Data processing: vehicles – navigation – and relative location – Vehicle control – guidance – operation – or indication – Vehicle subsystem or accessory control
Reexamination Certificate
1997-11-03
2003-08-19
Louis-Jacques, Jacques H. (Department: 3661)
Data processing: vehicles, navigation, and relative location
Vehicle control, guidance, operation, or indication
Vehicle subsystem or accessory control
C701S046000, C701S047000, C180S268000, C180S272000, C180S282000, C280S734000, C280S735000
Reexamination Certificate
active
06609053
ABSTRACT:
BACKGROUND OF THE INVENTION
Pattern recognition techniques, such as artificial neural networks are finding increased application in solving a variety of problems such as optical character recognition, voice recognition, and military target identification. In the automotive industry, pattern recognition techniques have now been applied to identify various objects within the passenger compartment of the vehicle, such as a rear facing child seat, as well as to identify threatening objects such as an approaching vehicle about to impact the side of the vehicle. See, for example, patent application Ser. No. 08/239,978 filed May 9, 1994 (now U.S. Pat. No. 5,563,462), Ser. No. 08/640,068 (now U.S. Pat. No. 5,829,782) and Ser. No. 08/247,760 (now abandoned) filed May 23, 1994 which are included herein by reference. Heretofore, pattern recognition techniques have not been applied to sensing automobile crashes for the purpose of determining whether or not to deploy an airbag or other passive restraint, or to tighten the seatbelts, cutoff the fuel system, or unlock the doors after the crash.
“Pattern recognition” as used herein means any system which processes a signal that is generated by an object, or is modified by interacting with an object, in order to determine which one of a set of classes the object belongs to. Such a system might determine only that the object is or is not a member of one specified class, or it might attempt to assign the object to one of a larger set of specified classes, or find that it is not a member of any of the classes in the set. The signals processed are generally electrical signals coming from transducers which are sensitive to either acceleration, or acoustic or electromagnetic radiation and, if electromagnetic, they can be either visible light, infrared, ultraviolet or radar.
To “identify” as used herein means to determine that the object belongs to a particular set or class. The class may be one containing all frontal impact airbag desired crashes, one containing all events where the airbag is not required, one containing all events requiring the passenger headrest to be moved into position, or one containing all events requiring the deployment of an airbag in the event of side impacts depending on the purpose of the system.
All electronic crash sensors currently used in sensing frontal impacts include accelerometers which detect and measure the vehicle accelerations during the crash. The accelerometer produces an analog signal proportional to the acceleration experienced by the accelerometer and hence the vehicle on which it is mounted. An analog to digital converter transforms this analog signal into a digital time series. Crash sensor designers study this digital acceleration data and derive therefrom computer algorithms that determine whether the acceleration data from a particular crash event warrants deployment of the airbag. This is usually a trial and error process wherein the engineer or crash sensor designer observes data from crashes where the airbag is desired and when it is not needed, and other events where the airbag is not needed. Finally, the engineer or crash sensor designer settles on an algorithm that seems to satisfy the requirements of the crash library, i.e., the crash data accumulated from numerous crashes and other events. The resulting algorithm is not universal and most such engineers or crash sensor designers will answer in the negative when asked whether their algorithm will work for all vehicles.
Several papers have been published pointing out some of the problems and limitations of electronic crash sensors which are mounted out of the crush zone of the vehicle, usually in a protected location in the passenger compartment of the vehicle, the crush zone being defined as that portion of the vehicle which has crushed at the time that the crash sensor must trigger deployment of the restraint system. These sensors are frequently called single point crash sensors. Technical papers which discuss these limitations along with discussions of the theory of crash sensing, which are relevant to this invention and which are included herein by reference, are:
1) Breed, D. S. and Castelli, V. “Problems in Design and Engineering of Air Bag Systems”, Society of Automotive Engineers Paper SAE 880724, 1992.
2) Breed, D. S., Castelli, V. “Trends in Sensing Frontal Impact”, Society of Automotive Engineers Paper SAE 890750, 1989.
3) Breed, D. S., Sanders, W. T. and Castelli, V. “A Critique of Single Point Crash Sensing”, Society of Automotive Engineers Paper SAE 920124, 1992.
4) Breed, D. S., Sanders, W. T. and Castelli, V. “A complete Frontal Crash Sensor System-I ”, Society of Automotive Engineers Paper SAE 930650, 1993.
5) Breed, D. S. and Sanders, W. T. “Using Vehicle Deformation to Sense Crashes”, Presented at the International Body and Engineering Conference, Detroit Mich., 1993.
6) Breed, D. S., Sanders, W. T. and Castelli, V., “A complete Frontal Crash Sensor System-II”, Proceedings Enhanced Safety of Vehicles Conference, Munich, 1994, Published by the US Department of Transportation, National Highway Traffic Safety Administration, Washington, D.C.
These papers demonstrate, among other things, that there is no known theory which allows an engineer to develop an algorithm for sensing crashes and selectively deploying the airbag except when the sensor is located in the crush zone of the vehicle. These papers show that, in general, there is insufficient information within the acceleration signal measured in the passenger compartment to sense all crashes. Another conclusion supported by these technical papers is that if an algorithm can be found which works for one vehicle, it will also work for all vehicles since it is possible to create any crash pulse in any vehicle. See in particular SAE paper 920124 referenced above.
In spite of the problems associated with finding the optimum crash sensor algorithm, many vehicles on the road today have electronic single point crash sensors. Some of the problems associated with single point sensors result in that an out-of-position occupant who is sufficiently close to the airbag at the time of deployment is likely to be injured or killed by the deployment itself Fortunately, systems are now being developed which monitor the location of occupants within the vehicle and can suppress deployment of the airbag if the occupant is more likely to be injured by the deployment then by the accident.
Since there is insufficient information in the acceleration data, as measured in the passenger compartment, to sense all crashes and since some of the failure modes of published single point sensor algorithms can be easily demonstrated using the techniques of crash and velocity scaling described in the above referenced technical papers, and moreover since the process by which engineers develop algorithms is based on trial and error, pattern recognition techniques such as neural network should be able to be used to create an algorithm based on training the system on a large number of crash and non-crash events which will be superior to all others. This in fact has proved to be true and is the subject of this invention.
Naturally, once any crash sensor has determined that an airbag should be deployed, the system should perform several other functions such as tightening the seatbelts for those vehicles which have seatbelt retractor systems, cutting off of the fuel system to prevent fuel spillage during or after the crash, and unlocking the doors after the crash to make it easier for the occupants to escape.
The use of pattern recognition techniques in crash sensors has another significant advantage in that it can share the same pattern recognition hardware and software as other systems in the vehicle. Pattern recognition techniques have proven to be effective in solving other problems related to airbag passive restraints. In particular, the identification of a rear-facing child seat located on the front passenger seat, so that the deployment of the airbag can be suppressed, has been demonstrated. Also, the use of patt
Automotive Technologies International Inc.
Louis-Jacques Jacques H.
Roffe Brian
LandOfFree
Method and apparatus for sensing a vehicle crash does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method and apparatus for sensing a vehicle crash, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method and apparatus for sensing a vehicle crash will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3096790