DE102005001770A1 - Method for determination of maximal braking power of anti-lock braking system, using extreme local data - Google Patents

Abstract

A set braking slip level is determined using the adaptive curve of friction forces and the braking slip at the individual wheel of a vehicle. In a first step a number of straight lines is created by estimation from a number of matched pairs of data. In a second step an optimal braking slip is determined using the level of inclination of the resulting line. This again is used for determining a set braking slip for a particular wheel.

Description

The The invention relates to a method for brake slip control of a Anti-lock braking system of a motor vehicle brake system.

Around When braking motor vehicles blocking the wheels and so that a slinging of the motor vehicle is prevented in modern automotive brake systems, a so-called anti-lock brake control system or an anti-lock braking system (ABS) used. This should, in particular even on slippery road surfaces, like For example, in snowy roads or wet, a Blocking the wheels prevent and thus the braking distance to a stop of the motor vehicle minimize or maximize braking performance.

One ABS usually includes a brake control device serving to apply the brake pressure settle with the one for each wheel of the motor vehicle used applied wheel brake cylinder is to thereby generate a braking force. Furthermore includes an ABS usually a wheel speed scanner for sampling the speed of each vehicle wheel or a number Raddreh- with signals, provided by the wheel speed scanner will calculate the wheel-specific slip and then depending from the calculated slip to the brake control device, a control signal to spend that, necessary is to a possible Blocking an associated one Avoid cycling.

there the brake pressure of a wheel brake cylinder is reduced when the ABS signs of it determines that a wheel could block. As an indication of the blocking a wheel is usually the crossing a predetermined threshold value used by the brake slip. The control device of an ABS therefore usually includes a so-called wheel slip control, which reduces the wheel speed during the Braking through the use of targeted ABS pressure build-up pulses, Pressure holding phases and pressure reduction pulses regulated. This wheel slip controller usually works according to a predefined control strategy or Usually philosophy.

aim This rule strategy is for every wheel one for the minimization of the braking distance optimal Sollbremsschlupf to determine and to specify this desired brake slip to the wheels via the control device.

Problematic in previous realizations of ABS is that the determination of a suitable optimal Sollbremsschlupfes complicated and very costly is. The difficulty is in particular, a wheel over the Phases of pressure build-up pulses, pressure hold phases and pressure release pulses always the possible to specify optimal brake slip to the braking power overall to maxi mize. Here is a theoretically optimal brake slip of various parameters such as the brake request, the speed of the wheels and the motor vehicle, the road surface or the lateral acceleration of the motor vehicle, where these parameters can change dynamically during a braking process. Besides that is a recording of these parameters comparatively difficult. For example, the determination of the vehicle speed is designed due to the case of an ABS braking occurring different large brake slip of the wheels accordingly difficult because a motor vehicle speed usually based on a wheel speed is determined.

Of the The invention is therefore based on the object of specifying a method with a simple way to the highest possible braking performance can be achieved with an anti-lock system.

These The object is achieved according to the invention by a desired brake slip of a wheel of the motor vehicle is determined, and this desired brake slip predefined a wheel of the motor vehicle is, with a desired brake slip from an adaptive curve of adhesion coefficient and brake slip of a wheel of the motor vehicle is determined.

there is especially for all wheels in each case a desired brake slip from an adaptive adhesion coefficient / brake slip curve a wheel calculated and then given the respective wheel.

The Invention is based on the consideration That's the main difficulty with an antilock system in it exists, on the one hand one for a high braking power as possible to calculate ideal brake slip, and on the other hand to ensure that the resulting Istbremsschlupf a wheel as accurate as possible corresponds to the predetermined target brake slip or an occurring Deviation possible low fails. This deviation can occur upwards and downwards. The The invention therefore further assumes that, since it is an ABS braking process is a dynamic process, in which an optimal brake slip of dynamically changing parameters over time, this dynamics Should be taken into account. However, should be from an investigation all for a determination of an optimal desired brake slip required brake parameters, such as the exact speed of the motor vehicle or the physical characteristics of the road surface, apart because this would be too complex and therefore too costly and time-consuming.

the across from but all go for a determination of an optimal brake slip necessary parameters also indirectly from the dynamic state of motion of a wheel, so that a brake slip in principle from a combination the currently applied brake slip of a wheel and the adhesion coefficient of road surface and wheel results. However, because of the dynamic rule problematic should a number of adhesion coefficient / brake slip pairs are not isolated to be viewed as. Rather, it should be used in temporal Intervals of a braking process a functional context of Adhesion coefficient / brake slip pairs exists. The adaptive Traction coefficient / brake slip curve represents the continuous time Detection of adhesion coefficient and brake slip.

With One can identify and interpret these functions calculate where the optimum brake slip of a wheel in temporal Course is and give them a wheel or a desired brake slip specify that is close to the optimal brake slip. It is at a deviation of the Istbremsschlupfes from the desired brake slip this Deviation virtually automatically in the calculation of the setpoint slip Accounted for by the deviation in the calculation of the target brake slip with flows.

For a calculation an optimal target brake slip are preferably a number local extreme values of the coefficient of adhesion brake slip slip curve determined. This is when reaching an extreme value of the adaptive Traction coefficient brake slip curve optimal brake slip reached, in which the coefficient of adhesion has a local maximum. The identification of these extreme values is used for calculation and readjustment the desired brake slip. about the knowledge of the extreme values or the optimum brake slip values a setpoint brake slip over time this optimal Brake slip values approximated become.

For the investigation a coefficient of adhesion of a wheel, this is advantageously from the wheel contact force and the circumferential force applied to the wheel calculated. A coefficient of adhesion of a wheel is calculated here from the quotient of peripheral force and wheel contact force of a Wheel.

Conveniently, will do it estimated the wheel tread force or calculated according to a model.

to wheel-individual consideration of the curve of adhesion coefficient and brake slip or to identify local extreme values of Adhesion coefficient / brake slip slip curve, are preferably linear sections of the coefficient of adhesion / brake slip slip curve determined. It can on the basis of this straight line or in particular on the slope of this Define the local extreme values of the coefficient of adhesion / brake slip slip curve be located.

For a determination the linear sections of the coefficient of adhesion / brake slip curve These are advantageously provided with a regression calculation, in particular with a regression calculation of the least squares method, estimated.

For a continuous time Calculation of a desired brake slip, is expediently a linear section the coefficient of adhesion / brake slip curve for each new sampling time from a selectable Number of last measured value pairs of adhesion coefficient over time and brake slip detected. For one limited sum of quadratic errors in the regression analysis is this number of measured value pairs of adhesion coefficient and Brake slip deliberately limited.

In order to determine whether an actual brake slip of a wheel is removed in the course of time from an approximately optimal brake slip course or an optimum target brake slip is passed, it is more expedient example determines the slope of an estimated linear section of the coefficient of adhesion / brake slip curve and these slopes preferably used to determine a desired brake slip.

Around the desired brake slip of a wheel when passing the optimum Setpoint brake slip to adapt accordingly, this is advantageous with a change of sign of the slopes of temporally successive following estimated linear sections of the coefficient of adhesion / brake slip curve recalculated. It is assumed that at a sign change the Straight line slopes an optimal brake slip is passed.

To A change of sign expediently calculates a desired brake slip from an arithmetic mean of a selectable number of times over time last brake slip values. The number of brake slip values corresponds preferably the number of adhesion coefficient / brake slip pairs an estimated linear section of the coefficient of adhesion / brake slip curve.

To a sign change of estimated linear sections from minus to plus and a recalculation of a setpoint slip or at the beginning of a braking, a desired braking slip is advantageously as long as gradually increased, until the determined slope of the estimated linear sections is the sign changes or falls below or exceeds a preset limit. For the gradual increase can a desired brake slip with a constant control gain factor be charged or in particular multiplied by this become.

To a sign change from linear sections from plus to minus is a desired brake slip expediently gradually reduced until the determined slope of the estimated linear sections, the sign changes or a given Limit value or below.

The The method described is preferably via a wheel slip controller executable, making it into an antilock brake system of a motor vehicle brake system integrate. The brake system is thus preferably controllable by the method.

The Method is advantageously carried out by a computer program product executed.

The particular advantage of the invention is that with the described method a calculation of a suitable Sollbremsschlupfes guaranteed is, which are given simultaneously a wheel of the motor vehicle can. The procedure is for the requirements occurring during ABS braking, in particular in terms of the high temporal dynamic control requirements, designed.

One Another advantage of the method is the possibility of a desired brake slip a wheel essentially only from the dynamic state of motion to be able to calculate a wheel. It allows the method at a devia tion of Istbremsschlupfes from the target brake slip this deviation simultaneously over to include the continuous recalculation of the setpoint slip. this leads to to a small deviation of the actual brake slip from the optimal Target brake slip and thus to a high braking performance of the anti-lock braking system.

Further Advantages of the method consist in the simple realization and in the stable operation and operation of the method in the motor vehicle.

An embodiment of the invention will be described with reference to 1 to 3 explained in more detail. It shows

1 1 schematically shows the sequence of a method for brake slip control of an anti-lock braking system on a wheel of a motor vehicle,

2 two straight line estimates of an adaptive traction coefficient (μ) / brake slip λ curve for two different time periods and

3 the braking parameters of an ABS braking as a function of time t.

The Explanations for the Abbreviations in the formulas, the list of reference numerals can be found.

In 1 schematically shows the flow of a method for Sollbremsschlupfermittlung and default to a wheel of a motor vehicle. It is a method for ABS braking and is shown in all figures of the embodiment, only exemplary of a wheel of the motor vehicle, since this is applied analogously for all wheels of the motor vehicle. How out 1 As can be seen, the method has a first control stage in which a number of straight lines y (k) are estimated from a number of measured value pairs of the adhesion coefficient μ and the brake slip λ. In a second stage, an optimum brake slip λ _{opt is} determined from the gradient c ₁ , this straight line y (k). A target brake slip λ _d (k) is calculated and a wheel of the motor vehicle from this optimum brake slip λ _opt specified in the third step of the process. The entire process is carried out by a wheel slip controller, not shown, of the motor vehicle brake system. The implementation of the determined target brake slip λ _d (k) with the hydraulic components of the brake system for a wheel is also not described in more detail, since this process works analogously to already known brake systems with ABS. In the following, the three stages of the method are explained in more detail.

ermittelt.In 2 are in the course of time continuously running a number of coefficient of adhesion μ / brake slip λ pairs in an ABS braking on asphalt shown. The pairs are each taken at a distance of a temporal sampling time T of the vehicle dynamics control. The continuous-time detection of the pairs takes place with the running parameter k, for which k ε N. The adhesion coefficients μ are estimated values of the wheel contact forces G and the circumferential force N at the wheel according to the context

determined.

For a calculation of an optimal brake slip, wheel-specific local extreme values or maxima of the adhesion coefficient μ / brake slip λ curve are determined. For a localization of these extreme points piecewise linear regions of the adhesion coefficient μ / brake slip λ curve are determined, wherein for the areas a number w of adhesion coefficient μ / brake slip λ pairs of a defined time window .DELTA.T be selected. In this case, w is 25 in the exemplary embodiment. 2 shows by way of example two straight lines estimated with a regression analysis μ (k) = c ₀ + c ₁ · λ (k) for two time segments k = 25 and k = 50 with t ε (0; 25 · T] and t ε (25 · T , 50 · T]. Regression analysis uses the method of least squares, which is based on a minimization of the quadratic equation errors.

For this purpose, the equation system y (k) = ψ ^T (k) · Θ is defined, where y k = μ (k), ψ (k) = [1 λ (k)] T and Θ = [c 0 c 1 ] T ,

The estimation of the parameter vector Θ then results in Θ = [Ψ ^T (k) · ψ (k)] ^-1 · Ψ (k) · y (k) with the output variable y (k) = [y (k). y (k - w + 1)] ^T and the input quantity Ψ (k) = [Ψ (k) ... Ψ (k - w + 1)] ^T. In 2 the two estimated straight lines y ₂₅ and y _{50 are} drawn, the first straight line y ₂₅ having a positive gradient c ₁ and the second straight line y ₅₀ having a negative gradient c ₁ . Comparable ABS braking, for example on loose snow, would result in a flatter course of the two estimated straight lines.

The estimated constant portion of the line c ₀ is of no further interest for further calculations. The sign of c ₁ or the course of c ₁ provides information about where the extreme values of the adhesion coefficient μ / brake slip λ curve are located. For the time-continuous case the coefficient of adhesion μ / brake slip λ-curve applies when a main or secondary maximum is reached:

In order to determine an optimal brake slip λ _opt , the sign change from the straight line slope c ₁ and the arithmetic brake slip mean value of the respectively last w = 25 number of brake slip λ values are used. The passage of an optimal brake slip λ _opt (k) is detected as soon as the condition sgn (c ₁ (k)) ≠ sgn (c ₁ (k-1)) is met or the sign of the straight line slope c ₁ changes and an extreme value passes becomes. The corresponding optimal brake slip λ _opt is then approximately calculated

The desired brake slip λ _d, which is given to the wheel, is calculated as:

von c₁ von kontinuierlich aufeinander folgenden geschätzten linearen Abschnitten von w = 25 Anzahl von Kraftschlussbeiwert μ/Bremsschlupf λ-Paaren.This formula for λ _d and the specification of the target brake slip of λ _d to the wheel is based on 3 explained in more detail. In the upper third of the figure is shown in gray the speed of the motor vehicle and in black the speed of the wheel. In the middle third, the target brake slip λ _{d of} the wheel is shown in gray and the measured actual brake slip of the wheel is shown in black. The lower third of the figure shows time continuously the amount of slopes

of c ₁ of continuously successive estimated linear sections of w = 25 number of adhesion coefficient μ / brake slip λ pairs.

negativ wird. Dies ist im Ausführungsbeispiel zum erstenmal bei 0,95 Sekunden der Fall. Dabei wird wie bei allen Vorzeichenwechseln von

der Sollbremsschlupf λ_d erneut berechnet und dem Rad vorgegeben. Deshalb tritt durch den neu berechneten optimalen Bremsschlupf, beispielsweise bei 0,95 Sekunden, ein Sprung im Sollbremsschlupf nach unten auf. Im folgenden wird der Sollbremsschlupf λ_d solange mit der Regelverstärkung k_λ reduziert, bis das Vorzeichen der Geradensteigung

bei 1,125 Sekunden wieder positiv wird. Dieser Regelvorgang bzw. die Vorgabe des Sollbremsschlupfes λ_d wird solange durchgeführt, bis das Kraftfahrzeug bei 3 Sekunden zum Stillstand gekommen ist. Wie sich aus dem zweiten Drittel von 3 erkennen lässt, verläuft der gemessene Istbremsschlupf für eine hohe Bremsleistung über die Bremszeit t näherungsweise entlang des Sollbremsschlupfes λ_d eines Rades.At the beginning of braking, the desired braking slip λ _{d is} increased with the control gain k _λ until the ascertained straight line gradient

becomes negative. This is the case in the embodiment for the first time at 0.95 seconds. It is like all sign changes of

the setpoint brake slip λ _{d is} calculated again and predefined for the wheel. Therefore, due to the newly calculated optimum brake slip, for example at 0.95 seconds, a jump in the desired brake slip occurs downwards. In the following, the desired braking slip λ _{d is} reduced with the control gain k _λ until the sign of the straight line gradient

becomes positive again at 1.125 seconds. This control process or the specification of the desired brake slip λ _d is carried out until the motor vehicle has come to a standstill at 3 seconds. As is apparent from the second third of 3 can be seen, the measured Istbremsschlupf for a high braking performance over the braking time t approximately along the target brake slip λ _{d of} a wheel.

k

running parameters

μ

coefficient of adhesion

λ

brake slip

λ _d

Target brake slip

λ _opt

optimal brake slip

Claims (16)

Method for brake slip control of an anti-lock brake system of a motor vehicle brake system, in which a desired brake slip (λ _d ) of a wheel of the motor vehicle is determined and this desired brake slip (λ _d ) is given to a wheel of the motor vehicle, wherein a desired brake slip (λ _d ) from an adaptive curve of adhesion coefficient ( μ) and brake slip (λ) of a wheel of the motor vehicle is determined. The method of claim 1, wherein a number of local Extreme values of the coefficient of adhesion (μ) / brake slip (λ) curve determined become. The method of claim 1 or 2, wherein a coefficient of adhesion (μ) one Rades from the Radaufstandskraft and the circumferential force applied to the wheel is calculated. The method of claim 3, wherein a wheel contact force estimated or calculated according to a model. Method according to one of claims 1 to 4, wherein linear Sections of the coefficient of adhesion (μ) / brake slip (λ) curve determined become. The method of claim 5, wherein a linear section the coefficient of adhesion (μ) / brake slip (λ) curve with a regression calculation, in particular with a regression calculation the method of least squares is estimated. Method according to one of claims 5 or 6, wherein an estimated linear section of the coefficient of adhesion (μ) / brake slip (λ) curve for each new sampling time (k) from a selectable number (w) the last measured value pairs of adhesion coefficient (μ) and brake slip (λ) are determined over time. Method according to one of claims 5 to 7, wherein the slope

an estimated linear sections is determined. The method of claim 8, wherein a gradients

an estimated linear portion for determining a target brake slip (λ _d ) is used. Method according to one of claims 8 or 9, wherein at a change of sign of the slopes

a desired braking slip (λ _d ) is recalculated from temporally successive estimated linear sections. Method according to Claim 10, in which a desired brake slip (λ _d ) is calculated from an arithmetic mean of a selectable number (w) of the last brake slip values (λ) over time. Method according to one of claims 1 to 11, wherein at the beginning of a braking or after a change of sign of the slope

of estimated linear sections from minus to plus, a target brake slip (λ _d ) is incrementally increased until the determined slope

of estimated linear sections, the sign changes or falls below a predetermined limit. Method according to one of claims 1 to 11, wherein after a change of sign of the slope

of the estimated linear portions from plus to minus a desired brake slip (λ _d ) is progressively reduced until the determined slope

of estimated linear sections, the sign changes or falls below a predetermined limit. Wheel slip control, about which a procedure after one of the claims 1 to 13 executable is. Computer program product for carrying out a method according to one of the claims 1 to 13. Braking system, with a procedure according to a the claims 1 to 13 is controllable.