Device and method for increasing occupant protection in a self-driving car

Abstract

Device (2) for increasing occupant protection in an ego vehicle (1), the device (2) comprising at least one reversibly adjustable actuating device for applying forces to a belt tensioner (3) for reversibly moving the belt tensioner (3) from a starting position to an impact position, and a control unit (5) for controlling the actuating device, further comprising a first sensor for detecting the traffic situation as first environmental data with a first quality and at least a second sensor for detecting the traffic situation as second environmental data with a second quality, wherein the control unit (5) is configured toto interweave the first environmental data as well as the at least second environmental data with each other as interweaving environmental data and to cascade the force exerted on the actuating device depending on the quality of the interweaving environmental data and a collision probability recognized in the interweaving environmental data, wherein the first sensor and the at least second sensor are designed as identical sensors, wherein the first environmental data are recorded at a first time and the at least second environmental data are recorded at a subsequent second time, and the control unit (5) is designed to interweave the first environmental data as well as the at least second environmental data with each other as interweaving environmental data.

Description

[0001] The invention relates to a device for increasing occupant protection in a self-driving car, the device comprising at least one reversibly adjustable actuating device for applying forces to a seatbelt tensioner to reversibly move the seatbelt tensioner from a starting position to an impact position, and a control unit for controlling the actuating device, further comprising a first sensor for detecting the traffic situation as first environmental data of a first quality and at least a second sensor for detecting the traffic situation as second environmental data of a second quality. The invention further relates to a method.

[0002] Vehicles usually feature reversible occupant restraint systems for passive safety, which can be activated by an automatic control mechanism.

[0003] Examples of reversible occupant restraint systems include reversible motorized seatbelt pretensioners, which tighten the seatbelts fastened by the vehicle occupants in dangerous situations to secure them in their seats. After a collision is detected, the seatbelt is tightened, thus restraint the buckled-up occupant. This is achieved by driving a spool, which winds and unwinds the seatbelt, a short distance in the winding direction, thereby tightening the seatbelt.

[0004] Other reversible occupant restraint systems include, for example, automatic seat adjustment, which allows the seats to be brought into an upright position in dangerous situations, as well as electric window lifters or devices with which an existing sunroof can be closed in collision situations.

[0005] Furthermore, in the future, irreversible occupant protection devices, such as airbags, which cannot be put into an inactive state by a corresponding automatic control mechanism after being triggered, can at least partially be implemented reversibly.

[0006] Such occupant protection devices are usually triggered upon detection of an accident, i.e., a collision, or only shortly before.

[0007] ISO 26262 ("Road vehicles - Functional safety") is an ISO standard for safety-related electrical / electronic systems in motor vehicles. ISO 26262 defines a process model along with required activities and work products, as well as methods to be applied in development and production. ISO 26262 mandates the performance of a hazard and risk analysis as a method for determining risk. In this analysis, the risk is considered and analyzed for each function of the component. This analysis then results in an ASIL (Advanced Safety Integrity Level) for the respective function. ASIL thus represents a scheme for risk assessment. There are four ASIL classes: ASIL A, ASIL B, ASIL C, and ASIL D. ASIL D represents the highest safety requirements and necessitates correspondingly stringent safety precautions.

[0008] The implementation of this standard is intended to ensure the functional safety of a system with electrical / electronic components in a motor vehicle. Therefore, the standard is an adaptation of IEC 61508 to the specific products in the automotive sector.

[0009] The state of the art also describes systems for optimizing triggering times. Triggering algorithms used to distinguish between triggering and non-triggering cases employ, for example, an evaluation of signals that are usually stronger in a triggering case than in a non-triggering case.

[0010] DE 10 2009 011 580 A1 discloses a method for warning a driver in a vehicle with at least one reversible belt tensioner, in which the belt tensioner, depending on a hazardous situation detected by a detector system and evaluated by an evaluation unit, is controlled to change the pretension of a belt as a warning signal for the driver, wherein parameters for triggering and influencing the warning signal are specified by the evaluation unit, which are set according to the driver's specifications via an input device connected to the evaluation unit in order to store a driver-specific warning profile.

[0011] EP 0 816 188 B1 discloses a triggering circuit for an occupant restraint system for motor vehicles with at least one belt tensioner for tightening a seat belt associated with a vehicle seat, wherein the triggering circuit evaluates the acceleration signal of an accelerometer and, upon detection of a critical impact, sends a triggering signal to the belt tensioner, wherein the triggering circuit has an evaluation channel for detecting a rear-end collision, wherein the evaluation channel comprises: a displacement detection stage with which, based on the acceleration signal occurring in a rear-end collision, a displacement signal characterizing the occupant displacement can be determined, and a displacement threshold switch for the displacement signal downstream of the displacement detection stage, which sets a high signal when a displacement threshold is exceeded by the displacement signal.In the event of a rear-end collision, this high-level signal is necessary for the release signal to the seatbelt pretensioner.

[0012] DE 10 2009 025 021 A1 discloses a method for controlling a restraint device for vehicle occupants. DE 10 2017 214 613 A1 discloses a method for protecting at least one occupant of a motor vehicle.

[0013] It is therefore an object of the invention to provide an improved device and a method for increasing the occupant protection of occupants of an ego vehicle.

[0014] The problem is solved by a device having the features of claim 1 and a method having the features of claim 11.

[0015] Advantageous embodiments of the invention are described by the dependent subclaims and the following description as well as the figures.

[0016] The problem is solved by a device for increasing occupant protection in an ego vehicle, the device comprising at least one reversibly adjustable actuating device for applying forces to a belt tensioner for reversibly moving the belt tensioner from a starting position to an impact position, as well as a control unit for controlling the actuating device, further comprising a first sensor for detecting the traffic situation as first environmental data with a first quality and at least a second sensor for detecting the traffic situation as second environmental data with a second quality, and wherein the control unit is configured toto intertwine the first environmental data as well as at least the second environmental data as intertwined environmental data and, depending on the quality of the intertwined environmental data and a collision probability recognized in the intertwined environmental data, to cascade and adjust the force exerted on the actuating device.

[0017] The probability of a collision can be viewed as a function of the trajectory of the ego-vehicle and the time.

[0018] The force exerted on the actuating device is the force that is applied to the actuating device.

[0019] It is known that reversible actuators can be activated even before a potential collision. Such reversible actuators could, for example, be a reversible drive for the seatbelt tensioner.

[0020] However, this requires the recognition of an existing traffic situation and its assessment with regard to the possibility of an impending collision and the type and severity of the collision; as well as the recognition of the probability of a collision with a collision time of high accuracy.

[0021] The accuracy of environmental sensing varies depending on the sensing principle. Broadly speaking, a distinction can be made between direct measurement of physical quantities such as acceleration, pressure, and rotational speed, and environmental sensors such as cameras, radar, lidar, etc. Directly measured physical quantities are independent of external influences and therefore generally have a higher accuracy than environmental sensors.

[0022] When measuring physical quantities, a direct measurement of the accident is only possible during or after the collision. However, critical traffic situations (skidding) can be detected with high accuracy using directly measured quantities even before the crash. With environmental sensors, such as cameras, radar, etc., these capabilities fluctuate depending on external influences like the weather. A camera's vision is impaired in darkness. Radar can detect certain materials and surfaces better than others. Furthermore, the information from environmental sensors must be interpreted as a prediction, which is potentially prone to error. In addition, the accuracy of the necessary prediction into the future is inherently limited.

[0023] Within the class of environment-sensing sensors, there are therefore further clear distinctions in quality.

[0024] According to the invention, it was therefore recognized that environmental perception has difficulties providing high-quality information early on for a variety of reasons. Weather conditions, as described above, are one example. For instance, a camera performs worse in darkness than in good lighting conditions, which makes both the probability and the timing of a collision less accurate.

[0025] According to the invention, it was further recognized that the controllability of the ego-vehicle must be ensured. The controllability of the ego-vehicle is influenced, among other things, by the force acting on the occupant, depending on the traffic situation, and decreases with increasing force. For this purpose, threshold values ​​for driver distraction are defined according to, for example, ISO 26262 for various traffic situations.

[0026] The device and the cascaded setting of the belt tensioner now make it possible, on the one hand, to trigger the actuating device early with low force and low probability of collision due to low quality of the environmental data, and on the other hand, to trigger it with high force shortly before the collision when a high probability of collision is detected due to high quality of environmental data.

[0027] This allows the occupants to be moved into or held in an optimal position in as many critical traffic situations as possible. The number of critical traffic situations is referred to as field coverage.

[0028] By considering the entire system and taking a holistic view of the details of the traffic situation, the effectiveness of the device can be optimized with lower requirements for early, high-quality detection.

[0029] The device according to the invention now makes it possible to trigger the actuating device earlier with lower force in the event of a collision that is only considered probable, even if the environmental detection by the sensors has only a low quality.

[0030] The device according to the invention now allows both the force exerted on the actuating device and the triggering time associated with the force to be cascaded and adjusted depending on the probability of collision as well as the quality of the interlocked environmental data.

[0031] For example, if a collision is detected at a late stage and has a low probability of occurrence (i.e., medium to low sensor quality), only a small force is exerted on the actuating device. This results in a shift in the reaction time. For instance, with a seatbelt pretensioner, the seatbelt is tightened slowly, bringing the driver into an upright, safe position. Such a slow adjustment also eliminates the need for high-quality early detection of a collision, including precise timing.

[0032] The potential side effects of forceful activation of the operating device, especially in the case of a false alarm, such as distraction or even loss of control over the vehicle at very high force levels, can thus be avoided.

[0033] For example, if a collision with a high probability of occurrence (i.e., high sensor quality) is detected at an imminent moment, a maximum force is exerted on the actuating device. This results in a shift in the deflection threshold. The quality of the environmental data correlates with time; for instance, the quality of the environmental data is generally high when a collision is imminent. By combining premature triggering with a high force applied shortly before the collision, the risk of a false triggering with a maximum force acting on the seatbelt pretensioner is reduced.

[0034] Such a device can reliably cover a wide range of traffic situations without placing the driver in a situation where, for example, a false alarm and subsequent seatbelt tightening could distract them and cause a collision. This increases safety in many traffic situations, resulting in broad coverage.

[0035] The device according to the invention allows the force to be cascaded depending on the time span from the triggering time to the time of collision, so that the belt tensioner reaches its final impact position shortly before or at least during the time of collision.

[0036] By considering the entire system, preferably taking into account the whole situation, or at least the environment, a high field coverage is achieved according to the invention despite conflicting requirements for the force level, by means of an interlinking of (sensor) information, thereby enabling a cascaded triggering of the actuating device. This can, for example, involve shifting the trigger point forward or adjusting the maximum force level to the situation.

[0037] Such entanglement generally improves the quality of the environmental data; this can involve temporal entanglement of a sensor or its environmental data, but also entanglement of different sensors or their environmental data.

[0038] This allows the occupants to be brought into or held in an optimal position in as many critical situations as possible. According to the invention, the control unit is designed to derive an adaptable escalation strategy in a critical traffic situation, depending on the probability of a collision and on other variables of varying quality that can be detected in a traffic situation. In order to bring or hold the occupants in an optimal position in as many situations as possible, the triggering strategy of the actuating device must be adapted for different traffic situations. An example of a triggering strategy is early triggering of the actuating device when information is available that the brake assist system will initiate emergency braking.A second example is the adjustment of the maximum force, which, however, is linked to the distraction threshold for an occupant, especially the driver, if the probability of collision is very high and if the quality of the entangled environmental data is very high and possibly if additional information such as lateral acceleration is known.

[0039] Depending on the quality and probability of collision, which describe the traffic situation and the possible further course of events, the force exerted on the actuating device is cascaded and adjusted.

[0040] The device according to the invention utilizes the operating time of the actuating device as well as the force level without risking a loss of controllability on the part of the driver, while reaching the impact position of the belt tensioner in a timely manner.

[0041] A reversible, early setting eliminates the need for high-quality environmental perception. Errors in environmental perception can therefore be safely accepted, as the actuation device is reversibly designed and a false triggering does not result in any consequential costs, such as repairs, and the device is immediately ready for use again. Furthermore, the cascaded setting, particularly with an early trigger point and low, non-critical force, prevents driver distraction or injury.

[0042] In the embodiment according to the invention, the first sensor and the at least second sensor are designed as identical sensors, wherein the first environmental data are recorded at a first time and the at least second environmental data are recorded at a subsequent second time, and the control unit is designed to entangle the first environmental data as well as at least the second environmental data with each other as entangled environmental data.

[0043] Furthermore, the first sensor and at least one second sensor can be designed as separate sensors.

[0044] Thus, horizontal interlocking (temporal interlocking) and vertical interlocking (multiple different sensors) can be performed.

[0045] Furthermore, multiple sensors of different types can be present, such as camera sensors or radar sensors, which are intertwined both with each other and in time. Sensors relating to the vehicle's dynamics, such as deceleration, lateral acceleration, and vehicle speed, which are measured, for example, by accelerometers, pressure sensors, and speed sensors, can also be incorporated into the intertwining process. This allows for efficient improvement of the quality of the intertwined environmental data. Environmental data from interior sensors can also be intertwined, for example, data from sensors that detect hand position on the steering wheel, driver attention, eye position, passenger seating position, and passenger characteristics such as age, weight, height, gender, etc.

[0046] In further training, the probability of collision includes at least the probability of a collision as well as the time of the collision.

[0047] Furthermore, the control unit can be configured as a cascaded force to set at least the triggering point for the force exerted on the actuating device, the force itself, and the retraction speed of the seatbelt tensioner. This cascaded force can, for example, be a dynamically stepped force. The retraction speed is the speed at which the actuating device retracts the seatbelt.

[0048] In further training, the control unit can be configured to cascade the force according to the time interval between the triggering time and the collision time, so that the seatbelt pretensioner reaches its final impact position shortly before or at least during the collision. In particular, the control unit can cascade the force according to the available calculated time interval between the triggering time and the collision time, so that the seatbelt pretensioner and the seatbelt reach their final impact position shortly before the collision time.

[0049] In a further embodiment, the control unit is designed to exert a small force on the actuating device when the quality level is below a first quality threshold in conjunction with a maximum first collision probability threshold, thus enabling a slow, smooth movement of the seatbelt pretensioner from its initial position towards the impact position. Smooth movement means movement without force. A small force is a low force, allowing for a soft, gentle, or fluid movement of the seatbelt pretensioner. A maximum first collision probability threshold indicates a collision probability that lies below the first collision probability threshold.

[0050] This force can vary depending on the seatbelt pretensioner itself. It can involve a prompt or immediate triggering of the release mechanism.

[0051] This smooth adjustment prevents driver distraction or injury caused by sudden movements. Such an initial collision probability threshold for a seatbelt pretensioner can, for example, be in the range below 95-100%; meaning that if the collision probability threshold is exceeded, a collision is almost certain.

[0052] In further training, the control unit is designed to perform haptic, visual, and / or acoustic preconditioning via the actuating device when the quality level falls below a second quality threshold in conjunction with a maximum second collision probability threshold that is lower than the first collision probability threshold. For example, haptic preconditioning can occur if only a 50% second collision probability threshold is detected. This could involve the seatbelt tensioner vibrating or flashing briefly. This allows the driver to be alerted to a potential collision and take appropriate action. A maximum second collision probability threshold indicates a collision probability that is lower than the second collision probability threshold.

[0053] In further development, the control unit is designed to exert a maximum force on the actuating device, particularly in the event of immediate triggering, when the quality of the belt tensioner exceeds a first quality threshold in conjunction with a minimum first collision probability threshold. This ensures rapid movement of the belt tensioner from its initial position towards the impact position. For example, this allows the maximum force, or the original maximum force level, to be increased in the event of an imminent and unavoidable accident that could not be foreseen, in order to protect the occupants more quickly. A minimum first collision probability threshold indicates a collision probability that lies above the first collision probability threshold.

[0054] In further development, the control unit is configured to exert a small force on the actuating device at a first triggering time when the quality level is below a first quality threshold in conjunction with a maximum first collision probability threshold, and to exert a maximum force at a collision probability level above the first quality threshold in conjunction with a first minimum collision probability threshold, triggering at a second triggering time, where the first triggering time precedes the second triggering time. A minimum first collision probability threshold indicates a collision probability that lies above the first collision probability threshold; a maximum first collision probability threshold indicates a collision probability that lies below the first collision probability threshold.

[0055] This is particularly relevant if the seatbelt pretensioner has not yet fully moved into the impact position, for example, because the probability of a collision or the associated collision time was incorrectly determined. This allows the pretensioner to be moved into the impact position with maximum force if it is currently positioned between its initial and impact positions. This increases occupant protection. This is especially advantageous because less distance needs to be covered to reach the final impact position than from the initial position, allowing the pretensioner to move into the impact position quickly.

[0056] In further development, the control unit is trained to also integrate vehicle dynamics sensor data into the integrated environmental data to improve accuracy, whereby the vehicle dynamics sensor data includes, in particular, at least the brake sensor data. Vehicle dynamics sensor data (parameters) include, for example, deceleration, lateral acceleration, vehicle speed, as well as the intervention steps of the vehicle's own warning and intervention safety systems, and the vehicle's own characteristics such as size, mass, stiffness, load, tire pressure / condition, number of seats, etc.

[0057] This can significantly improve the quality, enabling early detection of a collision.

[0058] In particular, the actuating device can be designed as a reversibly adjustable actuator or positioning device.

[0059] In further training, the control unit is trained to consider specific traffic situations detected in the interlocking environmental data when cascading the applied force. These situations can be selected from at least the following: high lateral acceleration above a predefined lateral acceleration value and / or high vehicle speed above a predefined speed value. This corresponds to traffic situations in which small steering angles pose an increased risk of collision, for example, when driving through construction zones. In these situations, for instance, a sudden jerk of the steering wheel, perhaps due to high seatbelt tension, can quickly lead to accidents. Therefore, the applied force can be adjusted to the driving situation, for example, reduced, to avoid these situations.

[0060] This adjustment reduces the risk of the driver being distracted by the seatbelt once a certain force or increase in force is reached. The risk of this distraction leading to an accident depends heavily on the traffic situation and the seatbelt pretensioner itself, and can potentially be reduced. In particular, the original maximum force level can be adjusted to the driving situation.

[0061] In further training, the control unit is designed to dynamically release the force on the actuating device if a collision does not occur or if a false diagnosis is detected. This can be achieved either by a rapid release when the belt tensioner is in the impact position or by a slow release if the belt tensioner has not yet moved into the impact position. A false diagnosis is defined as the incorrect detection of a collision when no collision actually takes place.

[0062] Furthermore, the problem is solved by a procedure for increasing occupant protection in an ego vehicle, comprising the following steps: - Providing at least one reversibly adjustable actuating device for applying forces to a belt tensioner for reversibly moving the belt tensioner from a starting position to an impact position, - Providing a control unit for controlling the actuating device, - Providing a first sensor to record the traffic situation as initial environmental data of a first quality and at least a second sensor to record the traffic situation as secondary environmental data of a second quality, - Intertwining the at least first environmental data as well as the at least second environmental data with each other as intertwined environmental data, - Cascaded adjustment of the force exerted on the actuating device depending on the quality of the entangled environmental data and a collision probability detected in the entangled environmental data, wherein the first sensor and the at least second sensor are designed as identical sensors, wherein the first environmental data are recorded at a first time and the at least second environmental data are recorded at a subsequent second time, and wherein the first environmental data as well as the at least second environmental data are entangled with each other as entangled environmental data.

[0063] The advantages of the device can then be transferred to the process.

[0064] Furthermore, the method is particularly suitable for being carried out on the device according to the invention. Moreover, the advantageous embodiments of the device can also be transferred to the method.

[0065] Furthermore, in addition to the interlocking environmental data, interior sensor data from an interior sensor can also be interlocked, as well as additional vehicle dynamics sensor data, whereby the vehicle dynamics sensor data includes at least the brake sensor data.

[0066] Further features and advantages of the present invention will become apparent from the following description with reference to the accompanying figures. These schematically illustrate: Fig. 1: an ego vehicle with a device according to the invention, Fig. 2: early triggering as qualitative field coverage depending on the belt force, Fig. 3: an effect of the distraction threshold on the trigger rate or field coverage.

[0067] In a rough classification, one can distinguish between direct measurement of physical quantities, such as acceleration, pressure, rotational speed, and environment-sensing sensors, such as camera 7, radar 7a, lidar sensors 7b, etc.

[0068] However, direct measurement of the accident is only possible at / after the collision. Nevertheless, critical traffic situations (skidding) can be detected with a high degree of accuracy using directly measured parameters even before the crash.

[0069] Directly measured physical quantities, such as those provided by environment-sensing sensors, are independent of external influences and therefore, in principle, have a higher quality than information from environment-sensing sensors.

[0070] With environmental sensors, such as a camera 7, a radar 7a, or lidar sensors 7b, these capabilities fluctuate depending on external influences like weather. A camera 7 has reduced visibility in darkness. A radar 7a can detect certain materials and surfaces better than others. The information from the environmental sensors must also be interpreted (perception), which is potentially prone to errors.

[0071] Furthermore, the accuracy of the necessary prediction into the future is inherently limited.

[0072] Within the class of environment-sensing sensors, there are therefore further clear distinctions in quality.

[0073] Fig. Figure 1 shows an ego-vehicle 1 with a device 2 according to the invention for increasing occupant protection in the ego-vehicle 1. The ego-vehicle 1 has several environment-sensing sensors, here, for example, cameras 7 and radar sensors 7a and lidar sensors 7b and / or ultrasonic sensors for generating environmental data depicting the environment of the ego-vehicle 1. The cameras 7 generate first environmental data, the radar sensors 7a second environmental data, and the camera lidar sensors 7b third environmental data, each depicting the environment.

[0074] The surrounding data can be used to estimate, for example, the probability of a collision and the possible severity of the collision, as well as to identify the lane width and the traffic volume as further environmental information.

[0075] The device 2 has at least one reversibly adjustable drive 9 for applying forces to the belt tensioner 3 for reversibly moving the belt tensioner 3 from a starting position to an impact position.

[0076] The belt tensioner 3 is designed to tighten a safety belt 4 assigned to a vehicle seat. The drive 9 can be designed as, or have, a reversible electrically operated actuator or control element.

[0077] Furthermore, the device 2 has a control unit 5. The control unit 5 can effect a reversible adjustment of the belt tensioner 3 and thus of the safety belt 4 by means of the drive 9.

[0078] The control unit 5 can interweave all environmental data. This interweaving can be vertical, i.e., all environmental data with each other, as well as horizontal, i.e., temporal. Such interweaving can improve the quality of the individual environmental data.

[0079] Furthermore, the control unit 5 can have an evaluation unit (not shown) for detecting a possible collision with a collision probability and a collision time in the entangled environment data.

[0080] Based on the collision probability and the quality of the interlocked environmental data, a cascaded activation of the electric drive 9 can now be enabled. This could, for example, involve shifting the activation point forward or adjusting the maximum force level to the traffic situation.

[0081] Depending on the quality and content of the information available to the device 2 from the intertwined environmental data, which describe the traffic situation and the possible further course, a trigger time, a force and a tightening speed of the belt tensioner 3 are adjusted.

[0082] Thus, depending on the probability of collision as well as the quality of the entangled environmental data, both the force exerted on the drive 9 and the triggering time associated with the force can be cascaded and adjusted.

[0083] Based on the collision probability and the quality of the interlocked environmental data, the control unit 5 can now cascade and adjust the force exerted on the belt tensioner 3 according to the time interval between the triggering time and the collision time, as well as the collision probability. In particular, the control unit 5 can cascade and adjust the force according to the available calculated time interval between the triggering time and the collision time in such a way that the belt tensioner 3 and the seat belt 4 reach their final impact position shortly before the collision time.

[0084] The device 2 and the cascaded adjustment of the belt tensioner 3 enable, on the one hand, early activation of the drive 9 in cases of low force and low probability of collision due to low quality of the interlocked environmental data, and on the other hand, activation with high force shortly before a collision in cases of a detected high probability of collision due to interlocked environmental data with high quality. This allows the occupants to be brought into or held in an optimal position in as many critical traffic situations as possible. The number of critical traffic situations is referred to as field coverage.

[0085] The device 2 according to the invention now makes it possible to trigger the drive 9 earlier with lower force in the event of a collision that is only considered probable, even if the environmental detection by the sensors has a very low quality.

[0086] Thus, the vehicle remains controllable at all times. The controllability of the Ego Vehicle 1 is influenced, among other things, by the force acting on the occupant, depending on the traffic situation, and decreases with increasing force, thereby raising the distraction threshold. These thresholds are defined according to ISO 26262 for various traffic situations.

[0087] Fig. Figure 2 schematically shows such an early triggering as qualitative field coverage depending on the belt force in a braked ego vehicle 1.

[0088] If a low quality level is present in conjunction with a maximum second collision probability threshold that is below a first collision probability threshold, haptic and / or visual and / or acoustic preconditioning can initially be achieved using drive 9. For example, if only a 50% second collision probability threshold is detected, haptic preconditioning can be performed. For instance, the seatbelt tensioner 3 can vibrate or flash briefly. This allows the driver to be alerted to a potential collision and appropriate measures to be taken. A maximum second collision probability threshold indicates a collision probability that is below the first collision probability threshold.

[0089] The quality of the data is directly related to the probability of collision; thus, a low quality of data usually does not result in a high probability of collision, as the data is too uncertain.

[0090] If the quality level is medium to low, the maximum first collision probability threshold is above the second collision probability threshold, and the collision occurs at a later time, then only a small force is exerted on the belt tensioner 3 by the drive 9. This results in a shift in the effective time. The quality level can be increased by temporally interlocking the environmental data or by interlocking several pieces of information of varying quality.

[0091] In addition to temporal merging, different sensor data can also be merging, for example, sensor data from sensors used for environmental perception and sensor data from sensors relating to vehicle dynamics, such as deceleration, especially braking, lateral acceleration, or vehicle speed. Examples of such sensors are acceleration sensors 8, pressure sensors 8a, and speed sensors 8b ( Fig. 1), which are incorporated into the entanglement accordingly. This allows the quality of the entangled environmental data to be efficiently improved. For example, environmental data can also be combined with the interior sensor data from an interior sensor 6 ( Fig. 1) are intertwined, for example sensor data for detecting a hand position on the handlebars, the driver's attention, an eye position, a seating position of the occupants and an occupant's constitution such as age, weight, height, gender, etc.

[0092] The control unit 5 also interweaves the environmental data, the interior sensor data and the vehicle dynamics sensor data over time, thereby continuously improving the quality.

[0093] This means that control unit 5 continuously evaluates the environmental data, taking into account the vehicle dynamics sensor data and the interior sensor data, and initiates, for example, the triggering of automatic braking. From the perspective of control unit 5, this increases the certainty that braking will occur.

[0094] Control unit 5 uses this information and knowledge about the brake latency times and triggers early with the necessary force level to bring the occupant into an upright position in the necessary time.

[0095] This results in a first cascade, namely the optical, acoustic or haptic preconditioning, and the second cascade, namely the slow reversible actuation of the electric drive 9 and thus the exertion of a small force by the drive 9 on the belt tensioner 3.

[0096] This allows many traffic situations to be covered without violating any ASIL safety requirements.

[0097] The diagram plots the percentage of field coverage against the force. It shows that a higher field coverage is achieved without changing the force level by triggering the drive earlier. This means that shortly before the emergency braking engages, the occupant is pulled into an upright position by the seatbelt without having to exert force against the deceleration (jump S1).

[0098] With a belt tensioner 3, the seat belt 4 is tightened slowly, bringing the driver into an upright, safe position. This slow adjustment eliminates the need for highly accurate early detection of a collision, including the precise moment of impact. The low force applied allows for a smooth and gentle movement of the belt tensioner 3.

[0099] Thus, if the probability of a collision is below a certain threshold, the control unit 5 exerts a small force on the belt tensioner 3, enabling a slow, non-jumping movement of the seat belt 4 from its initial position towards the impact position. This means that the belt tensioner 3 slowly retracts the seat belt 4, gradually tightening it to provide maximum restraint for the buckled-up vehicle occupant in the impact position.

[0100] This smooth adjustment prevents distraction or injury to the driver caused by sudden movements. The risk of the driver being distracted by the seatbelt (position 4) above a certain force / increase in force can thus be avoided.

[0101] The area in which a distraction occurs due to the fastened seatbelt 4 is specified by the ASIL (x) area, the area in which no distraction occurs is specified as the QM area.

[0102] Thus, an extended operating time of the belt tensioner 3 as well as the force level is utilized without risking a loss of controllability on the part of the driver, provided the impact position of the seat belt 4 is reached in time.

[0103] This makes it possible to trigger the belt tensioner 3 earlier in the event of a collision that is identified as likely, even if the quality of the collision detection (low collision probability) is low.

[0104] Fig. Figure 3 shows the effect of the deflection threshold on the trigger rate or field coverage.

[0105] Due to the requirements of the overall system, very high demands are placed on sensors and prediction: from the distraction threshold (ASIL range) onwards, both false triggering by sensors and the prediction of the collision probability must be reduced to a minimum. Therefore, the potential field coverage drops sharply from the distraction threshold onwards, i.e., with increasing force, to prevent the driver from being distracted by the seatbelt retraction above a certain force / force increase.

[0106] The control unit 5 is designed to exert a maximum force on the drive 9 when the quality is above a first quality threshold in conjunction with a minimum first collision probability threshold, so that a rapid movement of the belt tensioner from the starting position towards the impact position is achieved.

[0107] This allows, for example, the maximum force or the original maximum force level to be increased in the event of an imminent and unavoidable accident that could not be detected beforehand, in order to protect the occupants more quickly. A minimum first collision probability threshold indicates a collision probability that lies above the first collision probability threshold. The quality of the interlocked environmental data can be improved by interlocking the environmental data, for example, by detecting strong lateral acceleration, panic braking by occupants, or the initiation of automatic braking by a brake assist system, thus enabling the detection of an unavoidable collision. Since the quality of the sensor data also correlates with time, such an unavoidable collision can usually only be detected with a high degree of certainty immediately before or shortly before the accident.In such a traffic situation, the maximum force level can be increased. Prior to this, the occupant can undergo haptic, acoustic, or visual preconditioning.

[0108] The control unit 5 is also designed to take into account traffic situations in which small steering angles pose an increased risk of collision when cascading the adjustment of the applied force, namely by reducing the force level.

[0109] These include, in particular, the following traffic situations, which are recognized from the interlocking environmental data: high lateral acceleration exceeding a predefined lateral acceleration value and / or high vehicle speed exceeding a predefined speed value of the ego-vehicle. These are traffic situations in which small steering angles pose an increased risk of collision, for example, when driving through construction zones. In these situations, for instance, a sudden jerk of the steering wheel, perhaps due to excessive seatbelt tension, can quickly lead to accidents. Therefore, the force applied can be adjusted to the driving situation, for example, reduced, to avoid these situations.

[0110] This allows for optimization of the force level based on the traffic situation detected in the interlocking environmental data, thus ensuring maximum field coverage. This enables an expansion of the operating range into scenarios with higher fault tolerance.

[0111] The area in which a distraction occurs due to the fastened seatbelt 4 is specified by the ASIL area, and the area in which no distraction occurs is specified as the QM area.

[0112] Such a shift in the force level, here referred to as adjustment G1, can thus cause a shift in the deflection threshold.

[0113] By considering the interlocking environmental data, i.e., by considering the entire system, it is possible to take the entire traffic situation into account. Through the interlocking of (sensor) information, particularly from the environmental sensors, the vehicle dynamics sensors, and, if applicable, the interior sensors, a cascaded triggering of the drive 9 is possible. This allows for high field coverage, i.e., a large number of traffic situations, despite conflicting requirements for the force level, while maintaining the controllability of the ego-vehicle 1 for as long as possible. This could, for example, involve shifting the trigger point of the drive 9 forward or adjusting the maximum force level to the situation (G1).

[0114] This allows the occupant to be placed or held in an optimal position in as many critical traffic situations as possible.

[0115] In a critical traffic situation, the control unit 5 can derive an adaptable escalation strategy based on the probability of a collision and other parameters of varying quality detectable in the traffic situation. This allows the activation strategy of the drive 9 to be adjusted for different traffic situations. As a result, the occupants can be moved into or kept in an optimal position in as many traffic situations as possible. One activation strategy might involve earlier activation of the drive 9 if information is available indicating a likely collision with low to medium sensor accuracy and the occupant is out of position.

[0116] This results in a shift in the effective time. This eliminates the need to apply a high force against the deceleration (S1). Furthermore, the maximum force (G1), i.e., the distraction threshold, can be adjusted when additional information is available, such as lateral acceleration, i.e., high sensor quality is achieved through the integration of environmental data with the vehicle dynamics data. This results in a shift in the distraction threshold. This allows for optimization of the force level to the traffic situation and thus maximization of field coverage. Reference symbol list 1 Ego vehicle 2 Device 3 seatbelt tensioners 4 safety belt 5 Control unit 6 Interior sensor 7 Camera 7a Radar sensors 7a Lidar sensors 8 acceleration sensors 8a Pressure sensors 8b Speed ​​sensors 9 Drive

Claims

Device (2) for increasing occupant protection in an ego vehicle (1), the device (2) comprising at least one reversibly adjustable actuating device for applying forces to a belt tensioner (3) for reversibly moving the belt tensioner (3) from a starting position to an impact position, and a control unit (5) for controlling the actuating device, further comprising a first sensor for detecting the traffic situation as first environmental data with a first quality and at least a second sensor for detecting the traffic situation as second environmental data with a second quality, wherein the control unit (5) is configured toto interweave the first environmental data as well as the at least second environmental data with each other as interweaving environmental data and to cascade the force exerted on the actuating device depending on the quality of the interweaving environmental data and a collision probability recognized in the interweaving environmental data, wherein the first sensor and the at least second sensor are designed as identical sensors, wherein the first environmental data are recorded at a first time and the at least second environmental data are recorded at a subsequent second time, and the control unit (5) is designed to interweave the first environmental data as well as the at least second environmental data with each other as interweaving environmental data. Device (2) according to claim 1, characterized in that the first sensor and the at least one second sensor are designed as separate sensors. Device (2) according to one of the preceding claims, characterized in that the collision probability includes the probability of a collision as well as the time of collision. Device (2) according to one of the preceding claims, characterized in that the control unit (5) is designed to set, as a cascaded force, at least the triggering time for triggering the force exerted on the actuating device, the force and a tightening speed of the belt tensioner (3). Device (2) according to one of the preceding claims, characterized in that the control unit (5) is designed to exert a small force on the actuating device when the quality is below a first quality threshold in conjunction with a maximum first collision probability threshold, so that a slow movement of the belt tensioner (3) from the initial position towards the impact position is achieved. Device (2) according to claim 5, characterized in that the control unit (5) is configured to perform haptic and / or optical and / or acoustic preconditioning by means of the actuating device when the quality is below a second quality threshold in conjunction with a maximum second collision probability threshold which is below a first collision probability threshold. Device (2) according to one of the preceding claims, characterized in that the control unit (5) is designed to exert a maximum force on the actuating device when the quality is above a first quality threshold in conjunction with a minimum first collision probability threshold, so that rapid movement of the belt tensioner (3) from the initial position towards the impact position is achieved. Device (2) according to one of the preceding claims, characterized in that the control unit (5) is designed to additionally articulate vehicle dynamic sensor data along with the articulated environmental data. Device (2) according to one of the preceding claims, characterized in that the control unit (5) is designed to take into account certain traffic situations recognized in the interlocked environmental data when cascading the applied force, wherein the traffic situations can be selected from at least the following traffic situations: high lateral acceleration above a predetermined defined lateral acceleration value and / or high self-speed above a predetermined defined speed value of the ego vehicle (1). Device (2) according to one of the preceding claims, characterized in that the control unit (5) is configured to dynamically release the force on the actuating device if no collision occurs or if an incorrect diagnosis is detected. Method for increasing occupant protection in an ego vehicle (1) characterized by the steps: - Providing at least one reversibly adjustable actuating device for applying forces to a belt tensioner (3) for reversibly moving the belt tensioner (3) from a starting position to an impact position, - Providing a control unit (5) for controlling the actuating device, - Providing a first sensor for detecting the traffic situation as first environmental data with a first quality and at least a second sensor for detecting the traffic situation as second environmental data with a second quality, - Interlacing the at least first environmental data as well as the at least second environmental data with each other as interlaced environmental data, - Cascaded adjustment of the force exerted on the actuating device depending on the quality of the interlaced environmental data and a collision probability detected in the interlaced environmental data,wherein the first sensor and the at least second sensor are designed as identical sensors, wherein the first environmental data are recorded at a first time point and the at least second environmental data are recorded at a subsequent second time point, and wherein the first environmental data as well as the at least second environmental data are entangled with each other as entangled environmental data. Method according to claim 11, characterized in that, in addition to the entangled environmental data, interior sensor data from an interior sensor (6) are also entangled. Method according to claim 11 or 12, characterized in that vehicle dynamics sensor data are additionally integrated with the interlocked environmental data.

Images