DE102008002162B4 - Procedure for determining the coefficient of friction between wheel and road surface in a vehicle - Google Patents

Abstract

Method for determining the coefficient of friction between the wheel and the roadway in a vehicle by determining a corrected coefficient of friction (µkorr) from a raw value (µESP), which is determined from sensor signals, in that a local maximum (µESP ,max) is determined and the corrected coefficient of friction (µkorr) is set to the local maximum (µESP,max) of the raw value (µESP) until a subsequent local maximum (µESP,max) is reached.

Description

The invention relates to a method for determining the coefficient of friction between the wheel and the roadway in a vehicle.

State of the art

In the DE 10 2007 007 282 A1 a method for estimating friction coefficients in a vehicle is specified, whereby to improve the accuracy of the friction coefficient determination, a friction connection is estimated for each individual axle from the lateral forces and axle loads determined in a model, and the friction coefficient of the road is estimated by comparing the two adhesions. Since in critical driving situations or during highly dynamic lateral driving maneuvers, one axle of the vehicle always reaches the adhesion limit first before the other axle, the lateral adhesion of the front axle and the rear axle is calculated separately, from which the coefficient of friction can be determined, taking into account other parameters and conditions.

In the disclosure document DE 40 26 625 A1 a method for detecting the coefficient of friction is described. For this purpose, at least one wheel is turned and it is monitored whether the aligning torque has reached its maximum. The value of the aligning torque related to the wheel contact force is recorded when the maximum is reached and the current coefficient of friction is determined using a stored dependency.

The patent specification DE 41 02 301 C1 deals with a method for adapting response slip threshold values for a drive slip and/or brake slip control system to the tires of a motor vehicle. The measured operating data, such as fuel supply
and/or the supply of combustion air, engine speed and overall transmission ratio of the drive train of the vehicle determines the drive torque provided at the driven vehicle wheels and from this the utilized coefficient of adhesion is determined for a given or measured axle load at the driven vehicle wheels.

The disclosure document DE 10 2004 053 880 A1 relates to a method for determining the coefficient of friction between at least one vehicle tire and a road surface, the maximum coefficient of friction being determined by the fact that the instantaneous slip S and the instantaneous coefficient of friction µ are determined and in a plot of µ over S a straight line through the origin through the point determined in this way is intersected with a straight line parallel to the ordinate.

The Disclosure U.S. 2007/0150156 A1 relates to active chassis systems for estimating road wheel friction. In particular, the publication relates to a method for estimating the road surface friction coefficient. The method comprises the steps of: continuously estimating a road surface coefficient of friction using an algorithm based on a dynamic model of the vehicle, determining a range of the road surface coefficient of friction based on specific transient or static driving parameters of the vehicle, and resuming this algorithm so that the estimated coefficient of friction of the Road surface is adapted to the determined coefficient of friction range of the roadway.

A prerequisite for carrying out this method is that the vehicle is at least close to an unstable driving situation. However, such driving situations should generally be avoided.

It is also known to determine the coefficient of friction according to mathematical relationships from the normal force and the braking force and the lateral acceleration, for example in an ESP control system (Electronic Stability Programs) with which to
Vehicle stabilization interventions in the vehicle brakes and in the engine management are carried out. However, the coefficient of friction determined in this way is associated with a relatively large degree of uncertainty.

Disclosure of Invention

The invention is based on the object of providing a reliable estimated value for the coefficient of friction between the wheel and the road surface, measurement data from an ESP system preferably being used as input variables.

This object is achieved according to the invention with the features of claim 1. The subclaims indicate appropriate developments.

The coefficient of friction between a wheel and the road surface in a vehicle can be determined with the aid of the method according to the invention, starting from a raw value that is the starting point for a correction method. To improve the raw value, a local maximum is determined in the course of the raw value, with the improved or corrected coefficient of friction being set to the value of the local maximum until a new, local maximum is found in the further course of the raw value, at the value of which the corrected coefficient of friction is found is then set. It has been shown that the corrected Friction values correspond to the actual conditions better than the raw values supplied by an ESP system.

In principle, the sensors of an ESP system are sufficient to create a sufficient data basis for improving the raw value. This has the advantage that no additional sensors are required. However, it can also be expedient to carry out the method according to the invention independently of an ESP system, for example by evaluating sensor data outside of an ESP system. In principle, the longitudinal acceleration, the lateral acceleration in the vehicle and the braking force to be obtained from the brake pressure of a brake cylinder are sufficient for this; With knowledge of these values, which are determined by sensors, a raw value curve for the coefficient of friction between the wheel and the road can be formed, with the braking force relating to the brake cylinder of that wheel on which the coefficient of friction is to be determined.

A significant improvement is already achieved by setting the improved coefficient of friction to the local maximum in the course of the raw value. According to a further advantageous embodiment, it is provided that the corrected coefficient of friction is subjected to filtering, in particular PT ₁ filtering. A so-called plausibility-checked coefficient of friction is calculated via the filtering, which represents a further improvement of the raw value and is preferably further used in a vehicle dynamics control system. Filtering is preferably applied to the difference between a plausibility-checked coefficient of friction from a previous calculation cycle and the current, corrected coefficient of friction, with the filtered value then being subtracted from the plausibility-checked coefficient of friction, resulting in an updated, plausibility-checked coefficient of friction. The plausibility-checked coefficient of friction is approximated to the corrected coefficient of friction, which was determined via the local maxima of the raw value, via the filtering.

It can be expedient not to implement the time factor in the PT ₁ filtering as a constant, but as a variable, which preferably depends on the difference between the plausibility-checked coefficient of friction and the corrected coefficient of friction. The advantage of the variable time factor is that the increase or decrease in the course of the plausibility-checked coefficient of friction can be variably adjusted at different points in time, which means, for example, there is better adaptation in the event that the vehicle moves from a surface with a high coefficient of friction to a Subsoil with a low coefficient of friction changes.

According to a further advantageous embodiment, it is provided that the filtering is only carried out in the event that a vehicle variable to be checked lies within a plausible value range. If, on the other hand, the vehicle variable to be checked is outside the plausible range of values, the plausible coefficient of friction from a previous calculation cycle is used further; this value is retained until a new corrected coefficient of friction and, based on this, filtering has been carried out in a subsequent calculation cycle and the vehicle variable or variables to be checked are also within the assigned plausible value range. In this context, it can also be expedient to check several vehicle sizes, in which case both alternative and cumulative conditions can be specified which the vehicle sizes in question must meet.

For example, it is queried whether, in the event that the lateral acceleration exceeds an upper limit value, the coefficient of friction, in particular the corrected coefficient of friction, falls below a limit value. This coupled query is based on the fact that high lateral acceleration when the vehicle is in a stable driving state is only possible with a correspondingly high coefficient of friction. On the other hand, it is implausible that a high lateral acceleration is achieved on a roadway with a low coefficient of friction. If it is determined during this query that the coefficient of friction is not greater than the limit value despite high lateral acceleration, this indicates that the corrected coefficient of friction determined via the local maxima of the raw value is not reliable and the plausible coefficient of friction from a previous calculation cycle is used reused.

As a further condition, it can be checked whether the transverse acceleration falls below a lower limit value and at the same time the corrected coefficient of friction exceeds the filtered coefficient of friction that applied up to that point and was determined from a previous calculation cycle. It has been shown that these two conditions must also be met so that the calculated, corrected coefficient of friction is reliable and can then be subjected to filtering, from which the filtered, plausibility-checked value emerges, which can then be used further.

As a further condition that must be met so that the corrected coefficient of friction can be subjected to the filtering, the query is expediently started as to whether the wheel slip on a vehicle wheel is greater than an associated threshold value. For example, wheel slips on the front axle are determined via wheel-specific speed sensors, which are part of a standard ESP system in particular, and the slip on a front wheel is calculated therefrom. It has shown, that the corrected coefficient of friction is within a reliable value range if the front wheels slip by a few percent.

Furthermore, as a plausibility condition, it can be queried whether the corrected coefficient of friction is at least as large as the plausibility-checked coefficient of friction calculated from a previous cycle, which is identical to the filtered value.

With the exception of the coupled conditions, the aforementioned conditions do not have to be fulfilled cumulatively, but it is sufficient if these are alternatively fulfilled so that the calculated, corrected coefficient of friction can be taken into account as a reliable value. On the other hand, if none of these conditions are met or if one of the coupled conditions is not met, the corrected coefficient of friction is not filtered and a corresponding coefficient of friction from a previous cycle continues to be used instead of a current, plausibility-checked coefficient of friction.

• 1 eine grob schematisierte Darstellung eines Systems zur Ermittlung eines verbesserten Reibwertes,
• 2 zwei Diagramme mit dem Verlauf von verschiedenen Reibwerten als Funktion der Zeit sowie dem zugeordneten Verlauf eines logischen Flags, welches Bereiche mit erhöhter Zuverlässigkeit bzw. Plausibilität der verbesserten Reibwerte anzeigt,
• 3 ein Blockschaltdiagramm, das den Verfahrensablauf zur Ermittlung des verbesserten Reibwertes angibt.

Further advantages and expedient embodiments can be found in the further claims, the description of the figures and the drawings. Show it:

• 1 a roughly schematic representation of a system for determining an improved coefficient of friction,
• 2 two diagrams with the progression of different coefficients of friction as a function of time and the associated progression of a logical flag, which indicates areas with increased reliability or plausibility of the improved coefficients of friction,
• 3 a block circuit diagram that specifies the method sequence for determining the improved coefficient of friction.

1 shows a roughly schematic block diagram for obtaining an improved, plausible coefficient of friction μ _F from a coefficient of friction μ _ESP, which is determined from a standard ESP system (electronic stability program for vehicle stabilization). The ESP friction coefficient μ _ESP is to be regarded as a raw value that is associated with a relatively large degree of uncertainty. Since precise knowledge of the coefficient of friction between the tires and the road is of great importance for vehicle dynamics applications, a method for improving the coefficient of friction is carried out on the basis of the ESP raw value μ _ESP . This is preferably done solely via sensor signals that are available in a standard ESP system, in particular the slip values λ _FL/R on the left and right front wheels of the vehicle, which are obtained from wheel speed sensors, and also acceleration values in the vehicle, in particular the Longitudinal acceleration a _x and lateral acceleration a _y . In the block diagram, the input variables are processed according to a defined relationship described below, with the plausibility-checked, improved coefficient of friction µ _F and a logical flag bMueValid being supplied as output variables, which either has the value "1" or "true", which is what a reliable database on which the calculation of the improved coefficient of friction _µF is based, or the value "0" or "false" if the database is not to be used for the calculation of an improved coefficient of _friction µF.

wobei

FN

die Normalkraft

m

die Masse des Fahrzeugs

g

die Erdbeschleunigung

ax

die Längsbeschleunigung

ay

die Querbeschleunigung

hs

die Höhe des Fahrzeugschwerpunktes

Lx

den Achsenabstand vom Schwerpunkt

L

den Achsenabstand

bezeichnet.To determine the current ESP friction coefficient µ _ESP , the normal force F _N is calculated based on the measured longitudinal and lateral acceleration a _x and a _y : $$f_N=\frac{1}{2}mG\frac{L_x}{L}\pm \frac{1}{2}m{a}_y\frac{H_s}{b}\mp \frac{1}{2}m{a}_x\frac{H_s}{L},$$

whereby

FN

the normal force

m

the mass of the vehicle

G

the acceleration due to gravity

ax

the longitudinal acceleration

hey

the lateral acceleration

hs

the height of the vehicle's center of gravity

lx

the axis distance from the center of gravity

L

the axis distance

designated.

$${\mu }_{lot}=\frac{a_y}{g},$$

$${\mu }_{ESP}=\sqrt{{\mu }_{long}^2+{\mu }_{lot}^2},$$

wobei

µlοng

den Reibwert in Längsrichtung

µlat

den Reibwert in Querrichtung

µESP

den aus dem ESP-System gewonnenen, resultierenden Reibwert

FBr

die dem Bremsdruck im Radbremszylinder entsprechende Bremskraft

bezeichnet. Die Funktion f in Abhängigkeit der Bremskraft F_Br und der Normalkraft F_N ist eine nichtlineare, bekannte Funktion.The coefficient of friction between the vehicle wheel and the road surface is determined taking into account the normal force F _N according to the following relationship: $${\mu }_{lOnG}=ƒ\left(\frac{f_{B\mathrm{right}}}{f_N}\right),$$

$${\mu }_{lOt}=\frac{a_y}{G},$$

$${\mu }_{ESP}=\sqrt{{\mu }_{lOnG}^2+{\mu }_{lOt}^2},$$

whereby

µlong

the coefficient of friction in the longitudinal direction

µlat

the coefficient of friction in the transverse direction

µESP

the resulting coefficient of friction obtained from the ESP system

FBr

the braking force corresponding to the brake pressure in the wheel brake cylinder

designated. The function f as a function of the braking force F _Br and the normal force F _N is a non-linear, known function.

In the following, reference is made to the curves according to FIG 2 and the block diagram according to 3 . In 3 shows a detailed block diagram for determining the improved, plausible coefficient of friction µ _F based on the ESP raw value µ _ESP . First, the raw value μ _ESP is subjected to filtering in block 1, in particular to PT ₁ filtering. It is then determined in block 2 whether a maximum is present. If this is the case, according to block 4 a switch is set in such a way that the μ _ESP friction coefficient from the previous cycle is used further, which is shown in block 3 with Z ⁻¹ , which designates the previous calculation cycle. If, on the other hand, no maximum is determined in block 2, the previously valid value is retained.

The current ESP coefficient of friction μ _ESP and the value obtained in switch 4 are then fed to a block 5 in which a maximum query is carried out. The larger of the two values fed to block 5 is selected via the maximum query; on the output side, this is the corrected coefficient of friction μ _Korr .

This procedure can also be seen in the curves for the coefficients of friction μ according to 2 read off. The raw value μ _ESP obtained from the ESP system increases until a local maximum is reached; this increase is also completed by the corrected coefficient of friction μ _corr . The μ _ESP raw value then drops again, but this drop is not reproduced by the corrected coefficient of friction μ _corr ; instead, the previously valid value of the corrected coefficient of friction μ _corr is retained until a new local maximum is reached in the ESP raw value μ _ESP . In this case, the new, local maximum of the ESP raw value μ _ESP can be greater than, less than, or at the same level as the previous local maximum. In the embodiment according to 2 the second local maximum is at a higher value than the first local maximum, whereas the third local maximum has a lower value. Accordingly, the corrected coefficient of friction μ _corr increases to a higher value when the second maximum is reached and remains at this higher value until the third local maximum is reached, to the value of which the corrected coefficient of friction μ _corr then falls.

The following blocks 6 to 9 according to 3 represent the further processing of the corrected coefficient of friction μ _corr , which is subjected to filtering. This filtering is carried out in block 6, for example as a PT ₁ filter. The time constant T for the PT ₁ filter in block 6 is obtained from a preceding block 7, which contains a characteristic curve for determining the time constant T as a function of the difference between the corrected coefficient of friction μ _corr and the filtered, plausibility-checked coefficient of friction μ _F .

The next method step 9 contains a switch whose switching position depends on the value of the logical flag bMueValid, the plausibility of the filtered coefficient of friction μF being determined via the value of the logical flag _bMueValid . The logical flag bMueValid takes either the value "true" or "false", depending on whether a vehicle size to be checked or a combination of vehicle sizes is within or outside a plausible value range. The vehicle variables to be checked are, in particular, the wheel slip λ, preferably on the front wheels of the vehicle, the lateral acceleration a _y and the magnitude of the correction coefficient of friction μ _Korr .

In 3 the switch from switching block 9 is set to the value "true", so that the correction coefficient of friction μ _corr filtered in block 6 is accepted as the new, plausibility-checked coefficient of friction μ _F . This situation occurs when the condition complex that determines the value of the logical flag bMueValid is met, which physically means that the value range of the vehicle variables to be checked is within a plausible value range. The newly obtained, plausibility-checked coefficient of friction _μF is then retained until it is overwritten with a new calculated value in a subsequent cycle.

If, on the other hand, the query for the logical flag bMueValid shows that one or more of the vehicle variables to be checked are not within the plausible range of values, the flag bMueValid is set to the value "false" and the previously valid, plausible coefficient of friction µ _F is taken from a retained or taken over from the previous calculation cycle.

The course of values for the plausibility-checked coefficient of friction μF as a function of the value for the logical flag _bMueValid is also 2 refer to. As soon as the logical flag bMueValid is set to "true", filtering takes place and the plausibility-checked value µ _F is adjusted accordingly. The drop in the course of the plausibility-checked coefficient of friction μ _F comes about because the filter block receives the difference between the plausibility-checked coefficient of friction μ _F from a previous cycle and the current, corrected coefficient of friction μ _corr . As soon as the logical flag bMueValid is set to "true", the course of the plausibility-checked coefficient of friction μ _F approaches the course of the corrected coefficient of friction μ _corr . If, on the other hand, the logical flag bMueValid is set to "false", there is no adjustment in the course of the plausibility-checked coefficient of friction µ _F , the value of µ _F that was valid until then is retained.

Various conditions can be formulated for the vehicle variables to be checked, which influence the value of the logical flag bMueValid. A query is preferably carried out with regard to the wheel slip λ on the front wheels, the lateral acceleration and the corrected coefficient of friction μ _corr . For example, the wheel slip λ on a vehicle wheel, especially on the front wheels, must be greater than an assigned threshold value λ _limit so that the logical flag is set to the value “true” and the filtered value is thus accepted as the new, plausible coefficient of friction _µF . This is based on the experience that there are plausible coefficients of friction when slippage of a few percent prevails on the front wheels.

Furthermore, in order to determine the logical flag bMueValid, it is checked whether the lateral acceleration a _y exceeds an upper limit value a _y,max and at the same time the corrected coefficient of friction μ _corr exceeds an associated limit value μ _min . This is based on the fact that a high lateral acceleration a _y in the vehicle is only possible with stable driving if there is a correspondingly high coefficient of friction. If this coupled condition is not met, this indicates that the database on which the algorithm is based is not plausible, whereupon no new, updated coefficient of friction _μF is adopted, but the previously applicable coefficient of friction _μF is retained.

As a further condition, it is queried whether the lateral acceleration a _y falls below a lower limit value a _y,min and at the same time the corrected coefficient of friction μ _corr exceeds the filtered coefficient of friction n _F determined from a previous calculation cycle. If, on the other hand, the corrected coefficient of friction for the case of small lateral accelerations is below the plausible coefficient of friction from the previous calculation cycle, the logical flag bMueValid is set to the value "false".

Finally, it can also be checked whether the corrected coefficient of friction μ _corr is equal to or greater than the filtered coefficient of friction μ _F determined from a previous calculation cycle. This can also be formulated as a condition that must be met so that the logical flag bMueValid is set to the value "true".

Claims (14)

A method for determining the coefficient of friction between the wheel and the roadway in a vehicle by determining a corrected coefficient of friction (µ _corr ) from a raw value (µ _ESP ), which is determined from sensor signals, in _that a local Maximum (μESP _,max ) is determined and the corrected coefficient of friction ( _μkorr ) is set to the local maximum (μESP _,max ) of the raw value ( _μESP ) until a subsequent local maximum (μESP _,max ) is reached. procedure after claim 1 , characterized in that the corrected coefficient of friction (µ _corr ) is subjected to filtering, in particular PT ₁ filtering. procedure after claim 2 , characterized in that the difference between a plausibility-checked coefficient of friction (µ _F ) and the corrected coefficient of friction (µ _corr ) is subjected to filtering, the filtered value being subtracted from the plausibility-checked coefficient of friction (µ _F ). procedure after claim 3 , characterized in that the plausibility-checked coefficient of friction (µ _F ) represents the final coefficient of friction, which is supplied for further use. Procedure according to one of claims 2 until 4 , characterized in that the time factor (T) in the filtering is implemented as a variable which depends on the difference between the plausibility-checked coefficient of friction (µ _F ) and the corrected coefficient of friction (µ _corr ). Procedure according to one of claims 2 until 5 , characterized in that the filtering is only carried out if a vehicle variable to be checked (λ, a _y , µ _korr ) lies within a plausible value range. procedure after claim 6 , characterized in that in the event that the vehicle variable to be checked (λ, a _y , µ _korr ) is not within the plausible range of values, the corrected coefficient of friction (µ _korr ) or the plausible coefficient of friction (µ _F ) from a previous calculation cycle is used. procedure after claim 6 or 7 , characterized in that the vehicle variable to be checked is the wheel slip (λ) on a vehicle wheel, which must be greater than an associated threshold value (λ _limit ). Procedure according to one of Claims 6 until 8th , characterized in that the vehicle variables to be checked are the transverse acceleration (a _y ) and the coefficient of friction (µ _corr ). procedure after claim 9 , characterized in that the filtering is only carried out if, in the event that the transverse acceleration (a _y ) exceeds an upper limit value (a _y,max ), the corrected coefficient of friction (µ _corr ) also exceeds a limit value (µ _min ). procedure after claim 9 or 10 , characterized in that the filtering is only carried out if, in the event that the lateral acceleration (a _y ) falls below a lower limit value (a _y,min ), at the same time the corrected coefficient of friction (µ _corr ) is the filtered one determined from a previous calculation cycle Coefficient of friction (µ _F ) exceeded. Procedure according to one of Claims 6 until 11 , characterized in that the vehicle variable to be checked is the corrected coefficient of friction (µ _corr ), which must be at least as large as the filtered coefficient of friction (µ _F ) determined from a previous calculation cycle. Regulating or control unit for carrying out the method according to one of Claims 1 until 12 . Driving dynamics control system in a vehicle with a control or control unit Claim 13 .

Images