JPH05213153A - Expandable constraining system and system developing method - Google Patents

Abstract

PURPOSE: To improve performance by calculating a speed when an acceleration signal is beyond a threshold, and comparing the speed with a time dependent value on the speed obtained from the temporally changing boundary value of the speed, and outputting a system development enabling command when the speed is beyond the time dependent value in an expansible restraint system for a vehicle. CONSTITUTION: Acceleration along the longitudinal shaft of a vehicle is detected by an accelerometer 10, and an acceleration signal is transmitted to a filter 12. The filter 12 transmits an output ACCELX1 to a microprocessor 14 and an amplifier 18. Then, the amplifier 18 transmits an ACCELX8 to a microprocessor 14. The microprocessor 14 calculates the mean value of the nearest values of the ACCELX1 and X8, and then adjusts it so that a bias and scale coefficient can be generated, and compares it with a threshold. When this value is the threshold or more, it is integrated so that speech change can be calculated, and compared with a speed boundary curve. When this is beyond the curve, a development enabling command is outputted to an air bag developing system 22.

Description

Detailed Description of the Invention

[0001]

FIELD OF THE INVENTION This invention relates to inflatable restraint systems such as air bag systems for motor vehicles and methods for enabling deployment of such systems.

[0002]

Crash sensors process acceleration of sensor position in a vehicle to produce a signal for inflation of an air bag when the processed acceleration matches a specified reference value. For example, in one prior art system, a spring-restrained ball moves under the influence of an applied acceleration. When this ball movement (accumulated weighted integration) exceeds a specified amount, the electrical contacts are closed and an expansion signal is generated. Optimum performance is usually obtained by adjusting sensor parameters.

Systems have also been proposed that include a piezoelectric accelerometer that provides an acceleration signal that is integrated when the integrated acceleration (speed) signal exceeds a certain threshold speed to produce an air bag enable signal. For example, US Pat. No. 3,911,
See 391.

[0004]

SUMMARY OF THE INVENTION The present invention seeks to provide an improved inflatable restraint system and method of deploying such a system.

According to one characteristic of the invention, a method is provided which enables the deployment of an inflatable restraint system for a motor vehicle as claimed in claim 1.

According to another characteristic of the invention, there is provided an inflatable restraint system for a motor vehicle as claimed in claim 15.

According to the present invention, it is possible to approach the optimum performance of an air bag system in terms of processing the acceleration signal without regard to the final sensor implementation.

A velocity boundary approach can be used to discriminate between unfolding and unfolding collision events by defining a curve drawn for every unfolding event in the velocity versus time domain. The criterion used to initiate inflation of the air bag is based on changes in vehicle speed measured during the crash against previously undeployed events tested. The data forming the velocity boundary curve can be obtained from a crash testing a particular vehicle and is stored in a look-up table used as a function of time for a crash event. If the change in vehicle speed as indicated by the acceleration data exceeds the speed boundary curve, an air bag deployment enable command can be issued.

Accordingly, the present invention can provide an improved method of commanding the deployment of a vehicle's air bag, particularly based on the treatment of vehicle deceleration and the comparison of dependent speed thresholds.

One embodiment of the present invention is described by way of example only with reference to the accompanying drawings.

[0011]

DESCRIPTION OF THE PREFERRED EMBODIMENT In FIG. 1, multiple undeployed events for a particular vehicle are shown in relation to changes in velocity (km / h) as a function of time (milliseconds) after the start of the event. Certain events are of particular interest to collision detection systems, some in terms of discrimination, and others in terms of timing.
The non-deployment test shown in FIG. 1 is typically performed to assess the insensitivity of the collision detection system to actual events. The "bad road" plot shows test data from several different events of this kind. "Abrupt impact (abu
The "se impact)" plot represents multiple tests such as hammering, rough handling of doors and hoods. The "Deer Collision" plot represents the impact at 81 Km / H (50 MPH) with a simulated 50 Kg deer. "Collision collision" plot is 32.4 km / h
Fig. 6 represents a chassis collision with a curved railroad track or a large rock at (20MPH). Another non-expanded plot represents frontal obstacle collision at 14.6 Km / H (9 MPH).

The plot identified as the boundary is the line drawn near all unexpanded events in the velocity and time domain. Deployment event (in the case shown, 48.6 km
/ H (30MPH) obstacle test) crosses the boundary because it exhibits a greater change in velocity compared to the undeployed event. The air bag is deployed at a time T _f when the vehicle decelerates from its initial velocity by about 1.94 Km / H (1.2 MPH), which is the first severity indication of this deployment event.

Velocity boundary curve (velocity bo)
Undary curve (VBC) is defined as the curve that lies on an unexpanded composite curve at scheduled time intervals (based on velocity versus time graph), separated from the composite curve by a fixed or percentage amount. This compound curve is the locus of all maximum (or minimum) point characteristics of the set of collisions used. Typically, all collisions or other events that do not require air bag deployment are used to generate the maximum composite curve, and collisions requiring air bag deployment are used to obtain the minimum composite curve. .. These compound curves are (a) loaded with the net velocity change values for each crash or other event tested from an external file, the largest (unexpanded event) or from the loaded data as specified by the user or Any of the smallest (expansion event) velocity change values can be found at each point in time and (c) generated by a computer program that stores and writes the values thus obtained to separate the output files. Therefore, VBC is
A curve that meets the two criteria of not intersecting the maximum composite curve below and intersecting the minimum composite curve as soon as possible. Another constraint can be included in the VBC generation to further optimize the system.

Referring now to FIG. 2, accelerometer 10 is preferably mounted rigidly in the passenger compartment of the vehicle to produce a deceleration signal along the vehicle's longitudinal axis. This signal is provided to the filter 12, the output of which is represented by ACCELX1. This filter output ACCELX1 is sent to the microprocessor 14 via one of its analog / digital (A / D) inputs 16. This filter output ACCELX1 is also sent to an amplifier 18, which gives a predetermined gain to the signal ACCELX1 to produce a higher gain signal ACCELX8, which is the second analog / digital (A / D) of the microprocessor 14. )
Sent to input 20.

The analog provided by the filter 12
In addition to filtering, microprocessor 14 performs digital filtering as the signal is processed. A 2.5 ms moving average is achieved by sampling the analog / digital (A / D) input of the accelerometer every 500 ms and holding the last five values in memory. The average value of each accelerometer signal is then adjusted to produce a bias and scale factor. If the adjusted accelerometer signal ACCELX8 exceeds a threshold value, this is integrated to obtain the resulting change in vehicle speed. This velocity change is compared to a time-dependent threshold or velocity boundary curve, which in the preferred embodiment is indicated by a look-up table. If the change in vehicle speed exceeds the speed boundary curve, a deployment enable command is issued to the air bag deployment system 22.

Referring now to FIGS. 3-5, flowcharts of algorithms implemented by microprocessor 14 are shown. In steps 24 and 26, the microprocessor 14 causes the acceleration signal A to elapse every 500 milliseconds.
Sample each of CCELX1 and ACCELX8. In steps 28 and 30, accelerometer signal A
CCELX1 and amplified accelerometer signal ACCEL
Each of the 5 most recent values of X8 is saved. Signal A
CCELX1 and ACCELX8 are independently averaged in steps 32 and 34 to register 36, 38.
Used to calculate long-term bias values BIASX1 and BIASX8, respectively stored in

The output of accelerometer 10 has a bias or offset value equal to about half the analog / digital (A / D) dynamic range of the microprocessor. This allows both positive and negative acceleration measurements by the microprocessor A / D converter. The zero acceleration output of the accelerometer and subsequent signal processing circuits is subject to tolerance and thermal effects. To minimize these errors, zero acceleration values are measured at start-up (ie, when the ignition switch is turned to the on position) while the vehicle is stationary. This vl is validated against a permanently stored value and held as a bias value for collision discrimination in registers 36,38. This bias value is adjusted during the operating cycle to account for long-term changes.

Digital acceleration values are stored in bias registers 36, 38 as part of the power-up or initialization procedure. At the end of each subsequent 1 second interval, the difference between the most significant byte of the bias register and the acceleration signal as detected at steps 40 and 42 is the difference between the bias registers 36 and 38 at steps 44 and 46. Algebraically added to the least significant byte. The remaining difference between the acceleration value and the bias register is the bias register 3
Overflow the least significant bytes of 6,38 to the most significant byte, which causes the bias register value to converge to the accelerometer output. The small difference requires 4 minutes before the most significant byte begins to converge on the new steady state accelerometer output. Large changes propagate fast, but it still takes more than 2 seconds to cause a change in bias value. Thus, long-term changes are allowed during collision sequences with durations on the order of tenths of a second without changing the bias value.

The "calculation cycle" starts when the deceleration of the vehicle exceeds a threshold value. The amplified acceleration signal is used for this purpose to reduce the quantization error caused by the A / D converter. The accelerometer signal with gain ACCELX8 is adjusted in step 48 after adjustment for bias.
It is multiplied by a scale factor at 50 and compared at step 52 with the 1G threshold deceleration value shown at step 54, for example. If this accelerometer signal exceeds the threshold value, the calculation cycle begins at step 56. In the calculation cycle, the AND function step 57
Is enabled and the bias adjustment is interrupted by the INHIBIT line to prevent large sustained acceleration inputs occurring during a crash from changing the bias value.

As shown in step 55, for each millisecond in the calculation cycle, the five most recent samples of the amplified acceleration signal ACCELX8 from step 30 are all within limits, that is, the dynamics of the A / D conversion. A test is made at step 58 to see if it is within range.
Also, the millisecond counter 60 is incremented. Signal ACCE
If the 5 most recent samples of LX8 are within limits,
These values are added in step 62, and the bias amount BIASX obtained in step 66 is added in step 64.
Adjusted by subtracting a value equal to 8x5 and added to the INTEGX8 register in step 68.
If any of these values are out of range, the five most recent samples of the unamplified acceleration signal ACCELX1 are added at step 70 to give a value equal to the bias signal BIASX1 × 5 obtained at step 74 at step 72. Adjusted by subtraction and added to INTEGX1 in step 76. Therefore, one of the two registers 68 and 76 is updated every millisecond. The five adjusted values ACCELX8 are stored in register I
It is added to NTEGX8 or the five adjusted values ACCELX1 are added to the register INTEGX1. Therefore, the registers INTEGX1 and INTE
Each GX8 contains a portion of the change in velocity and the count in each has a different weight.

Registers INTEGX1 and INTEG
The values at X8 are step 78 and 80 with their appropriate scale factors in steps 48 and 82, respectively.
To obtain the total velocity change value. Starting at 5 ms after the start of the calculation cycle, whenever the counter 60 exceeds the reference value of 5 ms, the AND function step 8
6 and 88 are enabled from the "Compare" functional step 90. A delay of 5 ms after the start of the event avoids the effects of short duration high shock acceleration by integrating the initial high amplitude oscillation to zero before the first comparison is made. Next, in step 94, the register INTEGX
1 and the total velocity change calculated using the valve in INTEGX8 is compared with the corresponding value in the "Velocity Boundary Table 96" for every subsequent millisecond interval in the "calculation cycle" as determined by the millisecond counter 60. To be done. This table 96 contains velocity values for each millisecond of the time of the "velocity boundary curve" of FIG.

If this comparison indicates a change in vehicle speed equal to or greater than the table value, then a 20 millisecond timer is started at step 98 to indicate a "digestible command".
Issue 100. This timer is restarted in step 98 by each subsequent comparison that exhibits a sum equal to or greater than the "speed boundary table" value to expand the "digestible instruction". This "digestible command" allows the air bag to be inflated if the other crash sensors also indicate a crash condition. Without the "digestible command," it would prevent inflation of the air bag. Therefore, if the speed change continuously decreases below the speed boundary curve of FIG.
Is retracted.

Once the AND function step 88 is enabled (5 ms for the calculation cycle), the total speed change calculated in step 86 is less than the minimum value established by the minimum speed threshold given by step 102. , Which is detected by the "compare" function step 104 and causes a reset command by the "or" function step 106, which stops the comparison process and steps 68, 7
6. Reset the integration register in 6 and the time index register in step 60.

During the "calculation cycle", the counter is incremented at step 112 each time the acceleration falls below the threshold provided by step. This counter is reset in step 112 each time the acceleration rises above the threshold given in step 54, so that the contents of the counter represent the time during which this acceleration continuously falls below the threshold in the "calculation cycle". .. The content of the counter is compared in step 114 with the reset value provided by step 116, for example 20 milliseconds, and when this time is reached, step 68,
76 calculation processes and registers, and step 6
The zero counter is reset by the "or" function step 106. This counter is the "or" function step 10
It is also reset by 6.

The length of the velocity boundary table in step 96 is fixed and is checked in step 108. After reaching the limit of this table as determined in the comparison function step 110 (ie, when the millisecond count for "computation cycle" reaches the maximum cycle), the computation process is stopped and the integrator and time index registers are set to " It is reset to zero by the "or" function step 106.
This reset does not block the in-progress start command timeout that is in progress.

After the calculation process has been reset and before the calculation process can be restarted, the average deceleration is calculated in step 54.
Must be equal to or greater than the threshold given by

[Brief description of drawings]

FIG. 1 is a graph showing a velocity boundary line that separates a deployment event from various non-deployment events.

FIG. 2 is a block diagram of a portion of one embodiment of an inflatable restraint system.

3 is a flowchart of one embodiment of a deployment algorithm used in the system of FIG.

4 is a flow chart of one embodiment of a deployment algorithm used in the system of FIG.

5 is a flowchart of one embodiment of a deployment algorithm used in the system of FIG.

[Explanation of symbols]

 10 Accelerometer 12 Filter 14 Microprocessor 16 Analog / Digital (A / D) Input 20 Analog / Digital (A / D) Input 22 Air Bag Deployment System 36 Bias Register 38 Bias Register 60 Counter

Front Page Continuation (72) Inventor David Dexter Lynch, California 93111, Santa Barbara, Berkeley Road 5442 (72) Inventor Milos Macasek, California 93111, Santa Barbara, Paseo Cameo 5230

Claims (18)

[Claims] 1. A method for enabling deployment of an inflatable restraint system for an automobile, wherein an acceleration signal representative of the acceleration of the vehicle is generated, the acceleration signal is compared to a threshold value, and when the threshold value is exceeded, the acceleration signal of the vehicle is exceeded. Determining a speed and comparing the determined vehicle speed with a corresponding time-dependent value of the speed obtained from a time-varying boundary value of the speed, and when the speed value exceeds the time-dependent speed value, the restraint system A method comprising issuing an enabling directive to deploy a. 2. The method of claim 1, wherein the threshold value indicates the start of an event requiring deployment of the restraint system. 3. The method of claim 2, wherein the time dependent value is compared to a vehicle speed determined at the beginning of an event requiring deployment of the restraint system. 4. Method according to claim 1, 2 or 3, characterized in that the time-varying boundary value of velocity is obtained from experimental crash data. 5. The method according to claim 1, wherein the time-dependent boundary value is found from velocity data representing a plurality of unexpanded events. 6. The method of claim 5, wherein the time dependent boundary value is obtained from a boundary curve in a velocity versus time domain that limits velocity versus time test data for a plurality of unexpanded events. 7. When the acceleration signal exceeds a threshold value, a speed calculation cycle in which the vehicle speed is determined from a plurality of consecutive speed calculations is started, and each speed calculation in the cycle is obtained from the data of the corresponding time frame. 7. A time-dependent velocity boundary value that is compared with the predetermined time-dependent velocity boundary value.
The method according to any one of 1. 8. The method of claim 7, wherein the step of comparing the velocity calculation with the corresponding time-dependent velocity boundary value is performed a predetermined time following the start of the velocity calculation cycle. 9. The method of claim 7 or 8 including the step of resetting the speed calculation cycle if the speed comparison falls below a predetermined speed threshold after a predetermined time interval after the start of an event. the method of. 10. The step of obtaining the acceleration signal includes the step of generating an analog signal related to the acceleration of the vehicle and sampling the analog signal at a predetermined speed to obtain a discrete value of the acceleration. Claims 1 to 9
The method according to any one of 1. 11. The method of claim 10, wherein each discrete value of acceleration is compared to the threshold value. 12. A step of obtaining a moving average of discrete values of acceleration over a predetermined number of samples and adjusting the moving average in consideration of a bias of an accelerometer, wherein the threshold value is the adjusted moving average. 11. Method according to claim 10, characterized in that it is compared with a value. 13. The method according to claim 1, wherein the speed of the vehicle is determined by adding a predetermined number of discrete values of acceleration data. 14. Amplifying the analog signal to produce a plurality of discrete amplified analog signal values,
The vehicle speed may be calculated by adding a predetermined number of discrete values of the analog signal or a preset number of discrete amplified analog signals if any of the predetermined number of discrete values of the analog signal is outside a predetermined limit. 14. Method according to any of claims 10 to 13, characterized in that it is determined by adding a preset number of said discrete amplified analog signal values if all of the signal values are within predetermined limits. 15. An inflatable restraint system for a motor vehicle, comprising an inflatable device for protecting a vehicle passenger in a crash event and an accelerometer (1) mounted on the vehicle.
0) and monitoring the data from the accelerometer to detect the onset of an event requiring deployment of the inflatable device, and then from vehicle speed changes and multiple vehicle test events following the onset of the event. A processor (14) for determining whether the event justifies deployment of the device based on a comparison between a velocity boundary representing the obtained time-dependent velocity value,
Inflatable restraint, characterized in that the processor issues an enabling command to deploy the restraint system when the calculated velocity value is equal to or exceeds a velocity boundary in a predetermined time interval following the start of an event. system. 16. The inflatable restraint system according to claim 15, further comprising a memory for holding data in the form of an index table representing velocity values of the velocity boundary. 17. The processor (10) includes a single-byte memory and a 2-byte memory, the processor obtaining the acceleration value into the single-byte memory and the acceleration bias value into the 2-byte memory. Store and periodically add the difference between the acceleration value and the most significant byte of the acceleration bias value to the least significant byte of the 2-byte memory, thereby dynamically compensating for long-term changes in accelerometer bias. Inflatable restraint system according to claim 15 or 16, characterized in that 18. The velocity boundary does not intersect (a) a first composite curve formed from a trajectory of maximum velocity values at a predetermined time point for a plurality of non-expansion events, and (b) for a plurality of development events. 18. Inflatable restraint system according to any of claims 15 to 17, characterized in that it is in the form of a curve intersecting a second compound curve formed from the locus of the minimum velocity values at a given time.