Method and apparatus for controlling actuable protection devices with rough terrain and lift-off detection

Abstract

According to one aspect, a vehicle safety system includes an actuatable controlled restraint (ACR) including a seatbelt for restraining an occupant of a vehicle. The ACR is actuatable to control the extraction and retraction of the seatbelt. The system also includes a controller configured to determine an operating condition of the vehicle and to control actuation of the ACR in response to the determined operating condition of the vehicle. The ACR has a normal restraint condition and an enhanced restraint condition, and the controller is configured to actuate the ACR from the normal restraint condition to the enhanced restraint condition in response to a determination of an abnormal driving condition of the vehicle.

Description

TECHNICAL FIELD

[0001] The present invention relates generally to a method and apparatus for controlling an actuatable vehicle occupant protection device. More particularly, the present invention relates to a method and apparatus for detecting, and responsively controlling actuation of an actuatable controlled restraint device, such as an electrically powered seat belt retractor, in rough terrain and / or rollover vehicle conditions. BACKGROUND

[0002] Vehicle safety systems include a central control unit, sometimes referred to as an airbag control unit ("ACU"). The ACU utilizes sensors, both local sensors included with the ACU and remote sensors of the ACU, to detect the occurrence of a crash event involving the vehicle and determine whether these events warrant activation of actuatable restraint devices, such as airbags and seat belt retractors. The sensors used by the ACU can include accelerometers and other sensors, such as impact sensors, seat belt buckle switches, seat pressure switches, steering angle sensors, etc. Using data from these sensors, the ACU can determine the occurrence of a vehicle crash event and can execute a discrimination algorithm to classify the crash event as a particular type of event. The ACU can actuate the actuatable restraint devices in accordance with the particular type of crash event.

[0003] For vehicle safety systems, it is desirable to be able to discriminate between various crash events that a vehicle can be involved in. To "discriminate" a crash event means to classify the crash event as a particular type of crash event and distinguish it from other types of crash events. If the vehicle safety system can discriminate or recognize a crash event as a particular type, then the actuatable restraint devices can be actuated in a manner specific to that particular type of crash event.

[0004] As used herein, a "crash event" can be used to include various events involving a vehicle. For example, a crash event can be a crash or impact of a vehicle with a different type of structure, or otherwise engaging. These crash events can be a crash with a deformable barrier, such as another vehicle, or a crash with a non-deformable barrier, such as a tree or a utility pole. As another example, a crash event can also involve some event, such as a rollover event, in which a vehicle crashes as a result of the vehicle rolling over. A rollover event can result from a vehicle sliding sideways and hitting a curb, sliding off (or otherwise moving away from) a road embankment or a ditch, or sliding up (or otherwise moving away from) a slope, such as a hillside.

[0005] A vehicle safety system can be configured or adapted to distinguish between events in which deployment of an actuatable restraint device is desired ("deployment events") and events in which deployment of an actuatable restraint device is not desired ("non-deployment events"). Crash discrimination requires determination of the type of event, e.g., deformable barrier, non-deformable barrier, frontal crash, rear crash, side crash, oblique crash, offset crash, rollover, etc. Crash discrimination also requires determination of the severity of the crash and implementation of safety functions that serve as checks to ensure that the actuatable restraint device deploys in a safe manner.

[0006] Misuse events are events in which the vehicle is in an unusual manner. For example, the vehicle can be off-roaded over rough terrain, or airborne, i.e., "jumping" over a slope such as a hill. These misuse events present a challenge to the engineering of a vehicle safety system because the system needs to distinguish misuse events from crash events in order to properly control actuation of protection devices such as airbags. Moreover, some misuse events can be intentional because some drivers of vehicles off-road their vehicles for entertainment purposes.

[0007] In fact, some vehicles are modified for off-road use for utilitarian or entertainment purposes. During off-road use, the vehicle can be subjected to sudden movements, difficult or sudden starts / stops, steep angles, severe shaking in all directions, etc. However, during some off-road operations in which actuation of safety devices is not required, the vehicle movements can resemble crash events in which actuation of safety devices is required.

[0008] It can be appreciated, therefore, that it would be desirable to provide the ability to discriminate any vehicle condition that can be used to determine whether actuation / deployment of safety devices is required. Even if the event itself is not sufficient to make the actuation / non-actuation determination ("fire" or "not fire"), the event can be information that, together with other information, is used to make the fire / not fire determination.

[0009] Using this information, actuation and timing of actuatable restraints in a safety system can be controlled in response to the type and / or severity of the crash event to which the vehicle is subjected. Using information gathered from the discrimination algorithm, the safety system can implement a crash assessment process to discriminate the type of crash event and determine which, if any, occupant protection devices need to be actuated in response to the sensed crash event. If the identified crash event meets or exceeds a severity threshold and the safety function agrees, the actuatable restraint device can be actuated in a manner commensurate with the type of event that was discriminated. SUMMARY

[0010] According to a first aspect, a vehicle safety system includes an actuatable controlled restraint (ACR) including a seatbelt for restraining an occupant of a vehicle. The ACR is actuatable to control the extraction and retraction of the seatbelt. The system further includes a controller configured to determine an operating condition of the vehicle and to control actuation of the ACR in response to the determined operating condition of the vehicle. The ACR has a normal restraint condition and an enhanced restraint condition, the controller being configured to actuate the ACR from the normal restraint condition to the enhanced restraint condition in response to a determined abnormal driving condition of the vehicle.

[0011] According to a second aspect, in the normal restraint condition, the ACR can be configured to exert a relatively light retraction force sufficient to pull up the seatbelt webbing and tension the seatbelt across the occupant, while the ACR is configured to extract the seatbelt webbing in response to movement of the occupant. In the enhanced restraint condition, the ACR increases the retraction force exerted to the seatbelt webbing and increases the resistance to extraction of the seatbelt webbing in response to movement of the occupant.

[0012] According to a third aspect, alone or in combination with any of the preceding aspects, the abnormal driving condition of the vehicle includes at least one of a rough terrain vehicle condition and an airborne vehicle condition.

[0013] According to a fourth further aspect, alone or in combination with any of the preceding aspects, the controller can be an airbag controller (ACU) including:

[0014] • an ACU_X accelerometer for measuring a vehicle acceleration along an X axis of the vehicle and generating a signal indicative of the vehicle acceleration;

[0015] • an ACU_Y accelerometer for measuring a vehicle acceleration along a Y axis of the vehicle and generating a signal indicative of the vehicle acceleration;

[0016] • an ACU_Z accelerometer for measuring a vehicle acceleration along a Z axis of the vehicle and generating a signal indicative of the vehicle acceleration; and

[0017] • a roll (ROLL) sensor for measuring a vehicle roll acceleration about the X axis of the vehicle and generating a signal indicative of the vehicle roll acceleration;

[0018] The ACU can be configured to determine the abnormal driving condition of the vehicle in response to the signals from the ACU_X, ACU_Y, ACU_Z, and ROLL sensors.

[0019] According to a fifth aspect, alone or in combination with any of the preceding aspects, the controller can be configured to implement a lateral acceleration rough terrain classification algorithm to classify a rough terrain condition of the vehicle in response to a vehicle lateral acceleration measured along a Y axis of the vehicle. The controller can be further configured to implement a roll acceleration rough terrain classification algorithm to classify the rough terrain condition of the vehicle in response to a vehicle roll acceleration measured about an X axis of the vehicle. The controller can be further configured to implement a vertical acceleration rough terrain classification algorithm to classify the rough terrain condition of the vehicle in response to a vehicle vertical acceleration measured along a Z axis of the vehicle.

[0020] Additionally, according to the fifth aspect, the controller can be configured to determine the rough terrain condition of the vehicle in response to the lateral acceleration rough terrain classification algorithm, the roll acceleration rough terrain classification algorithm, and the vertical acceleration rough terrain classification algorithm classifying the rough terrain condition of the vehicle simultaneously.

[0021] According to a sixth aspect, alone or in combination with any of the preceding aspects, the controller can be configured to implement a lateral acceleration rough terrain metric that evaluates a vehicle lateral acceleration measured along a Y axis of the vehicle as a function of time. The lateral acceleration rough terrain metric classifies a rough terrain condition of the vehicle in response to the vehicle lateral acceleration measured along the Y axis of the vehicle crossing a threshold in both positive and negative directions over a predetermined time period. The controller can be further configured to implement a roll acceleration rough terrain metric that evaluates a vehicle roll acceleration measured about an X axis of the vehicle as a function of time. The roll acceleration rough terrain metric classifies the rough terrain condition of the vehicle in response to the vehicle roll acceleration measured along the X axis of the vehicle crossing a threshold in both positive and negative directions over a predetermined time period. The controller can be further configured to implement a vertical acceleration rough terrain metric that evaluates a vehicle vertical acceleration measured along a Z axis of the vehicle as a function of time. The vertical acceleration rough terrain metric classifies the rough terrain condition of the vehicle in response to the vehicle vertical acceleration measured along the Z axis of the vehicle crossing a threshold in a positive direction.

[0022] Additionally, according to the sixth aspect, to determine the lateral acceleration rough terrain metric, the controller can be configured to evaluate a moving average of the vehicle lateral acceleration measured along the Y axis of the vehicle as a function of time. To determine the roll acceleration rough terrain metric, the controller can be configured to evaluate a moving average of the vehicle roll acceleration measured about the X axis of the vehicle as a function of time. To determine the vertical acceleration rough terrain metric, the controller can be configured to evaluate a moving average of the vehicle vertical acceleration measured along the Z axis of the vehicle as a function of time.

[0023] According to a seventh aspect, alone or in combination with any of the preceding aspects, the controller can be configured to implement a longitudinal acceleration airborne classification algorithm to classify a vehicle airborne condition in response to a vehicle longitudinal acceleration measured along an X-axis of the vehicle. The controller can be further configured to implement a lateral acceleration airborne classification algorithm to classify the vehicle airborne condition in response to a vehicle lateral acceleration measured along a Y-axis of the vehicle. The controller can be further configured to implement a vertical acceleration airborne classification algorithm to classify the vehicle airborne condition in response to a vehicle vertical acceleration measured along a Z-axis of the vehicle. The controller can be further configured to implement a roll acceleration airborne classification algorithm to classify the vehicle airborne condition in response to a vehicle roll acceleration measured about the X-axis of the vehicle.

[0024] Additionally, according to the seventh aspect, the controller can be configured to implement a vertical acceleration airborne confirmation algorithm to confirm the vehicle airborne condition in response to the vehicle vertical acceleration measured along the Z-axis of the vehicle. Additionally, according to this aspect, the vertical acceleration airborne confirmation algorithm can be configured to determine a vehicle touchdown condition in response to the vehicle vertical acceleration measured along the Z-axis of the vehicle. Additionally, according to this aspect, the controller can be configured to determine the vehicle airborne condition in response to classifying the vehicle airborne condition and confirming the vehicle airborne condition. The controller can be configured to further determine the vehicle airborne condition in response to determining that the vehicle off-road condition has not been determined. The controller can be further configured to implement an R_ANGLE off-road classification metric on the vehicle off-road condition, the R_ANGLE off-road classification metric evaluating the vehicle R_ANGLE over time and classifying the vehicle off-road condition in response to the R_ANGLE crossing an upper threshold and a lower threshold in any order within a predetermined amount of time.

[0025] According to an eighth aspect, alone or in combination with any of the preceding aspects, the controller can be configured to implement a longitudinal acceleration airborne metric, the longitudinal acceleration airborne metric evaluating a vehicle longitudinal acceleration measured along an X-axis of the vehicle over time. The longitudinal acceleration airborne metric can define a threshold having a predetermined bandwidth extending above and below a zero point of the longitudinal acceleration airborne metric. The longitudinal acceleration airborne metric can be configured to classify a vehicle airborne condition in response to a magnitude of the metric value exceeding and falling back within the threshold. The longitudinal acceleration airborne metric can maintain the vehicle airborne condition classification when the metric value remains within the threshold.

[0026] According to the eighth aspect, the controller can be further configured to implement a lateral acceleration airtime metric that evaluates a vehicle lateral acceleration measured along a Y-axis of the vehicle as a function of time. The lateral acceleration airtime metric can define a threshold having a predetermined bandwidth extending above and below a zero point of the lateral acceleration airtime metric. The lateral acceleration airtime metric can be configured to classify a vehicle airtime condition in response to a magnitude of the metric value exceeding and falling within the threshold. The lateral acceleration airtime metric can maintain the vehicle airtime condition classification when the metric value remains within the threshold.

[0027] According to the eighth aspect, the controller can be further configured to implement a roll acceleration airtime metric that evaluates a vehicle roll acceleration measured about an X-axis of the vehicle as a function of time. The roll acceleration airtime metric can define a threshold having a predetermined bandwidth extending above and below a zero point of the roll acceleration airtime metric. The roll acceleration airtime metric can be configured to classify a vehicle airtime condition in response to a magnitude of the metric value exceeding and falling within the threshold. The roll acceleration airtime metric can maintain the vehicle airtime condition classification when the metric value remains within the threshold.

[0028] According to the eighth aspect, the controller can be further configured to implement a vertical acceleration airtime metric that evaluates a vehicle vertical acceleration measured along a Z-axis of the vehicle as a function of time. The vertical acceleration airtime metric can define a threshold having a predetermined bandwidth extending above and below a zero point of the vertical acceleration airtime metric. The vertical acceleration airtime metric can be calibrated to compensate for the effects of gravity. The vertical acceleration airtime metric can be configured to classify a vehicle airtime condition in response to the metric value crossing the threshold. The vertical acceleration airtime metric can maintain the vehicle airtime condition classification when the metric value remains within the threshold.

[0029] Additionally, according to the eighth aspect, the controller can be configured to classify a vehicle airtime condition in response to determining that all of the following are true simultaneously:

[0030] • the longitudinal acceleration airtime metric classifies a vehicle airtime condition;

[0031] • the lateral acceleration airtime metric classifies a vehicle airtime condition;

[0032] • the vertical acceleration airtime metric classifies a vehicle airtime condition; and

[0033] • the roll acceleration airtime metric classifies a vehicle airtime condition.

[0034] According to this aspect, the controller can be configured to maintain the vehicle airborne classification for a predetermined period of time. Additionally, according to this aspect, the controller can be configured to implement a vertical acceleration airborne confirmation metric that evaluates a vehicle vertical acceleration measured along the Z-axis of the vehicle as a function of time. The vertical acceleration airborne confirmation metric can define a threshold having a predetermined bandwidth selected such that the metric value passes through the bandwidth in response to an airborne post-landing, thereby confirming the vehicle airborne condition. According to this aspect, in this manner, the controller can be configured to determine the vehicle airborne condition in response to classifying the vehicle airborne condition and confirming the vehicle airborne condition.

[0035] According to a ninth aspect, the controller can be an airbag controller unit (ACU) configured to control deployment of one or more airbags and the ACR. BRIEF DESCRIPTION OF DRAWINGS

[0036] The above mentioned and other features and advantages of this application will be apparent from the following description of the application, when taken in connection with the drawings, wherein:

[0037] Figure 1 is a schematic diagram of a vehicle and signals obtained from a sensor architecture deployed therein.

[0038] Figure 2 is a schematic diagram illustrating a seated vehicle occupant with a vehicle safety system that includes an actuatable constrained restraint device (ACR).

[0039] Figure 3 is a block diagram illustrating a vehicle safety system.

[0040] Figure 4 is a block diagram illustrating metric calculations implemented in a vehicle safety system.

[0041] Figures 5-7 is a view illustrating discriminant metrics and algorithms implemented in a vehicle safety system for classifying a rough terrain event.

[0042] Figure 8 is a view illustrating a rough terrain deployment algorithm implemented in a vehicle safety system.

[0043] Figure 9 is a view illustrating discriminant metrics and algorithms implemented in a vehicle safety system for classifying an airborne event.

[0044] Figure 10 is a logic view illustrating an airborne deployment algorithm implemented in a vehicle safety system. DETAILED DESCRIPTION

[0045] The present invention relates to a vehicle safety system that implements a discrimination metric, a classification algorithm, and a deployment algorithm for controlling actuatable vehicle restraint devices. The metric and algorithms implemented by the vehicle safety system can detect rough road conditions of the vehicle and lift-off conditions of the vehicle. Advantageously, the vehicle safety system can react to the detection of these conditions by actuating one or more vehicle restraint devices.

[0046] In one particularly advantageous implementation, the vehicle safety system can control actuation of an actuated controlled restraint (ACR). An example ACR is a seatbelt retractor that includes a motor that can be used to actively control the extension and retraction of the seatbelt in response to a control signal. The ACR can also include other known features such as a pretensioner and / or a load limiter. In this implementation, the vehicle safety system can actuate the ACR to tension the seatbelt around the occupant in response to the detection of a condition that can lead to a vehicle crash for which it can be desirable to ensure deployment of additional safety devices, e.g., airbags. If a crash event then occurs, the seatbelt will pretension the occupant, thereby enhancing the occupant protection provided by the safety system. If the sensed condition does not lead to a crash event, the safety system can actuate the ACR to release or relax the seatbelt, e.g., a predetermined time after the initial detected condition.

[0047] Vehicle safety system

[0048] Reference is made to Figure 1 According to one example construction, the vehicle 12 includes a vehicle safety system 10 that includes a central control unit, here referred to as an airbag control unit (ACU) 50. The ACU 50 operatively actuates one or more actuatable restraint devices in the vehicle 12, such as an actuated controlled restraint (ACR), a frontal airbag, a curtain airbag, a thorax airbag, a side airbag, and a knee airbag.

[0049] The ACU 50 includes one or more sensors that operatively provide signals indicative of linear and / or angular acceleration and / or velocity of movement of the vehicle in different directions and relative to different vehicle axes. These sensors can be locally mounted in or on the ACU 50 itself or remote from the ACU and interconnected therewith, e.g., via wires. These vehicle axes include an X-axis that extends longitudinally in the vehicle in the direction of forward / rearward travel of the vehicle. A Y-axis of the vehicle extends laterally in the vehicle perpendicular to the X-axis. A Z-axis of the vehicle extends vertically in the vehicle perpendicular to both the X-axis and the Y-axis.

[0050] The X, Y, and Z axes are Figure 1The two axes intersect at ACU50. This is because ACU50 can include sensors for measuring the movement of vehicle 12 relative to the X, Y, and Z axes, i.e., acceleration. These movements occur in... Figure 1 The system uses a sign (+ / -) to indicate the sign (positive or negative) of the motion assigned by the safety system 10 along the axis. The ACU 50 may also include sensors that sense rotation about the X-axis (i.e., pitch), rotation about the Y-axis (i.e., roll), and rotation about the Z-axis (i.e., yaw). The vehicle safety system 10 can utilize these accelerations and / or rotations in different combinations to detect certain vehicle conditions.

[0051] like Figure 1 As shown, the vehicle safety system 10 can be configured to interpret motion along the X-axis as positive (acceleration) from front to back and negative (deceleration) from back to front. Motion along the Y-axis can be interpreted as positive from right to left and negative from left to right. Motion along the Z-axis can be interpreted as positive in the downward direction and negative in the upward direction. The vehicle safety system 10 can also be configured to interpret vehicle rotational motion (i.e., roll) about the X-axis as positive to the left and negative to the right. Vehicle rotational motion (i.e., pitch) about the Y-axis can be positive to forward / downward pitch and negative to backward / upward pitch. Vehicle rotational motion (i.e., yaw) about the Z-axis can be positive to the left (viewed from the forward angle) and negative to the right.

[0052] Actuatable controlled restraint (ACR)

[0053] Figure 2 This describes parts of vehicle 12 and vehicle safety system 10. For example... Figure 2 As shown, the occupant 20 of vehicle 12 is positioned on vehicle seat 30. Seat 30 includes seat base 32 for mounting the seat on floor 14 of vehicle 12. The base supports a base pad 34 on which the occupant 20 sits, a seat back 36 on which the occupant leans, and a headrest 38 for receiving the occupant's head.

[0054] like Figure 2 As shown, the safety system 10 includes a seatbelt 40, an ACU 50, and an actuable controlled restraint device (ACR) 60 associated with the seatbelt. Figure 2 In the example configuration shown, the seat belt 40 is a conventional three-point seat belt, which includes a lap belt portion 42 that extends over the occupant's knees and a shoulder strap portion 44 that extends over the occupant's shoulders and torso. A buckle 46 secures the seat belt 40 to the occupant. Figure 2 In the shown secure configuration, the D-ring 48 guides the seatbelt 40 to the ACR 60.

[0055] The ACR 60 serves the purpose of a conventional seat belt retractor. The ACR pulls the seat belt webbing out to allow the occupant 20 to extend his or her shoulders 44 and knees 42 over his or her body and in Figure 2 engagement with the buckle 46 in the illustrated belted and restrained condition. When the occupant 20 unbuckles the seat belt 40, the ACR also retracts the seat belt webbing to place itself in an un-restrained condition. Additionally, the ACR 60 can be configured to include components that allow the ACR to function as a pretensioner to retract the seat belt in response to a vehicle crash. The ACR 60 can further be configured to include components that allow the ACR to function as a load limiter to pull the seat belt webbing out in a vehicle crash event where the load on the seat belt 40 reaches or exceeds a predetermined amplitude.

[0056] The ACR 60 also includes a motor configured to actively control the pulling out and retraction of the seat belt webbing. The ACR 60 is operatively connected to the ACU 50, which is operable to control the ACR to actively pull out and retract the seat belt webbing. As described herein, the ACU 50 is configured to control the operation of the ACR 60 in response to vehicle conditions determined by evaluating data received from vehicle sensors.

[0057] When sensing a generally driving condition, as Figure 2 illustrated, the ACU 50 is configured to place the seat belt 40 in a generally restrained condition, which pulls against the occupant 20 with a light force just enough to remove slack from the seat belt and maintain the seat belt against the occupant's body. In the generally restrained condition, the ACR 60 functions as an ordinary seat belt retractor to spool the seat belt webbing with a light force so that the seat belt can be easily pulled out, for example, if the occupant leans forward. If a crash occurs, the ACU 50 is configured to sense the event and actuate the ACR to provide any desired pretensioning in response to the crash event in a conventional manner. Conventional load limiting performance can also occur here, but is generally passive in nature and provided mechanically through structures such as torsion bars / torsion springs.

[0058] Advantageously, when a driving condition is determined to be considered non- general, such as certain misuse events, the ACU 50 is configured to respond by actuating the ACR 60 in order to place the seat belt 40 in an enhanced restrained condition. In the enhanced restrained condition, the ACR 60 tensions the seat belt 40 around the occupant 20 to enhance the degree of restraint of the occupant in the vehicle seat 30. Once the driving condition is determined to be general, the ACU 50 can control the ACR 60 to place the seat belt 40 in a generally restrained condition.

[0059] According to one example implementation of the safety system 10, the ACU 50 is configured to determine a rough road condition and / or an empty vehicle condition that does not necessarily rise to the level of requiring a guaranteed deployment of an actuatable restraint device, such as an airbag. These events are considered dangerous misuse events that can lead to a vehicle crash that requires a guaranteed airbag deployment. However, these dangerous misuse events can also actually not lead to a vehicle crash event that requires a guaranteed airbag deployment. For example, this can be the case where a vehicle operator is operating the vehicle off-road for recreational purposes. For example, it can also be the case where a vehicle operator loses control of the vehicle, leaves (e.g., slides or skids) the road and encounters rough terrain and / or flies through the air.

[0060] Accordingly, in response to determining a rough road / empty condition, the ACU 50 is configured to control the ACR 60 to place the safety belt 40 in an enhanced restraint condition to increase the degree of restraint of the occupant in the vehicle. This is done as a precaution in the event that a crash event occurs. If a crash event subsequently occurs, the ACU 50 is configured to determine the crash event and actuate other protective devices, such as airbags, in accordance with known crash discrimination algorithms. If, for example, a crash event does not subsequently occur within a predetermined period of time after determining a rough road / empty condition, the ACU 50 is configured to actuate the ACR 60 to release the tension on the safety belt 40 to return the safety belt to the normal restraint condition.

[0061] Figure 3 Portions of the vehicle safety system 10 are schematically illustrated. Referring to Figure 3 , the ACU 50 includes internal sensors in the form of accelerometers for sensing Figure 1 some of the vehicle parameters identified in the Figure 3 . The ACU 50 utilizes the accelerometer 52 to sense the vehicle longitudinal (X-axis) acceleration (ACU_X), the accelerometer 54 to sense the vehicle lateral (Y-axis) acceleration (ACU_Y), the accelerometer 56 to sense the vehicle vertical (Z-axis) acceleration (ACU_Z), and the roll rate sensor 58 to sense the vehicle roll rate value (ROLL) (i.e., the roll rate about the vehicle X-axis). Although the vehicle safety system 10 can also include additional accelerometers and / or sensors, such as satellite sensors, pitch sensors, and yaw sensors, these values are not implemented in the algorithms disclosed herein and therefore are not shown in

[0062] It is desirable to position the sensors on or near the corresponding axes, along or around which they sense vehicle motion. Since the sensors can be locally mounted on the ACU 50, it is desirable to mount the ACU at or near the vehicle's center of gravity, through which the X, Y, and Z axes all pass. However, the location of the ACU 50 at or near the vehicle's center of gravity is not critical, and the ACU 50 can also be positioned elsewhere on the vehicle.

[0063] The hardware and software configurations for the ACU implemented in a vehicle safety system are known in the art. Therefore, a detailed description of the hardware construction of the ACU 50 is not required for those skilled in the art to understand and comprehend the vehicle safety system 10. Figure 1 The ACU50 includes a central processing unit (CPU), such as a microcomputer, which is configured to receive signals ACU_X, ACU_Y, ACU_Z, and ROLL from the corresponding sensors, thereby performing vehicle measurement calculations 70 based on these signals, and using the calculated measurements to perform rugged terrain and airborne recognition algorithms 80.

[0064] The vehicle measurements generated by calculation 70 include:

[0065] • Moving average of the vehicle's longitudinal X-axis acceleration (AMA_X).

[0066] • Moving average of the vehicle's lateral Y-axis acceleration (AMA_Y).

[0067] • Moving average of the vehicle's vertical Z-axis acceleration (AMA_Z).

[0068] • Vehicle roll difference rate, i.e., roll acceleration (D_RATE).

[0069] • Vehicle roll angle (ANGLE).

[0070] • Vehicle roll rate (RATE).

[0071] Rough terrain and airborne incident detection algorithm 80 are merely two events that the vehicle safety system 10 is configured to detect. These two situations are the only ones described here, as the determination of these situations is the subject of this paper. Of course, the safety system 10 can also be configured to determine many other situations, including vehicle collisions and rollovers. Which events the vehicle safety system 10 is configured to detect depends on various factors, such as manufacturer requirements and / or government and industry standards for producing the vehicles.

[0072] Figure 4 This describes the vehicle measurement calculation 70 performed by ACU 50. Figure 4 The elements of vehicle measurement calculation 70 are shown, referred to here as "functions" performed internally by ACU 50.

[0073] Roll rate

[0074] The ACU 50 employs internal signal conditioning, including an analog-to-digital converter (ADC), to digitize the ROLL, ACU_X, ACU_Y, and ACU_Z signals. Accelerometers 52, 54, and 56, and the tilt sensor 58 can perform the digitization of the ROLL, ACU_X, ACU_Y, and ACU_Z signals. The ACU 50 can also perform track checking and offset adjustment using the ROLL, ACU_X, ACU_Y, and ACU_Z signals.

[0075] like Figure 4 As shown, the digitized and biased slant ratio ROLL is passed to a high-pass filter (HPF) function 104, which can be selected, for example, to have a time constant that causes the filter function to reset after a predetermined time period, e.g., T = 8 seconds. The high-pass filtered slant ratio ROLL generated at HPF function 104 is passed to a moving average function 106, and then to a moving average function 108. Each moving average function 106, 108 can be, for example, tunable to select the number of samples in the moving average, e.g., 1-32 samples. Moving average functions 106, 108 smooth the changes in the slant ratio.

[0076] The high-pass filtered roll rate ROLL generated at HPF function 104 is passed to low-pass filter (LPF) function 112, which generates a roll rate metric R_RATE. This roll rate metric has a value indicating the vehicle's roll rate (i.e., angular velocity), which is used in rough terrain and airborne discrimination algorithm 80 (see...). Figure 3 The R_RATE signal is passed to the Integral High-Pass Filter (IHPF) function 114, which includes an integrator function and a dual-time-constant high-pass filter function. The IHPF function 114 integrates the R_RATE signal to produce a value indicating the determined relative roll angle of the vehicle. The IHPF function 114 also performs high-pass filtering on the R_RATE signal. The IHPF function 114 generates a metric R_ANGLE, which is implemented in the rugged terrain and airborne discrimination algorithm 80 (see [link to algorithm 80]). Figure 3 ).

[0077] R_ANGLE indicates the normalized roll angle of the vehicle, which is a measure of the relative angular rotation of the vehicle in response to the sensed roll rate. The IHPF function 114 can reset R_ANGLE based on the time constant of the high pass filter function such that R_ANGLE provides an indication of the angular rotation during the detected roll rate occurrence. Thus, R_ANGLE can not indicate the actual angular orientation of the vehicle with respect to the ground. As such, the determination of a vehicle rollover condition does not necessarily rely on a determination of the initial angular orientation of the vehicle with respect to the ground or road.

[0078] The roll rate moving average function 108 provides to the difference function 110, where the difference between the current sample and the previous sample is compared. This results in a differentiated roll rate metric D_RATE, which indicates the rate of change of roll rate, i.e., acceleration. This roll acceleration D_RATE is the angular acceleration of the vehicle about the vehicle X axis. The roll acceleration D_RATE is implemented in the rough road and airborne discrimination algorithm 80 (see Figure 3 ).

[0079] Longitudinal acceleration

[0080] As shown in Figure 4 , the digitized and biased longitudinal acceleration ACU_X is passed to a high pass filter (HPF) function 120, which can be selected, for example, to have a time constant that results in the filter function being reset after a predetermined period of time, e.g., T = 8 seconds. The high pass filtered longitudinal acceleration ACU_X produced at the HPF function 120 is passed to a low pass filter (LPF) function 122. The low pass filtered longitudinal acceleration ACU_X value produced at the LPF function 122 is passed to a moving average block 124, which produces a longitudinal acceleration metric AMA_X. The number of samples included in the moving average function 124 can be tuned within a predetermined range, such as within 1-32 samples. AMA_X is the longitudinal acceleration moving average value, which is implemented in the rough road and airborne discrimination algorithm 80 (see Figure 3 ).

[0081] Lateral acceleration

[0082] As shown in Figure 4As shown, the digitized and biased lateral acceleration ACU_Y is passed to a high-pass filter (HPF) function 130, which can be selected, for example, to have a time constant that causes the filter function to reset after a predetermined time period, e.g., T = 8 seconds. The high-pass filtered lateral acceleration ACU_Y generated at HPF function 130 is passed to a low-pass filter (LPF) function 132. The low-pass filtered lateral acceleration ACU_Y value generated at LPF function 132 is passed to a moving average block 134, which generates a lateral acceleration metric AMA_Y. The number of samples included in the moving average function 134 can be tuned within a predetermined range (e.g., 1-32 samples). AMA_Y is a moving average of lateral acceleration, which is implemented in the rough road surface and takeoff discrimination algorithm 80 (see...). Figure 3 ).

[0083] Vertical acceleration

[0084] like Figure 4 As shown, the digitized and biased vertical acceleration ACU_Z is passed to a high-pass filter (HPF) function 140, which can be selected, for example, to have a time constant that causes the filter function to reset after a predetermined time period, e.g., T = 8 seconds. The high-pass filtered vertical acceleration ACU_Z generated at HPF function 140 is passed to a low-pass filter (LPF) function 142. The low-pass filtered vertical acceleration ACU_Z value generated at LPF function 142 is passed to a moving average block 144, which generates a vertical acceleration metric AMA_Z. The number of samples included in the moving average function 144 can be tuned within a predetermined range (e.g., 1-32 samples). AMA_Z is a moving average of vertical acceleration, which is implemented in the rough road surface and takeoff discrimination algorithm 80 (see...). Figure 3 ).

[0085] Rough terrain discrimination algorithm

[0086] Figures 5-8 The rugged terrain discrimination algorithm implemented by the vehicle safety system 10 is explained. Figures 5-7 The algorithm implements metrics and Boolean logic to classify rugged terrain conditions based on signals obtained from the ACU_Y, ACU_Z, and ROLL sensors. In this description, the metrics and logic functions are described as on / off or triggered as on / off. The ON state is associated with a Boolean value of one in the Boolean logic of the illustrated algorithm, and the OFF state is associated with a Boolean value of zero in the Boolean logic of the illustrated algorithm. Figures 5-7 The algorithm classifies the situation for Figure 8The discrimination algorithm determines the rugged terrain conditions and activates ACR 60 to place seat belt 40 in an enhanced restraint condition.

[0087] ACU-Y rough terrain classification algorithm

[0088] The ACU 50 implements an algorithm for classifying vehicle terrain conditions in response to lateral vehicle acceleration along the Y-axis. (See reference...) Figure 5 The ACU 50 implements a lateral acceleration rugged terrain classification algorithm 150. This algorithm 150 utilizes the ACU_Y accelerometer 54 within the ACU 50 to classify the rugged terrain conditions of the vehicle 12. To this end, algorithm 150 implements AMA_Y rugged terrain measurements 160 and 170, which are... Figure 5 The diagram is used to illustrate this.

[0089] The AMA_Y rugged terrain metric 160 evaluates AMA_Y over time to determine a metric value 162. If the metric value 162 crosses a threshold 164 as shown in 166, the AMA_Y rugged terrain metric 160 is triggered ON and outputs a Boolean value of one to latch block 152. In response, latch block 152 is triggered ON, outputs a Boolean value of one, and this output is held (i.e., latched) for a predetermined amount of time, which is a tunable parameter of algorithm 150. Threshold 164 evaluates positive values ​​of AMA_Y, thus evaluating positive acceleration along the vehicle's Y-axis, i.e., the vehicle's leftward acceleration (see...). Figure 1 ).

[0090] Similarly, the AMA_Y rugged terrain metric 170 evaluates AMA_Y over time to determine a metric value 172. If the metric value 172 crosses a threshold 174 as shown in 176, the AMA_Y rugged terrain metric 170 is triggered ON and outputs a Boolean value of one to latch block 154. In response, latch block 154 is triggered ON and outputs a Boolean value of one, and this output is held (i.e. latched) for a predetermined amount of time, which is a tunable parameter of algorithm 150. Threshold 174 evaluates negative values ​​of AMA_Y, thus evaluating negative acceleration along the vehicle's Y-axis, i.e., the vehicle's rightward acceleration (see...). Figure 1 ).

[0091] The outputs of the latch blocks 152, 154 are provided to an AND block 156. If both latch blocks 152, 154 output a Boolean value of one, the AND block 156 triggers to ON and outputs a Boolean value of one and the ACU Y rough terrain classification condition is determined at block 158. As the outputs of the metrics 160, 170 are latched at the latch blocks 152, 154 for a predetermined time, it can be observed that the ACU Y rough terrain classification algorithm 150 classifies the ACU Y rough terrain classification 158 when the vehicle 12 experiences lateral accelerations of sufficient magnitude in opposite directions consistent with the vehicle driving on rough terrain for a predetermined period of time.

[0092] ROLL rough terrain classification algorithm

[0093] The ACU 50 implements an algorithm for classifying vehicle rough terrain conditions in response to roll acceleration about the vehicle X axis. Referring to Figure 6 , the roll acceleration rough terrain classification algorithm 180 is implemented by the ACU 50. This algorithm 180 utilizes the ROLL sensor 58 internal to the ACU 50 to classify the rough terrain conditions of the vehicle 12. To this end, the algorithm 180 implements D_RATE rough terrain metrics 190, 200 that are graphically illustrated in Figure 6 .

[0094] The D_RATE rough terrain metric 190 evaluates the D_RATE as a function of time to determine a metric value 192. If the metric value 192 crosses a threshold 194 as shown at 196, the D_RATE rough terrain metric 190 triggers to ON and outputs a Boolean value of one to a latch block 182. In response, the latch block 182 triggers to ON, outputs a Boolean value of one, and this output is held (i.e., latched) for a predetermined amount of time that is a tunable parameter of the algorithm 180. The threshold 194 evaluates positive values of D_RATE and thus positive roll about the vehicle X axis, i.e., roll to the right (see Figure 1 ).

[0095] Similarly, the D_RATE rough terrain metric 200 evaluates the D_RATE as a function of time to determine a metric value 202. If the metric value 202 crosses a threshold 204 as shown at 206, the D_RATE rough terrain metric 200 triggers to ON and outputs a Boolean value of one to a latch block 184. In response, the latch block 184 triggers to ON, outputs a Boolean value of one, and this output is held (i.e., latched) for a predetermined amount of time that is a tunable parameter of the algorithm 180. The threshold 204 evaluates negative values of D_RATE and thus negative roll about the vehicle X axis, i.e., roll to the left (see Figure 1 ).

[0096] The outputs of latches 182, 184 are provided to AND block 186. If both latches 182, 184 output a Boolean value of one, AND block 186 is triggered to ON and outputs a Boolean value of one and a ROLL rough terrain classification condition is determined at block 188. As the outputs of metrics 190, 200 are latched at latches 182, 184 for a predetermined time, it is observed that the roll rough terrain classification algorithm 180 classifies a ROLL rough terrain classification 188 when vehicle 12 experiences roll acceleration (D_RATE) in opposite directions of sufficient magnitude and for a predetermined period of time consistent with vehicle travel over rough terrain.

[0097] ACU-Z rough terrain classification algorithm

[0098] ACU 50 implements an algorithm for classifying vehicle rough terrain conditions in response to vertical vehicle acceleration along the Z-axis. Referring to Figure 7 , a vertical rough terrain classification algorithm 210 is implemented by ACU 50. Algorithm 210 utilizes ACU-Z accelerometer 54 internal to ACU 50 to classify rough terrain conditions of vehicle 12. To this end, algorithm 210 implements AMA_Z rough terrain metrics 220 that are graphically illustrated in Figure 7 .

[0099] AMA_Z rough terrain metrics 220 evaluate AMA_Z as a function of time to determine a metric value 222. If metric value 222 crosses threshold 224 as shown at 226, AMA_Z rough terrain metrics 220 are triggered to ON and output a Boolean value of one to latch 212. In response, latch 212 is triggered to ON, outputs a Boolean value of one, and this output is held (i.e., latched) for a predetermined amount of time that is a tunable parameter of algorithm 210. Threshold 224 evaluates positive values of AMA_Z and thus positive acceleration along the Z-axis of the vehicle, i.e., vehicle downward acceleration (see Figure 1 ).

[0100] Latch 212 is triggered to ON and outputs a Boolean value of one, which results in an ACU_Z rough terrain classification condition determination at block 214. As the output of metrics 220 is latched at block 212 for a predetermined time, it is observed that ACU_Z rough terrain classification algorithm 210 classifies an ACU_Z rough terrain classification 218 when vehicle 12 experiences downward acceleration of a magnitude consistent with vehicle travel over rough terrain.

[0101] Rough terrain determination

[0102] Referring to Figure 8The rough terrain determination algorithm 230 is implemented by the ACU 50. This algorithm 230 utilizes the ACU_Y rough terrain classification algorithm 150, the ROLL rough terrain classification algorithm 180, and the ACU_Z rough terrain classification algorithm 210 to determine whether the vehicle 12 is traveling on rough terrain. As shown, the ACU_Y rough terrain classification 158 Figure 8 Figure 5 Figure 6 Figure 7 inputs to an AND block 232. If all three conditions are met, the AND block 232 is triggered ON, outputs a Boolean value of one, and the rough terrain determination algorithm 230 determines rough terrain and issues a deploy ACR command, as shown by block 234.

[0103] From the above, it can be understood that the ACU 50 implementing the algorithms 150, 180, 210, and 230 determines the rough terrain conditions of the vehicle 12 and deploys the ACR when all of the following occur within a predetermined time period:

[0104] • the vehicle 12 experiences lateral acceleration (AMA_Y) along the vehicle Y axis in the opposite direction consistent with traveling on rough terrain.

[0105] • the vehicle 12 experiences roll acceleration (D_RATE) along the vehicle X axis in the opposite direction consistent with traveling on rough terrain.

[0106] • the vehicle 12 experiences vertical acceleration (AMA_Z) along the vehicle Z axis in the opposite direction consistent with traveling on rough terrain.

[0107] Deploying the ACR 60 in response to the rough terrain determination 234 results in the placement of the seatbelt 40 in an enhanced restraint condition. Once the driving conditions are determined to be typical, the ACU 50 can control the ACR 60 to place the seatbelt 40 in a typical restraint condition. For example, the transition from enhanced restraint to typical restraint can occur in response to the AND block 232 being triggered OFF and outputting a Boolean value of zero, which indicates the loss of one or more of the rough terrain classifications 158, 188, 214. As another example, the rough terrain determination 234 can be latched such that the transition from enhanced restraint to typical restraint occurs in response to the AND block 232 being triggered OFF, outputting a Boolean value of zero, and the timeout of a latched timer.

[0108] Vacation discrimination algorithm

[0109] Figures 9-10 A hang-up discrimination algorithm implemented by the vehicle safety system 10 is described. Figure 9 ​​​The ACU 50 implements an algorithm that measures and Boolean logic to classify the airborne condition based on signals obtained from the ACU_X, ACU_Y, ACU_Z, and ROLL sensors. Figure 10 The algorithm determines the airborne classification based on Figure 9 the algorithm and responsively actuates the ACR 60 to place the safety belt 40 in an enhanced restraint condition.

[0110] ACU vacation classification algorithm

[0111] Referring to Figure 9 The ACU airborne classification algorithm 250 is implemented by the ACU 50. This algorithm 250 utilizes the accelerometers internal to the ACU 50, namely ACU_X 52, ACU_Y 54, ACU_Z 56, and Roll 58 (see Fig. 2), to classify the airborne condition of the vehicle 12. To do this, the algorithm 250 implements an AMA_Z airborne classification metric 260, an AMA_Y airborne classification metric 270, an AMA_X airborne classification metric 280, and a ROLL airborne classification metric 290. Figure 2 The metrics 260, 270, 280, 290 are graphically illustrated in Fig. 3. Figure 9

[0112] AMA-Z vacation classification metric

[0113] The AMA_Z airborne classification metric 260 evaluates the AMA_Z over time to determine a metric value 262. The AMA_Z airborne classification metric 260 adds the effect of gravity (-1g) to the measured acceleration so that when the vehicle is airborne, the metric value 262 is zero. During normal driving on a smooth surface, the AMA_Z airborne classification metric value 260 will thus hover around -1g. When the vehicle 12 launches to become airborne, the vehicle experiences a sharp rise in AMA_Z, as shown on the left side of the metric value 262. Once airborne, the metric value 262 settles around 0g. When the vehicle 12 descends, the AMA_Z falls, as shown by the metric value 262 on the right side of the steady value. When the vehicle 12 hits the ground, a spike below -1g in the metric value 262 is shown, followed immediately by a bounce upward.

[0114] ​If the metric 262 exceeds the threshold 264 as shown in 266, the AMA_Z air classification metric 160 is triggered ON and outputs a Boolean value of 1 to the AND block 252. Once triggered ON, the Boolean value of 1 from the AMA_Z air classification metric 260 is maintained until the metric 262 falls below the threshold 264. The AMA_Z air classification metric 260 evaluates AMA_Z after subtracting the effect of gravity and determines the vehicle's air status when the vehicle's acceleration along the vehicle's Z-axis is zero. By tuning the bandwidth of the threshold 264, the AMA_Z air classification metric 260 can accurately classify the vehicle's air status.

[0115] AMA-Y vacation classification metric

[0116] The AMA_Y takeoff classification metric 270 evaluates AMA_Y over time to determine a metric value 272. This metric 270 implements a narrow threshold 274 centered at AMA_Y = 0, and has only a small bandwidth around this value. Under normal driving conditions, AMA_Y is expected to oscillate around the narrow bandwidth of threshold 274. When the metric value 272 crosses the threshold 274, the metric 270 triggers to ON and outputs a Boolean value of one, and remains ON while the metric value is within the bandwidth of the threshold.

[0117] exist Figure 9 In the example situation shown, metric 272 crosses threshold 274 and remains within its bandwidth for an exceptionally long period. These crossings into and out of threshold 274 are indicated at 276. During the duration that metric 272 remains within threshold 274, AMA_Y vacancy classification metric 270 is triggered ON and outputs a Boolean value of 1 to AND block 252. Once triggered ON, the Boolean value output of AMA_Y vacancy classification metric 270 is maintained until metric 272 leaves threshold 274.

[0118] It can be observed that the AMA_Y airborne classification metric 270 classifies vehicle airborne conditions in response to the vehicle's Y-axis acceleration remaining within a threshold 274 (which hovers around zero). This is consistent with expectations for airborne vehicles. When the vehicle is on the driving surface, the lateral (Y-axis) acceleration is as expected. When the vehicle is airborne, these accelerations are expected to decrease sharply due to the loss of contact between the vehicle and the driving surface. By tuning the bandwidth of threshold 274, the AMA_Y airborne classification metric 270 can accurately classify vehicle airborne conditions.

[0119] AMA-X vacation classification metric

[0120] The AMA_X takeoff classification metric 280 evaluates AMA_X over time to determine a metric value 282. This metric 280 implements a narrow threshold 284 centered at AMA_X = 0 and has only a small bandwidth around this value. Under normal driving conditions, AMA_X is expected to oscillate around the narrow bandwidth of threshold 284. When the metric value 282 crosses the threshold 284, the metric 280 triggers to ON and outputs a Boolean value of one, remaining ON while the metric value is within the bandwidth of the threshold.

[0121] exist Figure 9 In the example situation shown, metric 282 crosses threshold 284 and remains within its bandwidth for an exceptionally long period of time. These crossings into and out of threshold 284 are represented at 286. During the duration that metric 282 remains within threshold 284, AMA_X vacancy classification metric 280 is triggered ON and outputs a Boolean value of 1 to AND block 252. Once triggered ON, the Boolean value of AMA_X vacancy classification metric 280 is maintained until metric 282 leaves threshold 284.

[0122] As can be observed, the AMA_X airborne classification metric 280 classifies vehicle airborne status in response to the vehicle's X-axis acceleration remaining within a threshold 284 (which hovers around zero). This aligns with expectations for airborne vehicles. When the vehicle is on the driving surface, the longitudinal (X-axis) acceleration is as expected. When the vehicle is airborne, these accelerations are expected to decrease sharply due to the loss of contact between the vehicle and the driving surface. By tuning the bandwidth of threshold 284, the AMA_X airborne classification metric 280 can accurately classify vehicle airborne status.

[0123] D-RATE vacation classification metric

[0124] The D-RATE levitation classification metric 290 evaluates D_RATE over time to determine a metric value 292. This metric 290 implements a narrow threshold 294 centered at D_RATE = 0, and has only a small bandwidth around this value. Under normal driving conditions, D_RATE is expected to oscillate around the narrow bandwidth of threshold 294. When the metric value 292 crosses the threshold 294, the metric 290 triggers to ON and outputs a Boolean value of one, and remains ON while the metric is within the bandwidth of the threshold.

[0125] exist Figure 9In the example condition shown, the metric value 292 crosses the threshold 294 and remains within its bandwidth for an especially long amount of time. These crossings into and out of the threshold 294 are represented at 296. During the duration that the metric value 292 remains within the threshold 294, the D_RATE airborne classification metric 290 fires ON and outputs a Boolean value of 1 to the AND block 252. Once fired ON, the Boolean value output of the D_RATE airborne classification metric 290 remains so until the metric value 292 exits the threshold 294.

[0126] It can be observed that the D_RATE airborne classification metric 290 classifies the vehicle airborne condition in response to the vehicle roll acceleration remaining within the threshold 294, which hovers around zero. This is consistent with expectations for an airborne vehicle. When the vehicle is on a driving surface, roll acceleration about the X axis is expected. When the vehicle is airborne, these accelerations are expected to drop sharply due to the loss of contact between the vehicle and the driving surface. By tuning the bandwidth of the threshold 294, the D_RATE airborne classification metric 290 can accurately classify the airborne condition of the vehicle.

[0127] Vacation classification

[0128] The outputs of the AMA airborne classification metrics 260, 270, 280, 290 are provided to the AND block 252. As shown, in the case of a vehicle airborne condition, these metrics will turn ON and output a Boolean value of 1 at or about the same time, or at least partially overlapping times. Thus, in the case of a vehicle airborne condition, the AND block 252 will fire ON and output a Boolean value of 1 for at least a portion of the vehicle airborne event. Figure 9

[0129] The output of the AND block 252 is provided to the calibratable time classification block 254, which fires ON and outputs a Boolean value of 1 indicating that an airborne vehicle condition has been classified at block 256. The time classification block 254 remains on for a predetermined period of time, which can be tuned or calibrated to achieve a desired performance and / or response. This tuning can be vehicle platform specific, can be selected to meet desired manufacturer performance requirements, and can be selected in accordance with government and / or industry standards. The vehicle airborne classification at block 256 remains on for the calibration / tuned period of time of block 254.

[0130] Vacation confirmation algorithm

[0131] Referring to Figure 10 The ACU airborne confirmation algorithm 300 is implemented by the ACU 50. This algorithm 250 utilizes the ACU_Z accelerometer 56 (see Figure 2 ​) to confirm the airborne condition of the vehicle 12 classified by the airborne classification algorithm 250 (see Figure 9 ). To this end, the algorithm 300 implements AMA_Z airborne confirmation metrics 310, which are graphically illustrated in Figure 10 .

[0132] AMA-Z vacation confirmation metric

[0133] The AMA_Z airborne confirmation metrics 310 can be similar or identical to the AMA_Z airborne classification metrics 260 (see Figure 9 ), except for the determination of their threshold values. The AMA_Z airborne confirmation metrics 310 evaluate AMA_Z over time to determine a metric value 312. The AMA_Z airborne confirmation metrics 310 add a gravitational effect (-1g) to the measured acceleration, so that when the vehicle is airborne, the metric value 312 is zero. During a typical ride on a smooth surface, the AMA_Z airborne confirmation metric value 310 will thus hover around -1g. When the vehicle 12 launches to become airborne, the vehicle experiences a sharp rise in AMA_Z, as shown on the left side of the metric value 312. Once airborne, the metric value 312 settles around 0g. When the vehicle 12 descends, the AMA_Z falls, as shown by the metric value 312 on the right side of the steady value. When the vehicle 12 lands, a spike below -1g in the metric value 312 is shown, followed immediately by a bounce upward.

[0134] The AMA_Z airborne confirmation metrics 310 are configured to evaluate the metric value 312 to determine the landing of the vehicle after it is airborne. To this end, the AMA_Z airborne confirmation metrics 310 implement a threshold 314, which extends below the -1g AMA_Z value, and has a bandwidth selected to extend a predetermined amount below this -1g value, which is tunable. When the metric value 312 crosses the threshold, the AMA_Z airborne confirmation metrics 310 trigger as shown at 316. The metric value 312 that crosses the threshold 314 is initiated when the metric value first crosses the threshold, and is completed when the metric value passes below the threshold, again as shown at 316. When the AMA_Z airborne confirmation metrics 312 detect that the vehicle 12 has landed, the AMA_Z airborne confirmation metrics trigger ON, and output a Boolean value of one, which indicates that the airborne vehicle condition is confirmed, as shown at block 318. The vehicle airborne condition confirmation indication 318 is passed to the AND block 334. The vehicle airborne classification determination 256 (see Figure 9 ) is also passed to the AND block 334.

[0135] The airborne confirmation algorithm 300 also implements R_ANGLE off-road classification metrics 320 (at Figure 10(Illustrated graphically) to determine the off-road condition of vehicle 12. Determining that vehicle 12 is in an off-road condition will inhibit algorithm 300 from determining the vehicle's airborne condition. In other words, in order to determine the airborne condition, the vehicle must not be determined to be in an off-road condition.

[0136] R_ANGLE off-road classification metric 320 evaluates R_ANGLE over time to determine a metric value 326. If this metric value 326 crosses the upper threshold 322 in any order within a predetermined time period, as shown at 328, and crosses the lower threshold 324 as shown at 330, then R_ANGLE off-road classification metric 320 is triggered to ON and outputs a Boolean value of one, indicating that the vehicle off-road classification has been determined.

[0137] The output of R_ANGLE off-road classification metric 320 is passed to NOT block 332, whose output is passed to AND block 334. When the clearance classification 256 and clearance confirmation 318 are both triggered to ON, but the vehicle off-road classification determination is not triggered to ON (NOT block 332 is ON), AND block 334 is triggered to ON and outputs a Boolean value of 1, indicating that the vehicle clearance status has been determined and an ACR actuation command is issued, as shown in block 336. ACU 50 is configured to actuate ACR60 in response to the ACR actuation block 336 of clearance determination (see block 336). Figure 2 ).

[0138] From the above description of the invention, those skilled in the art will recognize improvements, variations, and modifications to the disclosed systems and methods, all of which fall within the spirit and scope of the invention. These improvements, variations, and / or modifications are intended to be covered by the appended claims.

Claims

1. A vehicle safety system, comprising: An actuable controlled restraint device includes a seat belt for restraining a vehicle occupant, the actuable controlled restraint device being actuable to control the pull-out and retraction of the seat belt; as well as A controller configured to determine the operating state of the vehicle and, in response to the determined operating state of the vehicle, control the actuation of the actuable controlled constraint device. The actuable controlled restraint device has a normal restraint state and an enhanced restraint state, and the controller is configured to actuate the actuable controlled restraint device from the normal restraint state to the enhanced restraint state in response to determining an abnormal driving condition of the vehicle. The controller is further configured to: implement a longitudinal acceleration airborne classification algorithm to classify the vehicle's airborne status in response to the vehicle's longitudinal acceleration measured along the vehicle's X-axis; implement a lateral acceleration airborne classification algorithm to classify the vehicle's airborne status in response to the vehicle's lateral acceleration measured along the vehicle's Y-axis; implement a vertical acceleration airborne classification algorithm to classify the vehicle's airborne status in response to the vehicle's vertical acceleration measured along the vehicle's Z-axis; and implement a roll acceleration airborne classification algorithm to classify the vehicle's airborne status in response to the vehicle's roll acceleration measured about the vehicle's X-axis.

2. The vehicle safety system according to claim 1, wherein, Under the normal restraint condition, the actuable controlled restraint device is configured to apply a relatively light retraction force sufficient to pull up the seat belt webbing and tension the seat belt across the occupant, while the actuable controlled restraint device is configured to pull out the seat belt webbing in response to occupant movement. as well as In the enhanced restraint condition, the actuable controlled restraint device increases the retraction force applied to the seat belt webbing and increases the resistance to the seat belt webbing being pulled out in response to occupant movement.

3. The vehicle safety system according to claim 1, wherein, The abnormal driving conditions of the vehicle include at least one of the following: vehicle condition on rough terrain and vehicle condition while airborne.

4. The vehicle safety system according to claim 1, wherein, The controller is an airbag control unit (ACU), comprising: The ACU_X accelerometer is used to measure vehicle acceleration along the X-axis of the vehicle and generate a signal indicating the vehicle acceleration along the X-axis of the vehicle. The ACU_Y accelerometer is used to measure the vehicle acceleration along the Y-axis of the vehicle and generate a signal indicating the vehicle acceleration along the Y-axis of the vehicle. An ACU_Z accelerometer, wherein the ACU_Z accelerometer is used to measure vehicle acceleration along the Z-axis of the vehicle and generate a signal indicating the vehicle acceleration along the Z-axis of the vehicle; and A ROLL sensor is used to measure the vehicle roll acceleration about the vehicle's X-axis and generate a signal indicating the vehicle roll acceleration. The ACU is configured to determine abnormal driving conditions of the vehicle in response to signals from the ACU_X, ACU_Y, ACU_Z and ROLL sensors.

5. The vehicle safety system according to claim 1, wherein, The controller is configured to implement a lateral acceleration rugged terrain classification algorithm to classify vehicle rugged terrain conditions in response to the vehicle's lateral acceleration measured along the vehicle's Y-axis. The controller is configured to implement a roll acceleration rugged terrain classification algorithm to classify vehicle rugged terrain conditions in response to vehicle roll acceleration measured about the vehicle's X-axis; and The controller is configured to implement a vertical acceleration rugged terrain classification algorithm to classify vehicle rugged terrain conditions in response to the vehicle's vertical acceleration measured along the vehicle's Z-axis.

6. The vehicle safety system according to claim 5, wherein, The controller is configured to determine the vehicle's rough terrain condition in response to the lateral acceleration rough terrain classification algorithm, the roll acceleration rough terrain classification algorithm, and the vertical acceleration rough terrain classification algorithm, which simultaneously classify the vehicle's rough terrain condition.

7. The vehicle safety system according to claim 1, wherein, The controller is configured to implement a lateral acceleration rugged terrain metric that evaluates the vehicle's lateral acceleration measured along the vehicle's Y-axis as changing over time. The lateral acceleration rugged terrain metric classifies the vehicle's rugged terrain condition in response to the vehicle's lateral acceleration measured along the vehicle's Y-axis exceeding a threshold in both positive and negative directions within a predetermined time period. The controller is configured to implement a roll acceleration rough terrain metric that evaluates the vehicle roll acceleration measured about the vehicle's X-axis over time. The roll acceleration rough terrain metric classifies the vehicle's rough terrain condition in response to the vehicle roll acceleration measured along the vehicle's X-axis exceeding a threshold in both positive and negative directions within a predetermined time period. The controller is configured to implement a vertical acceleration rugged terrain metric that evaluates the vehicle's vertical acceleration measured along the vehicle's Z-axis as changing over time, and classifies the vehicle's rugged terrain condition in response to the vehicle's vertical acceleration measured along the vehicle's Z-axis crossing a threshold in the positive direction.

8. The vehicle safety system according to claim 7, wherein, To determine the lateral acceleration rough terrain metric, the controller is configured to evaluate the moving average of the vehicle's lateral acceleration measured along the vehicle's Y-axis as a function of time. To determine the roll acceleration rough terrain metric, the controller is configured to evaluate a moving average of the vehicle roll acceleration measured about the vehicle's X-axis over time; and To determine the vertical acceleration rough terrain metric, the controller is configured to evaluate the moving average of the vehicle's vertical acceleration measured along the vehicle's Z-axis over time.

9. The vehicle safety system according to claim 1, wherein, The controller is configured to implement a vertical acceleration takeoff confirmation algorithm to confirm the vehicle's takeoff status in response to the vehicle's vertical acceleration measured along the vehicle's Z-axis.

10. The vehicle safety system according to claim 9, wherein, The vertical acceleration takeoff confirmation algorithm is configured to determine the vehicle's landing status in response to the vehicle's vertical acceleration measured along the vehicle's Z-axis.

11. The vehicle safety system according to claim 9, wherein, The controller is configured to determine the vehicle's vacancy status in response to classifying and confirming the vehicle's vacancy status.

12. The vehicle safety system according to claim 11, wherein, The controller is configured to further determine the vehicle's airborne status in response to determining that the vehicle's off-road condition has not yet been determined.

13. The vehicle safety system according to claim 12, wherein, The controller is configured to perform an R_ANGLE off-road classification metric on the vehicle's off-road condition, the R_ANGLE off-road classification metric evaluating the vehicle's R_ANGLE over time, and classifying the vehicle's off-road condition in response to the R_ANGLE crossing both an upper and lower threshold in any order within a predetermined time period.

14. The vehicle safety system according to claim 1, wherein, The controller is configured to implement a longitudinal acceleration takeoff metric that evaluates the vehicle's longitudinal acceleration measured along the vehicle's X-axis as varying over time. The metric is defined with a threshold having a predetermined bandwidth extending above and below the zero point of the longitudinal acceleration takeoff metric. The longitudinal acceleration takeoff metric is configured to classify the vehicle takeoff status in response to the magnitude of the metric value exceeding the threshold and falling back within the threshold. The longitudinal acceleration takeoff metric maintains the classification of the vehicle takeoff status when the metric value remains within the threshold. The controller is configured to implement a lateral acceleration take-off metric that evaluates a vehicle's lateral acceleration measured along the vehicle's Y-axis over time, the metric defining a threshold having a predetermined bandwidth extending above and below the zero point of the lateral acceleration take-off metric, wherein the lateral acceleration take-off metric is configured to classify vehicle take-off status in response to the magnitude of the metric value exceeding and falling back within the threshold, and to maintain the classification of vehicle take-off status when the metric value remains within the threshold; and The controller is configured to implement a roll acceleration takeoff metric that evaluates a vehicle roll acceleration measured about the vehicle's X-axis over time, the metric being defined by a threshold having a predetermined bandwidth extending above and below the zero point of the roll acceleration takeoff metric, wherein the roll acceleration takeoff metric is configured to classify a vehicle takeoff condition in response to the amplitude of the metric value exceeding and falling back within the threshold, and to maintain the classification of the vehicle takeoff condition when the metric value remains within the threshold; and The controller is configured to implement a vertical acceleration takeoff metric that evaluates the vehicle's vertical acceleration measured along the vehicle's Z-axis as varying over time. The metric is defined by a threshold having a predetermined bandwidth extending above and below the zero point of the vertical acceleration takeoff metric. The vertical acceleration takeoff metric is calibrated to compensate for the effects of gravity. The vertical acceleration takeoff metric is configured to classify the vehicle's takeoff status in response to the metric value crossing the threshold, and to maintain the classification of the vehicle's takeoff status while the metric value remains within the threshold.

15. The vehicle safety system according to claim 14, wherein, The controller is configured to classify the vehicle's vacant status in response to determining that all of the following conditions are true simultaneously: The longitudinal acceleration airborne measurement classifies the vehicle's airborne status; The lateral acceleration airborne measurement classifies the vehicle's airborne status; The vertical acceleration take-off metric is used to classify vehicle take-off status; and The lateral acceleration airborne metric is used to classify the vehicle's airborne status.

16. The vehicle safety system according to claim 15, wherein, The controller is configured to maintain the classification of the vehicle's vacant status for a predetermined period of time.

17. The vehicle safety system according to claim 15, wherein, The controller is configured to implement a vertical acceleration takeoff confirmation metric, which evaluates the vehicle's vertical acceleration measured along the vehicle's Z-axis as changing over time. The metric is defined with a threshold having a predetermined bandwidth, which is selected such that the metric value passes through the bandwidth in response to landing after takeoff. The controller is configured to confirm the vehicle's takeoff status in response to the metric value.

18. The vehicle safety system according to claim 17, wherein, The controller is configured to determine the vehicle's vacancy status in response to classifying and confirming the vehicle's vacancy status.

19. The vehicle safety system according to claim 1, wherein, The controller includes an airbag controller unit (ACU) configured to control the actuation of one or more airbags and the actuable controlled restraint device.

Images