle is located, this invention first determines the actual lane and the lateral center point of the lane based on the image captured by the camera. Then, it determines the center point of the vehicle [the vehicle center point generally refers to the geometric center of the vehicle's projection on the ground (such as the midpoint of the wheelbase or the symmetrical point of the track width), usually located directly above the center of the rear axle or at the midpoint between the front and rear axles]. Next, it determines the vertical zero coordinate of the coordinate system based on the vehicle's center point, meaning the vehicle's center point is always on the X-axis of the coordinate system. It also determines the lateral zero coordinate of the coordinate system based on the lateral center point of the lane in which the vehicle is located, meaning the lateral center point of the lane in which the vehicle is located is always on the Y-axis of the coordinate system. After establishing the coordinate system, the position coordinates of the target vehicle in the coordinate system can be determined. Different lane center line equations can also be established for different lane lines. By substituting the position coordinates into the corresponding lane center line equation, the lateral offset distance between the center point of the target vehicle and the center line of the lane in which the target vehicle is located can be calculated. Based on the above description, the steps for calculating the lateral offset distance between the center point of the target vehicle and the center line of the lane in which the target vehicle is located include:

[0078] The camera identifies the vehicle's current lane and the lateral center point of that lane.

[0079] Get the center point of the current vehicle;

[0080] A coordinate system is established based on the current vehicle's center point and lateral center point. The X-axis of the coordinate system is determined based on the current vehicle's center point, and the Y-axis is determined based on the lateral center point.

[0081] The target vehicle's position coordinates in the coordinate system are obtained through a camera;

[0082] Substitute the position coordinates into the preset lane centerline equation to determine the lateral offset distance between the center point of the target vehicle and the lane centerline of the lane where the target vehicle is located. Different lanes correspond to different lane centerline equations.

[0083] The equation for the lane centerline is expressed as:

[0084]

[0085] in, , , and These are the fitting parameters for the lane centerline equation. Different lane centerline equations correspond to different fitting parameters. y is the longitudinal position coordinate of the target vehicle, and x is the lateral position coordinate of the lane where the target vehicle is located at the longitudinal position coordinate.

[0086] The lateral offset distance between the center point of the target vehicle and the center line of the lane in which the target vehicle is located can be calculated using the lane centerline equation described above. For example,... Figure 3 As shown, a coordinate system is established based on the current vehicle. The X-axis of the coordinate system is determined based on the center point of the current vehicle, meaning the center point of the vehicle is always on the X-axis. Therefore, the coordinates of the current vehicle are ( The coordinates of target vehicle 1 are determined based on the coordinate system as (0), The coordinates of target vehicle 2 are ( Lane centerline equations 1, 2, and 3 are established from left to right. Since target vehicle 1 is in lane 1, its coordinates are substituted into lane centerline equation 1, and the result is then calculated. The difference is the lateral offset distance between the center point of target vehicle 1 and the center line of the lane where target vehicle 1 is located. Correspondingly, target vehicle 2 is located in lane 3, so the coordinates of target vehicle 2 are substituted into lane centerline equation 3, and the result is then calculated. The difference is the lateral offset distance between the center point of target vehicle 2 and the center line of the lane where target vehicle 2 is located. .in, , , and The values ​​of lane centerline equation 1 and lane centerline equation 3 are different.

[0087] The above steps accurately identify the relative position of the vehicle and the lane using a camera, and construct a dynamic coordinate system based on the current vehicle center point and the lane center point, thereby achieving real-time calibration of the target vehicle position and accurate calculation of the lane offset distance.

[0088] Step 102: Determine the initial judgment values ​​of different types of sensors for the driving state of the target vehicle based on different driving parameters.

[0089] This invention acquires different driving parameters using different types of sensors, and then determines the initial judgment values ​​for the target vehicle's driving state based on these driving parameters. These initial judgment values ​​represent the driving state of the target vehicle (including intelligent driving state and non-intelligent driving state) determined by the sensor during detection. Specifically, the initial judgment value for the target vehicle's driving state by the millimeter-wave radar includes a first initial judgment value and a second initial judgment value. In detail, step 102... Figure 4 As shown:

[0090] Step 1021: If the lateral offset distance is less than or equal to the preset distance threshold within the first preset time period, then the initial judgment of the camera on the driving state of the target vehicle is assigned to the target calibration value.

[0091] Step 1022: If the change in the longitudinal speed of the vehicle is less than the first speed change threshold, and the target vehicle has not entered the new road, then the millimeter-wave radar assigns the first initial judgment value of the target vehicle's driving state to the target calibration value.

[0092] Step 1023: If the change in the longitudinal speed of the target vehicle is greater than or equal to the second speed change threshold within a second preset time period after determining that the target vehicle has entered a new road or interacted with other traffic participants, then the second initial determination value of the millimeter-wave radar for the driving state of the target vehicle is determined to be the target calibration value.

[0093] Step 1024: Obtain the standard deviation of the lateral displacement within the third preset time period.

[0094] Step 1025: If the standard deviation is less than the standard deviation threshold, then the initial judgment value of the LiDAR on the driving state of the target vehicle is determined to be the target calibration value.

[0095] The first, second, and third preset durations can be set according to requirements, and the durations can be the same or different; this invention does not impose specific limitations on them. The initial judgment assignment can be 1 or 0. An initial judgment assignment of 1 means the target vehicle is considered to be in intelligent driving mode; if the initial judgment assignment is 0, the target vehicle is considered to be in non-intelligent driving mode. It should be understood that the above steps are based on driving parameters (lateral offset distance between the target vehicle's center point and the lane centerline of the target vehicle's lane, vehicle longitudinal speed, and lateral displacement) obtained from cameras, millimeter-wave radar, and lidar to determine the initial judgment assignment. In practice, other types of sensors can also be installed in the vehicle to obtain other types of driving parameters to determine the initial judgment assignment of the sensor for the target vehicle's driving state; this invention does not impose specific limitations on them.

[0096] Whether the target vehicle is in intelligent driving mode is determined by the lateral offset distance within a first preset time period. The basis for this determination is that when the vehicle is in intelligent driving mode, it maintains centering assistance during lateral control. If the lateral offset distance within the first preset time period is consistently less than or equal to a preset distance threshold, the camera's initial assessment of the target vehicle's driving state is set to 1. If the lateral offset distance exceeds the preset distance threshold at some point within the first preset time period or for a very short duration, the camera's initial assessment of the target vehicle's driving state is set to 0. The change in the vehicle's longitudinal speed is also used to determine whether the target vehicle is in intelligent driving mode. There are two scenarios: the first is when the target vehicle has not entered a new road. In this scenario, the basis for the determination is that when the vehicle is in intelligent driving mode, it maintains a constant speed during longitudinal control through functions such as adaptive cruise control, lane keeping assist, and navigation cruise assist. Therefore, the vehicle's longitudinal speed change is judged to be less than a first speed change threshold to determine whether the vehicle is in intelligent driving mode. The second scenario involves the target vehicle entering a new road or interacting with other road users. In this case, the judgment is based on the vehicle's operation in intelligent driving mode. In longitudinal control mode, functions such as automatic emergency braking, stop-and-go braking, and navigation cruise assist control the vehicle's rapid response. Therefore, the vehicle's longitudinal speed change is judged to be greater than or equal to a second speed change threshold within a second preset time period to determine if the vehicle is in intelligent driving mode. The standard deviation of lateral displacement within a third preset time period is used to determine if the target vehicle is in intelligent driving mode. In this scenario, the vehicle's centering assist is used in lateral control mode. Therefore, the standard deviation of lateral displacement within a third preset time period is judged to be less than a standard deviation threshold to determine if the vehicle is in intelligent driving mode.

[0097] The formula for calculating the standard deviation is as follows:

[0098]

[0099] in, This refers to the standard deviation of the lateral displacement within the third preset time period. It refers to the lateral displacement of the i-th sample. This refers to the average lateral displacement, and n refers to the total number of times the lateral displacement is sampled within the third preset time period.

[0100] The traffic participants mentioned in the above steps refer to all entities involved in traffic behavior in road traffic activities. Other traffic participants refer to other entities or obstacles in the road that affect the normal driving of the target vehicle, in addition to the target vehicle being detected. These can be various types of vehicles (bicycles, electric vehicles, motorcycles, cars and trucks, etc.), pedestrians, roadblocks, etc.

[0101] To facilitate understanding, an example is provided above. For instance, suppose the first preset duration is 2 minutes and the preset distance threshold is 15cm. Then, based on the coordinates of target vehicle 1, the lateral offset distance is calculated... If the distance is ≤15cm, the camera's initial assessment of the target vehicle's driving status will be assigned a value (which can be expressed as a parameter). A value of 1 indicates that, based on the data detected by the camera, the driving state of target vehicle 1 can be determined to be intelligent driving mode. If the lateral offset distance is calculated based on the coordinates of target vehicle 2... If the distance is ≤15cm, the camera will assign an initial value to the target vehicle's driving status. A value of 1 indicates that the driving status of the target vehicle 2 can be determined to be intelligent driving status based on the data detected by the camera.

[0102] If the first speed change threshold is 1 kph, and the change in the vehicle's longitudinal speed remains <1 kph before the target vehicle enters the new road, then the initial determination of the target vehicle's driving state by the millimeter-wave radar is assigned a value (which can be expressed as a parameter). A value of 1 indicates that the target vehicle's driving state can be determined to be intelligent driving state based on the data detected by the millimeter-wave radar.

[0103] If the second preset duration is 0.8s and the second speed change threshold is 3kph, and the target vehicle is detected entering a new road or interacting with other traffic participants, the change in the vehicle's longitudinal speed within 0.7s is 5kph. Since 0.7s < 0.8s and 5kph > 3kph, the millimeter-wave radar assigns a second initial judgment value (which can be expressed as a parameter) to the target vehicle's driving state. A value of 1 indicates that the target vehicle's driving state can be determined to be intelligent driving state based on the data detected by the millimeter-wave radar.

[0104] If the third preset duration is 3 minutes, the standard deviation threshold is 12, and the standard deviation of the lateral displacement within the third preset duration is 10, since 10 < 12, the initial judgment assignment of the lidar to the target vehicle's driving state (which can be expressed as a parameter) is... A value of 1 indicates that the target vehicle's driving state can be determined to be intelligent driving state based on the data detected by the lidar.

[0105] The above steps, based on indicators such as lateral offset distance, longitudinal speed change, and lateral displacement, distinguish between steady-state (such as straight-line cruising) and dynamic scenarios (such as lane changing and interaction), and specifically verify the reliability of cameras, millimeter-wave radar, and lidar. Moreover, through multi-dimensional verification, the perception accuracy and decision reliability of the target vehicle's driving status are significantly improved.

[0106] Step 103: Obtain the current weather type and the real-time confidence level of different types of sensors.

[0107] The weather types in this invention include sunny days, foggy days, rainy days, snowy days, and sandstorms. The criteria for determining the real-time confidence level differ for different types of sensors. When determining the real-time confidence level of a camera, the average brightness, image sharpness, and historical trajectory consistency of the image frames captured by the camera are considered. Specifically, when determining the average image brightness, a brightness threshold is set; when the average image brightness is greater than the brightness threshold, the image is considered to have good average brightness performance. Image sharpness is determined based on the grayscale variance of a single frame. Blurry image frames have smaller grayscale differences and smaller grayscale variances, so a threshold is set; when the grayscale variance is greater than the threshold, the image sharpness is considered to be good (it can also be judged based on gradient; the larger the average gradient, the higher the sharpness). Historical trajectory consistency is determined by whether the image content in multiple consecutive image frames is consistent or has a small range of variation (the range of variation is less than a preset value); if so, the historical trajectory consistency is considered to be good. If the average image brightness, image sharpness, and consistency of historical trajectories are good, the confidence equation coefficients are adjusted to make the camera's real-time confidence level approach 1. If the performance is poor, the confidence equation coefficients are adjusted to make the camera's real-time confidence level approach 0. The first preset linear relationship algorithm formula for the camera's real-time confidence level is as follows:

[0108]

[0109] in, , and These are the confidence equation coefficients, which can be dynamically adjusted. This refers to the average brightness of the image. This refers to image clarity. This refers to the consistency of historical trajectories. It is the real-time confidence level of the camera.

[0110] When determining the real-time confidence level of millimeter-wave radar, the signal-to-noise ratio (SNR) and speed-time consistency of the radar data are considered. For SNR determination, a threshold is set; a SNR greater than the threshold is considered good. Speed-time consistency is determined by whether the observed speed of the target vehicle remains consistent across multiple consecutive moments. Consistent speed is considered good speed-time consistency [or the ratio of the current speed to the average speed; the closer to 0 and less than a specified value (e.g., 0.2), the better the consistency]. If both SNR and speed-time consistency are good, the confidence equation coefficients are adjusted to make the real-time confidence level of the millimeter-wave radar approach 1. If the performance is poor, the confidence equation coefficients are adjusted to make the real-time confidence level approach 0. The second preset linear relationship algorithm formula for the real-time confidence level of millimeter-wave radar is as follows:

[0111]

[0112] in, and These are the confidence equation coefficients, which can be dynamically adjusted. SNR refers to the signal-to-noise ratio. This refers to the consistency of speed and time. It is the real-time confidence level of millimeter-wave radar.

[0113] When determining the real-time confidence level of a LiDAR system, the system considers the average intensity of the point cloud, the consistency of reflection intensity, and the inverse geometric fitting residual index. For the average point cloud intensity, a preset range is used; if the average point cloud intensity falls within this range, the system is considered to have good performance. Reflection intensity consistency is calculated by comparing the standard deviation of the point cloud within the target vehicle's outline to the point cloud with the highest reflection intensity after vehicle detection. A ratio closer to 1 indicates better consistency. For example, a ratio greater than 0.8 indicates good consistency. The inverse geometric fitting residual index is calculated by fitting a 3D bounding box after vehicle detection and calculating the mean residual between the point cloud within the target vehicle's outline and the bounding box. A ratio closer to 0.8m indicates better fitting quality. For example, a ratio less than 0.1 indicates good performance. If the point cloud average intensity, reflection intensity consistency, and geometric fitting residual inverse indices perform well, the confidence equation coefficients are adjusted to make the real-time confidence of the lidar approach 1. If the performance is poor, the confidence equation coefficients are adjusted to make the real-time confidence of the lidar approach 0. The third preset linear relationship algorithm formula for the real-time confidence of the lidar is as follows:

[0114]

[0115] in, , and These are the confidence equation coefficients, which can be dynamically adjusted. D refers to the average intensity of the point cloud, I refers to the consistency of reflection intensity, and F refers to the inverse index of the geometric fit residual. This refers to the real-time confidence level of the lidar. Therefore, the operational steps required before acquiring the real-time confidence level for different types of sensors include:

[0116] Acquire image frames captured by the camera, millimeter-wave radar data, and lidar data;

[0117] Obtain the average brightness, image sharpness, and historical trajectory consistency of the image frames;

[0118] Acquire signal-to-noise ratio and velocity-time consistency of millimeter-wave radar data;

[0119] Acquire the point cloud average intensity, reflection intensity consistency, and geometric fitting residual inverse index of lidar data;

[0120] The average brightness of the image, the image sharpness, and the consistency of the historical trajectory are substituted into the first preset linear relationship algorithm to generate the real-time confidence of the camera.

[0121] The signal-to-noise ratio and velocity-time consistency are substituted into the second preset linear relationship algorithm to generate the real-time confidence level of the millimeter-wave radar;

[0122] The average intensity of the point cloud, the consistency of the reflection intensity, and the inverse index of the geometric fitting residual are substituted into the third preset linear relationship algorithm to generate the real-time confidence level of the lidar.

[0123] The above steps determine the real-time confidence of different types of sensors through multiple parameters. The reliability weights of different types of sensors can be adjusted according to environmental changes when allocating weights in the subsequent process. Moreover, when a single sensor fails, the system can still rely on high-confidence data, thereby significantly enhancing the perception stability and safety of the autonomous driving system in complex scenarios (such as sudden changes in lighting, weather interference, and dynamic obstacles).

[0124] Step 104: Determine the target weather environment attenuation factor value corresponding to the current weather type.

[0125] This invention introduces a weather environment attenuation factor to assess the impact of weather. Because weather affects sensor recognition capabilities, the attenuation factor quantifies the degree of attenuation. The weather types included in this invention are sunny, foggy, rainy, snowy, and sandstorms. Different weather types are pre-calibrated with different weather environment attenuation factor values. The comparison relationship is shown in Table 1 below:

[0126] Table 1: Comparison of Attenuation Factor Values ​​for Different Weather Types and Different Weather Environments

[0127]

[0128] It should be understood that, The values ​​can be adjusted according to the sensor model or actual business needs, and the weather type can also be adjusted and expanded according to actual needs. This invention does not impose specific limitations here. The weather type can be determined by the sensor. For example, when using a camera (visible light camera), if sufficient light is detected, the image is clear, and shadows are obvious, the weather type is considered sunny. If the image is blurry, the contrast is low, and distant objects disappear, which may trigger white balance adjustment, the weather type is considered foggy. If water droplets are detected reflecting off the windshield, the road surface reflects light more intensely, and water mist or splashes are visible, the weather type is considered rainy. If white coverings are detected on the road or vehicle body, and snowflakes have dynamic trajectories, the weather type is considered snowy. If the image is yellowish / grayish, visibility drops sharply, and particulate matter is dynamically blurred, the weather type is considered a sandstorm. Because fog / rain / snow can cause abnormal point cloud density (signal attenuation or scattering), weather types such as foggy, rainy, and snowy days can be detected using lidar. Rain, snow, or dust storms weaken the echo signal of millimeter-wave radar and increase noise. Furthermore, millimeter-wave radar can analyze precipitation or particulate matter density through reflected signals, making it suitable for detecting weather types such as rain, snow, and dust storms. To improve accuracy, results from different sensors can be combined to comprehensively determine the weather type.

[0129] After pre-calibrating different weather environment attenuation factor values ​​for different weather types, this invention can determine the target weather environment attenuation factor value corresponding to the current weather type from the weather environment attenuation factor values ​​based on the pre-calibrated relationship when the current weather type is obtained.

[0130] Step 105: Determine the assigned weights for different types of sensors by using real-time confidence levels and target weather environment attenuation factor values.

[0131] The weighting of the different types of sensors in this invention is obtained by the following formula:

[0132]

[0133] in, This refers to the assigned weight of the i-th sensor. This refers to the real-time confidence level of the i-th sensor. This refers to the weather and environmental attenuation factor value, where n is the total number of sensors. This refers to the real-time confidence level of the k-th sensor.

[0134] As can be seen from the above formula, to determine the assigned weights of the sensors, it is necessary to obtain the real-time confidence level of the sensors and the current target weather environment attenuation factor value.

[0135] When setting the weather environment attenuation factor value, this invention considers not only the weather type but also the working status of different types of sensors. If a sensor malfunctions, the malfunctioning sensor is removed from the weight allocation.

[0136] For example, if there are three types of sensors A, B, and C, and sensor B fails, the calculated weight allocation could be 1 / 2 for sensor A and 1 / 2 for sensor C. Since sensor B is excluded from the weight allocation, its weight allocation would be 0.

[0137] Step 106: Determine whether the target vehicle is in intelligent driving mode by assigning initial values ​​and weights to different types of sensors.

[0138] This invention calculates the driving state determination value based on the initial judgment assignment and weight allocation of different types of sensors, as shown in the following formula:

[0139]

[0140] Where P is the driving status determination value of the target vehicle. It's about the weighting of the cameras. It is the initial judgment and assignment of the camera based on the driving status of the target vehicle. It is the allocation weight of millimeter-wave radar. This is the initial value assigned by the millimeter-wave radar to determine the driving status of the target vehicle. This is the second initial determination value assigned by the millimeter-wave radar to the target vehicle's driving status. It is the allocation weight of the lidar. It is the initial determination and assignment of the driving status of the target vehicle by the lidar.

[0141] As shown in the formula above, calculating the driving state determination value requires first obtaining the initial determination values ​​of the target vehicle's driving state from different sensors. These initial determination values ​​can be 0 or 1. It also requires obtaining the assigned weights for different types of sensors. However, when obtaining these weights, it's necessary to first determine the real-time confidence level of each type of sensor and the target weather environment attenuation factor value determined based on the current weather conditions. Different values ​​for these parameters will lead to different calculation results for the driving state determination value. Examples are shown in Table 2 below:

[0142] Table 2: Reference Table for Different Driving State Judgment Values ​​Corresponding to Different Parameters

[0143]

[0144] As shown in Table 2, when the weather environment attenuation factor is affected by severe weather and is low (0.4), it will reduce the allocation weight of cameras with low confidence and increase the allocation weight of millimeter-wave radars with high confidence. It should be understood that the values ​​in Table 2... Although the values ​​do not correspond to those in Table 1, as mentioned above... The values ​​will be adjusted depending on the sensor model or actual business needs, so the values ​​in Table 2... The value chosen is reasonable.

[0145] After calculating the driving state determination value, this invention can determine whether the target vehicle is in intelligent driving mode based on the driving state determination value. Therefore, step 106, as follows... Figure 5 As shown:

[0146] Step 1061: The initial judgment values ​​of different types of sensors are weighted and summed to obtain the driving state judgment value.

[0147] Step 1062: If the driving status determination value is greater than the preset value, the target vehicle is determined to be in intelligent driving state.

[0148] Step 1063: If the driving status determination value is less than or equal to the preset value, the target vehicle is determined not to be in intelligent driving mode.

[0149] For example, if the preset value is set to 60%, then according to Table 3, when the sensor parameter value is value 1, since 78.4% > 60%, the target vehicle can be determined to be in intelligent driving mode. When the sensor parameter value is value 2, since 54.15% < 60%, the target vehicle can be determined to be not in intelligent driving mode. When the sensor parameter value is value 3, since 73.58% > 60%, the target vehicle can be determined to be in intelligent driving mode.

[0150] The above steps integrate the advantages of different sensors through weight allocation, reduce the risk of misjudgment by a single sensor, and provide a clear decision basis for the current vehicle by quantitatively outputting a binary judgment of "intelligent driving / non-intelligent driving".

[0151] Step 107: If the vehicle is in intelligent driving mode, a warning reminder will be sent.

[0152] In this invention, there can be one or more target vehicles within a preset distance range from the current vehicle. When there are multiple target vehicles, they need to be assessed simultaneously. Based on the assessment results, if none of the target vehicles are in intelligent driving mode, no warning is issued, but the vehicle's display screen will show that no vehicles in intelligent driving mode are present nearby. If at least one of the target vehicles is in intelligent driving mode, an audible warning is issued, and the distribution of target vehicles is displayed on the vehicle's display screen, with the vehicles in intelligent driving mode specially marked. This allows the current driver to choose to avoid or move away from target vehicles in intelligent driving mode based on the displayed information, reducing driving safety hazards.

[0153] The driving state recognition method provided by this invention acquires different driving parameters of a target vehicle within a preset distance range using different types of sensors. This facilitates a comprehensive and accurate reflection of the driving state based on these parameters. The method uses different driving parameters to determine the initial judgment values ​​of different types of sensors for the target vehicle's driving state, reducing the error of single-sensor judgments. It also acquires the real-time confidence levels of the current weather type and different types of sensors, using these real-time confidence levels to evaluate sensor reliability and avoid reliance on failed data. Furthermore, by acquiring the current weather type, an appropriate weather environment attenuation factor value can be set, thus determining the target weather environment attenuation factor value corresponding to the current weather type. By considering the impact of weather on sensor accuracy, the subsequent judgment results are more accurate. By using real-time confidence levels and target weather environment attenuation factors, the assigned weights for different types of sensors are determined. These weights, calculated based on real-time confidence levels and target weather environment attenuation factors, can automatically reduce the weights of low-confidence sensors in adverse weather conditions, improving system robustness and the accuracy of subsequent judgments. Through initial judgment assignments and weight allocations for different types of sensors, the system determines whether the target vehicle is in intelligent driving mode. If it is, a warning is sent. By comprehensively judging the driving status of the target vehicle using different types of sensors, "information blind spots" can be avoided. Furthermore, in adverse weather conditions or when sensors are interfered with, the system combines the recognition parameters of multiple types of sensors for judgment, improving system robustness and the accuracy of the judgment results, avoiding misjudgments of the true driving status of other vehicles, and sending warnings reasonably and promptly to reduce safety hazards.

[0154] Reference Figure 6 The diagram shows a structural schematic of a driving state recognition device provided by the present invention, the device comprising:

[0155] The first acquisition module 201 is used to acquire different driving parameters of a target vehicle that is within a preset distance range from the current vehicle through different types of sensors.

[0156] The first determining module 202 is used to determine the initial judgment values ​​of different types of sensors on the driving state of the target vehicle through different driving parameters.

[0157] The second acquisition module 203 is used to acquire the current weather type and the real-time confidence level of different types of sensors.

[0158] The second determining module 204 is used to determine the value of the target weather environment attenuation factor corresponding to the current weather type.

[0159] The third determination module 205 is used to determine the allocation weights corresponding to different types of sensors by taking the values ​​of real-time confidence and target weather environment attenuation factor.

[0160] The judgment module 206 is used to determine whether the target vehicle is in intelligent driving state by assigning initial judgment values ​​and weights to different types of sensors.

[0161] The warning module 207 is used to send a warning reminder if the vehicle is in intelligent driving mode.

[0162] Optional, different types of sensors include cameras, millimeter-wave radar and lidar, and driving parameters include the lateral offset distance between the target vehicle's center point and the lane centerline of the lane in which the target vehicle is located, the vehicle's longitudinal speed and lateral displacement.

[0163] The first acquisition module 201 specifically includes:

[0164] The first acquisition submodule is used to acquire the lateral offset distance between the center point of the target vehicle and the center line of the lane where the target vehicle is located through the camera.

[0165] The second acquisition submodule is used to acquire the longitudinal speed of the target vehicle via millimeter-wave radar.

[0166] The third acquisition submodule is used to acquire the lateral displacement of the target vehicle through lidar.

[0167] Optionally, the first acquisition submodule specifically includes:

[0168] The recognition unit is used to identify the actual lane and the lateral center point of the current vehicle using a camera.

[0169] The first acquisition unit is used to acquire the center point of the current vehicle.

[0170] Establish a coordinate system unit to create a coordinate system based on the current vehicle's center point and lateral center point. The X-axis of the coordinate system is determined based on the current vehicle's center point, and the Y-axis is determined based on the lateral center point.

[0171] The second acquisition unit is used to acquire the position coordinates of the target vehicle in the coordinate system through the camera.

[0172] The determination unit is used to substitute the position coordinates into the preset lane centerline equation to determine the lateral offset distance between the center point of the target vehicle and the lane centerline of the lane where the target vehicle is located. Different lanes correspond to different lane centerline equations.

[0173] Optionally, the initial determination of the target vehicle's driving state by the millimeter-wave radar includes a first initial determination value and a second initial determination value.

[0174] The first determining module 202 specifically includes:

[0175] The first determining submodule is used to determine the initial judgment value of the camera on the driving state of the target vehicle as the target calibration value if the lateral offset distance is less than or equal to the preset distance threshold within a first preset time period.

[0176] The second determination submodule is used to determine the first initial judgment value of the target vehicle's driving state by the millimeter-wave radar as the target calibration value if the change in the vehicle's longitudinal speed is less than the first speed change threshold when it is determined that the target vehicle has not entered the new road.

[0177] The third determination submodule is used to determine the second initial judgment value of the millimeter-wave radar on the driving state of the target vehicle as the target calibration value if the change in the longitudinal speed of the vehicle is greater than or equal to the second speed change threshold within a second preset time period when the target vehicle enters a new road or interacts with other traffic participants.

[0178] The fourth acquisition submodule is used to acquire the standard deviation of the lateral displacement within a third preset time period.

[0179] The fourth determination submodule is used to determine the initial judgment value of the LiDAR on the driving state of the target vehicle as the target calibration value if the standard deviation is less than the standard deviation threshold.

[0180] Optionally, the driving status recognition device also includes:

[0181] The third acquisition module is used to acquire image frames captured by the camera, millimeter-wave radar data, and lidar data.

[0182] The fourth acquisition module is used to acquire the average brightness, image sharpness, and historical trajectory consistency of the image frame.

[0183] The fifth acquisition module is used to acquire the signal-to-noise ratio and velocity-time consistency of millimeter-wave radar data.

[0184] The sixth acquisition module is used to acquire the point cloud average intensity, reflection intensity consistency, and geometric fitting residual inverse index of the lidar data.

[0185] The first generation module is used to input the average brightness of the image, the image sharpness, and the consistency of the historical trajectory into the first preset linear relationship algorithm to generate the real-time confidence of the camera.

[0186] The second generation module is used to substitute the signal-to-noise ratio and velocity-time consistency into the second preset linear relationship algorithm to generate the real-time confidence level of the millimeter-wave radar.

[0187] The third generation module is used to substitute the point cloud average intensity, reflection intensity consistency and geometric fitting residual inverse index into the third preset linear relationship algorithm to generate the real-time confidence level of the lidar.

[0188] Optionally, the weighting of the different types of sensors is obtained by the following formula:

[0189]

[0190] in, This refers to the assigned weight of the i-th sensor. This refers to the real-time confidence level of the i-th sensor. This refers to the weather and environmental attenuation factor value, where n is the total number of sensors. This refers to the real-time confidence level of the k-th sensor.

[0191] Optionally, the judgment module 206 specifically includes:

[0192] The weighted submodule is used to perform weighted summation of the initial judgment values ​​of different types of sensors by assigning weights to obtain the driving state judgment value.

[0193] The first determination submodule is used to determine that the target vehicle is in intelligent driving mode if the driving state determination value is greater than the preset value.

[0194] The second determination submodule is used to determine that the target vehicle is not in intelligent driving mode if the driving state determination value is less than or equal to a preset value.

[0195] The driving state recognition device provided by this invention acquires different driving parameters of a target vehicle within a preset distance range from the current vehicle using different types of sensors. This facilitates a comprehensive and accurate reflection of the driving state based on these parameters. The device determines the initial judgment values ​​for the target vehicle's driving state using different types of sensors based on these different driving parameters. This preliminary judgment of the vehicle's driving state by different types of sensors reduces the error of a single sensor's judgment. The device acquires the real-time confidence levels of the current weather type and different types of sensors, using these real-time confidence levels to evaluate sensor reliability and avoid reliance on failed data. By acquiring the current weather type, an appropriate weather environment attenuation factor value can be set, i.e., determining the target weather environment attenuation factor value corresponding to the current weather type. By considering the impact of weather on sensor accuracy, the subsequent judgment results are more accurate. By using real-time confidence levels and target weather environment attenuation factors, the assigned weights for different types of sensors are determined. These weights, calculated based on real-time confidence levels and target weather environment attenuation factors, can automatically reduce the weights of low-confidence sensors in adverse weather conditions, improving system robustness and the accuracy of subsequent judgments. Through initial judgment assignments and weight allocations for different types of sensors, the system determines whether the target vehicle is in intelligent driving mode. If it is, a warning is sent. By comprehensively judging the driving status of the target vehicle using different types of sensors, "information blind spots" can be avoided. Furthermore, in adverse weather conditions or when sensors are interfered with, the system combines the recognition parameters of multiple types of sensors for judgment, improving system robustness and the accuracy of the judgment results, avoiding misjudgments of the true driving status of other vehicles, and sending warnings reasonably and promptly to reduce safety hazards.

[0196] Reference Figure 7 The present invention also provides an electronic device, such as Figure 7 As shown, it includes a processor 301, a communication interface 302, a memory 303, and a communication bus 304, wherein the processor 301, the communication interface 302, and the memory 303 communicate with each other through the communication bus 304.

[0197] Processor 301, memory 303 for storing processor-executable instructions;

[0198] The processor 301 is configured to execute the instructions to implement the driving state recognition method as described above:

[0199] Different driving parameters of target vehicles within a preset distance range are acquired using different types of sensors;

[0200] Different types of sensors are used to determine the initial values ​​for the target vehicle's driving state based on different driving parameters.

[0201] Obtain the current weather type and the real-time confidence level of the sensors for different types;

[0202] Determine the value of the target weather environment attenuation factor corresponding to the current weather type;

[0203] The allocation weights for different types of sensors are determined by the real-time confidence level and the target weather environment attenuation factor.

[0204] By assigning initial values ​​and weights to different types of sensors, it is determined whether the target vehicle is in an intelligent driving state.

[0205] If the vehicle is in intelligent driving mode, a warning reminder will be sent.

[0206] The communication bus mentioned above can be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus, etc. This communication bus can be divided into address bus, data bus, control bus, etc. For ease of illustration, only one thick line is used to represent it in the diagram, but this does not mean that there is only one bus or one type of bus.

[0207] The communication interface is used for communication between the aforementioned terminal and other devices.

[0208] The memory may include random access memory (RAM) or non-volatile memory, such as at least one disk storage device. Optionally, the memory may also be at least one storage device located remotely from the aforementioned processor.

[0209] The processors mentioned above can be general-purpose processors, including central processing units (CPUs), network processors (NPs), etc.; they can also be digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components.

[0210] In another embodiment of the present invention, a vehicle is also provided, which may specifically include the above-mentioned driving state recognition device.

[0211] In the above embodiments, implementation can be achieved, in whole or in part, through software, hardware, firmware, or any combination thereof. When implemented in software, it can be implemented, in whole or in part, as a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the processes or functions described in this invention are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium accessible to a computer or a data storage device such as a server or data center that integrates one or more available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., a solid-state drive (SSD)).

[0212] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0213] The various embodiments in this specification are described in a related manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, the system embodiments are basically similar to the method embodiments, so the description is relatively simple; relevant parts can be referred to the descriptions of the method embodiments.

[0214] The above description is merely a preferred embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention are included within the scope of protection of the present invention.

Claims

1. A driving state recognition method, characterized in that, The method includes: Different driving parameters of a target vehicle within a preset distance range are acquired by different types of sensors. These driving parameters include the lateral offset distance between the center point of the target vehicle and the center line of the lane where the target vehicle is located, the longitudinal speed of the vehicle, and the lateral displacement. The method involves determining the initial judgment values ​​of different types of sensors for the driving state of the target vehicle based on different driving parameters, including: determining the initial judgment values ​​of the corresponding sensors for the driving state of the target vehicle based on the lateral offset distance, the longitudinal speed of the vehicle, and the lateral displacement, respectively. The initial judgment values ​​are used to indicate whether the target vehicle is in an intelligent driving state or a non-intelligent driving state, and the initial judgment values ​​are Boolean values ​​of 0 or 1. Obtain the current weather type and the real-time confidence level of the sensors for different types; Determine the value of the target weather environment attenuation factor corresponding to the current weather type; The allocation weights for different types of sensors are determined by the real-time confidence level and the target weather environment attenuation factor. By assigning initial values ​​and weights to different types of sensors, it is determined whether the target vehicle is in an intelligent driving state. If the vehicle is in intelligent driving mode, a warning reminder will be sent.

2. The method according to claim 1, characterized in that, The different types of sensors include cameras, millimeter-wave radar, and lidar; The acquisition of different driving parameters of a target vehicle within a preset distance range from the current vehicle using different types of sensors includes: The camera is used to obtain the lateral offset distance between the center point of the target vehicle and the center line of the lane in which the target vehicle is located; The longitudinal velocity of the target vehicle is obtained using millimeter-wave radar; The lateral displacement of the target vehicle is obtained using lidar.

3. The method according to claim 2, characterized in that, The step of obtaining the lateral offset distance between the center point of the target vehicle and the center line of the lane where the target vehicle is located via a camera includes: The camera identifies the actual lane the vehicle is currently in and the lateral center point of that lane. Get the center point of the current vehicle; A coordinate system is established based on the current vehicle's center point and the lateral center point, wherein the X-axis of the coordinate system is determined based on the current vehicle's center point, and the Y-axis of the coordinate system is determined based on the lateral center point; The target vehicle's position coordinates in the coordinate system are obtained using a camera; Substitute the position coordinates into the preset lane centerline equation to determine the lateral offset distance between the center point of the target vehicle and the lane centerline of the lane where the target vehicle is located. Different lanes correspond to different lane centerline equations.

4. The method according to claim 2, characterized in that, The initial determination of the target vehicle's driving state by the millimeter-wave radar includes a first initial determination value and a second initial determination value. The initial determination of the target vehicle's driving state by different types of sensors based on different driving parameters includes: If the lateral offset distance is less than or equal to a preset distance threshold within a first preset time period, then the initial determination of the camera's driving state of the target vehicle is assigned to the target calibration value. If it is determined that the target vehicle has not entered the new road, and if the change in the longitudinal speed of the vehicle is less than the first speed change threshold, then the first initial determination of the driving state of the target vehicle by the millimeter-wave radar is assigned as the target calibration value. If the target vehicle enters a new road or interacts with other traffic participants, and the change in the vehicle's longitudinal speed is greater than or equal to a second speed change threshold within a second preset time period, then the second initial determination of the target vehicle's driving state by the millimeter-wave radar is assigned as the target calibration value. Obtain the standard deviation of the lateral displacement within a third preset time period; If the standard deviation is less than the standard deviation threshold, then the initial determination of the driving state of the target vehicle by the lidar is assigned to the target calibration value.

5. The method according to claim 1, characterized in that, Before obtaining the current weather type and the real-time confidence scores of the sensors for different types, the method further includes: Acquire image frames captured by the camera, millimeter-wave radar data, and lidar data; Obtain the average brightness, image sharpness, and historical trajectory consistency of the image frame; The signal-to-noise ratio and velocity-time consistency of the millimeter-wave radar data were obtained. Obtain the point cloud average intensity, reflection intensity consistency, and geometric fitting residual inverse index of the lidar data; The average brightness of the image, the image sharpness, and the consistency of the historical trajectory are substituted into the first preset linear relationship algorithm to generate the real-time confidence of the camera; The signal-to-noise ratio and velocity-time consistency are substituted into the second preset linear relationship algorithm to generate the real-time confidence level of the millimeter-wave radar; The average intensity of the point cloud, the consistency of reflection intensity, and the inverse index of geometric fitting residual are substituted into the third preset linear relationship algorithm to generate the real-time confidence level of the lidar.

6. The method according to claim 1, characterized in that, The weighting of the different types of sensors is obtained by the following formula: in, This refers to the assigned weight of the i-th sensor. This refers to the real-time confidence level of the i-th sensor. This refers to the weather and environmental attenuation factor value, where n refers to the total number of sensors. This refers to the real-time confidence level of the k-th sensor.

7. The method according to claim 1, characterized in that, The step of determining whether the target vehicle is in an intelligent driving state by assigning initial values ​​and weights to different types of sensors further includes: The driving state determination value is obtained by weighting and summing the initial judgment values ​​of different types of sensors according to the weighting assignment. If the driving state determination value is greater than the preset value, the target vehicle is determined to be in intelligent driving state; If the driving status determination value is less than or equal to the preset value, the target vehicle is determined not to be in intelligent driving mode.

8. A driving state recognition device, characterized in that, The device includes: The first acquisition module is used to acquire different driving parameters of a target vehicle that is within a preset distance range from the current vehicle through different types of sensors. The driving parameters include the lateral offset distance between the center point of the target vehicle and the center line of the lane where the target vehicle is located, the longitudinal speed of the vehicle, and the lateral displacement. The first determining module is used to determine the initial determination values ​​of different types of sensors for the driving state of the target vehicle through different driving parameters, including: determining the initial determination values ​​of the corresponding sensors for the driving state of the target vehicle through the lateral offset distance, the longitudinal speed of the vehicle and the lateral displacement, the initial determination values ​​are used to indicate whether the target vehicle is in intelligent driving state or non-intelligent driving state, and the value of the initial determination value is a Boolean value of 0 or 1. The second acquisition module is used to acquire the current weather type and the real-time confidence level of the sensors for different types; The second determining module is used to determine the value of the target weather environment attenuation factor corresponding to the current weather type; The third determining module is used to determine the allocation weights corresponding to different types of sensors by using the real-time confidence level and the target weather environment attenuation factor value; The judgment module is used to determine whether the target vehicle is in an intelligent driving state by assigning initial judgment values ​​and weights to different types of sensors. The warning module is used to send warning reminders if the vehicle is in intelligent driving mode.

9. An electronic device, characterized in that, include: processor; Memory used to store processor-executable instructions; The processor is configured to execute the instructions to implement the driving state recognition method as described in any one of claims 1 to 7.

10. A vehicle, characterized in that, include: The driving state recognition device according to claim 8.

Images