Vehicle safety system and method implementing a weighted active-passive crash mode classification

Abstract

The invention relates to a vehicle safety system implementing a method for assisting in protecting vehicle occupants in the event of a frontal crash. The method comprises determining a passive safety crash mode classification in response to a crash signal received in response to an occurrence of a crash event. The method further comprises determining an active safety crash mode classification in response to an active safety signal received prior to the occurrence of the crash event. The method further comprises determining an active safety confidence factor for the active safety crash mode classification. The method further comprises determining a weighted crash mode classification to be the active safety crash mode classification in response to the active safety confidence factor exceeding a predetermined confidence value. The method further comprises determining the weighted crash mode classification to be the passive safety crash mode classification in response to the active safety confidence factor not exceeding the predetermined confidence value. The invention further relates to a vehicle safety system and a vehicle.

Description

Technical Field

[0001] This invention relates to a method for helping to protect vehicle occupants in the event of a frontal collision. Furthermore, this invention also relates to a vehicle safety system and a vehicle including such a system. Background Technology

[0002] Modern vehicles include various systems designed to help ensure occupant safety. These vehicle safety systems can include passive safety systems and / or active safety systems. Generally, passive safety systems are reactive systems that provide occupant protection in response to the detection of events requiring occupant protection, such as a vehicle collision. On the other hand, active safety systems attempt to anticipate events requiring occupant protection and take proactive avoidance measures.

[0003] Passive safety systems include one or more passive restraint devices, such as airbags and seatbelt retractors, which can be actuated to help protect vehicle occupants. These vehicle safety systems utilize an airbag control unit operatively connected to the airbags and various collision sensors, such as accelerometers and pressure sensors. In response to determining a collision scenario based on information provided by the collision sensors, the airbag control unit can be operated to deploy the airbags by activating an inflator that directs inflation fluid into the airbags. When inflated, the driver and front passenger airbags help protect occupants from impacts with vehicle components such as the dashboard and / or steering wheel.

[0004] Active safety systems utilize sensors such as cameras, radar, lidar, and ultrasonic transducers to determine the conditions surrounding the vehicle. In response to the sensed conditions, vehicle warning systems can provide visual, audible, and tactile warnings to the driver. This is the case, for example, in blind spot detection, lane departure warning, front / rear object detection, intersection traffic detection, and pedestrian detection. Active safety systems can also use sensed conditions to proactively actuate vehicle controls, such as adaptive cruise control, active braking, and active steering in response to lane departure detection. Each sensor used in an active safety system has its own specific advantages.

[0005] Cameras are highly effective at object detection. When positioned to view from several angles, they provide information to the vehicle, which the vehicle's safety system's artificial intelligence algorithms can use to detect external objects along the roadside, such as other vehicles, pedestrians, or objects (like trees or trash cans). Cameras can precisely measure angles, allowing vehicle safety systems to identify early on whether an approaching object will enter the vehicle's path. By combining long-range and short-range zoom with varying degrees of wide and narrow field of view, cameras become crucial tools for safety features such as collision avoidance, adaptive cruise control, automatic braking systems, and lane-keeping assist.

[0006] Radar sensors use an echo system to detect objects, which is advantageous in situations where visibility is poor and camera effectiveness is reduced. Radar sensors emit electromagnetic waves and receive the "echoes" reflected back from surrounding objects. Radar sensors are particularly effective at determining the distance and speed of objects such as vehicles and pedestrians relative to vehicles. Radar sensors function regardless of weather, lighting, or visibility conditions, making them ideal for distance control, collision warnings, blind spot detection, emergency braking, and more.

[0007] LiDAR sensors also employ the echo principle, using laser pulses instead of radio waves. LiDAR sensors record distance and relative speed with accuracy comparable to radar. Furthermore, LiDAR sensors can identify object types and angles between objects with even higher levels of accuracy. Therefore, LiDAR sensors can be used to effectively identify more complex traffic situations, even in darkness. Unlike cameras and radar sensors, the field of view is not critical for LiDAR sensors because they can record a 360-degree environment around a vehicle. High-resolution 3D solid-state LiDAR sensors can even render pedestrians and smaller objects in three dimensions. Summary of the Invention

[0008] This invention relates to a vehicle safety system comprising both active and passive components. In this specification, "active safety" refers to technologies that help prevent collisions (i.e., "collision avoidance"), and "passive safety" refers to vehicle components (such as airbags, seat belts, and the vehicle's physical structure (e.g., crash buffers) that help protect occupants in response to the detection of a collision.

[0009] Passive safety systems include one or more sensors (such as accelerometers and / or pressure sensors) configured to sense the occurrence of a collision event. A controller is configured to receive signals from the sensors, determine or identify the occurrence of a collision based on these signals, and deploy one or more actuable restraint devices (such as airbags and / or seatbelt pretensioners / retractors) in response to the sensed collision.

[0010] Active safety systems are designed to prevent or reduce the severity of vehicle collisions by using radar (all-weather), laser (LIDAR), cameras (using image recognition), or a combination thereof to detect impending collisions. In response to the detection of an impending collision, active safety systems provide collision avoidance features (such as operator warnings (visual, audible, tactile)) and active safety measures (such as automatic emergency braking and / or automatic emergency steering) to help avoid or mitigate the collision.

[0011] Active safety systems can be standalone systems or subsystems that utilize components of another system, such as a driver assistance system (DAS). This DAS uses camera, radar, and LiDAR data to provide driver assistance functions such as adaptive cruise control, lane departure warning, blind spot monitoring, and parking assist. These components can even be used to provide autonomous driving capabilities.

[0012] According to the present invention, information obtained from an active safety system is combined with information obtained from a passive safety system to improve the classification of frontal collisions by determining a weighted collision mode classification.

[0013] According to one aspect, a method for assisting in the protection of vehicle occupants in the event of a frontal collision includes determining a passive safety collision mode classification in response to a collision signal received in response to the occurrence of a collision event. The method further includes determining an active safety collision mode classification in response to an active safety signal received prior to the collision event. The method also includes determining an active safety confidence factor for the active safety collision mode classification. The method further includes determining a weighted collision mode classification as the active safety collision mode classification in response to the active safety confidence factor exceeding the predetermined confidence value. The method further includes determining the weighted collision mode classification as the passive safety collision mode classification in response to the active safety confidence factor not exceeding the predetermined confidence value.

[0014] According to another aspect, alone or in combination with any other aspect, the method may further include: determining the occurrence of a frontal collision in response to the collision signal; and actuating vehicle occupant protection devices according to the weighted collision mode classification.

[0015] According to another aspect, either alone or in combination with any other aspect, determining the active safety collision mode classification may include estimating collision characteristics in response to the active safety signal.

[0016] According to another aspect, alone or in combination with any other aspect, estimating the impact characteristics may include identifying an object in the field of view of the active safety sensor, and for that object: determining the time of impact between the vehicle and the object; determining the relative velocity between the vehicle and the object; and determining the overlap ratio between the vehicle and the object.

[0017] According to another aspect, either alone or in combination with any other aspect, identifying objects in the field of view of the active safety sensor may include determining the object closest to the vehicle.

[0018] According to another aspect, alone or in combination with any other aspect, determining the impact time may include estimating the minimum impact time and the maximum impact time using predetermined acceleration values ​​of the vehicle and deceleration values ​​of the target.

[0019] According to another aspect, alone or in combination with any other aspect, determining the relative speed may include estimating the minimum and maximum relative speeds using predetermined acceleration values ​​of the vehicle and deceleration values ​​of the target.

[0020] According to another aspect, alone or in combination with any other aspect, determining the overlap rate may include: determining the width of the vehicle and the width of the object; determining the lateral distance between the longitudinal centerlines of the vehicle and the object; and determining the overlap as the sum of half the width of the vehicle, half the width of the object, and the lateral distance between the longitudinal centerlines of the vehicle and the object.

[0021] According to another aspect, determining the lateral distance between the longitudinal centerlines of the vehicle and the object, alone or in combination with any other aspect, may include estimating the minimum and maximum lateral distance between the longitudinal centerlines of the vehicle and the object.

[0022] According to another aspect, alone or in combination with any other aspect, estimating the minimum and maximum lateral distances between the longitudinal centerlines of the vehicle and the object may include estimating the variation in lateral distance based on the vehicle's speed, steering angle, and yaw rate.

[0023] According to another aspect, either alone or in combination with any other aspect, determining the overlap rate may include identifying the impact side of the vehicle as the left / driver's side or the right / passenger side of the vehicle.

[0024] According to another aspect, alone or in combination with any other aspect, the method may further include determining whether a collision is imminent in response to the collision time being less than a threshold.

[0025] According to another aspect, alone or in combination with any other aspect, the method may further include determining the active safety collision mode classification in response to determining an impending collision, the impact side of the vehicle, overlap classification, and speed classification.

[0026] According to another aspect, alone or in combination with any other aspect, the overlap classification may include one of asymmetric classification, symmetric (full overlap) classification, biased deformable obstacle (ODB) classification, and small overlap classification.

[0027] According to another aspect, alone or in combination with any other aspect, the speed classification may include one of a high-speed classification and a low-speed classification.

[0028] According to another aspect, either alone or in combination with any other aspect, determining the active safety collision mode classification may include determining at least one of the following in response to an active safety signal: the longitudinal distance between the vehicle and the object, the lateral distance between the centerlines of the vehicle and the object, the travel distance between the vehicle and the object, the approach angle between the vehicle and the object, and the speed of the vehicle relative to the object.

[0029] According to another aspect, alone or in combination with any other aspect, a vehicle safety system for helping to protect vehicle occupants in the event of a frontal collision may include an actuable safety device and a controller for controlling the actuation of the safety device according to the foregoing method.

[0030] According to another aspect, alone or in combination with any other aspect, the vehicle safety system may include: one or more passive sensors and one or more active sensors for providing the collision signal to the controller, the one or more active sensors for sensing objects in the vehicle's path and providing the active safety signal to the controller.

[0031] According to another aspect, alone or in combination with any other aspect, the one or more active sensors may include at least one of a camera, a radar sensor, and a lidar (LIDAR) sensor.

[0032] According to another aspect, alone or in combination with any other aspect, the controller may include an airbag controller unit (ACU).

[0033] According to another aspect, alone or in combination with any other aspect, the vehicle may include the vehicle safety system. Attached Figure Description

[0034] Figure 1 This is a schematic illustration of a vehicle including a vehicle safety system, based on an example configuration.

[0035] Figure 2 It is a schematic diagram of the vehicle relative to an object, and shows various parameters that can be sensed by the active safety system portion of the vehicle's safety system.

[0036] Figures 3 to 15 It is a schematic diagram illustrating the control algorithm implemented in a vehicle safety system. Detailed Implementation

[0037] In this manual, the left and right sides of the vehicle are sometimes referred to. These references should be understood as referring to the forward direction of the vehicle's movement. Therefore, a reference to the "left" side of the vehicle corresponds to the driver's side ("DS"). A reference to the "right" side of the vehicle corresponds to the passenger side ("PS").

[0038] Similarly, this specification provides some description of the vehicle's axes, specifically the X-axis, Y-axis, and Z-axis. The X-axis is the longitudinally extending central axis of the vehicle. The Y-axis is the laterally extending axis of the vehicle, perpendicular to the X-axis. The Z-axis is the vertically extending axis of the vehicle, perpendicular to both the X-axis and Y-axis. The X-axis, Y-axis, and Z-axis intersect at or near the vehicle's center of gravity ("COG"). Vehicle safety systems

[0039] refer to Figure 1 For example, vehicle 12 includes vehicle safety system 10, which includes passive safety system 20 and active safety system 100. Passive safety system 20 includes actuable occupant protection devices, which are schematically shown as 14. Protection device 14 may include any actuable occupant protection device, such as front airbags, side airbags, curtain airbags, knee pad airbags, and actuable seat belt pretensioners and / or retractors. Passive safety system 20 also includes an airbag electronic control unit (referred to herein as an airbag controller unit or “ACU”) 50 operatively connected to protection device 14. ACU 50 is operable to control actuation of protection device 14 in response to vehicle conditions sensed via one or more sensors operatively connected to ACU.

[0040] Passive safety system 20 includes several sensors (such as accelerometers and / or pressure sensors) for measuring certain conditions of vehicle 12, and determining whether to actuate vehicle occupant protection device 14 based on these conditions. These sensors may be installed at various locations throughout vehicle 12, selected to allow sensing of the specific vehicle conditions for which the sensors are intended to operate. In this specification, vehicle safety system 10 is described as including several different types and locations of collision sensors in vehicle 12. The collision sensors described herein are not necessarily a complete list of sensors included in vehicle safety system 10; these sensors are only those used by the present invention to detect the occurrence of a frontal impact. Therefore, those skilled in the art will understand that vehicle safety system 10 may include one or more other collision sensors of any type, any number, and any location in vehicle 12.

[0041] Passive safety system 20 is configured to detect a frontal vehicle impact using left crush zone sensor 60 and right crush zone sensor 62. Left crush zone sensor 60 and right crush zone sensor 62 are accelerometers configured to sense vehicle acceleration and transmit signals indicating that acceleration to ACU 50. ACU 50 is configured to determine whether the magnitude of the sensed acceleration meets or exceeds a threshold sufficient to indicate that a frontal collision event has occurred, and in response to this determination, actuates protective device 14.

[0042] exist Figure 1 In the compression zone, sensors 60 and 62 are uniaxial accelerometers, which are configured to detect compression in a region parallel to the longitudinal axis X. VEH The accelerations in the directions indicated are generally indicated by arrows LT_CZS and RT_CZS, respectively, shown in the schematic representation of the sensors. The left compression zone sensor 60 and the right compression zone sensor 62 are located at or near the left driver's side (DS) and right passenger side (PS) front corners of the vehicle 12, respectively. The left compression zone sensor 60 and the right compression zone sensor 62 may, for example, be mounted behind the front bumper 16 of the vehicle at these front corner locations. The ACU 50 includes an integrated 2-axis accelerometer 52 for sensing vehicle acceleration along the X and Y axes. These accelerations are indicated by CCU_1X and CCU_1Y, respectively.

[0043] Vehicle safety system 10 is implemented and configured to cooperate with other vehicle systems. For example, ACU 50 can be operatively connected to Body Control Module (BCM) 30 via a Vehicle Controller Area Network (CAN) bus. BCM 30 can communicate with other vehicle systems via the CAN bus, such as chassis control, stability control, traction / slip control, anti-lock braking system (ABS), tire pressure monitoring system (TPMS), navigation system, instrument cluster (speed, throttle position, brake pedal position, etc.), infotainment (“infotainment”) system, and other systems. Through these interfaces, ACU 50 can communicate with any of these external systems to provide and / or receive data.

[0044] Still refer to Figure 1The active safety system 100 may have a known configuration, including one or more active safety system components configured to provide active safety functions in a known manner. The active safety system 100 may utilize components of driver assistance systems (DAS), which, as the name suggests, provide assistance to the vehicle operator while driving. These components can help provide DAS functions such as the aforementioned active cruise control, lane departure warning, blind spot monitoring, parking assist, etc. These components can even be components used to provide autonomous driving functions, and therefore can use artificial intelligence (AI) and other machine learning techniques to provide a wealth of information about the vehicle's surroundings. For collision avoidance functions, the active safety system can provide collision warnings (auditory, visual, tactile), automatic emergency braking, and automatic emergency steering.

[0045] The active safety system 100 may include different components. Figure 1 In the example configuration, the active safety system 100 includes camera sensors, radar sensors, and lidar (LIDAR) sensors. The camera sensor 110 is mounted high on the windshield 18 in a forward-facing manner, for example, behind or in the area of ​​a rearview mirror. Multiple radar sensors 120 may be mounted forward, in the area of ​​the bumper 16, for example, in the vehicle grille. The lidar (LIDAR) sensor 130 may be mounted on or near the vehicle roof 22.

[0046] Camera sensor 110 is effective in providing a wide field of view and has the ability to identify various objects / obstacles with high accuracy. The camera can also determine whether an object / obstacle is in the path of vehicle 12. However, the camera requires good visibility and deteriorates in dark conditions, fog, rain, snow, etc. Radar sensor 120 does not deteriorate in poor visibility conditions and does provide accurate indications of time-of-collision (TTC). However, radar sensor 120 is less capable of distinguishing different types of objects / obstacles and is not as good as the camera in determining whether an object / obstacle is in the path of vehicle 12. LiDAR sensor 130 provides 3D sensing capabilities for TTC and vehicle path determination, offers good object / obstacle recognition, and is robust in both good and poor visibility conditions.

[0047] Camera 110, radar sensor 120, and LIDAR sensor 130 can be connected to a separate controller, such as DAS controller 140, which can communicate with ACU 50 via a CAN bus. Alternatively, both active and passive safety functions can be handled by a single controller, such as ACU 50, in which case camera 110, radar sensor 120, and LIDAR sensor 130 can be directly connected to ACU 50. These sensors monitor the area in front of vehicle 12 within the vehicle's predetermined field of vision and travel distance.

[0048] Active safety system sensors provide information (signals, data, etc.) that a controller (such as the ACU 50, DAS controller 140, or other controllers) can use to detect the presence of objects in the vehicle's path. By implementing known methods (such as artificial intelligence (AI) and other algorithms), the controller can determine information related to the detected object, such as the object type, distance from the vehicle, lateral position in the vehicle's path, time of impact with the vehicle, relative speed to the vehicle, the object's state (e.g., facing forward, facing backward, facing sideways, moving, stationary, etc.), and the probability of a collision. Parameters sensed by active safety systems

[0049] Figure 2 Some of the parameters that can be sensed by the active safety system 100 are shown. Figure 2 The parameters shown are those associated with object 24 and its position relative to vehicle 12. These parameters are sensed relative to an origin located at the front of vehicle 12 and centered on the vehicle's longitudinal axis. Active sensors may not be located at the origin. This is possible, for example, if the active sensor is a camera mounted on a rearview mirror.

[0050] Active sensors provide a field of view relative to the vehicle. Within this field of view, the active sensors can detect the presence of objects and provide parameters associated with those objects. These parameters include the longitudinal distance between the object and the vehicle's origin, and the lateral distance between the object and the vehicle's longitudinal axis. The object's range is the straight-line distance from the vehicle's origin to the object's centerline. When the object deviates from the vehicle's longitudinal axis, this range extends at an angle relative to the vehicle's longitudinal axis. The relative speed between the vehicle and the object is measured along this range.

[0051] Figures 3 to 15Control algorithms implemented by vehicle safety system 10 to aid in the protection of vehicle occupants in the event of a frontal collision (referred to herein as a frontal collision) with vehicle 12 are illustrated. These algorithms are implemented in a vehicle controller (such as ACU 50), which is operatively connected to safety device 14 and configured to actuate the safety device in response to the detection of a frontal collision. According to the invention, the control algorithms implemented in vehicle safety system 10 are configured to cause passive safety system 20 to adjust or regulate its response to a frontal collision based on information obtained from active safety system 100. Overview of Control Algorithms

[0052] Figure 3 An overview of a control algorithm 150 implemented by vehicle safety system 10 to help protect vehicle occupants in response to the detection of a frontal collision is presented. Control algorithm 150 implements a novel method for classifying detected frontal collisions. Control algorithm 150 is configured to determine a weighted collision mode classification flag 376 in response to the detection of a frontal vehicle collision. The weighted collision mode classification flag 376 is a frontal collision mode determination (i.e., symmetrical, asymmetrical, offset deformable barrier (ODB), small overlap, etc.) using information obtained from both active safety system 100 and passive safety system 20. As discussed herein, control algorithm 150 implements a unique weighting algorithm that combines active and passive collision mode determinations to determine the weighted collision mode classification.

[0053] like Figure 3 As shown, control algorithm 150 utilizes active safety signals 152 from active safety system 100. Control algorithm 150 also utilizes passive safety system collision discrimination signals 160, specifically frontal collision indication sign 162 and collision mode classification sign 164. Preset algorithm 160 includes collision estimation and classification algorithm 200 and weighted collision mode decision algorithm 390, which generate weighted collision mode classification sign 396. Weighted collision mode classification sign 396 is provided to frontal algorithm 400, which may also be informed by lateral algorithm 420. Frontal algorithm 400 uses weighted collision mode classification sign 396 to determine thresholds, misuse boxes, and delays, which are implemented to determine whether to deploy safety device 14 in response to a classified collision event. Overview of Collision Estimation and Classification Algorithms

[0054] Figure 4 An overview of the collision estimation and classification algorithm 200 portion of the control algorithm 150 implemented by the vehicle safety system 10 is shown. Figure 4As shown, active safety signal 152 is provided to active safety system signal converter 210. The converted active safety signal is provided to target tracking algorithm 230, which tracks the position of target 24 relative to vehicle 12 and generates an output for collision estimation algorithm 250. Vehicle signal 154 is provided to vehicle signal converter 220. The converted vehicle signal is provided to collision estimation algorithm 250. Frontal collision indication sign 162 is provided to frontal collision detection algorithm 240, and the output of frontal collision detection algorithm is also provided to collision estimation algorithm.

[0055] The frontal collision indication sign 162 can be obtained from a plurality of frontal collision discrimination algorithms implemented by the vehicle safety system 10. The frontal collision discrimination algorithms implemented by the vehicle safety system 10 can be, for example, one or more algorithms disclosed in U.S. Patent No. 9,650,006B2 to Foo et al., the disclosure of which is incorporated herein by reference in its entirety. Therefore, it should be understood that the vehicle safety system 10 may include components or portions thereof disclosed in the aforementioned U.S. Patent No. 9,650,006B2 to Foo et al.

[0056] The collision estimation algorithm 250 estimates the characteristics of a collision based on information obtained from the target tracking algorithm 230, the vehicle signal converter 220, and the frontal collision detection algorithm 240. The collision estimation algorithm 250 provides these estimated characteristics to the active safety collision mode classification algorithm 360, which classifies frontal collisions and provides an active safety collision mode classification flag 366 indicating the determined collision mode classification. Active safety signal converter

[0057] Figure 5 The diagram illustrates an active safety signal converter 200. The active safety signal converter 200 transforms vehicle- or platform-specific active safety signals 152, enabling collision estimation and collision mode classification functions to be performed using normalized values. This may include converting the vehicle coordinate system to the algorithm coordinate system, scaling, labeling (+ / -), signal range, units, etc. For example, some active safety systems use positive and negative values ​​(+ / -) to indicate values ​​on the driver's side / passenger side. This may be the case, for example, for lateral values ​​such as distance, speed, acceleration, angle, etc. However, depending on the vehicle platform, which side is positive (+) and which side is negative (-) may differ. The active safety signal converter 200 transforms these values ​​so that they are normalized and consistently follow the same rules to ensure the accuracy of the control algorithm 150.

[0058] Additionally, some vehicle platforms may not directly provide all the active safety system signals 152 required to implement control algorithm 150. In this example, active safety system converter 200 can be used to calculate the missing signals / values.

[0059] Figure 5 Some of the active safety signals 152 that can be converted by the active safety signal converter 210 are shown. These signals may include: number of objects, object ID, object category, longitudinal distance of objects, lateral distance of objects, relative velocity of objects, object angle, object angular rate, object width, and the state of the active safety system 100. Furthermore, as... Figure 5 As shown, the active safety signal converter 210 converts the active safety signal 152 to provide a corresponding signal that has been modulated, normalized, and standardized to comply with the rules desired by the control algorithm 150. Target tracking

[0060] Figure 6 The image shows target tracking algorithm 230. For example... Figure 6 As shown, the target tracking algorithm 230 receives converted active safety signals from the active safety signal converter 200. As illustrated, these active safety signals may include: number of objects, object ID, object classification, longitudinal distance of objects, lateral distance of objects, relative velocity of objects, object angle, object angular rate, object width, and the state of the active safety system 100.

[0061] When there is more than one object in the vehicle path, the target tracking algorithm 230 uses this information to perform nearest object calculation 232 to identify the nearest object in the vehicle path. For objects identified as the nearest, such as... Figure 6 As shown on the right, target tracking algorithm 230 can determine the following characteristics generally identified at 236: longitudinal acceleration, longitudinal distance, lateral distance, relative velocity, angle, angular rate, and object width. Target tracking algorithm 230 can also monitor and provide the status of the active safety system to ensure that the data used to track the identified object is up-to-date and accurate.

[0062] The target tracking algorithm 230 also includes an inference trigger 234 configured to calculate these characteristics when one or more of the aforementioned characteristics 236 from the vehicle signals cannot be directly obtained via the active safety sensors. This might be the case, for example, when an object is outside the field of view of the active safety system and less than a minimum distance from the vehicle / sensor. In other words, when it is determined that an object is so close to the vehicle that it is or may be outside the range of the active sensors, the inference trigger 234 will trigger the inference of characteristic 236. In this example, the inference trigger 234 may trigger the calculation of characteristic 236 of the object relative to the vehicle based on the vehicle signals. This vehicle signal converter

[0063] Figure 7 The image illustrates a vehicle signal converter 220. This vehicle signal converter 220 transforms vehicle signals 154 specific to a vehicle or platform, enabling collision estimation and collision mode classification functions to be performed using normalized values. This may include converting the vehicle coordinate system to the algorithm coordinate system, scaling values, markings (+ / -), signal ranges, units, etc. For example, some active safety systems use positive and negative values ​​(+ / -) to indicate values ​​on the driver's side / passenger side. This may be the case, for example, for lateral values ​​such as distance, speed, acceleration, angle, etc. However, depending on the vehicle platform, which side is positive (+) and which side is negative (-) may differ. This vehicle signal converter 220 transforms the vehicle signals 154 so that these signals are normalized and consistently follow the same rules to ensure the accuracy of the control algorithm 150.

[0064] Additionally, some vehicle platforms may not directly provide all the vehicle signals 154 required to implement control algorithm 150. In this example, the vehicle system converter 220 can be used to calculate the missing signals / values.

[0065] Figure 7 This demonstrates some of the vehicle signals 154 that can be converted by the vehicle signal converter 220. These signals may include: vehicle longitudinal velocity, vehicle lateral velocity, vehicle longitudinal acceleration, vehicle lateral acceleration, yaw rate, and steering angle. Furthermore, as... Figure 7 As shown, the vehicle signal converter 220 converts the vehicle signal 154 to provide a corresponding converted vehicle signal 222 that has been modulated, normalized, and standardized to comply with the rules desired by the control algorithm 150. Frontal collision detection algorithm

[0066] Figure 8The diagram illustrates a frontal collision detection algorithm 240. The frontal collision detection algorithm 240 is configured to provide a frontal collision sensing signal 242 in response to a collision mode classification flag 164 and a frontal collision indication flag 162 received from the passive safety system 20. Therefore, the frontal collision sensing signal 242 is an indication from the passive safety system 20 of: 1) a frontal collision has occurred, and 2) the type of frontal collision determined by the passive safety system.

[0067] As described above, the determination of the passive safety collision mode classification can be similar to or the same as that in U.S. Patent No. 9,650,006B2 to Foo et al. The classification of the collision mode classification flag 164 can include any one or more of the following classifications, each of which can have individually configurable and / or adjustable thresholds. Classifications can include, for example, fully overlapping symmetric, left / right (L / R) asymmetric, L / R small overlap, L / R low-speed angle / tilt, L / R high-speed angle / tilt, L / R low-speed offset deformable barrier (ODB), L / R high-speed ODB, and L / R offset moving deformable barrier (OMDB). The frontal impact indication flag 162 is a sensor signal indicating the occurrence of a frontal collision, such as a left crush zone sensor and / or a right crush zone sensor. These signals can be, for example, a CZS_3X signal from LT_CZS 60 or a CZS_4X signal from RT_CZS 62 (see [link to relevant documentation]). Figure 1 ). Collision estimation algorithm

[0068] Figure 9 The collision estimation algorithm 250 is shown in the example. Figure 9 As shown, the collision estimation algorithm 250 includes a time-to-collision (TTC) estimation algorithm 260, a relative velocity estimation algorithm 270, an overlap rate estimation algorithm 300, and a collision data identification algorithm 340. The algorithm, uniformly referred to as 252, performs calculations based on the active safety signal 212, the calculation trigger 238, the vehicle signal 222, and the frontal collision sensing indication 242. Algorithm 252 generates the TTC. min 274, TTC max 276. Relative velocity min 278. Relative velocity max 280. Overlap Rate min 312. Overlap Rate max Calculated values ​​for 314, the impact side 308, the impending impact 346, and the assessed impact data 350. The impact estimation algorithm 252 is discussed in detail in the following paragraphs. TTC and relative velocity estimation algorithm

[0069] Figure 10The diagram illustrates TTC estimation algorithm 260 and relative velocity estimation algorithm 270. (For example...) Figure 10 As shown, the TTC estimation algorithm 260 at point 262 is based on the relative velocity between the detected object 24 and the vehicle 12 (see [reference]). Figure 2 The system performs calculations based on the detected longitudinal distance between the object and the vehicle. Relative speed and longitudinal distance values ​​are obtained from the active safety signal 212. As shown, the TTC calculation result is obtained by dividing the longitudinal distance by the relative speed.

[0070] For both TTC and relative velocity, obtain the minimum and maximum values ​​(min / max). When the object is within the field of view of the active safety system (calculated trigger 238 = off), obtain the minimum and maximum values ​​(TTC). min / max V min / max Same as above. When the object is outside the field of view (calculation trigger 238 = on), use a calibrable minimum / maximum target deceleration level (target_deceleration). min and target deceleration max The minimum and maximum relative velocities are estimated using the vehicle's longitudinal acceleration values ​​(from vehicle signal 222) and the vehicle's own longitudinal acceleration values. This is specifically shown in box 272 within the relative velocity estimation algorithm 270, where: V min =V min -(Target_Deceleration) max +(machine longitudinal acceleration)*ΔT; and V max =V max +(Target_Deceleration) min +Machine_Longitudinal_Acceleration)*ΔT

[0071] Furthermore, when the object is outside the field of view, the estimated relative velocity is used to estimate the TTC. min and TTC max This is also shown in box 272, where: TTC min = (Vertical Distance - V) max *ΔT) / V max ;as well as TTC max = (Vertical Distance - V) min *ΔT) / V min like Figure 10 As shown, TTC estimation algorithm 260 and relative velocity estimation algorithm 270 generate TTC. min 274, TTC max 276. (Relative) V min 278 and (relative) V max The value is 280. Overlap rate estimation algorithm

[0072] Figure 11 The overlap rate estimation algorithm 300 is shown in the figure. Figure 11 As shown, the overlap rate estimation algorithm 300 performs calculations based on the active safety signal 212 and the vehicle signal 222. The active safety signal includes the target (object) width, the vehicle width, and the lateral distance. The vehicle signal includes the vehicle speed, the vehicle yaw rate, and the vehicle steering angle. A calculation trigger 238 is also utilized.

[0073] At box 302, the minimum and maximum lateral distances (min / max) between the vehicle 12 and the target object 24 are determined. When the target object is within the field of view of the active safety system 100, i.e., when the calculation trigger 238 is off, the minimum / maximum lateral distances are the same and equal to the lateral distance determined by the active safety system 100 (from the active safety signal 212). When the object is outside the field of view (when the calculation trigger 238 is on), the minimum and maximum lateral distances are estimated as follows: Horizontal distance min = Lateral distance - Δ lateral distance; and Horizontal distance max = Horizontal distance + Δ Horizontal distance; Specifically, the Δ lateral distance is calculated at box 304. The Δ lateral distance is the change in lateral distance between the vehicle and the target object caused by steering, and is calculated based on vehicle signal 222 (i.e., steering angle, yaw rate, and speed). Δ Lateral distance = f(steering angle, yaw rate, velocity).

[0074] At box 306, the overlap between the vehicle and the target object is calculated. More specifically, the minimum and maximum lateral distances calculated at box 302 are used to calculate the minimum and maximum left and right overlaps, as follows: Left overlap min =0.5*(HW+TW)-Horizontal Distance min ; Right overlap min =0.5*(HW+TW)+Horizontal Distance min ; Left overlap max =0.5*(HW+TW)-Horizontal Distance max ;as well as Right overlap max =0.5*(HW+TW)+Horizontal Distance max ; Where HW = local width, and TW = target width from active safety signal 212.

[0075] Based on calculations performed at box 306, the impact side 308 is determined based on the overlap markers, where a positive overlap value indicates left / driver's side overlap and a negative overlap value indicates right / passenger side overlap. Of course, this + / - rule can be reversed. This includes active safety signal converters (…). Figure 5 ) and the vehicle signal converter ( Figure 7 Examples of reasons that may be important, as they help maintain the fidelity of this and other similar rules.

[0076] At box 310, the minimum overlap value and the maximum overlap value are used to calculate the minimum overlap rate 312 and the maximum overlap rate 314, as shown below: Overlap rate min =100 * overlap min / HW; Overlap rate max =100 * overlap max / HW. Impact data identification and inspection

[0077] Figure 12 The collision data identification algorithm 340 of the collision estimation algorithm 252 is shown. The collision data identification algorithm 340 uses the frontal collision sensing flag 242 to identify active safety signals at the time of impact. At box 342, a check is performed to determine whether a collision is imminent. If the time to collision (TTC) is within a predetermined range, a collision is imminent. As shown, if the TTC... max 276 is less than the TTC threshold (which is configurable / adjustable) and TTC min If the value is zero, a collision is determined to occur (box 346). At box 344, the collision imminent indication 346 can be time-locked. At box 348, if the collision imminent AND frontal collision sensing flag 242 is triggered, the collision data is qualified (box 350). Active collision mode classification

[0078] refer to Figure 13 The Active Safety Collision Pattern Classification Algorithm 360 uses information obtained from the Active Safety System 100 to classify collisions. The Active Collision Pattern Classification Algorithm 360 is used when the collision data is acceptable (see...). Figure 12 This is based on ) and is carried out on the basis of ). Figure 13 As shown, the active collision pattern classification algorithm 360 utilizes the impact side 308 (see...). Figure 11 ), Maximum overlap rate 314 (see Figure 11 ) and maximum relative speed 280 (see Figure 10 To classify collisions.

[0079] The active collision pattern classification algorithm 360 implements an overlap rate threshold metric 362, which evaluates a maximum overlap rate 314 to classify collisions as symmetrical, offset deformable obstacle (ODB), or small overlap, and provides an output indicating the type of overlap after classification. The overlap threshold implemented in metric 362 can be configurable or adjustable to define different collision types in terms of overlap. The active collision pattern classification algorithm 360 also implements a relative velocity threshold metric 364, which evaluates a maximum relative velocity 280 to classify collisions as high-speed or low-speed. The velocity threshold implemented in metric 364 can be configurable or adjustable to define different collision types in terms of velocity.

[0080] like Figure 13 As shown, the active collision pattern classification algorithm 360 implements Boolean logic to classify collisions using information obtained from active safety signals and determinations made through metrics 362 and 364. The following table illustrates these classifications:

[0081] As shown in the table above, when the overlap rate threshold metric 362 indicates a symmetrical collision but does not indicate the impact side and is independent of vehicle speed, it indicates a fully overlapping symmetrical collision. When the impact side is indicated as left or right, but the overlap type is not categorized and is independent of vehicle speed, it indicates a left asymmetrical collision or a right asymmetrical collision. When the impact side is left or right, the overlap metric indicates ODB, and the speed metric indicates low speed, it indicates a left low-speed ODB collision or a right low-speed ODB collision. When the impact side is left or right, the overlap metric indicates ODB, and the speed metric indicates high speed, it indicates a left high-speed ODB collision or a right high-speed ODB collision. When the impact side is left or right, and the overlap metric indicates small overlap, it indicates a left small overlap collision or a right small overlap collision. Active safety confidence factor

[0082] refer to Figure 14 The Active Safety Confidence Factor Determination Algorithm 370 determines the Active Safety Confidence Factor (ASCF) 382. The Active Safety Confidence Factor 380 is a measure of confidence in the Active Safety Collision Mode Classification Flag 366 based on information obtained from the Active Safety System 100 prior to a collision event. Because the Active Safety Collision Mode Classification Flag 366 is based on an estimated classification, the Active Safety Confidence Factor 380 provides an indication of the likelihood that the estimated collision mode is correctly classified.

[0083] like Figure 14As shown, the active safety confidence factor determination algorithm 370 includes an overlapping uncertainty function box 372, which is based on the... Figure 11 The overlap estimation algorithm 300 determines the minimum / maximum values—overlap_rate_Min 312 and overlap_rate_Max 314—to determine the overlap rate uncertainty factor 374. The overlap rate uncertainty factor 374 can be, for example, a value in the range of zero to one (0-1), where zero indicates the minimum uncertainty and one indicates the maximum uncertainty.

[0084] The overlap uncertainty function 372 implemented at box 372 can be implemented in various ways. For example, the overlap uncertainty function box 372 can determine the overlap uncertainty factor 374 based on the range or increment between the minimum overlap value 312 and the maximum overlap value 314. In this example, the overlap uncertainty factor 374 may increase proportionally to the range / increment between the minimum value 312 and the maximum value 314 (i.e., the uncertainty may increase). Therefore, when the minimum / maximum range is small, the uncertainty is low, and the overlap uncertainty factor 374 is correspondingly low. Conversely, when the minimum / maximum range is large, the uncertainty is high, and the overlap uncertainty factor 374 is correspondingly high.

[0085] Moreover, as Figure 14 As shown, the active safety confidence factor determination algorithm 370 also includes a relative velocity uncertainty function box 376, which is based on the... Figure 10 The relative velocity estimation algorithm 270 determines the minimum / maximum values—relative_velocity_Min 278 and relative_velocity_Max 280—to determine the relative velocity uncertainty factor 378. The relative velocity uncertainty factor 378 can be, for example, a value in the range of zero to one (0-1), where zero indicates the minimum uncertainty and one indicates the maximum uncertainty.

[0086] The relative velocity uncertainty function implemented at box 376 can be implemented in various ways. For example, the relative velocity uncertainty function box 376 can determine the relative velocity uncertainty factor 378 based on the range or increment between the minimum relative velocity value 278 and the maximum relative velocity value 280. In this example, the relative velocity uncertainty factor 378 may increase proportionally to the range / increment between the minimum value 278 and the maximum value 280 (i.e., the uncertainty may increase). Therefore, when the minimum / maximum range is small, the uncertainty is low, and the relative velocity uncertainty factor 378 is correspondingly low. Conversely, when the minimum / maximum range is large, the uncertainty is high, and the overlap uncertainty factor 378 is correspondingly high.

[0087] The active safety confidence factor determination algorithm 370 also includes an active safety confidence factor function box 380, which determines the active safety confidence factor 382. For example... Figure 14 As shown, the active safety confidence factor function 380 is based on an overlap rate uncertainty factor 374, a relative velocity uncertainty factor 378, and an active safety collision mode classification flag 366. In the example implementation described herein, the relative velocity uncertainty factor 378 can be a value in the range of zero to one (0-1), where zero indicates the minimum confidence level of the accuracy of the active safety collision mode classification flag 366, and one indicates the maximum confidence level of the accuracy of the active safety collision mode classification flag.

[0088] The active safety confidence factor function 380 can be implemented in various ways. For example, the active safety confidence factor function 380 can determine the active safety confidence factor 382 based on the active safety collision mode classification label 366 and uncertainty factors 374 and 378. For example, this could be multiple lookup tables, where the table to be used is determined by the classification label 366, and the confidence factor 382 is looked up in the table based on a combination of uncertainty factors 374 and 378. The confidence factor associated with various combinations of uncertainty factors can be determined by testing performed on a specific vehicle platform on which the vehicle safety system 10 is implemented. Weighted collision mode classification

[0089] refer to Figure 15 The weighted collision mode decision algorithm 390 determines the weighted collision mode classification label 396 based on the active safety confidence factor (ASCF) 382. If the active safety confidence factor 382 reaches or exceeds the threshold confidence value, the active safety collision mode classification label 366 is implemented as the weighted collision mode classification label 396. If the active safety confidence factor 382 does not exceed the threshold confidence value, the passive safety collision mode classification label 164 is implemented as the weighted collision mode classification label 396.

[0090] The weighted collision mode decision algorithm 390 includes an active safety confidence threshold matrix 392, which implements threshold confidence values ​​for various combinations of collision mode classifications indicated by active safety collision mode classification flag 366 and passive safety collision mode classification flag 164. The threshold confidence values ​​indicate the confidence or probability that the active safety collision mode classification 366 is correct, and are assigned in the range of zero to one (0-1), where one is the highest confidence and zero is the lowest confidence. In matrix 392, each of the multiple collision mode combinations that can be indicated by active safety collision mode classification flag 366 and passive safety collision mode classification flag 164 is assigned a confidence value.

[0091] Figure 15 In the example configuration, matrix 392 includes threshold confidence values ​​for three different collision modes, which can be classified by active and passive safety systems: Offset Deformable Barrier (ODB), Small Offset (SO), and Symmetrical (SYM). The collision modes implemented in matrix 392 can vary. For each combination of collision modes, matrix 392 includes a threshold confidence value that the active safety collision classification must exceed to be output by algorithm 390 as a weighted collision mode classification flag 396. Each of these thresholds is configurable and adjustable, allowing the system to be customized for specific vehicle platforms and manufacturer requirements. It should be noted that when the active classification flag 366 and the passive classification flag 164 coincide (as shown by extending diagonally upwards and to the right in the matrix cell), no decision is required and the threshold classification is zero.

[0092] Various factors can influence the threshold confidence values ​​implemented in matrix 392. On any given vehicle platform, passive safety system 20 may outperform active safety system 100 in classifying certain collision modes and underperform it in classifying others. The threshold confidence values ​​in matrix 392 are set through crash testing and other studies.

[0093] The weighted collision mode decision algorithm 390 compares the active safety confidence factor 382 with the values ​​in matrix 392 corresponding to the combinations of active / passive collision mode classification labels 366 and 164 generated by the collision event. As shown in box 394, if the active safety confidence factor 382 is less than the confidence threshold from matrix 392, then the passive safety collision mode classification label 164 is implemented as the weighted collision mode classification label 396. Otherwise, that is, if the active safety confidence factor 382 is greater than or equal to the confidence threshold from matrix 392, then the active safety collision mode classification label 366 is implemented as the weighted collision mode classification label 396.

[0094] For example, consider the following collision event: Active safety collision mode classification flag 366 indicates a symmetrical (SYM) collision event, and passive safety collision mode classification flag 164 indicates an offset deformable barrier (ODB) collision event. In this scenario, if the active safety confidence factor (ASCF) 382 < 0.6, then passive safety collision mode classification flag 164 (i.e., ODB) is approved as weighted collision mode classification flag 396. Otherwise, i.e., if the active safety confidence factor 382 ≥ 0.6, then active safety collision mode classification flag 366 (i.e., SYM) is approved as weighted collision mode classification flag 396.

[0095] As another example, consider the following collision events: Active safety collision mode classification flag 366 indicates a small offset (SO) or offset deformable barrier (ODB) collision event, and passive safety collision mode classification flag 164 indicates a symmetrical (SYM) collision event. In either of these scenarios, if the active safety confidence factor (ASCF) 382 > 0, then active safety collision mode classification flag 366 (i.e., SO or ODB) is approved as weighted collision mode classification flag 396.

[0096] As another example, consider the following collision event: Active safety collision mode classification flag 366 indicates an ODB collision event, and passive safety collision mode classification flag 164 indicates an SO collision event. In this scenario, if the active safety confidence factor (ASCF) 382 < 0.8, then passive safety collision mode classification flag 164 (i.e., SO) is approved as weighted collision mode classification flag 396. Otherwise, i.e., if the active safety confidence factor 382 ≥ 0.8, then active safety collision mode classification flag 366 (i.e., ODB) is approved to be passed on as weighted collision mode classification flag 396.

[0097] Advantageously, the control algorithm 150 allows for reliable and accurate classification of collision modes using the active safety system 100. One advantage of this is that the active safety system 100 estimates / predicts the collision mode based on perceived conditions prior to a collision event. Therefore, the vehicle safety system 10, which implements the control algorithm 150 using the active safety system 100, can classify the collision mode earlier than if the passive safety system 20 were used alone. Once determined, the weighted collision mode classification flag 396 can be used to select the misuse boxes and delays implemented by the passive safety system 20 to control the triggering of the vehicle safety device 14 in response to a frontal impact.

[0098] From the above description of the invention, those skilled in the art will recognize improvements, changes, and modifications. The appended claims are intended to cover such improvements, changes, and / or modifications that fall within the scope of the art.

Claims

1. A method for assisting in protecting vehicle occupants in the event of a frontal collision, the method comprising: A passive safety collision mode classification is determined in response to a collision signal received in response to the occurrence of a collision event; Active safety collision mode classification is determined in response to active safety signals received prior to the occurrence of the collision event; The determination of the active safety collision mode classification includes estimating the impact characteristics in response to the active safety signal. Estimating the impact characteristics includes identifying an object in the field of view of the active safety sensor, and determining the relative speed between the vehicle and the object and the overlap rate between the vehicle and the object. Determining the relative velocity includes calculating the minimum and maximum relative velocities, and determining the overlap rate includes calculating the minimum and maximum overlap rates. Determine the active safety confidence factor for the active safety collision mode classification; The active safety confidence factor is determined based on the minimum and maximum overlap rates, as well as the minimum and maximum relative velocities. In response to the active safety confidence factor exceeding a predetermined confidence value, the weighted collision mode classification is determined to be the active safety collision mode classification; and In response to the active safety confidence factor not exceeding the predetermined confidence value, the weighted collision mode classification is determined to be the passive safety collision mode classification.

2. The method of claim 1, further comprising: The occurrence of a frontal collision is determined in response to the collision signal; as well as The vehicle occupant protection device is activated based on the weighted collision mode classification.

3. The method of claim 1, wherein, Estimating the impact characteristics includes identifying an object in the field of view of the active safety sensors and determining the time of impact between the vehicle and the object.

4. The method of claim 1, wherein, Identifying objects in the field of view of the active safety sensor includes determining the object closest to the vehicle.

5. The method of claim 3, wherein, Determining the impact time includes estimating the minimum and maximum impact times using the vehicle's predetermined acceleration and the target's deceleration values.

6. The method of claim 1, wherein, Determining the relative speed involves estimating the minimum and maximum relative speeds using the vehicle's predetermined acceleration and the target's deceleration values.

7. The method of claim 1, wherein, Determining the overlap rate includes: Determine the width of the vehicle and the width of the object; Determine the lateral distance between the longitudinal centerlines of the vehicle and the object; The overlap ratio is defined as the sum of half the width of the vehicle, half the width of the object, and the lateral distance between the longitudinal centerlines of the vehicle and the object.

8. The method of claim 7, wherein, Determining the lateral distance between the longitudinal centerlines of the vehicle and the object includes estimating the minimum and maximum lateral distances between the longitudinal centerlines of the vehicle and the object.

9. The method of claim 8, wherein, Estimating the minimum and maximum lateral distances between the longitudinal centerlines of the vehicle and the object includes estimating the changes in lateral distance based on the vehicle's speed, steering angle, and yaw rate.

10. The method of claim 3, wherein, Determining the overlap rate includes identifying the impact side of the vehicle as either the left / driver's side or the right / passenger side of the vehicle.

11. The method of claim 10, further comprising determining whether an impact is imminent in response to the impact time being less than a threshold.

12. The method of claim 11, further comprising determining the active safety collision mode classification in response to determining an impending collision, the impact side of the vehicle, overlap classification, and speed classification.

13. The method of claim 12, wherein, The overlapping classification includes one of the following: asymmetric classification, symmetric classification, biased deformable obstacle classification, and small overlap classification.

14. The method of claim 12, wherein, The speed classification includes one of two categories: high speed and low speed.

15. The method of claim 1, wherein, Determining the active safety collision mode classification includes determining at least one of the following in response to an active safety signal: the longitudinal distance between the vehicle and the object, the lateral distance between the centerlines of the vehicle and the object, the travel distance between the vehicle and the object, the approach angle between the vehicle and the object, and the speed of the vehicle relative to the object.

16. A vehicle safety system for assisting in protecting vehicle occupants in the event of a frontal collision, the vehicle safety system comprising: Actuable safety devices; as well as A controller for controlling the actuation of the safety device according to the method of claim 1.

17. The vehicle safety system of claim 16, further comprising: One or more passive sensors are used to provide the collision signal to the controller; as well as One or more active sensors are used to sense objects in the path of the vehicle and provide the active safety signal to the controller.

18. The vehicle safety system of claim 17, wherein, The one or more active sensors include at least one of a camera and a radar sensor.

19. The vehicle safety system of claim 16, wherein, The controller includes an airbag controller unit.

20. A vehicle comprising the vehicle safety system of claim 16.

Images