Patent Grant
9340211
The invention relates generally to tire monitoring systems for collecting measured tire parameter data during vehicle operation and, more particularly, to systems utilizing such tire sensor-based data in vehicle control systems.
It is desirable to ascertain and use road friction levels for adjusting vehicle control systems such as braking, anti-lock brake, steering and collision avoidance. Estimation of road friction, however, has proven problematic when road conditions are subject to constant change during vehicle operation. Accordingly, there remains a need for a road friction estimation system that is robust and accurate and which can adapt to changes to road conditions during operation of a vehicle.
In one aspect of the invention, a road friction estimation system and method includes tire-affixed sensors and on-board vehicle sensors, a model-based tire force estimator operable generating from sensor-input a model-derived tire force estimation, a vehicle observer generating an observer-derived tire force estimation and a friction estimator generating a road friction estimation from a comparison of the model-derived tire force estimation and the observer-derived tire force estimation.
In another aspect, the vehicle observer receives inputs based upon sensor-measured tire-specific information including a load estimation for the vehicle tire, a slip angle estimation for the vehicle tire and a cornering stiffness estimation for the tire. The cornering stiffness estimation for the tire receives as inputs from the tire-affixed sensors tire temperature change, tire pressure change, wear state of the tire tread and loading on the tire.
According to another aspect, the model-based tire force estimator employs and utilizes a slip-dependent friction function.
Pursuant to another aspect, the cornering stiffness adaptation input providing the loading on the tire is derived from a dynamic load estimator having as inputs the vehicle-specific information and the tire-specific information including sensor-measured changes in tire temperature and pressure.
“ANN” or “Artificial Neural Network” is an adaptive tool for non-linear statistical data modeling that changes its structure based on external or internal information that flows through a network during a learning phase. ANN neural networks are non-linear statistical data modeling tools used to model complex relationships between inputs and outputs or to find patterns in data.
“Aspect ratio” of the tire means the ratio of its section height (SH) to its section width (SW) multiplied by 100 percent for expression as a percentage.
“Asymmetric tread” means a tread that has a tread pattern not symmetrical about the center plane or equatorial plane EP of the tire.
“Axial” and “axially” means lines or directions that are parallel to the axis of rotation of the tire.
“Chafer” is a narrow strip of material placed around the outside of a tire bead to protect the cord plies from wearing and cutting against the rim and distribute the flexing above the rim.
“Circumferential” means lines or directions extending along the perimeter of the surface of the annular tread perpendicular to the axial direction.
“Dugoff Model” is an empirical tire model providing analytical relations for the longitudinal and lateral forces as functions of the slip angle and slip ratio. It accounts for the coupling between the side and longitudinal forces.
“Equatorial Centerplane (CP)” means the plane perpendicular to the tire's axis of rotation and passing through the center of the tread.
“Footprint” means the contact patch or area of contact created by the tire tread with a flat surface as the tire rotates or rolls.
“Groove” means an elongated void area in a tire wall that may extend circumferentially or laterally about the tire wall. The “groove width” is equal to its average width over its length. A grooves is sized to accommodate an air tube as described.
“Inboard side” means the side of the tire nearest the vehicle when the tire is mounted on a wheel and the wheel is mounted on the vehicle.
“Lateral” means an axial direction.
“Lateral edges” means a line tangent to the axially outermost tread contact patch or footprint as measured under normal load and tire inflation, the lines being parallel to the equatorial centerplane.
“Net contact area” means the total area of ground contacting tread elements between the lateral edges around the entire circumference of the tread divided by the gross area of the entire tread between the lateral edges.
“Non-directional tread” means a tread that has no preferred direction of forward travel and is not required to be positioned on a vehicle in a specific wheel position or positions to ensure that the tread pattern is aligned with the preferred direction of travel. Conversely, a directional tread pattern has a preferred direction of travel requiring specific wheel positioning.
“Outboard side” means the side of the tire farthest away from the vehicle when the tire is mounted on a wheel and the wheel is mounted on the vehicle.
“Peristaltic” means operating by means of wave-like contractions that propel contained matter, such as air, along tubular pathways.
“Piezoelectric Film Sensor” a device in the form of a film body that uses the piezoelectric effect actuated by a bending of the film body to measure pressure, acceleration, strain or force by converting them to an electrical charge.
“Radial” and “radially” means directions radially toward or away from the axis of rotation of the tire.
“Rib” means a circumferentially extending strip of rubber on the tread which is defined by at least one circumferential groove and either a second such groove or a lateral edge, the strip being laterally undivided by full-depth grooves.
“Sipe” means small slots molded into the tread elements of the tire that subdivide the tread surface and improve traction, sipes are generally narrow in width and close in the tires footprint as opposed to grooves that remain open in the tire's footprint.
“Tread element” or “traction element” means a rib or a block element defined by having a shape adjacent grooves.
“Tread Arc Width” means the arc length of the tread as measured between the lateral edges of the tread.
The invention will be described by way of example and with reference to the accompanying drawings in which:
Referring to
A vehicle 22 with on-board sensors generate tire slip angle and roll rate by means of conventionally deployed vehicle sensor systems. A chassis roll angle estimator 26 generates a roll angle estimate from the measured lateral acceleration and the measured roll rate at the center of gravity. A chassis pitch angle estimator 28 generates a pitch angle estimate from the vehicle-sensor measured longitudinal acceleration. The roll angle estimate, pitch angle estimate, measured lateral and longitudinal acceleration and measured roll rate at the center of gravity input into the dynamic tire load estimator 24. A static tire load estimator 30 generates a static load estimate input into the dynamic tire load estimator 24. Dynamic tire load estimates for the four vehicle tires are input into a tire force estimator with friction adaptation 32.
The acceleration measurement, yaw rate measured at the center of gravity and tire steer angle from the on-board sensors are processed through an axle force estimator (extended Kalman filter) 34 to generate front and rear axle force estimations Fyr and Fyf. A tire slip angle 36 is determined as variables of αf and αr for the front and rear tires. An axle cornering stiffness estimator 38 operating on recursive least squares receives as inputs αf, αr, Fyf, Fyr, vx, r, δ and generates cornering stiffness estimates Cyf and Cyr. A tire cornering stiffness adaptation model 40 receives as inputs the cornering stiffness estimates and the dynamic tire load estimates as well as pressure and temperature measurements 42 from the vehicle tire sensors and an indirect wear estimation 44. The model 40 outputs cornering stiffness determinations to the tire force estimator with friction adaptation 32 for each of the four vehicle tires, adapted for tire pressure, temperature, load and tire wear state.
A SMC (sliding mode control) observer based upon a friction update law, as explained below determines a friction coefficient estimation 48 that inputs into the tire force estimator 32 to be compared with an estimator friction estimation μ. The estimator 32 determines tire force estimations FYfl, FYfr, FYrl and FYrr determined from inputs into a modified dugoff model. The tire force estimations are a function of the cornering stiffness estimation Cy, the dynamic tire load estimator estimations Fz, the slip angle estimations a and friction μ. The tire force estimations 32 loop back to the observer 46. Error between the observer 46 and the friction used in the tire force estimator with friction adaptation 32 is reduced to yield a final estimation of road friction.
The state variables from the vehicle on-board sensors 22 are CAN bus available. The chassis roll angle estimator 26 and the chassis pitch angle estimator 28 are available through conventional means as taught in the dissertation: “Development of an Intelligent Tire Based Tire-Vehicle State Estimator for Application to Global Chassis Control”; Kanwar Bharat Singh, Master's Thesis, Department of Mechanical Engineering, Virginia Tech University, 2012, incorporated herein in its entirety by reference. The static tire load estimator 30 is taught in U.S. Pat. No. 8,844,346, issued Sep. 30, 2014, entitled TIRE LOAD ESTIMATION SYSTEM USING ROAD PROFILE ADAPTIVE FILTERING, incorporated herein in its entirety by reference.
The dynamic tire load estimator 24 is configured as presented in co-pending U.S. Patent Publication No. 2014/0278040, published Sep. 18, 2014, entitled VEHICLE DYNAMIC LOAD ESTIMATION SYSTEM AND METHOD, hereby incorporated herein in its entirety. The axle force estimator 34 is configured as taught in: Baffet, Guillaume, Ali Charara, and Daniel Lechner, “Estimation of Vehicle Sideslip, Tire Force and Wheel Cornering Stiffness”, Control Engineering Practice 17.11 (2009), Pages 1255 through 1264 and Doumiati, Moustapha, et al. “Observers for Vehicle Tyre/Road Forces Estimation: Experimental Validation”, Vehicle System Dynamics 48.11 (2010), Pages 1345 through 1378, incorporated herein in relevant part by reference. The axle cornering stiffness estimator (recursive least squares) 38 is as shown in Siena, C. et al. “Cornering Stiffness Estimation Based on Vehicle Lateral Dynamics”, Vehicle System Dynamics 44 supl (2006), Pages 24 through 38, incorporated herein in relevant part by reference. The tire slip angle (from intelligent tire) 36 used is seen in U.S. Pat. No. 8,886,395, issued Nov. 11, 2014, entitled DYNAMIC TIRE SLIIP ANGLE ESTIMATION SYSTEM AND METHOD incorporated herein in its entirety by reference. The tire cornering stiffness adaptation model 40, adapted for tire pressure and temperature 42, load and tire wear state 44 are taught in co-pending U.S. patent application Ser. No. 14/549,845, filed Nov. 21, 2014, entitled TIRE CORNERING STIFFNESS ESTIMATION SYSTEM AND METHOD incorporated herein in its entirety by reference.
The friction update law 48 used to realize tire-road friction estimation in the SMC observer 46 is seen in
The tread region 14 has a depth that decreases with tire wear through vehicle use. The subject friction estimation system requires an estimation be made of tread wear 44 (
Tire cornering stiffness characteristics fluctuate under varying operating conditions of the tire (temperature change, pressure change, tire wear state, load). Hence, adaptation is important for a good estimation of cornering stiffness. Sensitivity of cornering stiffness with load is on the order of a ten percent increase with a 200 pound increase in load. A 10 percent increase in cornering stiffness with a 4 psi decrease in tire pressure is likewise found. A 15 percent increase in stiffness with a 3 mm decrease in tread depth occurs, and a 30 percent drop in stiffness with a 25° C. increase in tire temperature is found. A modified dugoff tire model uses a slip dependent friction function and the general formulation is found below.
General Formulation
where λ is related to the tire/road friction coefficient, λ and f(λ) are defined as follows respectively:
The subject modified dugoff model is found in Ding, Neggen and Taheri, Saied, “A Modified Dugoff Tire Model for Combined-slip Forces”, Tire Science and Technology, TSTCA, Vol. 39, No. 3, July-September 2010, Pages 228 through 244, incorporated herein in relevant part by reference. The modified dugoff tire model performance is compared against a “magic formula” result. The “Magic Formula Tire Model”, defined and explained below, is a tire model conventionally used within the tire industry to calculate steady-state force and moment characteristics. Its use herein is solely for the purpose of validating the subject invention's performance which utilizes the modified dugoff model identified and incorporated herein by reference above. The “Magic Formula Model” is accordingly a validation tool and does not comprise a part of the claimed invention.
The Magic Formula Tire Model
The variation of the Magic Formula used herein is referred to as the “Pacejka Magic Formula”. This is a widely used semi-empirical tire model that is used to calculate steady-state force and moment characteristics of the tire. This model is called the “magic formula” because there is no particular physical basis for the structure of the model equations, but they fit a wide variety of tire constructions and operating conditions. The first versions of the model at TU-Delft in collaboration with Volvo [1,2].
The main assumption behind this tire model is that the steady-state force characteristics of the tire under pure and combined slip conditions can be represented by a set of empirical mathematical equations. Initial versions of the magic formula tire model only concentrated on steady-state model fitting to experimental data. Over the years, multiple improvements were made to the tire model, including adding additional degrees to polynomial fits, relaxation behavior in the contact patch [3], scaling factors for different surfaces, inflation pressure dependencies [4] and variation of rolling resistance with load.
Model Formulation
The magic formula tire model is based on representing the steady-state force curve in lateral, longitudinal directions and aligning moments through empirical mathematical equations. The basic equation for the magic formula is given by:
y=D sin [C arc tan {Bx−E(Bx−arc tan Bx)}]
Where
Y(X)=y(x)+Sv
x=X+SH
In this equation, the output quantity is represented by Y which can be either the lateral force, longitudinal force or aligning moment at a specific vertical load and camber for a given X which can be slip angle or slip ratio. The equation parameters are described as:
Model output, which can be either the lateral force, longitudinal force or aligning moment at a specific vertical load and camber
X—Which can be slip angle α or slip ratio K
SH—Horizontal shift
SV—Vertical shift
B, C, D, E, —Fit parameters
The “Magic Model” is described in further detail by the following treatises:
The dugoff model is compared against the “Magic Formula” estimation in
Simulation results are shown graphically in
In
From the foregoing, it will be appreciated that a robust and accurate estimation of the maximum tire-road friction coefficient is achieved by the system and method disclosed. The system and method are based on an observer 46 (
In the linear range of operation, the subject algorithm is not able to estimate maximum friction coefficient correctly. This can be attributed to the fact that the tire force curve is independent of friction coefficient in this region.
The subject road friction estimation system and method thus is described above to include tire-affixed sensors 18 mounted to the tires of a vehicle and on-board vehicle sensors providing vehicle-based parameter information via CAN bus. A model-based tire force estimator 32 is operable generating from both tire and vehicle sensor-input a model-derived tire force estimation. A vehicle observer 46 generates an observer-derived tire force estimation from the axle force estimator 34 based upon vehicle sensor generated information. A friction estimator in the form of the error-reducing loop between the tire force estimator 32 and the observer 46 generates a road friction estimation by comparing the model-derived tire force estimation 32 to the observer-derived tire force estimation 46.
It will be appreciated that the vehicle observer 46 receives inputs based upon sensor-measured tire-specific information including a load estimation for the vehicle tire obtained from the dynamic tire load estimator 24, a slip angle estimation for the vehicle tire obtained from tire slip angle estimator 36 and a tire cornering stiffness input from the tire cornering stiffness adaptation to pressure, temperature, load and tire wear state 40. Accordingly, the model-based tire force estimator 32 employs and utilizes a slip-dependent friction function. The cornering stiffness adaptation inputs include dynamic tire loading estimation from the dynamic tire load estimator 24 derived from the vehicle-based sensors 22 as well as tire-based sensors used in the static tire load estimator 30.
Variations in the present invention are possible in light of the description of it provided herein. While certain representative embodiments and details have been shown for the purpose of illustrating the subject invention, it will be apparent to those skilled in this art that various changes and modifications can be made therein without departing from the scope of the subject invention. It is, therefore, to be understood that changes can be made in the particular embodiments described which will be within the full intended scope of the invention as defined by the following appended claims.
| Number | Name | Date | Kind |
|---|---|---|---|
| 5260683 | Tanaka et al. | Nov 1993 | A |
| 5553491 | Naito et al. | Sep 1996 | A |
| 5826207 | Ohashi et al. | Oct 1998 | A |
| 6539295 | Katzen et al. | Mar 2003 | B1 |
| 6637276 | Adderton et al. | Oct 2003 | B2 |
| 6697726 | Takagi et al. | Feb 2004 | B2 |
| 6962075 | Bertrand | Nov 2005 | B2 |
| 7130735 | Brown et al. | Oct 2006 | B2 |
| 7240542 | Gustafsson et al. | Jul 2007 | B2 |
| 7404317 | Mancosu et al. | Jul 2008 | B2 |
| 7404319 | Poulbot et al. | Jul 2008 | B2 |
| 7415874 | Mancosu et al. | Aug 2008 | B2 |
| 7546764 | Morinaga | Jun 2009 | B2 |
| 7549327 | Breed | Jun 2009 | B2 |
| 7552628 | Mancosu | Jun 2009 | B2 |
| 7681960 | Wanke et al. | Mar 2010 | B2 |
| 7856871 | Mancosu et al. | Dec 2010 | B2 |
| 8155798 | Seiniger et al. | Apr 2012 | B2 |
| 8844346 | Singh et al. | Sep 2014 | B1 |
| 8886395 | Singh et al. | Nov 2014 | B2 |
| 20020059023 | Takagi et al. | May 2002 | A1 |
| 20020162389 | Yokota et al. | Nov 2002 | A1 |
| 20030058118 | Wilson | Mar 2003 | A1 |
| 20030121319 | Kojima et al. | Jul 2003 | A1 |
| 20030236603 | Lu | Dec 2003 | A1 |
| 20040199314 | Meyers et al. | Oct 2004 | A1 |
| 20040254707 | Lu et al. | Dec 2004 | A1 |
| 20050033486 | Schmitt et al. | Feb 2005 | A1 |
| 20050072223 | Fennel et al. | Apr 2005 | A1 |
| 20050085987 | Yokota | Apr 2005 | A1 |
| 20050150283 | Shick | Jul 2005 | A1 |
| 20050177296 | Brown et al. | Aug 2005 | A1 |
| 20070010928 | Brusarosco et al. | Jan 2007 | A1 |
| 20070017727 | Messih et al. | Jan 2007 | A1 |
| 20070255510 | Mancosu | Nov 2007 | A1 |
| 20080103659 | Mancosu | May 2008 | A1 |
| 20090055040 | Nagaya | Feb 2009 | A1 |
| 20100063671 | Fink et al. | Mar 2010 | A1 |
| 20100211256 | Takenaka | Aug 2010 | A1 |
| 20110060500 | Irth et al. | Mar 2011 | A1 |
| 20110199201 | Brusarosco et al. | Aug 2011 | A1 |
| 20130151075 | Moshchuk et al. | Jun 2013 | A1 |
| 20130211621 | Breuer et al. | Aug 2013 | A1 |
| Number | Date | Country |
|---|---|---|
| 19716586 | Aug 1998 | DE |
| 102008046269 | Dec 2009 | DE |
| 2301769 | Mar 2011 | EP |
| 2172760 | Aug 2012 | EP |
| WO2011054363 | May 2011 | WO |
