JP2006341843A - Classification method for collision incident - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a more reliable collision detection method of an automobile and a person. <P>SOLUTION: In a method for classifying a collision incident in which evaluation of an acceleration value outputted by acceleration sensors (10, 12) for a predetermined evaluation time is started for the collision incident and includes a step for determining a speed value and a distance value from the outputted acceleration value; a step for classifying the collision incident to one of a plurality of categories regarding each of the acceleration value, the speed value and the distance value; and a step for classifying the collision incident by the categories in which the collision incident is classified respectively regarding the acceleration value, the speed value and the distance value. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to a method for classifying a collision case, and more particularly, to a method for detecting a collision case between a person such as a pedestrian and a car by evaluating an acceleration value output by an acceleration sensor.

  A plurality of methods are known for detecting a collision between one automobile and another automobile using two acceleration sensors arranged in the area of the bumper of the automobile. By these methods, a collision can be evaluated, and a passive safety means for protecting a vehicle occupant can be activated by the collision evaluation.

  Newer safety standards for automobiles require active safety devices that are located outside the automobile and protect people who are impacted by the automobile. In order to prevent a fatal ending collision, an appropriate safety device reduces the force caused by a person's head, for example, hit by a car, colliding with the car. A safety device provided for this purpose can elevate the engine hood, for example in the event of a collision between a car and a person, thereby increasing the spacing between the engine hood and the engine block. The head is prevented from reaching the engine block.

  The conventional collision detection method is not suitable for operating a safety device that is located outside the automobile and protects a person who is collided by the automobile because the accuracy of the collision evaluation is insufficient.

  The lack of accuracy of the evaluation, on the other hand, means that the acceleration value measured at the time of a car-to-car collision is compared with the acceleration value at the time of a car-to-person collision referred to below as a pedestrian collision. Thus, due to the fact that it takes a substantially wide range, the corresponding detection method has a measurement resolution that is too coarse in the measurement range related to detection of pedestrian collision.

  In addition, airbags operate faster, especially in severe car / car crash events than less severe crashes. However, in pedestrian collision cases, the safety measures should only be activated when the car actually collides with people, such as metal pillars, trees, snowy mountains, supermarket shopping carts, or other vehicles. Should not be activated in the event of a collision with an object that is softer, harder or heavier than Conventional detection methods do not provide the necessary sensitivity in this relatively narrow range of all possible collision events.

  In addition, acceleration sensors of known collision detection systems can be used, for example, in the event of a collision at a specific position of a specific speed vehicle and / or vehicle bumper, or of a collision between a non-human object and a vehicle. At the same time, a signal similar to the signal output in the case of a pedestrian collision in a certain state is output. Therefore, the risk of malfunction of the safety device is unacceptably high with conventional detection methods.

  The object underlying the present invention is to provide a more reliable method of detecting a collision between a car and a person.

  To meet this object, a method having the characteristics of claim 1 is provided.

In the method according to the present invention, the evaluation of the acceleration value output during a predetermined evaluation time by the acceleration sensor is started by a collision incident,
Determining a velocity value and a distance value from the output acceleration value;
For each of the acceleration value, the velocity value, and the distance value, classifying the collision incident into one of several categories, particularly “slow collision”, “pedestrian collision”, and “violent collision”. And categorizing the collision incidents depending on categories classified respectively with respect to the acceleration value, the velocity value, and the distance value.

  By the method according to the present invention, a collision between an automobile and a person, for example a pedestrian, is detected from a collision between an automobile and an object other than a person, which can produce a signal output similar to that produced by a collision with a person, for example. It becomes possible to distinguish. Such objects are, for example, metal pillars, plastic tubes, trash cans, supermarket shopping carts, trees, snow mountains, and the like.

  According to the invention, particularly reliable detection of a car-person collision is possible as a result. That is, the collision is recognized as a pedestrian collision with particularly high reliability.

  In accordance with the present invention, not only can the risk of a safety device for protecting a person colliding with an automobile not be activated in this way, but also, as a result of a collision incident mistakenly classified as a pedestrian collision, The risk of unnecessary operation is also reduced.

  This ultimately improves road traffic safety by improving the protection of people collided by cars on the one hand and reducing the risk of driving the car driver by accidentally activated safety devices on the other hand. Not only is the performance increased, but unnecessary repair costs that may be incurred due to malfunction of the safety device are avoided.

  The method according to the invention can be carried out with the aid of the known arrangement of acceleration sensors arranged in the area of the bumper of a motor vehicle. In this way, the method according to the invention can be realized in an already existing system.

  In contrast to conventional collision detection methods that use only positive acceleration values to evaluate collision events, the method according to the present invention is based on the evaluation of the overall acceleration signal, regardless of the signs of acceleration in practice.

  Advantageous embodiments of the invention can be seen from the dependent claims, the best mode for carrying out the invention, and the figures.

  According to a preferred embodiment, acceleration, speed based on one acceleration value, speed value, distance value of at least two acceleration sensors arranged in the area of the bumper of the vehicle, particularly spaced laterally from each other Classification of collision incidents related to distances. The amount of data to be evaluated is thus reduced, thereby achieving a faster classification of collision events. As a result, the safety means relating to the collision between the car and the person can operate particularly fast.

  The classification of collision incidents with respect to acceleration, velocity, and distance is the acceleration sensor that responds first to the collision incident or, if both acceleration sensors respond simultaneously, the acceleration sensor whose acceleration value first exceeded a predetermined operating threshold. This is performed based on the acceleration value, speed value, and distance value.

  The classification of collision incidents is advantageously performed with respect to acceleration values during a portion of the predetermined evaluation time. The classification of collision incidents can be done in this way, for example, with respect to acceleration values during the first half of the evaluation time. This allows a particularly quick temporary estimation of whether the detected collision incident can be considered as a pedestrian collision in all cases. A pre-decision is made on the basis of this temporary estimation of whether to continue or cancel the evaluation.

Provisional estimation or pre-determination can be made in a few milliseconds. The length of the predetermined evaluation time is, for example, about 10 ms in total, and the classification of the collision incident regarding the acceleration value is performed during the first 5 ms. The acceleration value can be read from the acceleration sensor at a rate of 1 ms ⁻¹ .

  Collision incidents are either (i) classified in each case into the category “pedestrian collision” with respect to both speed and distance values, or (ii) collision incidents are classified into the category “ In each case, it is classified as “violent collision”, and when classified into the category “slow collision” with respect to acceleration values, it is preferably classified as a collision with a person. The classification of collision cases as pedestrian collisions according to these criteria has proven particularly reliable.

  According to another embodiment, the classification of collision incidents with respect to acceleration, velocity, and distance corresponds to acceleration value, velocity value, or minimum value of distance value, maximum value, and difference between minimum value and maximum value, respectively. This is performed depending on whether or not a threshold value to be exceeded is exceeded.

  The threshold level preferably depends on the position of the collision with respect to the acceleration sensor, in particular depending on whether it is a central or non-central collision. While collisions that occur in the end region of an automobile bumper are designated as non-center or “corner” collisions, it is understood that a center or “center” collision occurs in the central region of the bumper.

  The position of the collision with respect to the acceleration sensor can be determined with respect to the acceleration values of two acceleration sensors which are arranged in the region of the automobile bumper and are spaced apart from one another in the lateral direction. The determination of the position of the collision is advantageously made during a part of the predetermined evaluation time, for example within the first half of the evaluation time.

  The determination of the location of the collision is used to classify collision incidents and not only enables the verification of the thresholds described earlier, but also activates a local safety device that protects the person who has collided with the car. Is optionally possible. For example, in a right corner collision, it is sufficient to activate only a portion of the right side of the safety device. In this way, unnecessary activation of parts of the safety device that are not required can be avoided.

  (A) When the shapes of the signals output by the two acceleration sensors are not substantially different from each other; (b) The change in the acceleration value of one acceleration sensor is substantially different from the change in the acceleration value of the other acceleration sensor. And (c) when the difference between the acceleration value of one acceleration sensor and the acceleration value of another acceleration sensor is not within the region of the acceleration value itself, the collision is preferably categorized as a central collision.

  The first criterion is based on a qualitative analysis of the acceleration signal and is met, for example, when the signals of both acceleration sensors increase or decrease simultaneously. The second criterion is satisfied when the change in the signal of one acceleration sensor is less than or equal to about twice the change in the signal of the other sensor in total. Finally, when the difference between the value of one acceleration sensor and the acceleration value of the other acceleration sensor is approximately half of the value output by each acceleration sensor, for example, the signals of the two acceleration sensors are almost the same. When the third criterion is met. This is therefore a quantitative measure.

  According to another embodiment, the acceleration value output during the evaluation time is multiplied by a calibration factor depending on the vehicle speed and vehicle type, or bumper structure. The multiplication by the calibration factor is advantageously performed before the speed value and the distance value are determined from the acceleration value. For acceleration value calibration, vehicle speed and vehicle type are not considered in the evaluation of calibrated acceleration values, velocity values, and distance values following calibration of acceleration values. Therefore, the classification of collision incidents is substantially simplified and can be performed even faster.

  According to another embodiment, the acceleration values output during the evaluation time are averaged, in particular by subtracting the minimum value of the output acceleration values. The averaging is advantageously done in the same way before the speed value and the distance value are determined. The reliability of this method and in particular the safety of accurate identification of pedestrian collisions is further increased by averaging.

  Embodiments of the invention will now be described with reference to the preferred embodiments and the accompanying drawings.

  The sensor device includes two acceleration sensors arranged in the area of the front bumper 14 of the automobile. The bumper 14 extends in a direction transverse to the traveling direction of the automobile, and a metal carrier 18 provided with a side surface facing a foam made of an energy-absorbing foam material 20, for example, foamed polypropylene, in the traveling direction. The foam is covered with the outer skin 22.

  As can be seen, the acceleration sensors 10, 12 are attached to the metal carrier 18 of the bumper 14 such that the acceleration sensors 10, 12 are spaced laterally with respect to the travel direction 16. The acceleration sensors 10, 12 are connected to the evaluation unit 26 via the electrical leads 24, and the evaluation unit 26 serves to evaluate the signals output by the acceleration sensors 10, 12.

  The signals output by the acceleration sensors 10 and 12 correlate with the acceleration of the bumper 14, and thus may be the voltage values detected by the acceleration sensors 10 and 12 that correlate with the acceleration of the entire automobile. Signals output from the acceleration sensors 10 and 12 are also referred to as acceleration values here.

  The acceleration values output by the acceleration sensors 10, 12 are monitored with the aid of the evaluation unit 26 and are evaluated with respect to the collision between the vehicle and the object according to the method according to the invention. This evaluation is performed for the purpose of classifying a collision case, in particular to determine whether a pedestrian collision, i.e. a collision between a car and a person, e.g. a pedestrian, has occurred.

  In the event of a collision between a car and a person, the evaluation unit 26 outputs an activation signal to a safety device 28 that activates the appropriate safety means to protect the person collided by the car. These safety measures, for example, raise the engine hood of the automobile and increase the spacing between the engine hood and the engine block, thus preventing a person from reaching the engine block through the engine hood.

  The method for categorizing collision incidents according to the present invention is described in more detail below.

  The acceleration sensors 10 and 12 are read by the evaluation unit 26 at regular intervals to monitor the acceleration of the automobile. In the embodiment shown, the acceleration value is recorded every millisecond. However, in general, the acceleration sensors 10, 12 are read at lower or higher speeds depending on the requirements of the safety device 28, in particular depending on the speed of the actuator system. Therefore, the reading speed can be selected to be lower in the case of a higher-speed actuator system than in the case of a lower-speed actuator system, for example.

  The last read acceleration value is each compared with the previous acceleration value in the evaluation unit 26. If the current acceleration value is significantly different from the previous acceleration value, i.e. the difference between the current acceleration value and the previous acceleration value exceeds a predetermined calibrated operating threshold, a collision event is considered to have occurred. A further evaluation of the acceleration value is started during a predetermined evaluation time. In the embodiment shown, the evaluation time totals 10 ms. However, in general, the evaluation time may be shorter or longer depending on the requirements of the safety device 28, in particular depending on the speed of the actuator system. Therefore, the evaluation time may be longer for a high-speed actuator system than for a low-speed actuator system, for example.

  For further evaluation, the position of the collision with respect to the bumper 14, i.e. the position of the collision with respect to the acceleration sensors 10, 12, i.e. whether a collision has occurred in the central area of the bumper 14 or in the end area of the bumper 14. It is determined whether it has occurred. Thus, a collision is designated as a “center collision” or “corner collision”.

The determination of the collision position is performed for a part of the predetermined evaluation time, in the present embodiment, during the first half of the evaluation time, that is, during the first 5 ms after the start of further evaluation. This is done by evaluating the acceleration value. A collision is classified as a center collision if the following three conditions are met:
(A) The shapes of the evaluation values output by the two acceleration sensors 10 and 12 are not substantially different from each other, that is, when the acceleration value of one acceleration sensor 10 increases, the acceleration values of the other acceleration sensors 12 also increase. If it increases, and vice versa;
(B) A change in acceleration value of one acceleration sensor 10, that is, a change in signal intensity, is not substantially different from a change in acceleration value of another acceleration sensor 12. This is the case in this embodiment as long as the change in acceleration value of one acceleration sensor 10 is not more than about twice the change in acceleration value of other acceleration sensors 12 in total; and (c) one The difference between the acceleration value of the acceleration sensor 10 and the acceleration value of the other acceleration sensor 12, that is, the change in the signal between the sensors does not fall within the range of the acceleration value itself. As long as the difference is less than about half of the acceleration values of the acceleration sensors 10 and 12 in total, this is the case in the present embodiment. That is, this condition is satisfied when the acceleration values output from the acceleration sensors 10 and 12 are substantially equal in total.

  In all other cases, collisions are categorized as corner collisions.

  For the following evaluation steps, only the acceleration value of one of the two acceleration sensors 10, 12 is taken into account, and in fact only one acceleration of the first acceleration sensor 10, 12 that detected the collision event first. Value is taken into account. When the acceleration sensors 10 and 12 detect a collision at the same time, the acceleration values of the acceleration sensors 10 and 12 whose acceleration value changes exceed the calibrated minimum acceleration threshold value are used for further evaluation. The acceleration sensor whose acceleration value is used for further evaluation is hereinafter referred to as a first acceleration sensor 10.

  After the collision position is determined, the acceleration value output from the first acceleration sensor 10 during the evaluation time is multiplied by the calibration coefficient and stored. When classifying a crash, the method according to the invention is implemented to calibrate the calibration factor to the type of vehicle depending on the speed of the vehicle. By calibrating the acceleration value with a speed-dependent calibration factor, the speed of the vehicle with respect to the classification of the collision event no longer needs to be taken into account in the following method steps.

  Whether the detected collision event is considered to be a complete pedestrian collision, or the first half of the calibrated acceleration value of the first acceleration sensor 10, where the first five acceleration values are their minimum values, their maximum values, And its range, ie the difference between the minimum and maximum values, and the classification of the collision into one of the categories “slow collision”, “pedestrian collision” and “violent collision” is based on the evaluation result To be done.

  In this process, a collision case with a small minimum value and a small maximum value of acceleration, and having a narrow range of acceleration values, is classified as a `` slow collision '' and has a high and low maximum acceleration value. Collision incidents with or with a wide range of acceleration values are classified as “violent collisions”.

  Collision incidents where the minimum, maximum, and range of acceleration values fall between the corresponding thresholds of the categories “slow collision” and “violent collision” are, in contrast, classified into the category “pedestrian collision”. . In this process, the threshold levels for the categories “slow collision” and “violent collision” depend on whether the detected collision event is a center collision or a corner collision.

  As soon as all acceleration values output by the first acceleration sensor 10 are present in the evaluation unit 26 during a predetermined evaluation time, the minimum value of these values is determined and other acceleration values are determined. Deducted from. In this way, the acceleration value of the first acceleration sensor 10 output during the evaluation time is converted into a positive acceleration value, which is averaged to some extent.

  In this case, the acceleration values are averaged after the acceleration value calibration described above. However, it is also generally possible to first average acceleration values and then calibrate.

  The corresponding velocity value is determined by simple integration or addition from the calibrated and averaged acceleration values of the first acceleration sensor 10, and the corresponding distance value is determined by double integration or addition.

  In a manner similar to that for acceleration values, the determined velocity values are also evaluated for their minimum, maximum and range. Depending on the time over which the determined speed value exceeds a corresponding set of predetermined speed thresholds, the collision incident is classified into one of the categories “slow collision”, “pedestrian collision”, and “violent collision”. The level of the speed threshold varies in the same manner depending on whether the classified collision event is a center collision or a corner collision.

  In contrast, the determined distance values are also evaluated with respect to their minimum, maximum and range. Crash incidents that are categorized according to the time at which the distance value exceeds a set of predetermined distance thresholds are classified into one of the categories “slow collision”, “pedestrian collision”, and “violent collision”. The distance threshold level also depends on whether it is a center collision or a corner collision.

  After the classification of the crash incident to be evaluated has been made with respect to acceleration, velocity, and distance values, the final classification of the collision incident is based on the respective classification with respect to the acceleration, speed, and distance of the collision incident. Done.

In this process, a collision incident is categorized as a car-person collision when one of the following two conditions is met:
(I) The case where the collision incident is classified into the category “pedestrian collision” in both cases in terms of both speed and distance values.
(Ii) The case where the collision event is classified into the category “severe collision” with respect to the speed value and the distance value and the category “slow collision” with respect to the acceleration value in each case. This condition relates in particular to the case where an object collided by a car changes from an initial soft object to a hard object, for example in the case of a person who has subsequently collided with a car.

  In all other cases, the collision incident is categorized as a non-pedestrian collision.

  The decision to activate the optional safety means is made with reference to the last category of the collision incident. When classifying a collision case as a “pedestrian collision”, for example, the safety device 28 operates to protect a person collided with a vehicle, and the safety device has a function of raising the engine hood of the automobile.

1 is a schematic cross-sectional view showing a sensor arrangement for performing the method according to the present invention.

Explanation of symbols

DESCRIPTION OF SYMBOLS 10 Acceleration sensor 12 Acceleration sensor 14 Bumper 16 Travel direction 18 Metal carrier 20 Foam material 22 Outer skin 24 Electric lead 26 Evaluation unit 28 Safety device

Claims (11)

A method of classifying a collision case in which an evaluation of an acceleration value output by an acceleration sensor (10, 12) is started during a predetermined evaluation time due to the collision case,
Determining a speed value and a distance value from the output acceleration values;
For each of the acceleration value, the velocity value, and the distance value, classifying the collision event into one of several categories: “gentle collision”, “pedestrian collision”, and “violent collision”; And categorizing the collision incident according to the category into which the collision incident is classified with respect to the acceleration value, the velocity value, and the distance value.   2. The method according to claim 1, wherein the acceleration value, the velocity value and the distance value of one of at least two acceleration sensors (10, 12) arranged in the area of the bumper (14) of the vehicle. The above-mentioned classification of collision incidents is performed with respect to acceleration, speed, and distance based on the above.   3. A method according to claim 2, wherein the classification of the collision event is the first response to the collision event or the acceleration value of the acceleration sensor is initially a predetermined operating threshold with respect to acceleration, velocity and distance. The method is performed based on the acceleration value, the speed value, and the distance value of the acceleration sensor (10) exceeding the range.   4. A method according to any one of claims 1 to 3, wherein the classification of the collision event is performed with respect to the acceleration value during a portion of a predetermined evaluation time. A method according to any one of claims 1 to 4, comprising
(I) the collision incident is classified in each case in the category “pedestrian collision” with respect to both the speed value and the distance value; or (ii) the collision incident is the speed value and When the distance value is classified into the category “strong collision” in each case, and the acceleration value is classified into the category “slow collision”,
A method characterized by being classified as a collision with a person.   6. The method according to any one of claims 1 to 5, wherein the classification of the collision incident is such that the minimum value, maximum value, and / or range of the acceleration value, speed value, and distance value are each A method characterized in that it is performed on acceleration, speed and distance depending on whether a corresponding threshold is exceeded.   7. The method according to claim 6, wherein the threshold level depends in each case on the collision position relative to the acceleration sensor (10) and whether the collision is a central or non-central collision. A method characterized by.   The method according to any one of claims 1 to 7, wherein the position of the collision with respect to the acceleration sensor (10) is arranged in the region of a bumper (14) of an automobile and is laterally separated from each other. And determining the acceleration value of two acceleration sensors (10, 12) arranged in a row. The method according to claim 8, comprising:
(A) The shapes of signals output by the two acceleration sensors (10, 12) are not substantially different from each other;
(B) a change in the acceleration value of one acceleration sensor (10) is not substantially different from a change in the acceleration value of another acceleration sensor (12); and (c) one acceleration sensor (10 ) And the acceleration value of the other acceleration sensor (12) are not within the range of the acceleration value itself,
A method characterized in that a collision incident is categorized as a central collision.   10. The method according to any one of claims 1 to 9, wherein the acceleration value output during the evaluation time is multiplied by a calibration factor that depends on the speed of the vehicle and the type of vehicle. A method characterized by that.   The method according to any one of claims 1 to 10, wherein the acceleration value output during the evaluation time is averaged by subtracting a minimum value of the output acceleration value. A method characterized by that.

Images