DE10328028A1 - Method for determining motor-vehicle and/or trailer braking appliances braking characteristics, involves calculating actual braking characteristics by interpolation from families of characteristics - Google Patents

Abstract

A method for determining the actual braking characteristic of braking appliances of a motor vehicle in which discrete braking characteristics are ascertained depending on input data. Initially, at least two families of characteristics are generated from the discrete braking characteristics, and the actual braking characteristics are calculated, by interpolation, from the families of characteristics. Independent claims are included for (a) braking control for carrying out the method, for (b) a motor vehicle equipped with the braking control and for (c) a motor vehicle trailer or second car equipped with the braking control.

Description

The The invention relates to a method for determining the brake characteristic value of braking devices of a motor vehicle and / or a motor vehicle trailer, wherein discrete brake characteristics depending on be determined from input data. The invention also relates to a Brake control, a motor vehicle and a motor vehicle trailer.

The determination of discrete brake characteristics as a function of input data obtained during vehicle operation is known in principle from US Pat DE 196 48 936 A1 and the building on that DE 100 11 270 A1 , The exact knowledge of the brake characteristic is necessary in order to achieve different control targets of the brake system in motor vehicles. Such a control objective is, for example, the optimum adhesion during braking, with all wheels to be kept at the same slip for optimum adhesion utilization. Another goal is the optimal gradeability of the brake system to get a car-like pedal feel. For semitrailer tractors, the main goal is the mass-proportional distribution of braking work between the towing vehicle and the trailer. This has the meaning that the towing vehicle and trailer are braked with the same delay. Thus, it can be avoided, for example, that is pushed too strong deceleration of the towing vehicle this less braked trailer vehicle (so-called trailer overrun), which is the safety and ride comfort of the semitrailer strongly detrimental.

The brake characteristic C ^* is defined as the quotient of the circumferential force and the applied application force and is also referred to as the internal gear ratio of the brake. The brake characteristic depends, among other things, on the temperature of the brake disk and the brake pads, the brake pressure, the friction speed and thus the speed of the vehicle and the chemical composition, the condition and the thermal preload of the brake pads. For disc brakes, the brake characteristic C ^{* is} equal to twice the coefficient of friction μ _B between the brake pads and the brake disc. In practice, usual values for the coefficient of friction μ _B for disc brakes are between 0.3 and 0.6; Thus, the corresponding brake characteristics C ^* 0.6 to 1.2.

In the method known from the prior art, measurements are made during the driving operation in order to be able to determine brake parameters as a function of the measured input data. According to the DE 196 48 936 A1 In this case, equations are used which result from the equilibrium of forces between towing vehicle and trailer vehicle during a braking operation. According to the DE 100 11 270 A1 This principle is refined with the aid of wheel-specific equations in such a way that a wheel-specific determination of a brake characteristic value is also possible.

Problematic However, in the case of the above-mentioned methods, the determination of a Brake characteristic corresponding measurements to the input data to disposal and do not yet take these measurements at the beginning of a ride to disposal stand because no braking has been done yet. In addition, too not all braking for the Determination of a brake characteristic suitable and it may have several measurements carried out become.

Of these, Based on the object of the invention, a method for the determination the current brake characteristic and an on device for performing this Provide method with which brake characteristics relatively accurate also available at the beginning of the journey can be made.

The Invention solves this problem in that from the discrete brake characteristics at least two maps are generated, from which the current Brake characteristic value is calculated by interpolation. In principle, the discreet brake characteristics both on vehicle-remote test stands, on Test drives with the motor vehicle and during normal driving of the motor vehicle can be determined.

The determination of the discrete brake characteristics on vehicle-remote test stands has the advantage that the input data are relatively easy to control. Thus, the brake system can bring to a certain temperature and / or to certain brake pressures and bring the disc to a corresponding speed corresponding to a certain vehicle speed when using a disc brake. For example, brake pressures and brake temperatures can be used for certain speeds, in particular riiert and for this input data, a map are created in the brake characteristic values are shown as a function of brake pressure and braking temperature for a given vehicle speed. In a further step, additional maps (ie, for example, for other speeds) can be determined.

During the Driving can then depending the operating state of the vehicle and the brake, ie driving speed, Brake pressure and temperature from maps adjacent to the operating state Determined braking characteristics and these for determining the current Brake characteristic value are weighted. In principle, the current Brake characteristic thus from previously known, adjacent brake characteristics interpolated.

Of course, the discrete brake characteristics even during test drives with the motor vehicle be determined. The motor vehicle can be equipped with special measuring devices for this purpose be, for example, with load cells, with which the fitting Braking torque during a braking process can be determined, thereby involving of brake pressure and geometry of the brake the corresponding brake characteristic can be determined.

at Determination of discrete brake characteristics during normal driving it may be useful to "refresh" already existing discrete brake characteristics and thus to brake characteristics, on test stands, during a test drive or during a previous trip during the Normal operation, supplement or replace. By the replacement of only individual discrete brake characteristics is ensured that even when driving and after performing only a few measurements the brake characteristic value can be reliably determined by the interpolation according to the invention is.

The Input data can the vehicle speed, the brake pressure and / or the brake temperature affect. For example, you can for certain Vehicle speeds depending on determined by brake pressure and / or brake temperature brake characteristics be stored in the maps. Be for certain Vehicle speeds maps generated, the brake characteristics dependent on of brake pressure and braking temperature, it is necessary to determine the current brake characteristic of the knowledge of the vehicle speed, around the for select this operating point relevant adjacent maps. It also requires knowledge of brake pressure and temperature to get out the selected one Maps to determine the appropriate reference brake characteristics and these depending on the to weight the current vehicle speed.

It is useful that before implementation checked for interpolation whether the input data are within allowable intervals, to ensure, that at all corresponding brake characteristics are determined from the maps can, yes, only for Input data can be present at certain intervals.

The Maps can for specific Vehicle speeds are generated, for example, for 20, 40, 60, 80, 100 and 120 km / h. Of course, maps can coarser or be determined finer gradation.

For a good Evaluability of the maps, it may be useful that the maps be normalized to the readability by a vehicle computer to facilitate. The readability is also facilitated by that the determined maps do not directly from discrete brake characteristics be constructed, but described by mathematical functions with which the maps can be smoothed.

The Functions can be formed for example by polynomials, with which already a good approach to the map spanned by discrete brake characteristics map is possible. Of course you can too the coefficients of the polynomials are normalized to a lighter one To allow processability of the data in the vehicle computer.

The Maps can adjusted with correction constants and / or factors. This has the significant advantage that once determined maps essentially in their structure, however to current operating conditions are customizable. When applying a correction constant can be move a map in total and apply a correction factor stretch the map. Of course, both methods can be used combine with each other.

It makes sense that an adaptation of the maps only occurs when the calculated correction constants and / or factors within of admissibility intervals lie. This allows Once obtained brake characteristics are retained, if correction constants and / or factors outside an admissibility interval lie and the map are thus changed in an inadmissible way would.

To an embodiment The invention provides the correction constants and / or factors starting from at least two measurements calculated on three Sizes A, B and C are based. Here is a first measurement of the sizes A and B at a certain first state of size C and a second measurement the sizes A and B at a deviating from the first state of the size C second state. For example At a certain speed, the brake pressure and the Brake temperature determined and at one of the first speed deviating second speed again the brake pressure and the Brake temperature. about the balance of forces known in principle from the prior art and / or Radspezifischen equations can now be determined brake characteristics, in turn, the maps of the invention can update. For trucks it is necessary that three measurements be carried out when the semitrailer in total equipped with three axles, each of which must be braked is. For need more axes according to further measurements are carried out.

It It is not necessary that all input data is given by appropriate Sensors are provided. It may also be sufficient that For example, the brake temperature provided by a temperature model becomes.

at execution It makes sense to measure the resulting input data in a memory that uses the FIFO principle (First In / First Out) can work.

Consequently can non-usable measurements, for example, from previous measurements not enough deviate and thus only to a very erroneous determination of a Lead brake characteristic, discarded and other input data stored in memory used to determine the brake characteristic value.

The Invention also relates to a brake control, which is used to carry out the described method is suitable and sensors for detection the input data and / or a computing unit and / or a memory unit can have.

The The invention further relates to motor vehicles and motor vehicle trailers, which are equipped with a corresponding brake control.

Of the The following description and the drawings are exemplary embodiments of the subject invention.

For this embodiments demonstrate:

1 a program flow chart for determining the current brake characteristic value;

2 the equilibrium of forces in the horizontal direction for a two-axle vehicle;

3 the equilibrium of forces in the horizontal direction for a semitrailer; and

4 a program flow chart for the online adaptation of the correction factor.

definition the brake characteristic value

Algorithmus zur Ermittlung des Bremskennwerts aus gemessenen KennfeldernThe brake characteristic C ^* is defined as the quotient of the circumferential force F _U and the applied application force F _S

Algorithm for determining the braking characteristic from measured maps

angenähert werden, um ein Kennfeld in Abhängigkeit des Bremsdrucks p und der Temperatur ϑ zu generieren. Die Koeffizienten α_i werden gemäß dem Gauß'schen Ausgleichsprinzip so bestimmt, dass die Summe der Quadrate der Fehler der einzelnen Gleichungen minimal wird.For a speed, characteristic values determined from measurements, which are determined, for example, at the brake test stand, are to be determined by a function

be approximated to generate a map in dependence of the brake pressure p and the temperature θ. The coefficients α _i are determined in accordance with the Gaussian equation so that the sum of the squares of the errors of the individual equations becomes minimal.

oder ausgeschriebenFor f (θ, p) a biquadratic function is used, ie

or advertised

Thus, nine coefficients α _{j are} to be determined for each speed. The coefficients α ₁ to α ₈ are provided with units so as to calculate the dimensionless quantity C ^* .

• 1. Zuerst wird sichergestellt, dass die drei Parameter innerhalb der Maxima und Minima liegen, für die ein gesichertes Kennfeld vorliegt. Liegt ein Parameter außerhalb dieser Schranken, wird er gleich der Grenze gesetzt. FALLS ν < ν_min DANN ν ≔ ν_min (mit ν_min = 20 km/h) FALLS ν > ν_max DANN ν ≔ ν_max (mit ν_max = 100 km/h) FALLS ϑ < ϑ_min DANN ϑ ≔ ϑ_min (mit ϑ_min = 100°C) FALLS ϑ > ϑ_max DANN ϑ ≔ ϑ_max (mit ϑ_max = 500°C) FALLS p < p_min DANN p ≔ p_min (mit p_min = 0,70 bar) FALLS p > p_max DANN p ≔ p_max (mit p_max = 8,17 bar)
• 2. Über die Geschwindigkeit ν werden die Kennfelder ermittelt, die auszuwerten sind. Die Geschwindigkeit ν wird entweder zwischen zwei Kennfeldern liegen oder genau die Geschwindigkeit eines Kennfeldes haben. Dies kann durch die Überprüfung im ersten Schritt geschehen. Der Ausdruck int(ν/20) liefert nur das ganzzahliege Ergebnis der Division und ignoriert den Rest. FALLS int(ν/20) 1: Berechne C * / 1(ϑ, p) mit den Koeffizienten für ν = 20 km/h Berechne C * / 2(ϑ, p) mit den Koeffizienten für ν = 40 km/h 2: Berechne C * / 1(ϑ, p) mit den Koeffizienten für ν = 40 km/h Berechne C * / 2(ϑ' p) mit den Koeffizienten für ν = 60 km/h 3: Berechne C * / 1(ϑ, p) mit den Koeffizienten für ν = 60 km/h Berechne C * / 2(ϑ, p) mit den Koeffizienten für ν = 80 km/h 4: Berechne C * / 1(ϑ, p) mit den Koeffizienten für ν = 80 km/h Berechne C * / 2(ϑ, p) mit den Koeffizienten für ν = 100 km/h 5: Berechne C * / 1(ϑ, p) mit den Koeffizienten für ν = 100 km/h C * / 2(ϑ, p) = C * / 1(ϑ, p) da int(ν/20) = 5 nur für ν = ν_max gilt ENDE FALLS
• 3. Der letzte Schritt besteht in einer Interpolation, um den Bremsenkennwert
  

zu berechnen.In the following, the calculation of the brake characteristic value is described for the case where the characteristic map was approximated by a biquadratic polynomial at five speeds (20, 40, 60, 80 and 100 km / h). This is for illustrative purposes only and neither the selected speeds nor the equidistant step size are required for the evaluation. As input variables for the algorithm, the speed ν, the temperature θ and the brake pressure p are needed. The brake characteristic value is calculated in three steps (cf. 1 ):

• 1. First, make sure that the three parameters are within the maxima and minima for which there is a saved map. If a parameter is outside these limits, it is set equal to the limit. FALLS ν <ν _min THEN ν ≔ ν _min (with ν _min = 20 km / h) FALLS ν> ν _max THEN ν ≔ ν _max (with ν _max = 100 km / h) FALLS θ <θ _min THEN θ ≔ θ _min (with θ _min = 100 ° C) FALLS θ> θ _max THEN θ ≔ θ _max (with θ _max = 500 ° C) IF p <p _min THEN p ≔ p _min (with p _min = 0.70 bar) FALLS p > p _max THEN p ≔ p _max (with p _max = 8.17 bar)
• 2. Velocity ν is used to determine the maps that are to be evaluated. The speed ν will either lie between two maps or have exactly the speed of a map. This can be done by checking in the first step. The expression int (ν / 20) returns only the integer result of the division and ignores the rest. IF int (ν / 20) 1: Compute C * / 1 (θ, p) with the coefficients for ν = 20 km / h Calculate C * / 2 (θ, p) with the coefficients for ν = 40 km / h 2: Calculate C * / 1 (θ, p) with the coefficients for ν = 40 km / h Calculate C * / 2 (θ 'p ) with the coefficients for ν = 60 km / h 3: Calculate C * / 1 (θ, p) with the coefficients for ν = 60 km / h Calculate C * / 2 (θ, p) with the coefficients for ν = 80 km / h 4: Calculate C * / 1 (θ, p) with the coefficients for ν = 80 km / h Calculate C * / 2 (θ, p) with the coefficients for ν = 100 km / h 5: Calculate C * / 1 (θ, p) with the coefficients for ν = 100 km / h C * / 2 (θ, p) = C * / 1 (θ, p) since int (ν / 20) = 5 only for ν = ν _max is END FALLS
• 3. The last step is an interpolation to the brake characteristic
  

to calculate.

Adaptation of the approximated characteristic diagram

• – Luftwiderstand F_L,
• – Steigungswiderstand F_St,
• – Rollwiderstand F_R, und
• – Motorschleppmoment und die Bremswirkung des Retarders bei einem Fahrzeug mit Dauerbremsintegration (kurz DBI) F_DBI.

The braking force F _{B of} the wheel brakes on the front and rear axle is calculated according to F B = m ges ẍ - F L - F St - F R - F DBI (6) with the vehicle mass m _ges , the deceleration ẍ and the following resistances:

• - Air resistance F _L ,
• - slope resistance F _St ,
• - rolling resistance F _R , and
• - Motor drag torque and the braking effect of the retarder in a vehicle with continuous braking integration (DBI short) F _DBI .

des Fahrzeugs. Die Größen z_ist, z_R, z_L, z_St, z_DBI und m_ges liegen innerhalb der elektropneumatischen Bremsanlage vor (vgl. DE 44 43 522 A1 und DE 44 44 650 A1 ).Usually, the delays are given in percent of the gravitational acceleration g and denoted by z. The delay of the wheel brakes z B = z is - z R - z L - z St - z DBI (7) is calculated from the measured delay

of the vehicle. The variables z _are , z _R , z _L , z _St , z _DBI and m _tot are within the electropneumatic brake system (see. DE 44 43 522 A1 and DE 44 44 650 A1 ).

The problem with determining the brake characteristic values is that only the total braking force of the wheel brakes, ie the sum of the braking forces on the individual axles, can be determined from the deceleration. The deceleration of the wheel brake z _{B is} calculated according to equation (7) above and is generated by the braking forces of the wheel brake F _B on the front and rear axles.

In 2 the braking forces of the wheel brake and the resulting from the deceleration of the wheel inertia force are located. The forces in the vertical direction, such as the weight and the axle loads, are not included for clarity.

According to d'Alembert's principle, the equilibrium of forces in the horizontal direction results for a two-axle vehicle F BVA + F BHA = m ges z B g (9)

Out the braking pressure, the braking force of the wheel as unknown Parameters are determined. The brake characteristic is for each axis determined as the average value of the left and right brakes.

mit dem Wirkungsgrad η, der Übersetzung i, der Nennkraft F_N, dem Nenndruck p_N, dem Anlegedruck p_A, dem wirksamen Radius r_WS und dem dynamischen Reifenradius r_dyn. Eingesetzt in Gleichung (9) ergibt sich eine Gleichung

mit den Bremsenkennwerten an der Vorder- und Hinterachse als Unbekannte. Die Quasikonstanten a_VA und a_HA können auch für andere Bremsen bestimmt werden und das Verfahren ist nicht nur auf diese Scheibenbremse beschränkt.The braking forces at the front and rear axles are calculated, for example, for a disc brake

with the efficiency η, the ratio i, the nominal force F _N , the nominal pressure p _N , the application pressure p _A , the effective radius r _WS and the dynamic tire radius r _dyn . Substituted in equation (9) gives an equation

with the brake characteristics at the front and rear axle as unknown. The quasi-constants a _VA and a _HA can also be determined for other brakes and the method is not limited to this disc brake.

Adaptation by a correction factor

From the temperature of the brake disk θ, the brake pressure p and the speed ν of the two braking operations, the braking characteristics C ^* = f (θ, p, ν) for the front and rear axle are obtained from the characteristic map. The function f (θ, p, ν) can be the approximate characteristic map described above or any other function that describes the brake characteristic as a function of the parameters temperature, pressure and velocity. The brake pressures and temperatures at the front and rear axles are measured at two speeds or determined by a temperature model.

It It is now assumed that the brake characteristic value is the same as that determined Characteristic map behaves and differs only by a correction factor k from this.

This results in a system of equations with the two unknown correction factors

From two measurements, the correction factors k _VA and k _HA can be calculated according to equation (13). It is checked whether the correction factors are within limits k _min and k _max to be specified, k min ≤k≤k Max , (14)

So For example, the correction factors are only plausible if they are are positive. The correction factors thus obtained are collected, until a certain number has been calculated. Then the arithmetic Mean calculated and set as a new correction factor or weighted with the previous factor. This correction happens with one factor per axis. This is just a storage of one Correction factor per axis necessary while the coefficients of the on the brake test bench determined map unchanged stay.

Another possibility is the adaptation of the correction factor after each evaluation. The correction factor after the adjustment k ⁺ becomes from the correction factor before the adaptation k ^- with the weight w (0 ≦ w ≦ 1) and the calculated correction factor k after k + = wk - + (1 - w) k (15) determined. It is also useful for the correction factor after the adaptation k ⁺ limits k + / min and k + / max with k + min ≤ k + ≤ k + Max (16) set. While the limits k _min and k _max can be selected generously right, since averaging is performed on the weighting, the limits for the correction factor after adjustment k ⁺ should closely to get voted.

Adaptation of the Characteristic map by a correction constant

A another possibility for adapting the map is a correction constant c to use. The calculation of the correction constants happens analogous to the determination of the correction factor.

die in Gleichung (11) eingesetzt das lineare Gleichungssystem

ergeben. Die berechneten Korrekturkonstanten c_VA und c_HA werden nach einer Plausibilitätsprüfung gesammelt oder direkt, wie für den Korrekturfaktor beschrieben, zu einer neuen Korrekturkonstanten verarbeitet. Auch hier gilt, dass zwischen den Grenzen (c_min und c_max) der zur Korrektur verwendeten Konstanten und der. Grenzen (c + / min und c + / max) der korrigierten Konstanten unterschieden wird. Auch für die Adaption durch eine Korrekturkonstante ist je ein Wert pro Achse zu speichern, während die am Bremsenprüfstand ermittelten Koeffizienten unverändert bleiben. Im Folgenden wird der Begriff "Korrekturgröße" als Oberbegriff für den Korrekturfaktor und die Korrekturkonstante verwendet. Die Lösung des Gleichungssystems mit

kann direkt angegeben werden:

First, the brake characteristics for two different states are determined from the map. It is now assumed that the brake characteristic value behaves approximately like the determined characteristic diagram and differs from it only by a correction constant c. This results in the two conditions

in equation (11) used the linear system of equations

result. The calculated correction constants c _VA and c _HA are collected after a plausibility check or processed directly as described for the correction factor to a new correction constant. Again, between the limits (c _min and c _max ) the constants used for the correction and the. Limits (c + / min and c + / max) of the corrected constant is distinguished. Also, for the adaptation by a correction constant, one value per axis is to be stored, while the coefficients determined at the brake test stand remain unchanged. In the following, the term "correction quantity" is used as a generic term for the correction factor and the correction constant. The solution of the equation system with

can be specified directly:

More options for the adaptation of the characteristic field

For the adaptation of the map further possibilities are conceivable. So a combination of the two methods described above is possible, which is a correction of the map C * = kf (θ, p, ν) + c (21) by a factor k and a constant c permits. However, this would mean that a system of equations with four equations and four unknowns would have to be solved for two characteristics (two-axle vehicle). For a semitrailer, a system of equations with six unknowns would have to be solved, which leads to difficulties in implementation in the vehicle computer.

Likewise, the correction quantities could depend on one or more parameters and be calculated for different speeds ν, for example. Thus, it is only assumed that the brake characteristic value behaves the same in the temperature and pressure parameters except for the correction quantity. At constant speeds, the correction values are given C * = k (ν) f (θ, p, ν = const.) or C * = f (θ, p, ν = const.) + c (ν). (22)

This means that the two measurements must be performed at approximately the same speeds, which limits the readability. Of course, this limitation is increased if the correction quantities are adjusted depending on two parameters. Should the correction values be included For example, depending on the temperature and speed C * = k (θ, ν) f (θ = const., p, ν = const.) (23) or C * = f (θ = const., p, ν = const.) + c (θ, ν) (24) adjusted, two equations with approximately the same temperature and speed must be found. The next step, evaluating for a particular state, is not possible when using the equation system. For the brake pressures, or more precisely the ratio of the brake pressures, of the two measurements must be different in order to obtain a non-singular or almost singular coefficient matrix.

It It is also conceivable to use more than two measurements to calculate the Use correction factors or constants and the correction values an over-determined one Equation system with the linear compensation calculation. However, this is for the following reasons for the Implementation in the vehicle computer does not make sense. On the one hand, the Arithmetic operations for the implementation in the vehicle computer does not become too complicated, so that a simple algorithm is sought. The other is the system resulting from two measurements is poorly conditioned and this is reinforced by the linear adjustment calculation.

system of equations for the Tractor

The algorithms described for a two-axle vehicle are also applicable to a semitrailer, ie towing vehicle, by an extension of the system of equations. For a semitrailer, in turn, only the entire braking force is known and it must be distributed to the front and rear axles of the towing vehicle, as well as the trailer vehicle (see 3 ): F BVA + F BHA + F Banh = m ges z B G. (25)

The deceleration of the wheel brake z _{B is} calculated as in the tractor alone according to equation (7). The braking forces on the n _Anh semi-trailer _{axles are} combined to form a force F _BAnh , whereby the same equipment is assumed on the trailer _axles .

mit drei unbekannten Bremsenkennwerten für den Sattelzug. Aus drei Messungen ergibt sich ein Gleichungssystem

mit neun unbekannten Bremsenkennwerten. Damit werden sechs zusätzliche Bedingungen benötigt, damit das Gleichungssystem lösbar ist. Wird die oben diskutierte Annahme eines Korrekturfaktors k oder einer Korrekturkonstante c eingeführt, ergibt sich analog zum zweiachsigen Fahrzeug ein Gleichungssystem mit drei Gleichungen und drei Unbekannten.This results in an equation

with three unknown brake characteristics for the semitrailer. Three measurements result in a system of equations

with nine unknown brake characteristics. This requires six additional conditions for the system of equations to be solvable. If the above-discussed assumption of a correction factor k or a cor introduced rectification constant c, analogously to the biaxial vehicle results in a system of equations with three equations and three unknowns.

für die drei unbekannten Korrekturfaktoren der einzelnen Achsen. Erfolgt die Adaption des Kennfeldes durch eine Korrekturkonstante für jede Achse mit C*(i) = f(ϑ(i), p(i), ν(i)) + c (30)führt dies auf folgendes Gleichungssystem:

The assumption C * (I) = kf (θ (I) , p (I) , ν (I) ) (28) leads to a system of equations

for the three unknown correction factors of the individual axes. If the adaptation of the characteristic field is carried out by a correction constant for each axis C * (I) = f (θ (I) , p (I) , ν (I) ) + c (30) this leads to the following system of equations:

Algorithm for online adaptation through the equation system

für den Einfluss von Fehlern ΔA in der Koeffizientenmatrix A und Δb in der rechten Seite b auf den relativen Fehler der Lösung eines linearen Gleichungssystems. Diese Abschätzung gilt nur für die betragsgroßen Komponenten von x. Die relativen Fehler betragskleiner Komponenten von x können jedoch viel größer sein.The estimation applies

for the influence of errors ΔA in the coefficient matrix A and Δb in the right side b on the relative error of the solution of a system of linear equations. This estimate applies only to the amount-sized components of x. However, the relative error amounts of small components of x may be much larger.

from that follows for the mapping matching algorithms that the condition number of the Coefficient matrix as possible must be small. This requirement corresponds to different pressure conditions in the measurements to be evaluated. The relative errors in the coefficient matrix correspond to relative errors in determining the brake pressure. To minimize the relative error in the coefficient matrix To keep small, the brake pressure must be well above the application pressure of the diaphragm cylinder lie. Errors of the right side are errors in the vehicle mass and the calculation of the delay through the wheel brake. Therefore, when calculating the delay must be on it be respected that one enough greater Share of the wheel brake is applied and not by others Shares, such as those of the Dauerbremsintegration.

For the individual measurements, this results in a condition for the brake pressure p on each axis p> p min (33) and a condition for the deceleration by the wheel brake z B > z Bmin , (34)

The requirement for a small conditioning is to be tested for a combination of two measurements: cond (A) <cond Max , (35)

This corresponds to a redistribution of brake pressures and thus the braking forces between the axes.

For the implementation in the vehicle computer can not as many measurements are collected and against each other be evaluated. Therefore, a sequential storage of the stationary Braking phases according to the FIFO principle advantageous. It becomes a field created, with only the beginning of data removed and inserted at the end can. This field is called a queue or queue.

The Conditions for the brake pressures and the delay the wheel brake can for every single stationary Braking phase to be checked while for the Conditioning always a combination of two stationary braking phases checked becomes. So only stationary Brake phases saved the conditions for brake pressures and the delay fulfill.

To Beginning of the ride contains the queue is not an item and the weighted correction values are initialized with k + / start = 1 or c + / start = 0. The first stationary braking phase, the conditions for the brake pressures and the delay Fulfills, is saved. If a second stationary braking phase is found, the pressure and delay criteria enough, this is stored and the conditioning is checked. Is the condition number less than the required limit, the system of equations becomes Calculation of the correction quantities for the first time evaluated. The correction values are adjusted with the weight w. Will the nth stationary braking phase stored, become n - 1 Combinations with the already found stationary braking phases checked and, where appropriate evaluated. For Each evaluated combination is weighted by the Correction variables.

The maximum number of stationary braking phases stored is limited to n _max . If the field is occupied by n _max stationary braking phases and another is found, the braking phase is deleted at the beginning according to the FIFO principle and the new one added at the end. Then the n _max - 1 combinations with the other stationary brake phases are checked for readability. If the equation system is evaluated, a weighted adaptation of the correction quantities takes place. The storage of the stationary braking phases is found or not found for the two cases, maximum number of stationary braking phases, as shown below.

On the one hand, the number n _{max of} the maximum stored stationary braking phase must not be too small, since otherwise not enough evaluable combinations will be found. On the other hand, not too many stationary brake phases must be stored, since n _max - 1 combinations must be checked for readability. The algorithm can be summarized as follows (see 4 for the adaptation of the characteristic field by a correction factor):

Initialize the weighted correction values

With This algorithm allows the calculation of the correction quantities in the Vehicle computer can be implemented. For the construction of the coefficient matrix A is the evaluation of the calculation of the correction factor Kennfeldes necessary, as previously described. At the calculation the correction constant is done the evaluation of the map for the Construction of the right side b.

Kombinationen zu überprüfen, ob die Konditionierungsbedingung erfüllt ist.For the semitrailer, this algorithm can be used in slightly modified form. In addition to the brake pressure at the front and rear axle, the brake pressure on the semi-trailer must also be above the required minimum pressure. If three steady-state braking phases have been found that meet the conditions, the first evaluation takes place. For each new found stationary braking phase n are

Combinations to check if the conditioning condition is met.

Claims (30)

Method for determining the current brake characteristic of brake devices of a motor vehicle and / or a motor vehicle trailer, wherein discrete brake characteristics are determined in dependence on input data, characterized in that the following steps are performed: a) at least two maps are generated from the discrete brake characteristics, b) off the maps, the current brake characteristic value is calculated by interpolation. Method according to claim 1, characterized in that that the determination of the discrete braking characteristics on remote vehicle test stations he follows. Method according to claim 1 or 2, characterized that the determination of discrete braking characteristics during test drives done with the motor vehicle. Method according to one of claims 1 to 3, characterized that the determination of discrete brake characteristics during the normal driving operation of the motor vehicle takes place. Method according to one of the preceding claims, characterized characterized in that the input data is the vehicle speed, pertain to the brake pressure and / or the brake temperature. Method according to one of the preceding claims, characterized characterized in that before implementation checked for interpolation determines whether the input data is within allowable intervals. Method according to one of the preceding claims, characterized characterized in that maps for specific vehicle speeds be generated. Method according to claim 8, characterized in that that the maps for specific vehicle speeds Brake parameters depending on represent brake pressure and brake temperature. Method according to one of the preceding claims, characterized characterized in that the maps are normalized. Method according to one of the preceding claims, characterized characterized in that the determined maps by mathematical Functions are described. Method according to claim 10, characterized in that that biquadratic functions are used. Method according to claim 10 or 11, characterized that the functions are formed by polynomials. Method according to claim 12, characterized in that that the coefficients of the polynomials are normalized. Method according to one of claims 12 or 13, characterized that coefficients are stored in a matrix. Method according to one of the preceding claims, characterized characterized in that the maps with correction constants and / or - factors be adjusted. Method according to claim 15, characterized in that that a weighting of the correction constants and / or factors he follows. Method according to claim 15 or 16, characterized that an adaptation of the maps only occurs when the calculated Correction constants and / or factors within admissibility intervals lie. A method according to any one of claims 15 to 17, characterized in that the correction constants and / or factors based on at least two measurements based on three sizes A, B and C, wherein a first measurement of the sizes A and B takes place at a certain first state of size C and a second Measurement of the sizes A and B at a different from the first state of the size C second state he follows. Method according to claim 18, characterized that the sizes A, B, C or A, C, B or B, A, C or B, C, A or C, A, B or C, B, A the three sizes brake pressure, Brake temperature, vehicle speed correspond. Method according to claim 18 or 19, characterized that for the evaluation of the measured quantities A, B and C conditions are checked the conditions are dependent of limits for the measured quantities and / or the condition number of the coefficient matrix for a combination of several Measurements are defined. Method according to one of claims 18 to 20, characterized that the input data resulting from the measurements in one Memory are stored. Method according to claim 21, characterized that the filing of the data takes place according to the FIFO principle. Method according to one of the preceding claims, characterized marked that for Tractors three Measurements performed become. Method according to one of the preceding claims, characterized characterized in that a temperature model is used. Method according to one of the preceding claims, characterized in that the formula for calculating the current brake characteristic value C ^* is:

with C ₁ : reference braking characteristic at a first vehicle speed, C ₂ : reference braking characteristic at a second vehicle speed, ν: current vehicle speed and s: increment between the maps for the first and second vehicle speed. int (): integer result of the division Brake control used to carry out the method after a of the preceding claims is. Brake control according to claim 26, characterized in that that the brake control sensors to capture the input data and / or a computing unit and / or a memory unit. Motor vehicle following with a brake control Claim 27 equipped is. Their trailers for a Motor vehicle according to claim 28. Its trailers, which is equipped with a brake control according to claim 27.