Communications: electrical – Land vehicle alarms or indicators – Internal alarm or indicator responsive to a condition of the...
Reexamination Certificate
2000-04-03
2001-04-24
Hofsass, Jeffery (Department: 2736)
Communications: electrical
Land vehicle alarms or indicators
Internal alarm or indicator responsive to a condition of the...
C340S443000, C340S444000, C340S466000
Reexamination Certificate
active
06222444
ABSTRACT:
FIELD OF THE INVENTION
The invention relates to a method of detecting when the air pressure in a tire has fallen below a predetermined pressure level, based on the measured wheel speeds.
BACKGROUND OF THE INVENTION
The advent of anti-lock brake systems (ABS) and the placement of speed sensing devices at each of the wheels has sparked efforts to develop reliable methods for detecting tire deflation based on wheel speeds measured during driving. Theoretically, four equally inflated tires on a vehicle will have substantially the same rolling radius and will therefore each rotate at substantially the same speed during straight-line driving on a dry, flat and uniform surface. When a tire becomes deflated, its rolling radius is reduced and the wheel speed increases to compensate for the smaller radius. Numerous methods have been created that monitor the wheel speeds and detect variations that may be attributable to tire deflation.
A recurring problem associated with these deflation detection methods involves the ability to filter or eliminate the collection of faulty data points caused by a variety of factors such as built-in tire radius variations, vehicle maneuvering, road conditions and drive slip. Without eliminating or filtering out these faulty readings, the detection methods are prone to improper warnings of tire deflation or failure to detect actual tire deflation. False warnings are troublesome and annoying to drivers, while failure to detect deflation is dangerous and could result in tire blow-outs. Methods that accurately detect tire deflation while avoiding false detection have become the utmost priority.
Filtering or eliminating faulty data caused by drive slip has been a significant obstacle. Simply stated, a vehicle's driven wheels slip due to the torque being applied to the axle. This slippage results in higher wheel speeds for the driven wheels than for the non-driven wheels. For example, on a rear wheel drive vehicle travelling at highway speeds, the rear wheels may rotate approximately one percent faster than the non-driven front wheels. This one percent variation is unacceptably high in light of the small wheel speed variations (typically 0.1 percent to 0.5 percent) caused by an actual deflated tire. If faulty data is not filtered or eliminated, drive slip error can wash out the influence of tire pressure and thus severely degrade the ability to detect tire deflation. Drive slip becomes even more problematic during acceleration, uphill/downhill driving and driving on non-uniform surfaces (such as dirt, sand or gravel).
Several attempts have been made to filter or eliminate the effect of drive slip on collected data. U.S. Pat. No. 5,760,682, issued Jun. 2, 1998, applies an analysis of the variance (ANOVA) statistical technique to the data collected from all four wheel speeds. The statistical method incorporated provides more accurate results than the more common average value comparison methods (where the data collected for each wheel is simply averaged before being used in a comparison algorithm) typically used. Filters are used to eliminate data collected during acceleration/deceleration, uphill/downhill driving, tuning/cornering fluctuation and rough road driving, but drive slip occurring during straight-line driving would tend to fool the analysis of variance technique since it is not able to distinguish the increase in wheel speed due to drive slip from the increase in wheel speed due to tire deflation. A false detection may occur.
U.S. Pat. No. 5,578,984, issued Nov. 26, 1996, discloses a system where drive slip is learned and compensated for with a correction factor designated as the front/rear wheel ratio Z. Such a learning process is not robust since data resulting from such learning only applies to the surface on which it was learned. For example, if such learning occurred on dry asphalt, the resulting correction factor will be wrong for data collected while driving on wet asphalt. Additionally, if such learning occurred on a level road surface, the resulting correction factor will be wrong for data collected while driving uphill since uphill driving requires more power at the drive wheels, causing more drive slip. False detection or failure to detect may result.
U.S. Pat. No. 4,876,528, issued Oct. 24, 1989, and U.S. Pat. No. 5,591,906, issued Jan. 7, 1997, disclose methods wherein angular velocities of two diagonally opposed wheels are added together and then compared (using various techniques) with the sum of the angular velocities of the other two diagonally opposed wheels. While this method should be resistant to drive slip error, there are other limitations associated with the formula, such as sensitivity to the diagonal component. For example, the front left wheel and rear right wheel could each only be slightly deflated (perhaps only ten percent), but the sum of that diagonal would appear the same as the case where only one of the wheels were significantly deflated (perhaps twenty-five percent). This results in undesired sensitivity since the objective is to detect twenty-five percent or greater deflation on one tire. Again, false detection may result.
Another problem with combining data from diagonally opposed wheels is discussed in U.S. Pat. No. 5,578,984. In many high performance sports cars, different sized tires are used for the front and rear axles. When this occurs, the critical threshold values used to detect deflation will be different for front and rear tires. Using diagonally opposed front and rear tires commingles the data so that the front and rear wheels can not be treated independently.
SUMMARY OF THE INVENTION
The present invention provides an improved method that substantially remedies the above-identified problems associated with attempts to eliminate error due to drive slip. Unlike the prior art methods that utilize data taken from all four tires in various combinations (i.e., comparing values from all four tires individually, opposing sets of diagonal tires or front tires and rear tires), the method of the present invention recognizes that utilizing combinations of data taken from all four tires cannot adequately prevent errors associated with drive slip.
As such, the present invention comprises a method for detecting a deflated tire wherein only the two wheels of one axle are considered at any one time to determine whether a tire is deflated. Since both wheels on a given axle rotate at substantially the same speed during straight-line driving, the comparison is inherently immune to drive slip and errors associated with using data from diagonally opposed tires. Even though the driven wheels tend to rotate faster than the non-driven wheels and may be larger than the non-driven wheels, the relative speed difference between the two driven wheels or the two non-driven wheels is zero. The present invention therefore analyzes the difference in speed between two wheels on a single axle to detect tire deflation. On the other hand, all four wheel speeds are still used together to detect rough road conditions and vehicle maneuvering conditions that provide other instances when false data should be filtered or eliminated. The result is a method of detecting a deflated tire that recognizes instances when data is faulty and should not be used, and additionally utilizes an algorithm that is inherently immune to drive slip error.
More specifically, the present invention is an improvement over U.S. Pat. No. 5,760,682, which is hereby incorporated by reference. In the present invention, a new statistical method suited for comparing data variability between two wheels replaces the ANOVA statistical analysis, which is suited for comparing data variability between all four wheels. The present invention utilizes a paired t-test statistical analysis, and more preferably a modification of the paired t-test, to consider data variability. Use of the modified paired t-test provides deflation detection that is more accurate and reliable than the previous ANOVA method which was already more accurate than calculations using the more common average v
Hofsass Jeffery
Michael & Best & Friedrich LLP
Robert Bosch Corporation
Tang Son
LandOfFree
Method for detecting a deflated tire on a vehicle 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 for detecting a deflated tire on a vehicle, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method for detecting a deflated tire on a vehicle will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2526655