Method for controlling the release of passenger restraint systems

Abstract

In a method for controlling the release of a passenger restraint system in a vehicle, an acceleration signal is measured and integrated with respect to time to obtain a velocity signal. A release threshold value for the velocity signal is determined. If the velocity signal then falls below the release threshold value, thus indicating a vehicle collision, the passenger restraint system is released. The release threshold value is controlled depending on the type of accident situation and upon the operating parameters of the vehicle to increase the release sensitivity of the passenger restraint system. For example, the release threshold value is adjusted based on the value of the velocity signal. If the velocity signal decreases in value, then the release threshold is lowered to a more sensitive value.

Description

FIG. 20 is a flow chart illustrating an exemplary embodiment of the method according to the present invention. situation, and that the passenger restraint system should be rapidly activated. Therefore, based on the slope of the line TA1, the DV threshold (DVG) is decreased to a more sensitive level (DV2), as indicated for example by the solid line curve in FIG. 9. The slope of the acceleration curve can be determined, for example, based on two successive acceleration signals, in a manner known to those skilled in the art. Then, if the slope exceeds a threshold value, the DV threshold (DVG) is adjusted accordingly, so as to increase the release sensitivity of the passenger restraint system.

It should also be noted that the reaction behavior of a passenger restraint system can be further improved by performing a frequency analysis on the acceleration curve. Based on the frequency analysis, it is possible to recognize and suppress natural oscillations in an acceleration transducer system. It is also possible to recognize noise oscillations with relatively high amplitude values, which are not caused by collision situations, such as the impact stress of a vehicle driving over a rocky surface, a bump, or a pothole. Likewise, a pattern recognition can be compared to a stored ideal pattern in order to determine whether an actual collision situation is occurring.

Moreover, output signals generated by additional acceleration sensors located in different locations on a vehicle can also be analyzed to determine whether to lower the DV threshold (DVG). For example, switching signals generated by a rear axle switch, the seatbelt locks, the seat contacts, the brake switch, a gear switch, a contact switch, or a mechanical acceleration switch, can be generated and combined with the output signal of a central acceleration sensor, in order to determine whether, and how to decrease the DV threshold (DVG) for a particular vehicle. Moreover, signals generated by other control devices, such as brake regulators, the engine control system, or the navigation system of a vehicle, can also be evaluated.

In FIG. 10, another embodiment of the method of the present invention is illustrated, wherein the DVI curve corresponds to another typical oblique impact collision. At each point in time T1, T2, T3, T4, T5, and TA, when the DVI increases in value, the DV threshold (DVG) is lowered from its relatively high initial value, in order to increase the release sensitivity of the passenger restraint system. However, immediately after each point in time, when the value of the DVI levels off or maintains a constant value until the next point in time, the DVG immediately returns to its initially relatively high value. Then, at time TA, when the DVI exceeds the lowered DVG value, the passenger restraint system is activated.

In FIG. 11, another embodiment of the method of the present invention is illustrated. Like the embodiment described above with reference to FIG. 10, when the DVI increases at times T1, T2, T3, T4, and TA, the DV threshold (DVG) is lowered to a more sensitive level. Then, when the DVI levels off or maintains a constant value after each point in time, the DV threshold (DVG) is raised back to its original level, but only after a time delay (DT). Thus, the DVG curve is phase shifted with respect to the DVI curve by the time delay DT. By delaying the return of the DVG to its initial relatively high level, the restraint system can be released more rapidly when the deceleration maintains a constant value for only a short moment, and then inoreases again, as shown in FIG. 11. Therefore, immediately after time TA and during the time delay DT, when the DVI reaches the lower DVG threshold, the passenger restraint system is rapidly released. If, on the other hand, at time TA the DV threshold is not delayed in returning to its initially high value, but is immediately raised to that value, the passenger restraint system would not be as quickly released.

In FIGS. 12 and 13, another embodiment of the method of the present invention is illustrated. In FIG. 12, the integrator range DV1-DV2 is maintained at a constant value, and the integration values (RI) of a reference integrator are illustrated. FIG. 13, on the other hand, illustrates the integration values of a release integrator (AI).

In FIG. 12, the reference integration (RI) reaches the lower threshold DV1 at time T1, and then increases within the integrator range DV1-DV2 until time T2, when it reaches and thereafter exceeds the upper threshold DV2. In FIG. 13, the value of the release integration (AI) is determined as a function of the value of the reference integration (RI) illustrated in FIG. 12. Therefore, after time T1, when the reference integration (RI) reaches the value DV1, the output signals generated by the acceleration sensor are adjusted based on a factor corresponding to the time elapsed since time T1. The adjusted acceleration values are then integrated by the release integrator to achieve the release integration values (AI) shown in FIG. 13. Thus, at time T2, when the reference integration (RI) exceeds the upper threshold DV2, as shown in FIG. 12, the release integration (AI) in FIG. 13, which is based on the value of the reference integration (RI), exceeds the DV threshold (DVG) and, therefore, the passenger restraint system is released.

As can be seen, the release integration values (AI) are substantially influenced when the DV reference integration (RI) is within the integrator range DV1-DV2, as illustrated in FIG. 12. The release integration values (AI) are obtained by multiplying the acceleration values by a factor proportional to the amount of time elapsed since time T1, when the reference integration (RI) reached the lower threshold DV1. The release integration values (AI) may likewise be obtained by adding a set value to the acceleration values.

Therefore, when the reference integration (RI) is maintained within the integrator range DV1-DV2 for a relatively long period of time (T1-T2), the value of the release integration (AI) is increased rapidly. The reference integration curve (RI) illustrated in FIG. 12, is typical of integration values experienced during oblique collisions, for example, a collision occurring at about 30° with respect to the longitudinal axis of a vehicle. Such collisions are frequently encountered in traffic accidents, offset crashes, pole crashes, etc. Accordingly, the release threshold (DVG) is reached rapidly and thus the passenger restraint system can be released quickly during a collision to effectively protect the passengers. It should be noted that the reference integration curve (RI) is different for oblique collisions as opposed to direct front end or direct rear end collisions. Therefore, a different integrator range DV1-DV2 would be employed for a front end or rear end collision. Accordingly, the release sensitivity of the passenger restraint system can be varied depending upon the type of collision.

In a further embodiment of the method of the present invention, a time interval is initiated when the reference integration value (RI) exceeds a lower threshold, for example, DVI in FIG. 12. Control of the DV threshold (DVG) value is then determined based on the length of the time interval. Depending upon the length of the time interval, the DV threshold (DVG) is lowered accordingly. A counting circuit can be used, for example, to initiate the time interval. The counting circuit can be reset each time the reference integration (RI) falls below the lower threshold (DV1). Furthermore, the DV threshold (DVG) can be controlled depending upon the amplitude of the acceleration values, as described above. As a result, it is possible to suppress fluctuations in the acceleration signals generated by an acceleration sensor, that might be the result of noise or vibrations on the sensor that are not caused by a collision. Noise vibrations may occur, for example, by a hammer hitting the vehicle's chassis, rocks hitting the bottom of the vehicle, or when the vehicle drives over railroad tracks.

Therefore, an inappropriate release of the passenger restraint system is avoided by employing the method of the present invention, without diminishing the sensitivity of the passenger restraint system in responding to oblique collisions. Acceleration signals stemming from such non-collision impact stresses characteristically start with a very high amplitude value and then fall very rapidly to a comparatively low value. Therefore, by delaying a decrease in the DV threshold (DVG) by a time interval, in response to changes in the acceleration values, a premature release of the passenger restraint system in response to a non-collision impact can be avoided.

FIG. 17 illustrates an accelertion curve corresponding to typical acceleration values generated by an acceleration sensor when a vehicle crosses railroad tracks. As can be seen, the curve corresponds to a bipolar type signal, wherein there is an initially very high amplitude of approximately 30 g, which then diminishes very rapidly. After approximately 30 ms, the amplitude values are again very low.

In FIG. 18, a DVI curve corresponding to a typical non-collision type of impact is illustrated. The DV threshold (DVG) is controlled as a complex function of time. Thus, at time T0, the DVG is maintained at a relatively high constant value, but is then lowered to the value DVI to increase the sensitivity of the passenger restraint system after about 15 ms. The DVG is maintained at the value DV1 for about 15 ms, and is then momentarily raised again to its initial value (DVG). Thereafter, the DVG is decreased linearly for about 7.5 ms, momentarily increased to its initial value, decreased linearly again for about 7.5 ms, and then increased and maintained at its initial value (DVG) thereafter. As can be seen, the DVG follows a sawtooth type curve for about 15 ms. However, after about 13.5 ms, the DVI exceeds the lower threshold DV1. As a result, the passenger restraint system would be released during a non-collision situation, possibly endangering the passengers in the vehicle.

In FIG. 19, a DVI curve corresponding to a non-collision situation, which is similar to the curve in FIG. 18, is illustrated with reference to another embodiment of the present invention. As can be seen, the DV threshold (DVG) is not lowered to the value DVI, as shown in FIG. 18, but is maintained at its initial value (DVG) throughout the occurrence. In accordance with the method of the present invention, a counting circuit is initiated at the beginning of the occurrence, which is about 13.5 ms after time 0. Here, the counting circuit is set to run for a minimum time of about 25 ms. Therefore, although the DVI would exceed the DVG if it were decreased to the lower threshold DV1, the DVG cannot be decreased until after the 25 ms time interval. As a result, because the change in DVI is due to a non-collision type of impact, after the 25 ms time interval, the DVI again falls below the lower threshold DV1, and thus does not give rise to a decrease in the DVG value. Accordingly, because of the time delay, the passenger restraint system is not be prematurely released by a non-collision type of impact on the vehicle.

FIG. 20 is a flow chart illustrating an exemplary embodiment of the method according to the present invention. In particular, in step 201, the acceleration of the vehicle is determined. In step 203, the acceleration is integrated. In step 205, it is determined whether the integrated acceleration is within the range defined by two threshold values. If it is not, step 201 follows. If it is, the release threshold value is determined and modified, or the acceleration is modified (in relation to a measured time) and then integrated, in step 207. In step 209, the passenger restraint system of the vehicle is released as a function of a comparison of the integrated acceleration to the release threshold value.