Procedures for determining road grip classes

The method improves road grip classification accuracy by analyzing circumferential force and tire slip pairs, calculating tire slip stiffness, and using statistical evaluation to classify road surfaces, addressing limitations in existing technologies and enhancing adaptability to tire changes and environmental conditions.

DE102016203545B4Active Publication Date: 2024-11-07CONTINENTAL AUTOMOTIVE TECHNOLOGIES GMBH
View PDF 9 Cites 0 Cited by

Patent Information

Application Number
DE102016203545
Authority / Receiving Office
DE · DE
Patent Type
Patents
Current Assignee / Owner
Filing Date
2016-03-03
Publication Date
2024-11-07
Estimated Expiration
2036-03-03

AI Technical Summary

Technical Problem

Existing methods for determining road grip classes are limited in accuracy and applicability, particularly during vehicle acceleration and braking, and do not effectively utilize self-learning capabilities to adapt to tire changes and environmental conditions.

Method used

A method for determining road grip classes by analyzing pairs of circumferential force and tire slip, calculating tire slip stiffness, and using statistical evaluation to classify road surfaces based on high and low friction conditions, incorporating vehicle behavior and external data sources for improved accuracy.

Benefits of technology

Enhances the accuracy of road grip classification by enabling classification during acceleration and braking, adapting to tire changes and environmental conditions through self-learning, and utilizing multiple data sources for precise identification of high and low friction road surfaces.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure 00000000_0000_ABST
    Figure 00000000_0000_ABST
Patent Text Reader

Abstract

Method for determining road grip classes from measured values ​​of a circumferential force on at least one wheel of a vehicle traveling on a roadway and the tire slip caused by this circumferential force on the same wheel, comprising the following steps: a) Determining pairs of values ​​from the value of a circumferential force acting on the vehicle wheel and a tire slip generated by this circumferential force on the vehicle wheel, by determining the tire slip from a measured wheel rotation speed of the vehicle wheel and the measured vehicle speed, and by determining the circumferential force from an engine drive torque acting on the vehicle wheel or on a vehicle axle of the vehicle, from an engine drag torque acting on the vehicle wheel or on a vehicle axle of the vehicle, from a braking torque on the vehicle wheel or from a braking torque acting on a vehicle axle of the vehicle, b) Determining a group of values ​​indicating a high-friction road surface from the determined pairs of circumferential force and tire slip, whereby a high-friction road surface is indicated at the time of determination of these pairs of values ​​of the vehicle dynamics parameters greater than a predetermined value, while at the same time there is no intervention by a vehicle stability system and / or optical sensors of the vehicle and / or external data sources of the vehicle. c) Calculating tire slip stiffness values ​​from the pairs of values ​​by forming the ratio of circumferential force to wheel slip, d) Defining a road grip class indicating a high-friction road surface by determining an average tire slip stiffness from the tire slip stiffness values ​​of the pairs of values ​​in the value group and the dispersion of these pairs of values ​​around the average tire slip stiffness using a statistical evaluation method, and e) Classifying a current tire slip stiffness value as indicating a high-friction road surface if this value is in the road grip class indicating the high-friction road surface, otherwise as indicating a low-friction road surface.
Need to check novelty before this filing date? Find Prior Art

Description

[0001] The invention relates to a method for determining road grip classes from measured values ​​of a circumferential force on at least one vehicle wheel of a vehicle traveling on a road and the wheel slip caused by this circumferential force.

[0002] For improved functionality of a modern driver assistance system, such as an emergency braking system, it is necessary to estimate a coefficient of friction between tires and road surface.

[0003] In manual driving, the driver is responsible for estimating the coefficient of friction. A key part of the driving task is assessing the road conditions and adjusting the driving style accordingly. With the introduction of more highly automated systems and driving functions, the driving task gradually shifts from the driver to the vehicle.

[0004] A key prerequisite for this is to determine and describe the driving environment and conditions as precisely as possible and make this information available to the system. The maximum coefficient of friction between the tires and the road surface defines the physical limit for the vehicle's achievable driving dynamics. Therefore, information about the coefficient of friction forms an essential part of the environmental representation. With regard to automated driving, reliable information about the road surface condition, or, derived from this, the available maximum coefficient of friction, is indispensable for ensuring the safety of the driving function. Furthermore, in the area of ​​so-called Driver Assistance Systems and Advanced Driver Assistance Systems, knowledge of the available maximum coefficient of friction results in a significant safety benefit.

[0005] From DE 40 10 507 C1, a method for monitoring the frictional contact between the road surface and the tires of driven vehicle wheels is known. This method assumes that a specific coefficient of friction with a corresponding maximum braking torque can be assigned to different initial gradients. Thus, the wheel slip λ and simultaneously an effective circumferential force of the driven wheels are continuously determined during steady-state driving conditions. These pairs of values, consisting of the determined circumferential force and the wheel slip, are stored to form a wheel circumferential force-wheel slip characteristic curve field, so that the linear part of the slip characteristic curve is known. Knowledge of the slope of the currently effective slip characteristic curve also provides knowledge of the maximum effective coefficient of friction.the order of magnitude of the maximum possible wheel circumferential force in this driving condition, since according to this known method there is an essentially fixed relationship between the linear increase and the curve maximum of the slip characteristics.

[0006] WO 2014 / 199328 A1 also describes a method for estimating the coefficient of friction and is based on the finding that slip characteristics with a low maximum coefficient of friction (also called µ-peak) leave the linear range at a lower slip value than is the case with slip characteristics with a higher maximum coefficient of friction.

[0007] Furthermore, DE 10 2012 217 772 A1 also describes a method for determining the maximum coefficient of friction by determining the initial slope of a slip characteristic curve and calculating the maximum coefficient of friction from this. For this purpose, the current coefficient of friction at the vehicle wheel and the corresponding slip are first determined by measurements, and the value of the initial slope is calculated from this. By comparison with reference lines through the origin, the range for a maximum coefficient of friction is determined.

[0008] In the initial slope of a slip characteristic curve, both tire forces and tire slip are very low. To obtain sufficient accuracy under steady-state or quasi-steady-state conditions, a robust and, above all, high-resolution signal is required for both the braking torque / force and the slip. If the variation is very large and / or the resolution is low for only one of the two signals, it is no longer possible to achieve an accurate and, more importantly, reproducible result.

[0009] Furthermore, DE 10 2004 044 788 B4 discloses a method for determining the coefficient of friction between a vehicle tire and a road surface, in which the wheel speed and the absolute speed of the vehicle are measured, the slip is determined from this wheel speed and the absolute speed, a drive or braking torque acting on the vehicle is then measured, and finally a coefficient of friction between the tire and the road surface is determined using an algorithm in which the slip and the drive or braking torque are taken into account as input variables.

[0010] DE 102 08 815 A1 discloses a method for determining a maximum coefficient of friction between a vehicle tire and road surface from force information generated during contact between the tire and the road surface. DE 10 2012 204 671 A1 discloses a method for determining a friction ratio between at least one wheel of a motor vehicle and a road surface. DE 10 2014 103 843 A1 teaches a method for determining a coefficient of friction between at least one tire of a vehicle and a road surface, wherein the coefficient of friction is calculated as a function of at least one driving condition of the vehicle and environmental conditions. US 2007 / 0 016 354 A1 discloses a method for determining a coefficient of friction value representing the coefficient of friction between the underlying surface and a vehicle tire.

[0011] The object of the invention is to provide a method for determining road grip classes, with which the road surface on which the vehicle is moved can be classified with regard to the coefficient of friction and the information required for this purpose is generated not only during an acceleration process of the vehicle, but in particular also during engine drag torque operation or braking process of individual wheels of the vehicle.

[0012] This problem is solved by a method for determining road grip classes with the features of claim 1.

[0013] In the inventive method for determining road grip classes from measured values ​​of a circumferential force on at least one vehicle wheel of a vehicle traveling on a road and the tire slip caused by this circumferential force, the following steps are carried out: a) Determining pairs of values ​​from the value of a circumferential force acting on the vehicle wheel and a tire slip generated by this circumferential force on the vehicle wheel, by determining the tire slip from a measured wheel rotation speed of the vehicle wheel and the measured vehicle speed, and by determining the circumferential force from an engine drive torque acting on the vehicle wheel or on a vehicle axle of the vehicle, from an engine drag torque acting on the vehicle wheel or on a vehicle axle of the vehicle, from a braking torque on the vehicle wheel or from a braking torque acting on a vehicle axle of the vehicle, b) Determining a group of values ​​indicating a high-friction road surface from the determined pairs of circumferential force and tire slip, whereby a high-friction road surface is indicated at the time of determination of these pairs of vehicle dynamics parameters greater than a predetermined value, while simultaneously lacking intervention from a vehicle stability system and / or optical sensors of the vehicle and / or external data sources of the vehicle. c) Calculating tire slip stiffness values ​​from the pairs of values ​​by forming the ratio of circumferential force to wheel slip, d) Defining a road grip class indicating a high-friction road surface by determining an average tire slip stiffness from the tire slip stiffness values ​​of the value pairs of the value group and the dispersion of these value pairs around the average tire slip stiffness using a statistical evaluation procedure, and e) Classifying a current tire slip stiffness value as indicating a high-friction road surface if this value lies in the road grip class indicating the high-friction road surface, otherwise as indicating a low-friction road surface.

[0014] In this method according to the invention, a reference standard indicating a high coefficient of friction is first generated by producing a data series of tire slip stiffnesses, calculated as quotients of pairs of values ​​of the circumferential force of the vehicle wheel and the associated wheel slip, during the detection of a high-friction road surface. This data series is subjected to a statistical evaluation procedure to determine an average tire slip stiffness, and the corresponding value dispersion is calculated with respect to the average tire slip stiffness, i.e., the spread is determined. This defines a road grip class indicating a high-friction road surface with a coefficient of friction µ of preferably µ ≥ 0.5.Thus, all currently determined tire slip stiffness values ​​with values ​​within the range of the average tire slip stiffness indicate such a high friction value; in the other case, if these values ​​lie outside this road grip class, they indicate a low friction value.

[0015] The circumferential force of a vehicle wheel can be determined using this method according to the invention both during an acceleration process and during a braking process of individual vehicle wheels or axles, particularly during engine braking. This increases the scope of available information, thereby also increasing the accuracy of the statistical evaluation method and thus also the accuracy of determining the road grip class.

[0016] By comparing the current tire slip stiffnesses against a reference standard indicating a high friction value, tire-specific relationships, parameters and especially changes can be identified and eliminated, since the reference standard is continuously adjusted when a reference signal is present, thus realizing a self-learning behavior of the system over the tire's life, especially in connection with tire changes and tire wear, even under changed environmental conditions.

[0017] To determine a value group indicating a high-friction road surface from the measured wheel slip-wheel torque value pairs, the behavior of the vehicle control system is observed at the times when these value pairs were determined. The behavior of the vehicle control system allows us to deduce whether the vehicle is moving on a high-friction or low-friction road surface.

[0018] Displaying a high-friction road surface can be achieved using various data sources. For example, vehicle dynamics parameters can be used to generate a reference signal when a predefined acceleration or deceleration value is exceeded. This information can also be provided by evaluating road surface conditions detected by a camera (as an optical sensor) or another type of optical sensor, such as a laser sensor scanning the road surface, for instance, when a dry road surface is detected. Furthermore, external data sources providing local weather or road surface condition data (e.g., a backend connection, Car-to-X communication, or dynamic safety maps) can be networked to generate such a reference signal.

[0019] An advantageous embodiment of the invention provides for the following process steps, in which: d1) from each data series of temporally successive pairs of values ​​of circumferential force and wheel slip, whose tire slip stiffness values ​​indicate a low-friction road surface, an initial slope is determined on the basis of a regression line as a linear part of the slip-circumferential force characteristic curve, and d2) the slope of each regression line is defined as the mean tire slip stiffness of a road grip class indicating a low friction road surface with a dispersion of the pairs of values ​​around the determined regression line determined by means of a statistical evaluation procedure.

[0020] In these process steps d1 and d2, the tire slip stiffness values ​​lying outside the road grip class indicating a high-friction surface are classified by generating further road grip classes. For this purpose, a data series of temporally successive value pairs is used, whereby the slope of a regression line representing the mean tire slip stiffness in the circumferential force-slip diagram and the associated standard deviation are determined for the tire slip stiffness values ​​using a statistical evaluation method. This allows a road grip class indicating a low-friction surface to be defined for each regression line generated in this way, whose mean tire slip stiffness corresponds to the slope of the regression line. In this way, a road grip class indicating a snow-covered road surface and a road grip class indicating an icy road surface can preferably be defined.This allows the tire slip stiffness values ​​to be assigned to one of these road grip classes, which indicate, for example, a snowy road or an icy road, if it is not in the road grip class indicating a high friction road.

[0021] A particularly preferred embodiment of the invention consists in taking into account the curve shape, approximated to the corresponding data series by a mathematical adjustment calculation, to classify a tire slip stiffness value indicating a low-friction road surface. For example, a curvature indicates a snow-covered road surface, while discontinuities, e.g., sudden breaks, indicate an icy road surface.

[0022] Furthermore, it is particularly advantageous if, in accordance with the training guidelines, the procedural steps d, d1 and d2 are only carried out when there is a positive absolute slip gradient and the vehicle is in driving conditions without intervention from the vehicle control systems. This ensures that hysteresis phenomena caused by the increase and decrease of tire slip are excluded.

[0023] According to a further preferred embodiment of the invention, during the execution of process steps d, d1 and d2, the classification of the tire slip stiffness values ​​is validated based on the intervention behavior of a vehicle control system, by validating a road grip class indicating a high-friction road surface when there are no interventions by the vehicle control system above predetermined longitudinal and / or lateral accelerations, and by validating a road grip class indicating a low-friction road surface when there are interventions by the control system below predetermined longitudinal and / or lateral accelerations.

[0024] According to a further advantageous embodiment of the invention, process steps d, d1, and d2 are performed during both acceleration and deceleration of the vehicle, with the average tire slip stiffness in the same road grip class being equated for both acceleration and deceleration. This is possible because the initial slope in the tire slip-circular force characteristic curve is the same for both the drive and braking cases. This allows the accuracy of the calculation of the average tire slip stiffness to be improved by using multiple data points.

[0025] Furthermore, it is advantageous to use the magnitudes of the circumferential forces and the slip to calculate tire slip stiffness, as recommended in advanced training. This allows the evaluation to be performed in the same quadrant of the wheel slip-circular force diagram, thus increasing the point density through the superposition of drive, braking, or drag torque cases. This makes the results more accurate and improves availability.

[0026] According to a further advantageous embodiment of the invention, the process steps d, d1 and d2 are performed within a predetermined slip window between a slip value λ. min and a slip value λ max carried out, with a slip greater than λ maxControl system interventions by a vehicle control system take place, and these interventions are simultaneously used to validate the road grip classification. Within the defined slip range, the slip is continuously determined, allowing for a more precise determination of the slope of the regression line that defines a road grip class. This is because, with excessively low absolute values ​​for wheel torque and wheel slip, the ratio calculation for determining slip stiffness becomes inaccurate, considering finite measurement accuracy. Conversely, excessively high slip values ​​would negatively affect the classification due to the curve shape and potential control system interventions.

[0027] Finally, it is advantageous, according to further training, to determine the tire slip for both wheels of the vehicle on an axle subjected to a circumferential force in order to detect µ-split situations, and to calculate the tire slip stiffness values ​​by the quotient of the circumferential force and the tire slip of the wheel with the most slip, or by the average value of all wheels on a vehicle axle. In the case of axle-wise application of drive or braking torque, the torque is, to a first approximation, distributed evenly between the drive wheels. The slip value required to determine the slip stiffness quotient is either averaged over all wheels subjected to a force, or the higher slip value of the two wheels is used for the calculation.

[0028] According to a final advantageous embodiment of the invention, the averaging and / or filtering of the pairs of values ​​from circumferential force and tire slip is carried out over a full wheel revolution or over integer multiples of individual wheel revolutions.

[0029] The method according to the invention is suitable for use in vehicles that have an engine control unit, a brake control unit, a speed detection device for the vehicle wheels, a vehicle dynamics control system and a device for determining a wheel torque or a braking torque as a circumferential force of the vehicle wheel.

[0030] The method according to the invention is described below with reference to exemplary embodiments and the accompanying figures. These show: Fig. 1 a wheel torque-wheel slip diagram with plotted measured values, Fig. 2 a tire slip stiffness-time diagram with the measured values ​​from Fig. 1, Fig. 3 a wheel torque-wheel slip diagram with recorded measured values ​​classified according to road surface condition, Fig. 4 a tire slip stiffness-time diagram with the measured values ​​from Fig. 3, Fig. 5 a wheel drag torque-wheel slip diagram with measured values ​​recorded and entered during drag torque operation, Fig. 6 a wheel torque-wheel slip diagram for determining a slip offset, and Fig. 7 a wheel torque-wheel slip diagram with equal wheel slip stiffnesses for the drive and braking cases.

[0031] The detection of road surface grip (coefficient of friction) according to the inventive method is achieved by determining road surface grip classes from the tire slip stiffness of a vehicle wheel and the wheel circumferential force-wheel slip characteristic of the vehicle wheel. The inventive method enables differentiation between a high-friction road surface with a coefficient of friction of, for example, ≥ 0.5 and a low-friction road surface, whereby a low-friction road surface exhibits low grip, such as on snow or ice. The coefficient of friction of snow is approximately < 0.5, that of ice is approximately ≤ 0.2, and the corresponding value for dry asphalt is approximately 1.0.

[0032] The wheel circumferential force (hereinafter referred to as circumferential force) includes both the drive torque during an acceleration phase of the vehicle and the braking torque during the use of the friction brake or a recuperation process for energy recovery, for example via an electric motor, or during a drag torque operation of the vehicle.

[0033] To determine the wheel circumferential force-wheel slip characteristic, pairs of values ​​are calculated from the value of a circumferential force acting on the vehicle wheel and the tire slip generated by this circumferential force. This is done by determining the tire slip from a measured wheel rotational speed and the measured vehicle speed, and by determining the circumferential force from an engine drive torque, an engine drag torque, or a braking torque. To determine the circumferential force from the engine drag torque, the engine torque at the vehicle axle of the wheel is calculated first, followed by the torque acting on that wheel. The same procedure applies to determining the circumferential force from the braking torque.

[0034] The drag torque can be detected, for example, directly via the engine torque signal of the engine, or read from a speed / gear-speed drag torque characteristic curve.

[0035] By averaging and / or filtering each wheel slip / circular force value pair, a highly accurate slip determination in the per mille range is achieved over a complete wheel revolution or over integer multiples of individual wheel revolutions, thus enabling precise results even with only low engine drag torques. Furthermore, fluctuations in the tire slip values ​​can be compensated for by classifying the tire slip. The wheel slip / circular force value pairs are distributed in a wheel slip / circular force diagram according to the Fig. 1 and Fig. 5. In doing so, Fig. 1. The situation is that the circumferential force is a wheel torque T, namely a driving or braking torque, while Fig. 5 represents the drag torque operating mode with a wheel drag torque T. Since the Fig. 1. Since both drive and braking torques are recorded as wheel torque, the magnitudes |T| of wheel torque T and |λ| of wheel slip λ are entered. The clusters H of the wheel slip-wheel drag torque value pairs in Fig. Five values ​​result from different gear ratios and speeds. However, so-called slip reverses in the direction of the coordinate origin during decreasing vehicle acceleration or deceleration are not taken into account.

[0036] Furthermore, the tire slip stiffness B is calculated from the wheel slip-wheel torque value pairs as the quotient of wheel torque |T| and wheel slip |λ|. The representation of these tire slip stiffness values ​​B along the time axis is shown in the diagram below. Fig. 2 shown.

[0037] The vehicle is equipped with an electronic stability control system, such as ABS, TCS or AYC, generally called ESC (hereinafter referred to as vehicle control system).

[0038] To determine a value group indicating a high-friction road surface from the measured wheel slip-wheel torque value pairs, the behavior of the vehicle control system is observed at the times when these value pairs were determined. The behavior of the vehicle control system allows us to deduce whether the vehicle is moving on a high-friction or low-friction road surface.

[0039] If the behavior of vehicle dynamic parameters, such as longitudinal and / or lateral acceleration, exceeds a predefined value over an observation period, and no control system interventions occur, this behavior of the vehicle control system indicates a high-friction road surface. The wheel slip-wheel torque value pairs determined over this observation period are grouped together to form the value group indicating a high-friction road surface.

[0040] In contrast, control system interventions would indicate a low coefficient of friction during low acceleration, deceleration, or lateral acceleration. Thus, a road grip class indicating a low coefficient of friction can be validated based on such control system interventions.

[0041] Displaying a high-friction road surface can also be achieved using optical sensors, such as a mono or stereo camera on the vehicle. In this case, the road surface is classified by analyzing the images captured by the camera. Another way to indicate a high-friction road surface is by accessing external data sources, such as weather data, from which, for example, the absence of snow or ice can be inferred. Vehicle control system interventions such as ABS, TCS, or AYC (ESC) can also be used to determine the actual coefficient of friction between the tires and the road surface, thus displaying a high-friction road surface.

[0042] One such group of values ​​indicating a high-friction road surface consists of the wheel slip-wheel torque value pairs determined up to a time t1 of an observation period, whose tire slip stiffness values ​​B in Fig. 2 are shown.

[0043] Using a statistical evaluation method (statistical distribution function ΔB1 = f(σ) (B1) A mean value ΔB1 with its corresponding standard deviation is determined from the pairs of values ​​in the value group, thus defining a road surface friction class A. The class boundaries determined by the standard deviation are in Fig. 2 with ΔB 1o as upper limit and with ΔB 1u referred to as the lower limit. The mean value ΔB1 represents in Fig. 1 represents the slope of a regression line g1, as does the lower limit ΔB 1u and the upper limit ΔB 1o as the slope of a straight line g 1u or as the slope of a straight line g 1o in Fig. 1 shown.

[0044] Out of Fig. 1. It is evident that these pairs of values ​​indicating the road grip class A have only a small dispersion around this regression line g1.

[0045] This allows all tire slip stiffness values ​​(B) and wheel slip-wheel torque value pairs to be classified. If these values ​​or value pairs fall within road grip class A, the vehicle is traveling on a high-friction road surface; otherwise, if these values ​​or value pairs do not fall within road grip class A, they indicate a low-friction road surface. Therefore, road grip class A represents a reference standard indicating a high coefficient of friction.

[0046] By comparing the current tire slip stiffnesses against a reference standard indicating a high friction value, tire-specific relationships and parameters can be identified and eliminated, since this reference standard can be continuously adjusted and thus a self-learning behavior of the system is achievable, especially in connection with tire changes and tire wear, even under changed environmental conditions.

[0047] When determining the road grip class A, it is advantageous to consider the tire slip stiffness B only within predetermined slip limits λ. min and λ max according to the representation Fig. 1 to determine. By defining these slip limits, the working range of the method according to the invention is also determined. Thus, λ should max These values ​​are in the range of wheel slip values ​​at which electronic stability control systems, such as ABS, TCS, and ESC, become active on a road surface with low grip, such as one covered in snow or ice. Such control system interventions are used to validate the road grip classification. Therefore, considering such interventions by stability control systems, such as wheel slip control systems, leads to increased accuracy in defining a road grip class. The quantity λ minThe value should not be chosen too close to zero in order to avoid jeopardizing the mathematical division operation.

[0048] The above in connection with the Fig. 1 and Fig. The procedure described in section 2 is also carried out in the engine drag torque operating case. The wheel slip-wheel drag torque value pairs recorded in such an operating case are in Fig. 5 is shown. The tire slip stiffness B is calculated from the pairs of values ​​generated during a reference signal indicating a high-friction road surface and using a statistical evaluation function (ΔB = f(σ)). (B) )) a mean value ΔB and the corresponding variance are determined. The corresponding regression line g1 together with the class boundaries g 1u and the y-axis determine the road grip class A.

[0049] This allows all tire slip stiffness values ​​B or wheel slip-circular force value pairs to be classified, where the circular force is present as a driving, braking, or drag torque. If these values ​​or value pairs fall within road grip class A, the vehicle is traveling on a high-friction road surface; otherwise, if these values ​​or value pairs do not fall within road grip class A, they indicate a low-friction road surface.

[0050] The road grip class A thus represents a comparison standard based on high friction coefficient for the tire slip stiffness values ​​B or wheel slip-wheel circumferential force value pairs (|λ|, |T|).

[0051] By applying a statistical evaluation function, the tire slip stiffness values ​​B or the wheel slip circumferential force value pairs (|λ|, |T|) lying outside the road grip class A can be classified not only as a value indicating a low friction road surface, but also as indicating whether the vehicle is driving on a snow-covered or ice-covered road surface, for example.

[0052] Thus, according to Fig. 4. The tire slip stiffness values ​​B determined after time t1 are evaluated as data series of successive tire slip stiffness values ​​B using a statistical evaluation method (ΔB2 = f(σ)). (B2) and ΔB3 = f(σ (B3) )) a mean value ΔB2 with associated standard deviation and a mean value ΔB3 with associated standard deviation were determined, thus defining two further road grip classes B and C with the class boundaries ΔB 2uas lower limit and ΔB 2o as the upper limit for a road grip class B and ΔB 3u as lower limit and ΔB 3o as the upper limit for a further road grip class C according to Fig. 4 defined. These mean values ​​ΔB2 and ΔB3 are the initial slopes of regression lines g2 and g3 as slip-circular force characteristic curves in Fig. 3, where the class boundaries ΔB 2u and ΔB 2o the area limiting road grip class B and the class boundaries ΔB 3u and ΔB 3o which define the area limiting road grip class C and in Fig. 3 as straight lines g 2u and g 2o as well as g 3u (x-axis) and g 3o are registered.

[0053] Furthermore, it is from the Fig. 3 and Fig. 4 evident that the class boundaries ΔB 1u and ΔB 2u as well as the class boundaries ΔB 2u and ΔB 3ocoincide.

[0054] Road grip class B indicates a snow-covered road surface, while road grip class C indicates a snow-covered road surface after time t2 (see...). Fig. 4) indicates an icy road surface.

[0055] The above in connection with the Fig. 3 and Fig. The procedure described in section 4 is also carried out in the case of wheel slip torque operation. The wheel slip-wheel drag torque value pairs recorded in such an operating case are in Fig. Figure 5 illustrates this. From the wheel slip-wheel drag torque value pairs lying outside road grip class A, road grip classes B for snow-covered roads and road grip class C for ice-covered roads are determined in the same way. The corresponding regression lines g2 and g3 together with the class boundaries g 2u and g 2o or g 3o and g 3u (x-axis) determine the road grip classes B and C.

[0056] This allows all tire slip stiffness values ​​(B) or wheel slip-circular force value pairs to be classified, where the circumferential force is present as a driving, braking, or drag torque. If these values ​​or value pairs fall within road grip class A, the vehicle is traveling on a high-friction road surface; otherwise, if these values ​​or value pairs do not fall within road grip class A, they are either in road grip class B, indicating a snow-covered road surface, or in road grip class C, indicating an ice-covered road surface.

[0057] From the Fig. 3 and Fig. Figure 4 shows that the wheel slip-wheel torque value pairs for road grip classes B and C, and thus also the corresponding tire slip stiffness values, vary considerably more than the values ​​for road grip class A. This is also evident from the wheel slip-wheel drag torque diagram according to Fig. 5. This behavior can be advantageously used to assign the value pairs to a road grip class. Furthermore, it can be advantageously used that the curve shape of the corresponding data series from the wheel slip-circular force value pairs is also taken into account when classifying a value indicating a road grip class B or C (either as a wheel slip-circular force value pair or as a tire slip stiffness value). For example, a curvature indicates a snow-covered road surface, while discontinuities, such as a sudden break, indicate an icy road surface.

[0058] The due to a drive or braking torque according to Fig. 3 in inverse representation or due to a wheel drag moment according to Fig. Five specific regression lines g1, g2, and g3 with their associated limits are each combined into a single regression line; that is, the average tire slip stiffnesses in the same road grip class are equated, as shown below. Fig. 7 is explained.

[0059] This equation is possible because the initial slopes of the tire slip-circular force characteristic curve of a tire for the driving case and the braking case - as shown by the drag torque representation according to Fig. Figure 7 shows that the values ​​are the same. This fact is used to improve the accuracy of the calculation of the mean tire slip stiffness by using multiple support points, while simultaneously ensuring that the regression line does not necessarily pass through the zero point. A zero-point shift d, for example due to inaccurate or missing slip calibration – such as after adjusting tire pressure or changing tires – is not a hindrance with this method. This is a significant advantage that makes the method robust against calibration deviations.

[0060] According to this diagram Fig. Figure 7 shows the first quadrant for the drive case with the average tire slip stiffnesses for road grip classes A, B, and C, designated as regression lines g1, g2, and g3. The third quadrant shows the drag torque case, representing the general braking case, where the corresponding regression lines have the same slope as those in the first quadrant and are therefore also designated as regression lines g1, g2, and g3, and thus represent the same average tire slip stiffnesses.

[0061] It is also possible to determine these regression lines g1, g2 and g3 separately for a driven or braked vehicle wheel.

[0062] Finally, to detect µ-split situations, the tire slip is determined for both vehicle wheels on one vehicle axle subjected to a circumferential force, and the tire slip stiffness values ​​are calculated by the quotient of the circumferential force and the tire slip of the most slipping vehicle wheel.

[0063] The slip offset S0 is determined during both torque-free and (quasi-)stationary driving maneuvers. The slip offset results, for example, from different rolling circumferences of the tires, e.g., caused by manufacturing tolerances, different air pressures, or the wear of individual tires, and is not slip in the true sense. The slip offset occurs during torque-free driving maneuvers and therefore cannot usually be determined directly during normal driving. To collect pairs of values ​​for wheel slip and wheel torque or circumferential force for the inventive method, a regression line can be continuously determined from these pairs. The intersection of the regression line with the slip axis corresponds to the slip offset S0. Fig. Figure 6 shows the calculation of the slip offset using linear regression.

[0064] A vehicle for carrying out the method according to the invention comprises the following components: - an engine control unit, - a brake control unit, - a speed detection device for the vehicle's wheels, - a vehicle dynamics control system, and - a device for determining a wheel torque or a braking torque as a circumferential force of the vehicle wheel.

Claims

[1] Method for determining road grip classes from measured values ​​of a circumferential force on at least one vehicle wheel of a vehicle traveling on a road and the tire slip on the same vehicle wheel caused by this circumferential force, comprising the following steps: a) Determining pairs of values, each from the value of a circumferential force acting on the vehicle wheel and a tire slip on the vehicle wheel generated by this circumferential force, by determining the tire slip from a measured wheel speed of the vehicle wheel and the measured vehicle speed and determining the circumferential force from an engine drive torque acting on the vehicle wheel or on a vehicle axle of the vehicle, from an engine drag torque acting on the vehicle wheel or on a vehicle axle of the vehicle, from a braking torque on the vehicle wheel or from a braking torque acting on a vehicle axle of the vehicle, b) Determining a group of values ​​indicating a high friction road surface from the determined value pairs of circumferential force and tyre slip, whereby at the time of determination of these value pairs of driving dynamics variables of the vehicle is greater than a predetermined value and at the same time there is no intervention by a driving stability system and / or by optical sensors of the vehicle and / or by external data sources of the vehicle, a high friction road surface is or are indicated, c) Calculating tire slip stiffness values ​​from the pairs of values ​​by forming the quotient of circumferential force and wheel slip, d) Defining a road grip class indicating a high friction road surface by determining an average tyre slip stiffness from the tyre slip stiffness values ​​of the value pairs of the value group and the dispersion of these value pairs around the average tyre slip stiffness by means of a statistical evaluation method, and (e) Classifying a current tyre slip stiffness value as a value indicative of a high friction road surface if that value lies in the road grip class indicative of a high friction road surface, and otherwise as a value indicative of a low friction road surface. [2] Method according to claim 1, in which d1) from a data series of temporally successive pairs of values ​​of circumferential force and wheel slip, whose tire slip stiffness values ​​indicate a low friction road surface, an initial gradient is determined using a best-fit straight line as a linear part of the slip-circumferential force characteristic curve, and d2) the gradient of each best-fit line is defined as the mean tyre slip stiffness of a road surface grip class indicating a low-friction road surface with a scatter of the value pairs around the determined best-fit line determined by means of a statistical evaluation procedure. [3] Method according to claim 1 or 2, in which the tyre slip stiffness values ​​are assigned to a road surface grip class indicating a low friction road surface if its value lies in this road surface grip class. [4] Method according to one of the preceding claims, in which, for the classification of a tyre slip stiffness value indicating a low friction road surface, the curve shape approximated to the associated data series by a mathematical adjustment calculation is taken into account. [5] Method according to one of the preceding claims, in which the method steps d, d1 and d2 are carried out only in the presence of a positive slip gradient and driving conditions of the vehicle without vehicle control system interventions. [6] Method according to one of the preceding claims, in which, when carrying out method steps d, d1 and d2, the classification of the tyre slip stiffness values ​​is checked for plausibility on the basis of the intervention behaviour of a vehicle control system, in that, in the absence of vehicle control system interventions above predetermined longitudinal and / or lateral accelerations, a road surface grip class indicating a high friction coefficient road surface is checked for plausibility and, in the case of control system interventions below predetermined longitudinal and / or lateral accelerations, a road surface grip class indicating a low friction coefficient road surface is checked for plausibility. [7] Method according to one of the preceding claims, in which the method steps d, d1 and d2 are carried out both during acceleration and during deceleration of the vehicle, wherein for the acceleration case and the deceleration case the mean tyre slip stiffnesses of the road surface grip class are equated. [8] Method according to one of the preceding claims, in which the amounts of the circumferential forces and the slip are used to calculate the tire slip stiffnesses. [9] Method according to one of the preceding claims, in which the method steps d, d1 and d2 are carried out within a predetermined slip window between a slip value λ min and a slip value λ max be carried out, whereby for a slip greater than λ maxControl system interventions of a vehicle control system take place and at the same time such control system interventions are used to check the plausibility of the road surface grip classification. [10] Method according to one of the preceding claims, in which, in order to detect µ-split situations, the tire slip is determined for both vehicle wheels of the vehicle on a vehicle axle which are subjected to a circumferential force, and the tire slip stiffness values ​​are calculated by the quotient of the circumferential force and the tire slip of the vehicle wheel which slips the most or the average value of all wheels of a vehicle axle. [11] Method according to one of the preceding claims, in which the averaging and / or filtering of the value pairs of circumferential force and tire slip takes place over a full wheel revolution or over integer multiples of individual wheel revolutions.

Citation Information

Patent Citations

  • Methods for determining the coefficient of friction between a vehicle tire and a road surface

    DE102004044788B4

  • Method for determining frictional relationship between road surface and wheel of motor car, involves determining adhesion frictional relationship between wheel and road surface based on predetermined allocation function

    DE102012204671A1

  • Method for determining maximum traction coefficient between vehicle tire and road surface for safe driving of e.g. motor car, involves determining momentary traction coefficient based on condition of value tuple in predetermined regions

    DE102012217772A1

  • Method and device for determining the coefficient of friction in a vehicle

    DE102014103843A1

  • Maximum coefficient of friction determination method for automobile calculates grip between tyres and road surface in longitudinal and / or transverse directions

    DE10208815A1