Patent Application
20100063685
The invention relates to a motor vehicle having a preventive action protection system of the type defined in more detail in the preamble of patent claim 1.
Motor vehicles, in particular passenger cars, are equipped usually with active and passive safety devices which permit the driver to control hie vehicle better even in critical situations, and as a result possibly avoid the vehicle being involved in an accident. If a collision occurs, such safety devices also contribute to reducing the severity of the accident.
Safety devices which are already active in a preventive fashion even before a possible collision and use what is referred to as a pre-crash phase, i.e. a period of time starting from the detection of a high probability of a collision by appropriate detection systems in the vehicle up to the actual impact, to enhance the vehicle occupant protection by means of additional safety measures and thus lessen the severity of an accident, are referred to as preventive action protection systems or so-called PRE-SAFE™ systems. In order to detect possible accident situations, preventive action protection systems make use of information which is made available by various sensor devices of the motor vehicle. The sensor devices are in particular a component of another electronic driving stability program and/or a component of a distance sensor system. Depending on the detected situation, conclusions are drawn about a possible accident, and appropriate measures, relating to restraint systems for vehicle occupants and possible protection devices for other parties in an accident, such as pedestrians, are initiated in order to condition the vehicle for the imminent accident.
An example of the actuation of a reversible vehicle occupant protection means in a motor vehicle is described in German patent DE 101 21 386 C1. The motor vehicle here has a reversible vehicle occupant protection system which can be activated before a collision time and thus moved into an effective position. For this purpose, a sensor system is used to acquire driving state data which is monitored for any possible emergency braking, any possible oversteering and any possible understeering. If emergency braking, oversteering and/or understeering is detected, the vehicle occupant protection system is activated, it being possible to trigger the vehicle occupant protection system only if a minimum velocity is exceeded.
Important input variables for a decision as to whether safety devices such as a seatbelt pretensioners or airbags are to be placed in a state of increased readiness or triggered is provided in practice by a vehicle surroundings detection device by means of which information relating to the positions and relative velocities of objects in the vicinity of the vehicle can be acquired.
Radar sensors are used most frequently to monitor the vicinity of the vehicle.
However, in practice it is also possible to apply optoelectronic sensors of a wide variety of types which operate, for example, with infrared radiation, ultraviolet radiation and microwave radiation. Approaches with image sensors for monitoring the surroundings of the vehicle are also known.
Furthermore, in practice, further input variables which are acquired, for example, for controlling safety and comfort systems such as an electronic driving stability system, are compared with triggering thresholds of the preventive action protection system. Such input variables can be, for example, information which is output by a steering angle sensor, a pedal travel sensor, a brake pressure sensor, by wheel speed sensors, acceleration sensors and by a yaw rate sensor.
Variables such as, for example, acceleration values which can in turn represent what is referred to as a feature which indicates a hazard situation or emergency situation can be determined from this information. Such a feature can be a specific variable per se or a driving state or a driver reaction, for example excessive steering maneuver or emergency braking, formed from a plurality of variables.
Even if selective triggering of suitable safety devices and thus considerable protection for the vehicle occupant is possible with such protection systems which are also used in practice, the large amount of information which is received in parallel causes difficulties in the data processing and control device in terms of the detection of data which is decisive for a possible crash situation.
In particular, when a collision object is sensed by means of a vehicle surroundings detection device, reliable definitive information about the imminent accident event is necessary in the shortest possible time before the time when the accident occurs. The often large degree of time-consuming computational complexity with which a plurality of input variables are compared with triggering thresholds in existing systems prevents, under certain circumstances, safety devices for vehicle occupants or other parties to a collision from being actuated in a way which is appropriate for the situation before the vehicle experiences an impact.
The invention is thus based on the object of providing a motor vehicle with a preventive action protection system of the type specified in the introduction which improves the preventive action protection system further in terms of actuation in a way which is appropriate for the situation.
In a motor vehicle having a preventive action protection system which comprises safety devices which are actuated as a function of features which are formed from input variables from a safety sensor system in a data evaluation and control device, wherein, when a critical driving state is detected, the data evaluation and control device actuates at least one safety device which is assigned to this driving state, this object is achieved according to the invention in that the features are each assigned a specific weighting relating to the criticality of the driving state.
In this way, the accident criticality can be detected with a corresponding predictive driving state sensor system early before the contact time when there is, for example, a frontal impact or rear impact, if the features which relate to such a collision have correspondingly high specific weighting and thus can be easily differentiated from data which is less relevant for the criticality of the driving state.
Such weighting of features advantageously takes up very little storage space and computing power in the data evaluation and control device, as a result of which definitive information is possible very quickly and correspondingly time is gained for the triggering of the suitable safety devices.
Such actuation of safety devices can advantageously be implemented with very little financial outlay in a triggering algorithm for safety devices if said algorithm operates in the vehicle as a function of a predictive sensor system.
In one advantageous embodiment of the invention, a superordinate criticality is formed from the specific weightings of a plurality of features and is compared with a triggering threshold. The superordinate criticality can represent here, for example, the criticality in the longitudinal direction of the vehicle if the features which are decisive for this, for example a remaining time up to the collision, are taken into account, or else a criticality in the lateral direction if the superordinate criticality is formed from the weightings of the features which are decisive for this, for example lateral deviation from a potential collision object.
In one advantageous embodiment of the invention, the superordinate criticality represents the overall criticality of a sensed collision object, in which case when there is a plurality of sensed collision objects a corresponding number of superordinate criticalities can be formed and in turn compared with one another.
The features are preferably formed from input variables which are stored in a memory over a defined period of time, in which case profiles of the input data or input variables are observed as which, for example, positions and speeds of physical objects serve relative to the vehicle.
In order to make available the input data, the safety sensor system can comprise a driving state sensor system which is configured in a variety of ways and which is equipped, for example, with a steering angle sensor, a pedal travel sensor, a brake pressure sensor, a wheel speed sensor, an acceleration sensor, a yaw rate sensor and/or a distance sensor.
Important input data for controlling safety devices always represents the data which is made available by a vehicle surroundings detection device which is associated with the safety sensor system. In one expedient configuration, the vehicle surroundings detection device can operate on a radar basis even if other optoelectronic or image-processing systems can be applied to implement the preventive action protection system which is configured according to the invention.
In one simple configuration of the safety sensor system, the vehicle surroundings detection device can have two radar sensors on the front of the vehicle and two on the rear of the vehicle, said sensors covering the region in front of the vehicle and the rear part behind the vehicle, respectively. The radar sensors can usually operate in the 24 GHz range and determine not only the x position but also the y position of an object in front of or behind the motor vehicle. The y position can be determined directly here by means of angular sensing by means of a pattern detection method.
If, for example, ten targets are sensed per measurement and per sensor by each wheel sensor, when there are two sensors on the front of the vehicle the vehicle's surroundings detection device supplies twenty angle data items and twenty distance data items for the ten targets per measurement, as a result of which, given a sampling rate in a range of, for example, less than 30 milliseconds, sufficient accuracy for predicting crash conditions is provided.
In one preferred embodiment of the invention, such radar sensors are used to determine, from position data of a collision object in a defined period of time, a direction vector by means of which the position of the collision object is predicted. The direction vector can be estimated here from the stored position data using simple mathematical methods, for example by means of a regression line.
The feature obtained in this way relating to the position of the collision object can then be weighted according to the invention.
As an alternative to this, the position of a sensed object can also be determined from the measured distances by means of triangulation. To do this it is necessary for an object whose precise position is to be determined to lie in the overlapping range of at least two radar sensors. In this context, the area in which an object can be sensed by a radar sensor depends on what is referred to as a radar cross section (RCS) which can be considered as the reflectivity of an object for radar waves.
A further possible way of sensing the position of an object by means of radar sensors is to track the time profile of the position of a sensed object, referred to as a tracking method. Such a method which is described, for example, in DE 199 49 409 A1, supplies good results in particular when there is approximately constant movement of the sensed objects without excessively large dynamic changes.
In critical driving situations with a highly dynamic behavior, a device for sensing the position of objects in the surroundings of a vehicle, such as is described in German patent application DE 103 26 431 which is published after the priority date of the present document and to whose entire contents reference is made, is also advantageous.
In said document it is proposed that position information relating to objects in the vicinity of the vehicle are derived by means of a comparison of input values which are supplied by sensors with data records which are stored in a memory unit. The input values contain, for example, distance data and Doppler speeds. Doppler speeds are the speeds of an object relative to a sensor, which the sensor itself determines from a Doppler measurement and outputs. The data stored in the memory unit comprises reference data records which represent objects in a defined spatial area in the vicinity of the vehicle with their precise positions. In order to determine accurately the position of an object sensed by the sensors, a comparison of the input values as supplied by the sensors with the reference data records is performed within the scope of a classification process. By means of the position of the object relative to the vehicle which is determined in this way, it is possible to decide whether a sensed object is in an area for which a collision with the object is to be expected. In particular it becomes possible to differentiate whether an obstacle is expected to be passed or hit.
The input variables obtained in this way are preferably stored, together with other input variables such as position components and speed components, steering wheel angle, etc. in a memory of the data evaluation and control unit and used to calculate variables derived therefrom, which variables are obtained from the development of the respective input variable over an observed time window.
The data evaluation and control device of the preventive action protection system can be a data processing device of a driving stabilization system of the motor vehicle which is frequently present in any case in particular in modern motor vehicles. As an alternative, it is of course also possible to use a separate data processing device.
Features which relate to a specific driving state or a driver reaction can be determined from the input data. For example, a number of specific features which are relevant to the accident criticality are preferably formed for each potential collision object found by the vehicle surroundings detection device.
Such features include, in particular, a position of a collision object which can be predicted, for example, by means of a direction vector, a remaining time up to the impact, an offset of a collision object in the lateral direction, or y direction, of the motor vehicle, and a relative velocity between a collision object and the motor vehicle.
The relative velocity, which is preferably calculated from the change in distance between two measurement cycles, with the distance or Doppler speed being measured in each measurement cycle, can itself be a feature or serve as an input variable for other features such as, for example, the remaining time up to the collision.
After inventive weighting of the correspondingly selected features and determination of a superordinate criticality, said criticality can be subjected to filtering in the data evaluation and control device over an adjustable time window before the comparison with a triggering threshold, in order to minimize the risk of incorrect triggerings.
After a triggering decision, the actuation of the corresponding safety device can be made more precise and adapted to the respective situation if it takes place as a function of physical variables of a vehicle occupant which are determined. These include, in particular, the size of vehicle occupants and their weight. These data items can be determined by means of a weight detection device and a body size detection device connected to the data evaluation and control device, it being possible to embody the weight detection device in such a way that it is integral with a seat occupation detection device, and to embody the body size detection device in such a way that it is integral, for example, with a seat position sensor system and a, for example optical, head position determining device.
With such devices, which are to a certain extent already installed on a series production basis, it is also possible to determine the position of a vehicle occupant in the vehicle, which information is also used to actuate the safety device in one advantageous configuration of the invention.
It is basically possible here for specific safety devices to remain activated only from a specific vehicle velocity and for their actuators to remain actuated until the vehicle velocity has reached a very low value of, for example, 3 km/h.
Further advantages and advantageous refinements of the subject tatter according to the invention can be found in the description, the drawing and the patent claims.
The drawing comprises basic illustrations of an exemplary embodiment of a motor vehicle which is configured according to the invention and which is explained in more detail in the following description.
In the drawing:
The preventive action protection system 2 is connected at the input end to a safety sensor system 3 which comprises a vehicle surroundings detection device 4 and a driving state sensor system 6.
The vehicle surroundings detection device 4 is equipped here with two radar sensors 4A, 4B on the front of the vehicle and two further radar sensors 4C, 4D on the rear of the vehicle. These are what are referred to as short distance radar sensors which operate with a carrier frequency of 24 GHz and cover, for example, a range of 20 meters and are configured to sense an angle in the y direction.
Such radar sensors can also be components of an electronic parking aid and also be installed in relatively large numbers in the region of a bumper. Furthermore, in a further embodiment, a long distance radar sensor with a range of, for example, 150 meters, such as is also used in systems for adaptive distance control which are also referred to as adaptive cruise control (ACC), can also be provided in particular in a central position on the front of the vehicle.
The driving state sensor system 6 is configured to acquire longitudinal dynamic and lateral dynamic driving state information, inter alia for detecting emergency braking and panic braking, oversteering and/or understeering of the motor vehicle 1. For these purposes, use is made, inter alia, of information or data which is supplied by wheel speed sensors 8, 9, 10 and 11, by a steering angle sensor 13 arranged in the region of a steering wheel 12, by a longitudinal acceleration sensor 14 and by a lateral acceleration sensor 15.
The driving state sensor system 6 communicates here with a driving dynamics functionality 7 such as, for example, an anti-lock brake system and/or an electronic stability program, into which a data evaluation and control device 16 of the preventive action protection system 2 is integrated.
During normal operation of the motor vehicle, the driving state sensor system 6 can, for this purpose, analyze further important driving dynamics variables such as, for example, a vehicle velocity, a yaw rate, a spring compression-travel and spring extension travel, the level of the vehicle, an accelerator pedal movement, an accelerator pedal position, a brake pedal position, brake pedal movement, a steering wheel speed and/or a steering wheel acceleration. In this context, actual values of these variables are compared with predefined setpoint values and threshold values. On the basis of these comparisons, the anti-lock brake system and/or the electronic stability program, for example, are activated and these have the function of supporting the driver of the motor vehicle in critical driving situations in order to avoid an accident.
In order to use or activate the preventive action protection system 2, the information obtained from the safety sensor system 3 is evaluated in the data evaluation and control device 16 in such a way that a comparison with a triggering threshold value S takes place, at least one selection of existing safety devices 17 being activated when the triggering threshold value S is exceeded.
In the present case, the safety devices 17 comprise, in particular, a reversible seatbelt pretensioner 18, restraining upholstered elements 19 and a means for actuating an electronic seat adjustment device 20, it being possible to place the vehicle seats or their components, as a function of their occupation, in an orientation which provides the greatest possible safety for the respective vehicle occupant in any accident which the motor vehicle 1 may be involved in.
The safety devices can in principle be all known controllable vehicle occupant protection means and also protection means for another party to the collision, for example pedestrians.
The triggering algorithm which is provided for the safety devices 17 and which is stored in the data evaluation and control device 16 is described by way of example below with reference to
These input variables I are, in a subsequent step 22, stored in a memory of the data evaluation and control device 16 and evaluated in terms of their history and thus form input variables I which serve, in a subsequent step 23, to form features which relate here to a criticality of a collision object 30 illustrated in a basic fashion in
In the present exemplary embodiment, a position POS of the collision object 30, a remaining time TTC up to the impact of the collision object 30, an offset ΔY of the collision object 30 in the lateral or y direction of the motor vehicle 1 and a relative velocity v between the collision object 30 and the motor vehicle 1 are formed, as features, from input variables stored in the memory of the data evaluation and control device 16 over a defined period of time.
The features are formed with simple mathematical methods, for example the position POS of the collision object 30 is predicted by means of a direction vector D which is estimated here by means of a regression line from position data POS_1, POS_2, POS_3 which is represented in a basic form in
As is apparent from the flowchart in
For example, referring to
Referring to
In
In the embodiment shown, the specific weighting P_POS for the accident criticality, which is assigned to the position POS of the collision object 30, is dependent on whether the predicted position POS lies, on the one hand, in a predefined fuzzy region X and, on the other hand, deviates from a measured position POS_M. The weighting P_POS of the criticality of the position POS of the collision object 30 is increased if the predicted position POS lies in the fuzzy region X and/or the greater the degree of deviation of the measured position POS_M of the collision object 30 from its predicted position.
In the present embodiment, the fuzzy region X is a function of the distance from the collision object 30, the fuzzy region X being greater the closer the collision object 30 is to the motor vehicle 1. In this way, with the weighting P_POS of the criticality of the position POS of the collision object 30, it is also possible to take into account faults in the sensing of objects such as can be caused, for example, by humid air, reflections, etc.
The weighting of the direction vector D or of the position POS which is determined therefrom is carried out directly in the embodiment shown, i.e. without comparison of the direction vector D with other vectors stored in a memory so that the weighting requires only very little expenditure in terms of computation and time.
If the selected features are assigned their specific weighting P_POS, P_TTC, P_ΔY, P_v, it is possible, in a further step 25 of the triggering algorithm shown in
P
—
abs=P
—
POS*P
—
TTC*P
—
ΔY*P
—
v.
The superordinate criticality P_abs obtained in this way corresponds here to the overall criticality of the collision object 30 for a possible collision.
If a number of collision objects are sensed in a measurement cycle, only the collision object with the maximum overall criticality P_abs is tracked further here. As an alternative to this, a plurality of objects, for example three of ten sensed objects which have the maximum overall criticality of the ten sensed objects can also be tracked, with a correspondingly higher computational expenditure.
In order to eliminate brief faults which can occur, for example, due to noise in the transmission of data which is dependent on the weather or on the state of the road, the calculated superordinate criticality or overall criticality P_abs is filtered in a further method step 26 during an adjustable time window or over a specific number of measurement cycles, a mean value filter of four measurement cycles being used here.
The overall criticality P_abs which is now filtered is then compared with a triggering threshold S in a differentiating function 27, in which case, when the triggering threshold S is exceeded, the assigned safety device 17 is triggered. This triggering threshold 8, which can have here, for example, a value of 0.7 (=70%) can vary depending on the type of vehicle and the current operating and ambient conditions.
| Number | Date | Country | Kind |
|---|---|---|---|
| 10 2004 037 704.9 | Aug 2004 | DE | national |
| Filing Document | Filing Date | Country | Kind | 371c Date |
|---|---|---|---|---|
| PCT/EP2005/008289 | 7/30/2005 | WO | 00 | 11/14/2008 |
