The present invention is directed to a method FOR side impact sensing in a vehicle.
Acceleration measurements are used for side impact sensing, decentralized peripheral acceleration sensors (PAS) in particular. They are mounted in the physical proximity of the impact site to compensate for signal delays and signal attenuation. Typical installation sites include the rocker panel, the seat cross member, or the B pillar.
The side impact sensing method in a vehicle according to the present invention has the advantage over the related art that the acceleration signals from the left and right sides of the vehicle are intelligently combined by the control unit. Thus only one signal is processed to detect a side impact, so that the computing complexity for two signals becomes superfluous. Vehicle vibrations which occur in the event of hard driving maneuvers, known as misuse, contain no side motion components, which increases the difference from the triggering signals. In this case, the effect of the vehicle motion in the event of a side impact is utilized in particular. The side motion is detectable by all y acceleration sensors, i.e., sensors which detect side motion, and is superimposed on the corresponding side due to the intrusion in the event of a side impact. The motion of the vehicle and a filtered intrusion signal are visible on the opposite side and in the control unit. Thus, triggering reliability vis-a-vis non-triggering events is increased by combining the signals.
It is particularly advantageous that the differential acceleration signal is filtered prior to forming the threshold in order to avoid in particular signals that would cause triggering in a non-triggering situation, i.e., in the event of misuse.
Finally, it is also advantageous that a device for side impact sensing, which has appropriate acceleration sensors and the control unit, is provided to implement the method.
Side impact sensing is particularly critical due to the short deformation zone between the occupants and the impacting object in the event of a side impact; yet, it is important that erroneous triggering be avoided in this case.
Side impact situations are recognized according to the present invention by acceleration sensors, the acceleration signals from the acceleration sensors on the left and right sides of the vehicle being subtracted from one another to form a differential acceleration signal. A differential speed signal is determined from the differential acceleration signal by integration, and the triggering threshold is formed as a function of the differential acceleration signal. If the differential speed signal exceeds the threshold, a side impact is recognized, and restraining means may possibly be deployed; if it is below the threshold, no side impact requiring deployment of restraining means exists.
The acceleration signals are transmitted by acceleration sensors 2 through 5 to control unit 6 as digital data. For this purpose, each of sensors 2 through 5 has a measuring amplifier, a measuring filtering device, and an analog/digital converter. As an alternative, the analog/digital conversion may also take place in control unit 6, for example; electromagnetic interference signals must be taken into account in this case.
Control unit 6 computes the triggering algorithm as a function of the signals of acceleration sensors 2 through 5. In addition, control unit 6 itself has acceleration sensors for the x and y directions to lest the acceleration signals from the peripheral acceleration sensors for plausibility.
Filtering is performed in two stages. First the slope of the acceleration signal is limited to eliminate hard driving maneuvers that exhibit acceleration signals having very high slopes as misuse. Second, a zero-frequency component is subtracted from the acceleration differential signal to prevent the threshold from becoming excessively high. The zero-frequency component and the slope limitation are performed as a function of time.
The algorithm to be used is freely selectable and does not depend on the method. Opposite sensors 2 through 5 and 3 through 4 form pairs which are combined. Each pair is then evaluated by an algorithm of its own according to FIG. 1. This makes it possible to perform the plausibility test in adjacent pairs or in control unit 6.
Number | Date | Country | Kind |
---|---|---|---|
101 03 661 | Jan 2001 | DE | national |
Filing Document | Filing Date | Country | Kind | 371c Date |
---|---|---|---|---|
PCT/DE01/04761 | 12/14/2001 | WO | 00 | 1/13/2004 |
Publishing Document | Publishing Date | Country | Kind |
---|---|---|---|
WO02/05896 | 8/1/2002 | WO | A |
Number | Name | Date | Kind |
---|---|---|---|
5608628 | Drexler et al. | Mar 1997 | A |
6236308 | Dalum | May 2001 | B1 |
6584386 | Feser et al. | Jun 2003 | B2 |
Number | Date | Country |
---|---|---|
39 24 507 | Aug 1990 | DE |
198 06 836 | Sep 1999 | DE |
101 08 849 | Oct 2001 | DE |
02 14113 | Feb 2002 | WO |
Number | Date | Country | |
---|---|---|---|
20040117089 A1 | Jun 2004 | US |