JPH10505034A - Matching filter for vehicle collision discrimination - Google Patents

Abstract

(57) [Summary] The matched filter (14) for the vehicle collision discrimination system (10) uses at least one reference factor setting stored in a storage device (24). The number of elapsed times is combined with the setting of the acceleration data sample (22a-n). The numerical values of the n reference factors are selected in accordance with the execution of the target vehicle collision discrimination plan. The scaled or filtered samples are then summed (26) to generate an output signal suitable for use in crash discrimination analysis. The matched filter (14) of the present invention can optimize the flexibility and reconfiguration of the system design by selecting the corresponding reference factors to store in storage, thereby performing different planning runs. Can be easy.

Description

Description: BACKGROUND OF THE INVENTION The present invention relates to a vehicle collision discrimination system used to control the operation of an occupant safety restraint, such as an airbag, and more particularly to a vehicle collision discrimination system. Optimizing the work of In general, to sense vehicle collisions and sudden vehicle deceleration, to deploy vehicle safety equipment such as airbags, to lock a seat belt, or to activate a pretensioner for a seat belt retractor. Various systems and methods have been used. Typically, these safety devices are triggered by the system detecting an impact or deceleration above its threshold, or are otherwise placed in a protective position. These known systems and methods are generally designed around a predefined collision discrimination plan that detects collisions in response to variously designed physical measures, such as vehicle speed or speed changes [jerk values]. Have been. However, to be effective, most crash discrimination plans and their intended thresholds are narrow for particular vehicle products and types and / or corresponding to particular vehicle crash types. It has a tendency to be tailored within the range. Thus, known collision discrimination systems perform a targeted collision discrimination plan using specific hardware and / or firmware designed for the narrow range of specific specifications required. Thus, these systems are particularly difficult, if not impossible, to redesign or reconfigure for use in different crash plans or vehicle types. A rigidly integrated design into this system does not simply facilitate easy and inexpensive modification, updating, and / or reconfiguration of the system. SUMMARY OF THE INVENTION Accordingly, it is an object of the present invention to provide a matched filter having a structure that optimizes reconfigurability and flexibility in the design of a vehicle crash discrimination system. It is also an object of the present invention to provide a matched filter for applying a predetermined scaling factor s to vehicle acceleration data to perform crash discrimination analysis or planning. It is another object of the present invention to optimize the design flexibility of a vehicle collision discrimination system by providing a matched filter, which allows the vehicle collision discrimination system to have a reference stored therein. By changing the setting of the factor, it can be easily reconstructed. In accordance with aspects of the present invention, a matched filter for vehicle collision discrimination processes data representative of vehicle acceleration into a discrimination output value suitable for determining whether to deploy an occupant safety restraint. The filter is an input means for organizing acceleration data into n elapsed time sequential data samples from the current sample to the nth old sample, and for storing at least one setting of n pre-defined reference factors. Storage means and means for combining each of the n data samples with a predetermined one of the n reference factors. Each of the n reference factor values is predetermined for the execution of a particular vehicle collision discrimination plan. The means for combining the n reference data samples provides a discrimination output value. The first filter structure includes at least one n number of discrete output values, such that the oldest acceleration data sample is exponentially converted to obtain a discrimination output value representative of vehicle speed combined with braking memory of past vehicle acceleration. Select each numerical value of the reference factor setting of. The second filter structure uses two n reference factor settings, thereby achieving long-term low-speed collision discrimination. The first setting of the reference factor is selected so as to obtain a discrimination output value representing the vehicle speed, and the second setting of the reference factor is selected so as to obtain the discrimination output value representing the jerk value. The present invention will be more fully understood from the following detailed description of the preferred embodiments when read in conjunction with the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a schematic diagram of a closed circuit illustrating a vehicle collision discrimination system according to the present invention. FIG. 2 is a schematic diagram of a closed circuit of the matched filter according to the present invention. FIG. 3 is a graph illustrating the setting of the reference factor of the first filter structure. FIG. 4 is a graph illustrating the setting of the reference factor of the second filter structure. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS As shown in FIG. 1, a collision discrimination system 10 according to the present invention receives raw vehicle acceleration data from at least one accelerometer in an input data preprocessing stage 12. This preprocessing stage 12 can include signal processing circuits such as A / D converters, filters, natural logarithmic calculation means and autocorrelators, as described in US Pat. No. 5,282,134; The description of this patent is incorporated herein by reference. The output generated by this preprocessing step 12 is then applied to collision discrimination matched filter steps 14 (a) and 14 (b), where at least one of the corresponding decision steps 16 (a) and 16 (b) Is processed into output data suitable for comparison with a predefined threshold value. The decision discrimination process responsive to this corresponding threshold comparison includes an AND gate 18 that controls the operation of a vehicle occupant safety restraint such as an airbag. The present invention achieves optimal design flexibility by performing an appropriate vehicle collision discrimination analysis with a matched filter 14 when executing a target collision discrimination plan. As shown in FIG. 2, the matched filter 14 includes sequentially ordered output slots that input to corresponding multiplying or summing junctions 22 (a)-(n), respectively. And a FIFO register having the following. A storage device 24 such as a ROM applies predetermined n reference factor settings to the multiple zygotes 22 (a) to 22 (n), and performs collision discrimination processing on the received vehicle acceleration data samples. The n scaled acceleration data samples are then added in a summer unit 26 and output to the corresponding threshold comparison stage 16. Alternatively, the scaled acceleration data sample can be stored in an accumulator. In accordance with aspects of the invention, a comprehensive crash discrimination analysis is performed on the received acceleration data samples by calculating each of the n reference factor settings. As an example, FIG. 3 shows a first setting of n reference factors that allow the matched filter to determine the vehicle speed while damping the older acceleration sample. More specifically, the older data values of the n data samples are exponentially smaller, and the velocity is calculated using the 100% value of the most recent n data samples. Thus, the newer sample has a reference factor of 1.0, and the older sample has an exponentially smaller reference factor. This design allows for speed calculations that incorporate a "memory" of past events. FIG. 4 shows a second structure of n criterion factors, which provides collision discrimination for long term slow collisions. This is described in Serial No. 07 / 874,431, filed on April 27, 1992, entitled "Method for Discriminating Long-Period, Low-Velocity Crashes", filed on April 27, 1992. As described above (incorporated herein by reference), the wide gray area between the "ignition" and "no ignition" crash situations can accurately discriminate slowly occurring types of vehicle crashes. Have difficulty. The method described in Serial No. 07 / 874,431 compares these vehicle speed data with a threshold value that does not activate vehicle safety equipment such as airbags and is approximately equal to half the maximum permissible low speed collision speed. The type of collision can be determined accurately. If this speed threshold is exceeded, the speed gradient, ie the jerk value, is calculated and then compared to the slope threshold value. If this slope threshold is exceeded, the safety gear is activated or ignited. As shown in FIG. 4, according to the present invention, under the control of a feedback loop 28 and a controller 30, by using two separate settings of selectively applied reference factors. This kind of collision planning structure is obtained in a matched filter. More specifically, the first setting of the reference factor (indicated at 32) comprises a first setting (1.0) applied to the accelerated sample for speed determination. Alternatively, a braking structure such as that shown in FIG. 3 may be employed. If the feedback 28 / controller 30 determines that the speed threshold corresponding to half of the maximum allowable slow collision speed has been exceeded, a second setting of the reference factor (shown at 34) ranging from -1.0 to 1.0. Is applied to the acceleration data samples to allow the calculation of the accumulator slope. Thus, by using one or more matched filters as a means of performing a predefined collision discrimination process, the present invention provides the advantage of system design flexibility. Various different collision discrimination processes or plans can be easily implemented by changing the reference factor stored in the storage device 24. Furthermore, the present invention has been described as employing at least two matched filters, or it will be apparent to those skilled in the art that a number of different matched filters can be used in a single impact discrimination system. The foregoing preferred embodiments of the present invention are shown by way of example only, and the various structures disclosed herein may be used without departing from the spirit and scope of the invention as defined in the appended claims. It should be understood that many modifications of the operating characteristics can be made.

[Procedure amendment] Patent Law Article 184-8 [Submission date] September 26, 1996 [Amendment] The purpose of the present invention is to apply a predetermined scaling factor (s) to vehicle acceleration data. And a matched filter for performing collision discrimination analysis or planning. It is another object of the present invention to optimize the design flexibility of a vehicle collision discrimination system by providing a matched filter, which allows the vehicle collision discrimination system to have a reference stored therein. By changing the setting of the factor, it can be easily reconstructed. In accordance with aspects of the present invention, the matched filter of the vehicle collision discrimination system processes data representing vehicle acceleration into a discrimination output value suitable for determining whether to deploy an occupant safety restraint. The filter is an input means for organizing the acceleration data into n successive elapsed time data samples from the current sample to the nth old sample, and for storing at least one setting of n pre-defined reference factors. Storage means and means for combining each of the n data samples with a predetermined one of the n reference factors. Each of the n reference factor values is predetermined for the execution of a particular vehicle collision discrimination plan. The means for combining the n reference data samples provides a discrimination output value. The first filter structure includes at least one n number of discrete output values, such that the oldest acceleration data sample is exponentially converted to obtain a discrimination output value representative of vehicle speed combined with braking memory of past vehicle acceleration. Select each numerical value of the reference factor setting of. The second filter structure uses two n reference factor settings, thereby achieving long-term low-speed collision discrimination. The first setting of the reference factor is selected so as to obtain a discrimination output value representing the vehicle speed, and the second setting of the reference factor is selected so as to obtain the discrimination output value representing the jerk value. The present invention will be more fully understood from the following detailed description of the preferred embodiments when read in conjunction with the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a schematic diagram of a closed circuit illustrating a vehicle collision discrimination system according to the present invention. Figure 2 Claims 1. A matched filter that processes data indicative of vehicle acceleration and provides a discrimination output value suitable for threshold comparison to determine whether to deploy an occupant safety restraint device, the nth oldest sample from the current sample Input means for organizing acceleration data into n time-lapse sequential data samples up to; storage means for storing at least one setting of n predetermined reference factors; Means for combining each with one of the predetermined n reference factors, each value of said n reference factors being predetermined to execute a predetermined vehicle collision discrimination plan. Means for combining the n reference data samples to obtain a discrimination output value. 2. The value of each of the at least one setting of the n reference factors being selected such that the oldest acceleration data sample is exponentially converted to obtain a discriminated output value that has been braked with respect to past vehicle acceleration. Item 3. The matched filter according to Item 1. 3. The at least one setting of the n reference factors comprises first and second settings of the n reference factors, the first setting reference factor being selected to provide a discrimination output value representing vehicle speed. 2. The matched filter of claim 1, wherein the second set criterion factor is selected to provide a discrimination output value representative of a jerk value.

Claims (1)

[Claims] 1. Process data indicating vehicle acceleration to deploy occupant safety restraint device A filter for providing a discrimination output value suitable for determining whether     N time lapse sequences from the current sample to the nth oldest sample Input means for organizing the acceleration data during the data sample;     Storage means for storing at least one setting of n predefined reference factors ;     Combine each of the n data samples with one of the n predefined reference factors Means for matching, each value of the n reference factors being a predetermined vehicle collision Predetermined to carry out the discrimination plan; and     Means for combining the n reference data samples to obtain a discrimination output value;   The filter comprising: 2. The numerical value of at least one setting of the n reference factors is the oldest acceleration data. The vehicle speed incorporated in the braking memory of past vehicle acceleration 2. The filter according to claim 1, wherein the filter is selected to obtain a discriminative output value. . 3. The setting of at least one of the n reference factors is based on the first and the n reference factors. And the first setting criterion factor is that the discrimination output value representing the vehicle speed is And the second set criterion is a valve representing the jerk value. The filter according to claim 1, wherein the filter is selected to obtain another output value.