A vehicle collision detection method, device and unmanned vehicle

Abstract

The application relates to a vehicle collision detection method, device and unmanned vehicle, which comprises the following steps: a sensor collects first driving data and second driving data of a vehicle, the first driving data comprises at least one of a first acceleration, a first speed and a first throttle opening degree; the second driving data comprises at least one of a second acceleration, a second speed and a second throttle opening degree; a first driving state of the vehicle is judged according to the first driving data; and whether the vehicle has a collision is judged by comparing the first driving data and the second driving data. The method is based on vehicle acceleration, combined with throttle opening degree and speed, to comprehensively judge a vehicle collision accident type. The method uses a magnitude threshold value and a time threshold value as a collision judgment standard, analyzes a data mode of multi-source information, and thus completes dynamic parameter identification when a collision accident occurs, so as to accurately identify whether the vehicle collides.

Description

Technical Field

[0001] This invention relates to the field of vehicle collision detection technology, specifically to a vehicle collision detection method, device, and unmanned vehicle. Background Technology

[0002] Autonomous vehicles and mobile robots are collectively referred to as unmanned equipment, and collision detection is one of the key technologies for unmanned equipment. Especially when operating in complex urban environments, collisions between unmanned equipment and surrounding road users are inevitable. Therefore, it is necessary to promptly identify whether a collision has occurred and to control the unmanned equipment to stop in time to avoid secondary injuries. However, most existing collision detection methods use mechanical collision detection plates for identification and detection. These mechanical collision detection plates can only be installed at a certain location on the unmanned equipment and cannot cover all areas. Therefore, the unmanned equipment lacks the ability to detect collisions in areas where collision detection plates are not installed, resulting in detection blind spots. Furthermore, collision detection plates are often obstructed by other foreign objects, causing them to become stuck and malfunction, thus failing to detect collisions properly.

[0003] Collision detection panels are positioned at the front and rear of the vehicle. When a side collision occurs, the unmanned vehicle (UAV) may not be detected. In cases where a real collision occurs but the UAV cannot detect it, it may continue driving. This poses a significant safety hazard. For example, if the UAV's functional components are deformed and unable to function, but the driving mechanism is still operational, the UAV will continue driving. Even if it reaches its destination, it will have failed to function properly, resulting in wasted equipment. Furthermore, continued driving after a collision may cause debris to be scattered, creating new traffic hazards. Moreover, in the event of a three-way collision, the UAV may flee the scene after the impact.

[0004] Therefore, how to enable unmanned equipment, especially unmanned vehicles, to detect collisions in a timely manner without relying on collision detection boards has become one of the urgent issues to be solved in this field. Summary of the Invention

[0005] To overcome the shortcomings of the prior art, the present invention provides a vehicle collision detection method, device, and unmanned vehicle, specifically adopting the following technical solution:

[0006] A vehicle collision detection method includes the following steps:

[0007] The vehicle acceleration and / or rate of change of acceleration are acquired based on onboard IMU data. When the acceleration and / or rate of change of acceleration reaches a preset condition, first driving data within a first time period before that moment and second driving data within a second time period after that moment are collected. The first driving data includes at least one of a first acceleration, a first speed, and a first throttle opening. The second driving data includes at least one of a second acceleration, a second speed, and a second throttle opening.

[0008] Based on the first driving data, the second driving data, and the vehicle acceleration and / or rate of change of acceleration, collision information of the vehicle is obtained, including collision type and collision location;

[0009] Based on the collision information, control the vehicle's operation and / or send the collision information to a remote control center.

[0010] By comparing vehicle driving data at different times, the current driving status of the vehicle can be determined, and it can be determined whether the vehicle has been involved in a collision, thus achieving automatic detection of vehicle status.

[0011] Optionally: Based on the first driving data, the second driving data, and the vehicle acceleration and / or rate of change of acceleration, the collision information of the vehicle is obtained, including:

[0012] If the absolute value of the first speed is less than the first speed threshold, then the first driving state of the vehicle is determined to be a stationary state.

[0013] If it is determined that the first acceleration of the vehicle is less than a first acceleration threshold and the second acceleration is greater than the first acceleration threshold, then the throttle opening is detected.

[0014] If the first throttle opening and the second throttle opening are both less than the first throttle opening threshold, then the collision type is determined to be a stationary collision.

[0015] When the vehicle is stationary, by analyzing the values ​​of vehicle acceleration and throttle opening, it is possible to determine whether a collision has occurred, thus achieving real-time collision detection when the vehicle is stationary.

[0016] Optionally: Based on the first driving data, the second driving data, and the vehicle acceleration and / or rate of change of acceleration, the collision information of the vehicle is obtained, including:

[0017] If the first speed is greater than the first speed threshold, then the first driving state of the vehicle is determined to be a moving state;

[0018] If the direction of the first acceleration of the vehicle is opposite to the direction of the first velocity, and the first acceleration is greater than or equal to the second acceleration threshold, then it is determined that the vehicle has undergone emergency braking.

[0019] If it is determined that the direction of the second acceleration of the vehicle is the same as the direction of the second vehicle speed, and the second acceleration is greater than or equal to the third acceleration threshold, then the throttle opening is detected.

[0020] If both the first throttle opening and the second throttle opening are less than the second throttle opening threshold, then the collision type is determined to be a rear-end collision.

[0021] By analyzing the acceleration values ​​and directions of the vehicle while it is in motion, and combining this with the throttle opening values, real-time collision detection during emergency braking can be obtained, thus improving driving safety.

[0022] Optionally: After determining that the vehicle has experienced a stationary collision or a rear-end collision, the vehicle control sensing device determines whether a third vehicle exists around the vehicle within a third time period; and analyzes the positions of the vehicle and the third vehicle to obtain the first position of the collision based on the position. Image information can be used to obtain information about surrounding objects at the time of the collision, so as to promptly determine the position of the colliding object, obtain more vehicle collision information, and facilitate subsequent accident handling.

[0023] Optionally: When the vehicle is involved in a stationary collision or a rear-end collision, the vehicle control sensing device determines whether there is a third vehicle around the vehicle within a third time period;

[0024] If there is a third vehicle around the vehicle that is less than a first threshold distance from the vehicle, then the stationary collision or rear-end collision is confirmed to have occurred, and the second location of the collision is obtained.

[0025] If there is no third vehicle within a distance of less than a first threshold distance around the vehicle, it is determined that the stationary or rear-end collision has not occurred, and an abnormal information is output, which is then stored or sent to the remote control center.

[0026] By identifying and judging third vehicles within a certain distance range around the vehicle, the vehicle collision detection results can be further verified, thereby improving the accuracy of vehicle collision detection, avoiding false alarms, and ensuring accurate results.

[0027] Optionally: If there is no third vehicle in the vicinity of the vehicle that is less than a first threshold distance away from the vehicle;

[0028] Then, extract the vehicle's surrounding images from the first time period temporarily stored in the memory, and analyze whether there is a third vehicle in the vehicle's surrounding images that is less than a first threshold distance from the vehicle.

[0029] If a third vehicle is present in the vehicle's surrounding image within the first time period, and the distance between the third vehicle and the vehicle is less than a first threshold distance, then it is determined that the vehicle has experienced a stationary collision or a rear-end collision.

[0030] To further improve the accuracy of detection results and avoid omissions, the image information collected from the vehicle can be further analyzed. By increasing the collision distance and expanding the detection range, the objects around the vehicle before the collision can be further analyzed and identified to avoid omissions and reduce the accuracy of the detection results.

[0031] Optionally: When setting the first acceleration threshold, the second acceleration threshold, and the third acceleration threshold, the mass of the vehicle is first obtained. The mass of the vehicle includes the mass of the vehicle itself and the mass of the objects transported inside. The first acceleration threshold, the second acceleration threshold, and the third acceleration threshold are obtained through the mass of the vehicle and a preset correspondence.

[0032] By monitoring vehicle mass in real time and adjusting the vehicle acceleration threshold, the accuracy of vehicle collision detection can be improved, and detection errors caused by changes in vehicle mass can be avoided.

[0033] Optionally: After determining that the vehicle has experienced a stationary collision or a rear-end collision, extract a first internal image and a second internal image of the vehicle's cargo box;

[0034] The first internal image is a vehicle interior image acquired at the first time, and the second internal image is a vehicle interior image acquired at the second time.

[0035] By comparing the first internal image and the second internal image, it is determined whether there is a transported object moving relative to the carriage and colliding with the inner wall of the carriage.

[0036] If the object being transported inside the carriage moves relative to the carriage and collides with the inner wall of the carriage, and the direction of movement of the object relative to the carriage is opposite to the direction of the second velocity, the collision is confirmed as a stationary collision.

[0037] If the object being transported inside the carriage moves relative to the carriage and collides with the inner wall of the carriage, and the direction of movement of the object relative to the carriage is opposite to the direction of the second velocity, the collision is confirmed as a rear-end collision.

[0038] If an object being transported inside the carriage moves relative to the carriage and collides with the inner wall of the carriage, and the direction of movement of the object relative to the carriage is the same as the direction of the second velocity, when it is determined to be a rear-end collision, the vehicle control sensing device determines whether there is a third vehicle around the vehicle within a third time period.

[0039] If there is a third vehicle around the vehicle that is less than a first threshold distance from the vehicle, then the rear-end collision is confirmed.

[0040] If there is no third vehicle within a distance of less than a first threshold distance from the vehicle, it is determined that the rear-end collision did not occur, and an abnormal information is output, which is then stored or sent to a remote control center.

[0041] For the collision detection process, the motion state of objects inside the vehicle can be combined to further eliminate erroneous information, which can avoid erroneous detection results caused by abnormal sensors inside the vehicle, thereby improving the accuracy of the detection process.

[0042] Furthermore, this invention discloses an unmanned vehicle that applies the vehicle collision detection method described above.

[0043] The present invention also discloses a vehicle collision detection device applying the above-described vehicle collision detection method, wherein the system comprises at least:

[0044] The sensor module is used to collect vehicle driving data, which includes at least acceleration, vehicle speed, throttle opening, positioning information, radar information, and impact information.

[0045] The driving status judgment module is used to compare vehicle driving data and analyze vehicle driving status and collision information;

[0046] The control module is used to control the vehicle based on the results of the driving status judgment module and to remotely transmit collision information;

[0047] The remote control center is used to receive collision information transmitted by the control module and remotely control the vehicle's subsequent driving actions.

[0048] Optionally: The sensor module includes:

[0049] An acceleration sensor is used to collect vehicle acceleration data and transmit it to the driving status judgment module;

[0050] The vehicle speed sensor is used to collect vehicle speed data and transmit it to the driving status judgment module;

[0051] Throttle opening sensor is used to collect vehicle throttle opening data and transmit it to the driving status judgment module;

[0052] Positioning sensors are used to collect vehicle location information and transmit it to the control module;

[0053] The camera is used to collect image information about the vehicle's surroundings and transmit it to the control module;

[0054] Radar is used to collect distance information of obstacles around the vehicle and transmit it to the control module.

[0055] Optional: The system also includes a warning device, which is connected to the control module. After receiving the result from the driving status judgment module, the control module sends a control signal to the warning device, causing the warning device to issue a warning signal to the surrounding area of ​​the vehicle.

[0056] Beneficial effects

[0057] The technical solution of the present invention achieves the following beneficial effects:

[0058] The vehicle collision detection method of this invention utilizes multi-source information provided by built-in sensors. It comprehensively determines the type of vehicle collision accident based on data measured by an acceleration sensor, combined with the device's throttle opening and speed. This method uses amplitude and time thresholds as collision judgment criteria. By analyzing the data patterns of multi-source information, it identifies dynamic parameters at the time of a collision, thereby accurately determining whether a vehicle collision has occurred. Attached Figure Description

[0059] Figure 1 This is a schematic diagram illustrating the workflow of the vehicle collision detection method in this invention.

[0060] Figure 2 This is a schematic diagram of the data characteristics when a stationary collision occurs in this invention.

[0061] Figure 3 This is a schematic diagram of data characteristics when a rear-end collision occurs during sudden braking in this invention. Detailed Implementation

[0062] The present invention will be further described below with reference to the accompanying drawings. The following embodiments are only used to more clearly illustrate the technical solutions of the present invention and should not be construed as limiting the scope of protection of the present invention. It should be noted that the following detailed descriptions are exemplary and intended to provide further explanation of this application.

[0063] Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains. It should be noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the exemplary embodiments according to this application. As used herein, the singular form is intended to include the plural form as well, unless the context clearly indicates otherwise. Furthermore, it should be understood that when the terms “comprising” and / or “including” are used in this specification, they indicate the presence of features, steps, operations, devices, components, and / or combinations thereof.

[0064] Currently, most vehicle collision detection systems use mechanical collision detection plates for identification and detection. However, traditional detection plates are generally only installed on the front and rear sides of the vehicle. When a collision occurs, the detection plate presses against the pressure sensors located on its inner side, generating an electrical signal to identify the collision status. Since the detection plate cannot cover all parts of the vehicle, it generally lacks the ability to detect collisions in areas where a collision detection plate is not installed, such as the sides of the vehicle. This creates a detection blind spot. In the event of a real collision, if the collision occurs within this blind spot, it will not be detected, and the vehicle may continue driving, creating a greater safety hazard. However, installing a detection plate on the entire vehicle would incur excessive costs, and the limited placement of the doors and headlights prevents the installation of detection plates, thus hindering comprehensive vehicle collision detection.

[0065] Therefore, to solve the above problems, this embodiment improves upon the traditional method of detection using a detection board by analyzing the vehicle's own acceleration, throttle opening, and speed data. The analysis of this data is then used to determine whether a collision has occurred. Combined with... Figure 1 As shown, this embodiment discloses a vehicle collision detection method, which includes the following steps:

[0066] The vehicle acceleration and / or rate of change of acceleration are acquired based on onboard IMU data. When the acceleration and / or rate of change of acceleration reaches a preset condition, first driving data within a first time period before that moment and second driving data within a second time period after that moment are collected. The first driving data includes at least one of a first acceleration, a first speed, and a first throttle opening. The second driving data includes at least one of a second acceleration, a second speed, and a second throttle opening.

[0067] Based on the first driving data, the second driving data, and the vehicle acceleration and / or rate of change of acceleration, collision information of the vehicle is obtained, including collision type and collision location;

[0068] Based on the collision information, control the vehicle's operation and / or send the collision information to a remote control center.

[0069] In this embodiment, the first time and the second time can be understood as stage time, meaning that both the first time and the second time are a certain time period. The corresponding first driving data is the average value of all first accelerations, all first speeds, and all first throttle openings collected by the sensor within the first time period, etc., while the second driving data is the average value of all second accelerations, all second speeds, and all second throttle openings collected by the sensor within the second time period, etc. For example, in this embodiment, the time period within 400ms before the collision occurs is set as the first time period, and the time period within 100ms after the collision occurs is set as the second time period. Then, when collecting driving data within the first time period, this embodiment is essentially collecting the average value of all driving data within 400ms before the collision occurs; and when collecting driving data within the second time period, it is essentially collecting the average value of all driving data within 100ms after the collision occurs. Optionally, in this embodiment, when collecting driving data in the first time period and the second time period, the first driving data is the peak value of all first accelerations, the peak value of all first speeds, and the peak value of all first throttle openings collected by the sensor in the first time period, and the second driving data is the peak value of all second accelerations, the peak value of all second speeds, and the peak value of all second throttle openings collected by the sensor in the second time period.

[0070] It should be noted that sensors located inside the vehicle typically collect vehicle driving data at fixed intervals. For example, if the collection period is set to 10ms, the sensor will collect vehicle driving data every 10ms, thus forming a dataset. In this embodiment, when selecting the first and second time periods, the selection interval and range of the first and second time periods can be appropriately expanded according to actual needs. Generally, the selection interval of the first and second time periods can be determined based on the sensor's collection frequency, and it needs to change with the data source's collection frequency. The specific threshold can be set according to the actual situation.

[0071] Furthermore, it should be noted that this embodiment performs collision detection for both stationary and moving vehicles. Generally, vehicle collisions occur at the front, rear, left, and right sides. When a vehicle is parked in a parking lot or on the roadside, it is highly susceptible to collisions from the front, rear, and sides due to its stationary state. When a vehicle is in motion, collisions typically occur when the vehicle brakes suddenly, causing a rear-end collision. Therefore, this embodiment detects two types of collisions: one is when a stationary vehicle encounters another object, and the other is when a moving vehicle brakes suddenly and is rear-ended by another vehicle. Either of these collision types constitutes a collision accident.

[0072] Specifically, in this embodiment, different judgment processes are performed based on whether the vehicle is stationary or in motion:

[0073] For the static state:

[0074] If the absolute value of the first speed is less than the first speed threshold, then the first driving state of the vehicle is determined to be a stationary state.

[0075] If it is determined that the first acceleration of the vehicle is less than a first acceleration threshold and the second acceleration is greater than the first acceleration threshold, then the throttle opening is detected.

[0076] If the first throttle opening and the second throttle opening are both less than the first throttle opening threshold, then the collision type is determined to be a stationary collision.

[0077] By analyzing the vehicle's acceleration at the first and second time points, it can be determined whether the vehicle exhibits a motion trend at either time point. For example, if the vehicle's acceleration is 0 at the first time point but suddenly reaches 3 m / s² at the second time point... 2 The acceleration change is significantly different from the normal acceleration process of a vehicle (during normal acceleration, the second acceleration at the second time point is still less than the first acceleration threshold), which initially suggests that a collision has occurred.

[0078] At this point, the collision information is determined by combining the changes in vehicle throttle. If both the first and second throttle openings are less than the first throttle opening threshold, the collision information is considered a stationary collision. If the vehicle experiences a significant change in acceleration, but analysis shows that the vehicle itself did not increase throttle in the first and second moments, a stationary collision can be confirmed. By analyzing the vehicle's acceleration and throttle opening values ​​while the vehicle is stationary, it can be determined whether a collision has occurred, achieving real-time collision detection even when the vehicle is stationary.

[0079] For the state of motion:

[0080] If the first speed is greater than the first speed threshold, then the first driving state of the vehicle is determined to be a moving state;

[0081] If the direction of the vehicle's first acceleration is opposite to the direction of its first velocity, and the first acceleration is greater than or equal to a second acceleration threshold, then the vehicle is determined to have undergone emergency braking. Analysis of the acceleration at the first instant reveals that the vehicle's acceleration and velocity are in different directions, indicating that the vehicle is performing a braking operation. Further analysis reveals that when the vehicle performs the braking operation, its acceleration is significantly higher than the acceleration under normal braking conditions, indicating that the vehicle underwent emergency braking at the first instant.

[0082] If the direction of the second acceleration of the vehicle is the same as the direction of the second speed, and the second acceleration is greater than or equal to the third acceleration threshold, then the throttle opening is detected. After the vehicle brakes suddenly, if the current vehicle is rear-ended, the direction of the current vehicle's acceleration will change due to the external force, and will generally be the same as the speed direction. Therefore, if the vehicle's acceleration changes in the second time while braking, it indicates that the vehicle is affected by an external force.

[0083] To more accurately determine whether a rear-end collision caused the change in vehicle acceleration, the throttle position is also considered. If both the first and second throttle openings are less than a second throttle opening threshold, the collision type is determined to be a rear-end collision. Generally, a significant change in the direction of acceleration between the first and second moments can be caused by two reasons: either the vehicle accelerates again after braking, or the vehicle is subjected to an external force after braking. Analyzing the vehicle's throttle position can determine if the vehicle itself accelerated. If both the first and second throttle openings are less than the second throttle opening threshold, it indicates that the vehicle did not accelerate, and the change in acceleration direction was due to an external force, i.e., a collision. Therefore, it can be determined that the vehicle experienced a rear-end collision after sudden braking.

[0084] It should be noted that when the vehicle's mass changes, the vehicle's acceleration will change. To ensure the accuracy of the setting values ​​of the first acceleration threshold, the second acceleration threshold, and the third acceleration threshold in this embodiment, when setting the first acceleration threshold, the second acceleration threshold, and the third acceleration threshold, the mass of the vehicle is first obtained. The mass of the vehicle includes its own mass and the mass of the internal transported objects. The first acceleration threshold, the second acceleration threshold, and the third acceleration threshold are obtained through the vehicle's mass and the preset correspondence.

[0085] Specifically, in this embodiment, the setting methods for the first acceleration threshold, the second acceleration threshold, and the third acceleration threshold can be as follows:

[0086] The first acceleration threshold, the second acceleration threshold, and the third acceleration threshold are tested in relation to the mass of the vehicle to obtain the set values ​​of the first acceleration threshold, the second acceleration threshold, and the third acceleration threshold for the vehicle under different masses.

[0087] The set values ​​of the first acceleration threshold, the second acceleration threshold, and the third acceleration threshold are curve-fitted with the vehicle mass to obtain the mass calibration curves of the first acceleration threshold, the second acceleration threshold, and the third acceleration threshold, respectively.

[0088] When comparing the first driving data and the second driving data, the mass of the vehicle is obtained, and the mass of the vehicle includes the mass of the vehicle itself and the mass of the objects transported inside.

[0089] Based on the mass calibration curves for the first, second, and third acceleration thresholds, the set values ​​for the first, second, and third acceleration thresholds are obtained according to the vehicle's mass. The acceleration thresholds corresponding to different masses can be obtained based on the mass calibration curves. During vehicle collision detection, the vehicle's own mass can be monitored and updated in real time to ensure the accurate setting of the corresponding acceleration thresholds, thereby guaranteeing accurate detection results.

[0090] Furthermore, in this embodiment, based on the determined collision type, the vehicle's driving state is controlled and / or the vehicle collision information is sent to a remote control center: For stationary collisions, the vehicle does not need to adjust its driving state; it only needs to send collision information such as the current vehicle model, location, surrounding image information, and collision type to the remote central control platform. For rear-end collisions involving sudden braking, the vehicle will immediately brake and stop, and then send collision information such as the current vehicle model, location, surrounding image information, and collision type to the remote central control platform, where the remote control center takes over the vehicle. Furthermore, for vehicles involved in collisions, manual takeover and remote control can be used to move the vehicle to a safe area for subsequent processing. For example, when a collision is detected in a parked vehicle, the vehicle will automatically send its current model, location, and surrounding image information to the remote control platform for storage. The platform will also alert the operator to take over remotely. Once the operator takes over, the vehicle's cameras and audio equipment will be automatically activated to receive images of the surrounding environment, allowing the operator to manually control the vehicle or wait for on-site personnel to respond. For a moving vehicle, once a collision is detected, it will immediately capture images of the surrounding environment and send an alarm to the remote control platform. After remote control is initiated, the operator will assess the situation and remotely control the vehicle to park on the side of the road, awaiting on-site personnel.

[0091] It should be further explained that, in this embodiment, after determining that the vehicle has experienced a stationary collision or a rear-end collision, the vehicle control sensing device collects information about the vehicle's surroundings within a third time period and determines whether a third vehicle exists; it analyzes the positions of the vehicle and the third vehicle, and obtains the first position of the collision based on these positions. It should be noted that the sensing device in this embodiment may include a camera, radar, infrared distance sensor, etc., and can identify the distance to objects around the vehicle using visual recognition technology, electromagnetic waves, or infrared technology. It should be noted that the third time period in this embodiment can be understood as a phase time, i.e., a certain time range. In this embodiment, when determining whether a third vehicle is around the vehicle within the third time period, all collected data within the third time period will be analyzed. For example, when the vehicle uses a camera to collect images, all images collected within the third time period will be analyzed. These images can be multiple static intermediate frames or continuous dynamic images.

[0092] Furthermore, in this embodiment, when the vehicle is involved in a stationary collision or a rear-end collision, the vehicle control sensing device determines whether there is a third vehicle around the vehicle within a third time period.

[0093] If there is a third vehicle around the vehicle that is less than a first threshold distance from the vehicle, then the stationary collision or rear-end collision is confirmed to have occurred, and the second location of the collision is obtained.

[0094] If no third vehicle exists within a distance of less than a first threshold distance from the vehicle, it is determined that the stationary or rear-end collision has not occurred, and an anomaly message is output, which is then stored or sent to a remote control center. In this embodiment, image information acquired by a camera is preferably used to verify the vehicle detection results to avoid anomalies during the detection process. Furthermore, in this embodiment, the vehicle collision detection results can also influence the route planning of surrounding vehicles. When a collision is detected, the vehicle transmits the collision information to the remote control center. The remote control center analyzes the driving routes and road conditions of surrounding vehicles based on the current vehicle collision location and collision information, and adjusts the driving routes of other vehicles related to the current vehicle's location to avoid overlapping routes and exacerbate road congestion.

[0095] Furthermore, in this embodiment, after a vehicle collision, some vehicles may retract after the collision, meaning that even if a collision occurs, no third vehicle is detected around the vehicle within the first distance threshold. To ensure detection accuracy, data from before the collision (i.e., the first moment) can be further analyzed, using images captured by a camera as an example:

[0096] If no third vehicle is found in the images captured by the camera of the vehicle within the third time period, and the distance between the vehicle and the third vehicle is less than the first threshold distance;

[0097] Then, extract the vehicle's surrounding images from the first time period temporarily stored in the memory, and analyze whether there is a third vehicle in the vehicle's surrounding images that is less than a first threshold distance from the vehicle.

[0098] If a third vehicle is present in the vehicle's surrounding image within the first time period, and the distance between the third vehicle and the vehicle is less than a first threshold distance, then it is determined that the vehicle has experienced a stationary collision or a rear-end collision.

[0099] Alternatively, this embodiment can analyze the images captured by the cameras at the first, second, and third time points:

[0100] The images of the vehicle's surroundings at the first and second time points, temporarily stored in memory, are compared with the image at the third time point. Other objects in the images at the first, second, and third time points whose distance from the vehicle is less than a second distance threshold are extracted. If any other objects with a distance less than the second threshold exist, the time-gap images between the first, second, and third time points are retrieved and analyzed. If any other objects in the time-gap images are found to be less than a first threshold distance from the vehicle, it is determined that the vehicle has experienced a stationary collision or a rear-end collision. In this embodiment, the distance between the vehicle and other surrounding objects is divided into two stages. Objects with a distance less than the first threshold distance are identified as collision targets. For objects with a distance greater than the first threshold distance but less than the second threshold distance, the time-gap images between the first, second, and third time points are further extracted. By reducing the time interval, it can be analyzed whether any objects with a distance less than the first threshold distance exist within the interval. If so, it is determined that the vehicle has experienced a stationary collision or a rear-end collision.

[0101] Furthermore, in this embodiment, the detection results can be verified by the orientation to eliminate abnormal results. When the vehicle experiences a stationary collision or a rear-end collision, the second acceleration is decomposed along the longitudinal and lateral directions of the vehicle to obtain the longitudinal component and the lateral component of the second acceleration. Based on the directions of the longitudinal and lateral components of the second acceleration, the third position of the collision is obtained. Based on the first orientation, the second orientation, and the third position, the true collision orientation is confirmed. Through the above detection process, this embodiment can actually obtain three positions: a first position, a second position, and a third position. Further analysis of the states of these three positions can verify the vehicle collision position: when the results of the first, second, and third positions are all the same, the true collision position of the vehicle can be further determined; however, if the first, second, and third positions are significantly different, it indicates that the collision is abnormal and requires further manual analysis. For example, if the third position is significantly different from the first and second positions, it indicates that other external force factors may have existed during the collision, such as two or more vehicles colliding, and some vehicles displacing excessively after the collision, causing them to be undetected during image analysis. Analyzing and verifying the three positions can further improve the accuracy of the detection results.

[0102] It should be further explained that in this embodiment, the detection results can also be verified by combining the movement state of objects inside the vehicle. At the same time, combining the movement state of objects inside the vehicle can also eliminate false judgments caused by collisions of objects inside the vehicle. Furthermore, when the sensor malfunctions, erroneous results can be eliminated by detecting the movement state of the transported object, thereby improving the accuracy and reliability of the detection.

[0103] Specifically, after the vehicle experiences a stationary collision or a rear-end collision, a first internal image and a second internal image of the vehicle's interior are extracted; the first internal image is the vehicle's interior image acquired at the first time, and the second internal image is the vehicle's interior image acquired at the second time.

[0104] By comparing the first internal image and the second internal image, it is determined whether there is a transported object moving relative to the carriage and colliding with the inner wall of the carriage.

[0105] If the object being transported inside the carriage moves relative to the carriage and collides with the inner wall of the carriage, and the direction of movement of the object relative to the carriage is opposite to the direction of the second velocity, in a stationary state, if the vehicle suddenly accelerates and displaces after being hit by the collision, the object being transported inside the vehicle will generate the opposite motion under the action of inertia, thereby confirming whether the vehicle has experienced a stationary collision.

[0106] If the object being transported inside the carriage moves relative to the carriage and collides with the inner wall of the carriage, and the direction of movement of the object relative to the carriage is opposite to the direction of the second velocity, in the motion state, if the vehicle is hit after sudden braking, the vehicle's acceleration will also change in a very short time. For example, if the rear of the vehicle is rear-ended, it will cause the vehicle to move towards the front of the vehicle in an instant. At this time, the object being transported inside the vehicle will move towards the rear due to inertia, thus confirming that a rear-end collision has occurred.

[0107] If an object transported inside the carriage moves relative to the carriage and collides with the carriage's inner wall, and the direction of the object's movement relative to the carriage is the same as the second velocity direction, when it is determined to be a rear-end collision, the vehicle control sensing device determines whether there is a third vehicle around the vehicle within a third time period. If there is a third vehicle around the vehicle at a distance less than a first threshold distance, the rear-end collision is confirmed. If there is no third vehicle around the vehicle at a distance less than the first threshold distance, it indicates that the detection result is abnormal, which can be identified in time to avoid false alarms. At the same time, it is determined that the rear-end collision did not occur, and abnormal information is output and stored or sent to the remote control center.

[0108] Combination Figure 2 As shown, this illustrates a data pattern of a vehicle after a collision while stationary. The line with larger fluctuations represents horizontal acceleration, the line with smaller fluctuations represents vertical acceleration, and the line representing throttle opening and speed is a single, overlapping straight line. From... Figure 2 As can be seen, when a vehicle collides while stationary, there are significant fluctuations in acceleration along both the horizontal and vertical axes. Since the vehicle is stationary at this time, both the throttle opening and speed are 0, especially the speed, which is 0. Therefore, based on the above data characteristics, a stationary collision is detected. When the vehicle is stationary, its initial state (non-collision state) has almost zero speed and acceleration. Once a collision occurs, the vehicle's acceleration will instantly reach a certain peak value. In this embodiment, the first acceleration threshold is set to 2 m / s². 2 When the vehicle acceleration is greater than 2m / s 2 At this point, a stationary collision can be initially suspected. Further analysis of the throttle opening reveals that the vehicle's throttle opening was consistently zero, meaning the vehicle did not accelerate at the time of the collision. This acceleration was caused by an external force, further confirming that the collision occurred while the vehicle was stationary.

[0109] It should be further explained that the collision occurring in this embodiment when the vehicle is stationary is not simply a longitudinal or lateral collision; it could also be a lateral collision, meaning the force generated by the collision has components in both the lateral and longitudinal axes. Typically, vehicle acceleration sensors collect acceleration data in three axes, especially the lateral and longitudinal axes. The vehicle acceleration can be obtained by synthesizing the accelerations in the lateral and longitudinal axes. Specifically, the steps for calculating the vehicle acceleration in this method are as follows: Based on the measured lateral and longitudinal accelerations, the vehicle acceleration is synthesized according to the following formula:

[0110]

[0111] Where acc_xy refers to the vehicle's acceleration, acc_x refers to the vehicle's lateral acceleration, and acc_y refers to the vehicle's longitudinal acceleration. The longitudinal acceleration refers to the component of acceleration parallel to the direction of travel, and the lateral acceleration refers to the component of acceleration perpendicular to the direction of travel. Figure 2 As shown, in this embodiment, the lateral acceleration refers to the acceleration in the horizontal axis direction, and the longitudinal acceleration refers to the acceleration in the vertical axis direction.

[0112] It should be emphasized that the above steps can also be used for collisions caused by vehicles changing lanes or turning. Since the speed of vehicles changing lanes or turning is relatively slow, the detection process in a stationary state can also be used for vehicles with a speed below a certain value in this embodiment. In the specific analysis process, speeds below a certain value can be regarded as speeds of 0.

[0113] Analyzing vehicle acceleration can provide a preliminary assessment of whether a collision has occurred. However, to confirm that the vehicle is stationary and that the sudden acceleration was not caused by a sudden acceleration, further verification based on the collision type is required. Therefore, this embodiment will determine whether the throttle opening of the first and second driving data meets a preset throttle opening threshold. When the throttle opening meets the threshold, it is determined that a stationary collision has occurred.

[0114] This embodiment can statistically analyze throttle opening data before and after a collision, for example, throttle opening data within 400ms before and 100ms after the collision. Generally, it is sufficient to statistically analyze the throttle opening data between the first and second driving data. In actual testing, the statistical range can be appropriately expanded for different vehicle models, so the specific statistical range is not limited here. For example, if the maximum throttle opening in the driving data is less than 15%, it is basically determined that the driver did not press the accelerator, and the collision is considered a stationary collision. It should be noted that the vehicle collision process is relatively complex. In actual situations, the interaction forces between vehicles usually fluctuate within a very short time. Therefore, even if there are multiple collisions within 3 seconds, this embodiment generally only considers one collision to have occurred within 3 seconds, thereby reducing the amount of data analysis.

[0115] Combination Figure 3 As shown, this illustrates a data pattern where a rear-end collision occurs within 2 seconds after a vehicle suddenly brakes. The top broken line represents throttle opening data, indicating that the throttle opening is 10% during normal driving, but drops to 0 instantaneously after sudden braking. The broken line passing through 0 represents longitudinal acceleration, showing that after sudden braking, the vehicle's acceleration instantaneously produces an acceleration of -6.5 m / s², which is opposite to the vehicle's velocity. 2 The vehicle was subsequently struck, and it experienced an acceleration of 2.2 m / s² in the same direction as its initial velocity. 2 The broken line between the throttle opening and the longitudinal acceleration represents the speed, indicating that after the vehicle brakes suddenly, the speed gradually changes from 3m / s to 0. Based on the above data characteristics, it can be detected that the vehicle was rear-ended during sudden braking.

[0116] Furthermore, regarding the emergency braking and rear-end collision scenario, most rear-end collisions occur in the longitudinal direction of the vehicle. Therefore, when the initial vehicle speed is not zero, the emergency braking and rear-end collision detection process is applied. When an emergency braking action occurs, the vehicle will generate acceleration in the opposite direction of the velocity. Since the braking performance of different vehicles varies, this second acceleration threshold needs to be set separately for different vehicle models. In this embodiment, the second acceleration threshold can be set to -3 m / s². 2 Within a certain timeframe after an emergency braking maneuver, if a vehicle is rear-ended, it will tend to move forward due to the impact's inertia. At this point, the vehicle's acceleration direction is the same as its velocity direction, and the acceleration will reach a first acceleration threshold. Experience shows that rear-end collisions generally occur in the longitudinal direction of the vehicle; this embodiment only considers acceleration data in the longitudinal axis direction. Based on the measured longitudinal acceleration, the vehicle's acceleration is synthesized using the following formula:

[0117]

[0118] Where acc_xy refers to the vehicle's acceleration, acc_x refers to the vehicle's lateral acceleration (where acc_x is 0), and acc_y refers to the vehicle's longitudinal acceleration.

[0119] Next, it is determined whether the throttle opening of the first and second driving data is less than the throttle opening threshold. If the throttle opening meets the threshold, it is determined that the vehicle was rear-ended during emergency braking. This embodiment can collect throttle opening data before and after the collision as needed, for example, throttle opening data within 400ms before and 100ms after the collision. Generally, it is sufficient to collect the throttle opening data of the first and second driving data. In actual testing, the statistical range can be appropriately expanded for different vehicle models, so the specific statistical range is not limited here. For example, if the throttle opening in the driving data is less than 15%, it is basically determined that the driver did not press the accelerator, and the collision is considered to be a rear-end collision due to emergency braking. It should be noted that this embodiment generally only considers a rear-end collision due to emergency braking to occur once within 2 seconds, thereby reducing the amount of data analysis.

[0120] Furthermore, this embodiment also discloses a vehicle collision detection system applying the above-described vehicle collision detection method, wherein the system includes at least:

[0121] The sensor module is used to collect vehicle driving data, which includes at least acceleration, vehicle speed, throttle opening, positioning information, radar information, and impact information.

[0122] The driving status judgment module is used to compare vehicle driving data and analyze vehicle driving status and collision information;

[0123] The control module is used to control the vehicle based on the results of the driving status judgment module and to remotely transmit collision information;

[0124] The remote control center is used to receive collision information transmitted by the control module and remotely control the vehicle's subsequent driving actions.

[0125] Furthermore, the sensor module described in this embodiment includes:

[0126] An accelerometer is used to collect vehicle acceleration data and transmit it to the driving state judgment module. In this embodiment, the vehicle acceleration is measured using a multi-axis inertial sensor. A multi-axis inertial sensor is a device with x, y, z-axis acceleration and angular velocity measurement functions, which can detect collisions from any position and direction of the vehicle to improve the measurement angle and measurement range. Since a multi-axis inertial sensor is used in this embodiment, the measured vehicle acceleration data includes lateral acceleration and longitudinal acceleration.

[0127] The vehicle speed sensor is used to collect vehicle speed data and transmit it to the driving status judgment module; for example, by monitoring the vehicle's current speed through the vehicle's speedometer.

[0128] A throttle opening sensor is used to collect vehicle throttle opening data and transmit it to the driving status judgment module. The throttle opening is generally measured by an electronic throttle sensor. Specifically, in traditional fuel vehicles, the amount of fuel entering the vehicle can be controlled by the accelerator pedal, and the electronic throttle sensor collects the vehicle's throttle opening and displays it through the vehicle's tachometer. For electric vehicles, the voltage is generally controlled by the accelerator pedal, which in turn controls the motor speed, and it can also be monitored through the vehicle's tachometer.

[0129] Positioning sensors are used to collect vehicle location information and transmit it to the control module;

[0130] The camera is used to collect image information around the vehicle and transmit it to the control module. Optionally, in this embodiment, the vehicle is equipped with cameras in the front, rear and left and right directions. When the vehicle is collided, the camera will automatically turn on and take pictures of the surrounding environment. By analyzing the photos or videos taken by the camera, on the one hand, accident evidence can be preserved to facilitate subsequent liability determination, and on the other hand, it can be compared with the aforementioned collision type to further confirm the collision.

[0131] Radar is used to collect distance information of obstacles around the vehicle and transmit it to the control module.

[0132] Furthermore, the system also includes a warning device connected to the control module. After receiving the result from the driving status judgment module, the control module sends a control signal to the warning device, causing it to issue warning signals to the surrounding area of ​​the vehicle. This warning module can use driving lights, headlights, or speakers. When a collision occurs, it will control the driving lights or headlights to turn on or flash, or it may use the speaker to play a warning recording to alert surrounding vehicles or pedestrians and prevent secondary injuries.

[0133] This disclosure also provides a computer-readable storage medium, wherein when the instructions in the computer-readable storage medium are executed by a processor of an electronic device, the electronic device is able to perform the methods disclosed in this disclosure.

[0134] The computer-readable storage medium in this disclosure can be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, apparatus, or device. The aforementioned computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specifically, the aforementioned computer-readable storage medium may include electrical connections based on one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0135] The aforementioned computer-readable medium may be included in the aforementioned electronic device; or it may exist independently and not assembled into the electronic device.

[0136] Furthermore, embodiments of this disclosure also provide an electronic device, including: at least one processor; a memory for storing processor-executable instructions; wherein the at least one processor is configured to execute the instructions to implement the methods disclosed in the embodiments of this disclosure.

[0137] The aforementioned processor can also be called a central processing unit (CPU), which can be an integrated circuit chip with signal processing capabilities. Each step in the method disclosed in this embodiment can be implemented by integrated logic circuits in the processor's hardware or by software instructions. The aforementioned processor can be a general-purpose processor, a digital signal processor (DSP), an ASIC, a field-programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. A general-purpose processor can be a microprocessor or any conventional processor. The steps of the method disclosed in this embodiment can be directly implemented by a hardware decoding processor, or implemented by a combination of hardware and software modules in the decoding processor. The software modules can be located in memory, such as random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, registers, or other mature storage media in the art. The processor reads information from the memory and, in conjunction with its hardware, completes the steps of the above method.

[0138] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the technical principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A vehicle collision detection method, characterized in that, Includes the following steps: The vehicle acceleration and / or rate of change of acceleration are acquired based on onboard IMU data. When the acceleration and / or rate of change of acceleration reaches a preset condition, first driving data within a first time period before that moment and second driving data within a second time period after that moment are collected. The first driving data includes a first acceleration, a first speed, and a first throttle opening. The second driving data includes a second acceleration, a second speed, and a second throttle opening. Based on the first driving data, the second driving data, and the vehicle's acceleration and / or rate of change of acceleration, collision information of the vehicle is obtained, including collision type and collision location. If the absolute value of the first speed is less than the first speed threshold, then the first driving state of the vehicle is determined to be a stationary state. If it is determined that the first acceleration of the vehicle is less than a first acceleration threshold and the second acceleration is greater than the first acceleration threshold, then the throttle opening is detected. If the first throttle opening and the second throttle opening are both less than the first throttle opening threshold, then the collision type is determined to be a stationary collision. If the first speed is greater than the first speed threshold, then the first driving state of the vehicle is determined to be a moving state; If the direction of the first acceleration of the vehicle is opposite to the direction of the first velocity, and the first acceleration is greater than or equal to the second acceleration threshold, then it is determined that the vehicle has undergone emergency braking. If the direction of the second acceleration of the vehicle is the same as the direction of the second velocity, and the second acceleration is greater than or equal to the third acceleration threshold, then the throttle opening is detected. If both the first throttle opening and the second throttle opening are less than the second throttle opening threshold, then the collision type is determined to be a rear-end collision. Based on the collision information, control the vehicle's operation and / or send the collision information to a remote control center.

2. The vehicle collision detection method according to claim 1, characterized in that, After determining that the vehicle has experienced a stationary collision or a rear-end collision, the vehicle control and sensing device determines whether there is a third vehicle around the vehicle within a third time period; and analyzes the positions of the vehicle and the third vehicle to obtain the first position of the collision based on the position.

3. The vehicle collision detection method according to claim 1, characterized in that, When the vehicle is involved in a stationary collision or a rear-end collision, the vehicle control sensing device determines whether there is a third vehicle around the vehicle within a third time period. If there is a third vehicle around the vehicle that is less than a first threshold distance from the vehicle, then the stationary collision or rear-end collision is confirmed to have occurred, and the second location of the collision is obtained. If there is no third vehicle within a distance of less than a first threshold distance around the vehicle, it is determined that the stationary or rear-end collision has not occurred, and an abnormal information is output, which is then stored or sent to the remote control center.

4. The vehicle collision detection method according to claim 2 or 3, characterized in that: If there is no third vehicle around the vehicle that is within a distance of less than a first threshold distance; The vehicle's surrounding images are extracted from the first time period temporarily stored in the memory, and it is analyzed whether there is a third vehicle in the vehicle's surrounding images that is less than a first threshold distance from the vehicle. If a third vehicle is present in the vehicle's surrounding image within the first time period, and the distance between the third vehicle and the vehicle is less than a first threshold distance, then it is determined that the vehicle has experienced a stationary collision or a rear-end collision.

5. The vehicle collision detection method according to claim 1, characterized in that: When setting the first acceleration threshold, the second acceleration threshold, and the third acceleration threshold, the mass of the vehicle is first obtained. The mass of the vehicle includes the mass of the vehicle itself and the mass of the objects transported inside. The first acceleration threshold, the second acceleration threshold, and the third acceleration threshold are obtained through the mass of the vehicle and a preset correspondence.

6. The vehicle collision detection method according to claim 1, characterized in that: Once it is determined that the vehicle has experienced a stationary collision or a rear-end collision, a first internal image and a second internal image of the vehicle's cargo box are extracted. The first internal image is a vehicle interior image acquired at the first time, and the second internal image is a vehicle interior image acquired at the second time. By comparing the first internal image and the second internal image, it is determined whether there is a transported object moving relative to the carriage and colliding with the inner wall of the carriage. If the object being transported inside the carriage moves relative to the carriage and collides with the inner wall of the carriage, and the direction of movement of the object relative to the carriage is opposite to the direction of the second velocity, the collision is confirmed as a stationary collision. If the object being transported inside the carriage moves relative to the carriage and collides with the inner wall of the carriage, and the direction of movement of the object relative to the carriage is opposite to the direction of the second velocity, the collision is confirmed as a rear-end collision. If an object being transported inside the carriage moves relative to the carriage and collides with the inner wall of the carriage, and the direction of movement of the object relative to the carriage is the same as the direction of the second velocity, when it is determined to be a rear-end collision, the vehicle control sensing device determines whether there is a third vehicle around the vehicle within a third time period. If there is a third vehicle around the vehicle that is less than a first threshold distance from the vehicle, then the rear-end collision is confirmed. If there is no third vehicle within a distance of less than a first threshold distance from the vehicle, it is determined that the rear-end collision did not occur, and an abnormal information is output, which is then stored or sent to a remote control center.

7. A vehicle collision detection device, characterized in that, The device employs the vehicle collision detection method as described in any one of claims 1-6, wherein the vehicle collision detection device comprises at least: The sensor module is used to collect the vehicle's driving data, which includes at least acceleration, vehicle speed, throttle opening, positioning information, and distance information. The driving status determination module is used to compare the driving data of the vehicle and analyze the driving status and collision information of the vehicle. The control module is used to control the vehicle based on the result of the driving status judgment module and to remotely transmit collision information; The remote control center is used to receive collision information transmitted by the control module and remotely control the subsequent driving actions of the vehicle.

8. An unmanned vehicle, characterized in that... The unmanned vehicle uses the vehicle collision detection method as described in any one of claims 1-6.

Images