EP2576310A1

EP2576310A1 - Method and device for detecting undesired drive train reactions of a motor vehicle having at least one drive unit

Info

Publication number: EP2576310A1
Application number: EP11712844.7A
Authority: EP
Inventors: Jens-Werner Falkenstein
Original assignee: Robert Bosch GmbH
Current assignee: Robert Bosch GmbH
Priority date: 2010-06-04
Filing date: 2011-04-05
Publication date: 2013-04-10
Also published as: US20130138290A1; EP2660118B1; WO2011151094A1; US8983715B2; DE102010029706A1; EP2660118A3; EP2660118A2

Abstract

The invention relates to a method for detecting undesired drive train reactions of a motor vehicle having at least one drive unit, in which method at least one input variable (M_Des) of the motor vehicle and/or of the drive unit (1) is input into the drive train and at least one output variable (ω_EIM) is measured at the motor vehicle and/or the drive unit (1). In order to detect a fault situation at an early time, the at least one input variable (M_Des) is fed to a dynamic model which at least partially models the drive train of the motor vehicle, wherein the dynamic model determines, on the basis of the at least one input variable (M_Des), at least one model output variable (ω_EIMObs) which is compared with the at least one measured output variable (ω_EIM), wherein when there is a difference between the measured output variable (ω_EIM) and the model output variable (ω_EIMObs) it is inferred that an undesired drive train reaction has occurred.

Description

_

description

Method and device for detecting unwanted driveline reactions of a motor vehicle with at least one drive unit. The invention relates to a method for detecting unwanted driveline reactions of a motor vehicle with at least one drive unit, wherein at least one input of the motor vehicle and / or the drive unit is input to the driveline and at least one output variable is measured on the motor vehicle and / or the drive unit and a device to carry out the process.

In drive trains of motor vehicles, direct measurement of currently generated drive torques is usually not provided for reasons of cost. Due to the use of an electronic accelerator pedal, there is no mechanical connection to the drive unit. In software errors or, for example, due to disturbances in the signal transmission between the accelerator pedal and the drive unit driving the control unit, which receives the driver desired desired torque and passes it to the drive unit, but it may come to unintentional torque or power outputs of the drive unit. In particular, in hybrid, hydraulic or electric vehicles, a frictional connection between a drive unit and the drive wheels can be present even when the vehicle is stationary. These unwanted moments or

Power output of a power plant, e.g. due to an error in the data communication or a component failure can then cause a safety-critical movement of the vehicle, without the driver wants it.

Such unwanted torque and power outputs are not only in hybrid, electric or hydraulic vehicles but also in vehicles with conventional engine drive and must therefore be monitored in these.

Disclosure of the invention The inventive method for detecting unwanted

Drive train reactions of a motor vehicle with at least one

Drive unit with the features of claim 1 has the advantage that by means of the dynamic model early recognition of a

Torque and / or power output by the drive unit is possible. Characterized in that at least one input variable of the motor vehicle and / or the drive unit (for example, a torque or a power) is supplied to the drive train of the motor vehicle at least partially replicating dynamic model, wherein the dynamic model based on the input determines at least one model output variable, which with is compared with at least one measured output, wherein a difference between the measured output and the

Model output on an unwanted driveline reaction is closed, a detection and possibly a minimization of unwanted

Drive train reactions of the motor vehicle take place. This will be negative

Impacts such as avoiding or at least minimizing unwanted motion of the vehicle, thereby increasing vehicle safety. By using the dynamic model, errors in the

Drive train very quickly, for example, in the millisecond range, are detected. As a result, an immediate reaction of the motor vehicle, for example, by operating the brakes on these errors is possible to safely prevent the movement of the motor vehicle.

In one embodiment, in the presence of a plurality of measured output variables and a plurality of model output variables one of the measured

Output variables compared with an assigned model output.

This increases the variety of possible comparisons of the measured variables, which leads to the fact that the reliability of the comparison is increased. Advantageously, a disturbance of the drive train of the motor vehicle is estimated in the dynamic model. Errors in the real drive beach have an impact on the real drive train. Since disturbances acting on the real system of the drive train can not be measured on the real system of the drive train, it is possible in the present invention to evaluate the disturbance variable via the dynamic model, wherein a

Determine the magnitude of the disturbance reliably. Since the disturbance variable acting only on the real system of the drive train makes a difference between that measured on the drive train, on the motor vehicle and / or on the drive unit Output variable and the model output of the dynamic model generated, the disturbance can be determined mathematically.

In one embodiment, the value of the disturbance variable of the drive train of the

Motor vehicle in the dynamic model by methods of

Disturbance monitoring estimated. With this in itself from the measuring and

Control technology known disturbance estimation becomes the reality of

Driveline modeled or estimated so that the

dynamic model with reality in its dynamic dimensions

matches.

In a further development, the value of the disturbance variable of the drive train of

Motor vehicle in the dynamic model compared with at least one threshold, wherein when exceeding a first threshold or

Falling below a second threshold by the value of the disturbance at least one drive unit of the motor vehicle is affected or turned off. Through the disturbance estimation in the dynamic model can

safety-critical situations are reliably detected in the vehicle. In one variant, the threshold value depends on an operating state of the drive train. For example, at very dynamic and / or magnitude high target torque / target power at detected road bumps, when detected frosted road surface, recognized driveline vibrations, braking, ABS or ESP interventions or increases in gear shifts the first threshold.

Furthermore, the estimated value of the disturbance variable is fed back to the part of the dynamic model which is at least partially replicating the drive train of the motor vehicle. Through this feedback, an approximation between the dynamic model and the real behavior of the drive train takes place.

Advantageously, a torque and / or a power of

Drive unit used as input and / or estimated as a disturbance used. Torque and power of the drive unit are typical

Parameters for an intended change in the behavior of the

Drive unit, which is transmitted to the motor vehicle. Become one

Torque or an estimated power used, then the Momentenbzw. Power influence of the drive train or a shutdown of the at least one drive unit or an activation of a brake system whenever a deviation of the torque or the power from a target torque or a target power becomes greater than an allowable deviation.

In one embodiment, the torque and / or the power of the act

Drive units on the inertia of the moving parts of the at least one drive unit. For example, an air gap moment acts on the rotor of an electric machine or moments due to the gas and friction forces on the crank mechanism of an internal combustion engine. For these variables, setpoints are usually specified by a controller, which are suitable for direct comparison with determined or estimated actual values.

In a further development, the dynamic model uses or estimates

Target torque, a target power, a calculated actual torque and / or a calculated actual power of the drive unit. The dynamic model is therefore designed so that it can record and reproduce variables recorded or emitted by the real drive train dynamically. This gives an identity between the real powertrain and the dynamic model.

In another variant, the dynamic model simulates a drive dynamics of the at least one drive unit and / or further units. The other units may be additional drive units such as electric machines, hydraulic and internal combustion engines or units such as clutches, torque converters, lock-up clutches and gearboxes that respond dynamically to a control.

Furthermore, the dynamic model simulates the aggregate limits of the at least one drive unit and further units.

In particular, the dynamic model forms time delays in one

electronic signal transmission after and / or compensates for this. Thus, the measured output as well as the model output determined by the dynamic model can be compared as if they had occurred at the same time.

Advantageously, the dynamic model simulates torques and / or performances of mechanically driven ancillary components. To such Ancillaries include power steering pumps, air conditioning compressors,

Coolant pumps and the like. The consideration of the dynamic

Behavior of these ancillaries allows a very accurate replica of the vibration behavior of the drive train of the motor vehicle.

In a further embodiment, the dynamic model mass inertia of the rotating parts of the mechanically driven ancillaries after. These replicas also contribute to a realistic representation of the dynamic model.

In a further development, the dynamic model forms elasticities and / or mechanical play in the suspensions of the drive unit and / or in the suspensions of further units and / or in a chassis and / or in the chassis

Driveline at least partially after. The consideration of, for example, mechanical play and elasticities in the suspensions of the aggregates or

Wheel suspensions and elasticities of the tires improves the simulation of the vibration behavior of a motor vehicle.

Advantageously, the dynamic model or the parameters of the dynamic model based on a time course of at least one

Input and / or adapted at least one measured output. This takes account of wear and aging effects or temperature dependencies of the parameters. A development of the invention relates to a device for detecting unwanted driveline reactions of a motor vehicle with at least one drive unit, wherein at least one input of the

Motor vehicle and / or the drive unit is input to the driveline and at least one output on the motor vehicle and / or the

Drive unit is measured. In order to detect an unwanted drivetrain reaction at an early stage and thus reliably prevent a significant vehicle acceleration due to the unwanted driveline reaction, means are provided which feed the input quantity to a dynamic model at least partially replicating the drivetrain of the motor vehicle, wherein the dynamic model determines a model output variable based on the input variable , which is compared with the measured output, wherein an unwanted driveline response is inferred with a difference between the measured output and the model output. This is the dynamic model calculated from the real drive train. The dynamic model receives this

Input variables and output variables of the motor vehicle and / or the

Drive units, i. magnitudes determined on the real driveline, e.g.

Target torques and / or engine speeds and / or wheel speeds. In error-free operation, the dynamic model behaves just like the real one

Driveline, i. the state of vibration on the real drivetrain coincides with the vibrational state of the dynamic model. Due to the replica of the real drive train through the dynamic model, the error case is detected very early, which is possible, for example, within 5 ms. At this time, the motor vehicle still experiences no appreciable acceleration, so that a movement or unwanted acceleration of the

Motor vehicle is safely prevented by initiation of security measures.

Advantageously, a first control device comprising the dynamic model is connected to a second control device via a communication device transmitting at least one input variable and / or at least one measured output variable. With two control units, each of the two can initiate a safety measure in the event of a fault. This results in redundancy and a high degree of safety, even in the case of a defect in one of the two control units.

In another variant, the first control device comprises a first dynamic model and the second control device comprises a second dynamic model, wherein the first control device is connected to a power supply, while the second control device is connected to the drive unit. This has the advantage that two Abschaltpfade are independent of each other, creating a high degree of redundancy and security. In particular, when a control unit is disturbed by a software or hardware error.

In one variant, the first control unit is designed as a vehicle control unit and the second control unit as a drive unit control unit. As a result, a redundancy is present, wherein in case of disturbances of the vehicle control unit, the determination of the dynamic model can be interrupted, whereby the drive unit control unit can control the drive unit without adverse effect and regulate or switch off.

The invention allows numerous embodiments. One of them will be explained in more detail with reference to the figures shown in the drawing. It shows

Figure 1: a simplified model of a drive train

FIG. 2: Replica of the drive train according to FIG. 1

FIG. 3 shows a signal flow diagram for an embodiment of the invention

inventive method

FIG. 4: simulation results for a starting process of the

motor vehicle

Figure 5: an embodiment of the device according to the invention.

Identical features are identified by the same reference numerals.

FIG. 1 shows a simplified model of a drive train which comprises an electric machine 1 which drives the drive wheels 5, 6 of the motor vehicle via a transmission 2 and side shafts 3, 4. The gear 2 is not switchable, so it contains only one gear and provides permanent adhesion between the

Electric machine 1 and the drive wheels 5, 6 ago. At a standstill

Drive wheels 5, 6 is the rotor of the electric machine. 1

The basic idea of the present invention is that a dynamic model or partial model of the oscillatory drive train of the

Motor vehicle is calculated.

A vibratory replica of the electric drive train of Figure 1 is shown in Figure 2. The rotational inertial masses of the electric machine 1 and the transmission 2 are modeled in the replacement torque EIM. The

Ersatzdrehmasse EIM is a torsion spring / damper element 7 with a

Ersatzdrehmasse 8 of the motor vehicle connected, which the reduced inertia of powertrain parts, wheels and the translationally moving

Vehicle mass simulates. The torsion spring / damper element 7 simulates elasticities in the drive train, which are usually characterized by the stiffnesses of the side shafts 3, 4. For the sake of simplicity, all inertias, angular velocities and torques as well as other parameters with the real translations of the Gear 2 and the tire radii converted and reduced, so that in Figure 2 of translations i = 1 is assumed.

The replacement torque EIM of the electric machine 1 and of the transmission 2 shown in FIG. 2 rotates at the angular speed CUEIM, which is determined by a speed sensor, not shown, and reported to a vehicle control. The Ersatzdrehmasse EIM has the moment of inertia J _E IM, which is composed of the inertia of the rotor of the electric machine 1 and rotating parts of the transmission 2. The replacement rotational mass EIM is affected by the air-gap torque M _E IMAG generated by the electric machine 1 and by the torsion spring

/ Damper element 7 currently transmitted torque M _S D-

The Ersatzdrehmasse 8 forms with the moment of inertia J _{Veh of} the motor vehicle further drive train parts, wheels and the translationally moving vehicle mass and rotates with the angular velocity cc _eh - The angular velocity o _eh can in real drive, for example from the signals of Raddrehzahlgeber at the two drive wheels by averaging and under consideration of

Gear ratio can be calculated. The angular velocity cc _eh is available in the vehicle control. On the Ersatzdrehmasse 8 act of the

Torsion spring / damper element 7 currently transmitted torque M _S D and a

Travel resistance torque M _D , which simulates the roll, air and pitch resistance.

FIG. 3 shows a signal flow diagram for an embodiment of the method according to the invention. In the upper part A of Figure 3 is the real oscillatory

Drive train simulated. A block 9 specifies a desired torque M _Des , which is determined from the position of an accelerator pedal and / or predetermined by a driver assistance system or a driving stability system. In general, the desired torque M _{Des can} also be influenced by an idling control or an automated transmission. The desired torque M _Des can depend on

Operating state of the drive train shaped, e.g. low-pass filtered or

gradient-limited so as not to excite driveline vibrations.

In block 10, dynamic effects in the driving behavior of the electric machine 1 are stored, e.g. resulting from a current control and inductances. in the

Below is assumed by a first-order delay transmission (PT1) of a target torque M _DeS EiM on the air gap torque M _E IMAG. The Air gap torque M _E IMAG corresponds to the real-acting actual torque of the electric machine 1 and follows the setpoint torque M _DeS EiM delayed.

Acting on the compensation rotational mass EIM generated by the electric machine 1 the air gap torque M _E IMAG and the currently transmitted from Torsional / damper element 7 moment M _S D _E M IMAG Both torques and M _S D are in

Node 1 1 summarized. By dividing by the moment of inertia J _E IM of the spare rotational mass EIM in block 12 and integration in block 13, the angular velocity CÜEIM of the replacement rotational mass EIM is obtained. A corresponding procedure in the blocks 16 and 1 7 leads to the angular velocity cc _eh the

Replacement rotary mass 8. The currently transmitted from Torsional / damper element 7 moment M _S D is determined in block 14, the running resistance torque M _D is simulated in block 15th The lower part B of Figure 3 shows a possible embodiment of

inventive method with the dynamic model or partial model of the drive train. The determined on the real drive train

Angular velocities CÜEIM of the replacement torque EIM and o _eh the

Spare torque 8 and the desired torque M _Des are reported to the dynamic model. In block 1 8, the drive dynamics of the electric machine 1 as

First order delay modeled. The thus calculated modeled

Air gap torque M _E iMAGObs is summarized with a torque M _S DObs of the torsion spring / damper element modeled in block 19 at node 20 and divided by the estimated moment of inertia J _E iMOb _{S of} the replacement rotational mass EIM (block 21). An integrator 22 calculates therefrom an estimate cctiMObs for the

Angular velocity CC ^ IM of the replacement torque EIM. The estimated

Angular velocity cctiMObs of the replacement torque EIM and the

Angular velocity cc _before replacement torque 8 are used by a model of the torsion spring / damper element shown in block 19 to determine the modeled moment M _S DObs.

In ideal conditions fall the determined on the real drive train

Angular velocity CC ^ IM and the estimated angular velocity cctiMObs of the replacement rotational mass EIM together. In the case of undesired torque delivery of the electric machine 1, for example due to an error in the data communication or a software or component error, the two variables cetiM and cetiMObs deviate from one another. An undesired torque output of the electric machine 1 acts only on the real drive train and thus on the determined angular velocity CO ^ M, but not on the dynamic model and therefore not on the estimated

Angular velocity (OeiMObs.) An undesirable moment release of the

Electric machine 1 is in Figure 3 by a disturbance torque M _z in block 23rd

simulated. The disturbance torque M _z acts on the real drive train, but is not measured there.

In order to estimate the disturbance torque M _z , the difference Acc bs of the two variables CUEIM and coeiMObs in the node 24 is formed in the sense of a disturbance observer and via a proportional-integral feedback 25 as disturbance observer correction torque AM ₀ b _S the modeled air gap moment M _e iMAGObs in

Junction 28 switched. Thus, the estimated angular velocity GtiMObs is introduced to the determined angular velocity CÜEIM. The disturbance observer correction torque AM ₀ b _S required for this purpose then corresponds to an estimated value for the disturbance torque M _z .

It is assumed that the disturbance torque M _z on the real drive train with the drive dynamics of the electric machine 1 (block 10) acts on the air gap torque MEIMAG, so it is switched before the block 10 in the node 27. The disturbance observer correction torque AM ₀ bs is intended to provide an estimate for the

Represent disturbance torque M _z , so approximately follow the disturbance torque M _z . Since the disturbance torque M _z acts in front of the drive dynamics of the electric machine 1, the disturbance observer correction torque AM ₀ b _{S is} for the most part (for example 80%) applied to the model at node 28 before the modeled drive dynamics (block 18). The two quantities M _z and AM ₀ b _S can be compared directly in the simulations.

To avoid swinging of the disturbance observer, a small part of the disturbance observer correction torque AM ₀ b _S (for example 20%) behind the modeled drive dynamics after block 18 at node 26 becomes

Disturbance observer path switched on. This has a dampening effect. In order to avoid stability problems by the interaction of an integrating behavior in the model of the torsion spring / damper element (block 19) with the integral part of the disturbance observer correction torque AM ₀ b _S , a stabilizing feedback (block 29) is used for the integral part.

The disturbance observer correction torque AM ₀ bs is compared with torque thresholds and leads to the torque reduction or shutdown of Electric machine 1, if it exceeds an upper threshold above or below a lower threshold. An intervention in the brake system is possible in this case. The torque thresholds are modified depending on the operating state of the drive, eg widened at very dynamic and / or magnitude high target torque M _Des .

In the dynamic model, the setpoint torque M _{Des is} limited to the torque limits of the electric machine 1, since the real electric machine 1 can only set torques within its aggregate limits. Such a limitation is not shown in FIG. 3 for the sake of simplicity.

In the present example, the vibration capability of the real drive is utilized. In the case of an undesirable torque output of the electric machine 1, the angular velocity CUEIM determined on the real drive train reacts even before there are appreciable effects on the angular velocity cc _eh

Ersatzdrehmasse 8 or the vehicle movement result. This allows an early reaction. The method is particularly effective at low speeds or in vehicle standstill and large gradients or very dynamic behavior of the disturbance torque M _z , which represent a high security risk. On a slowly rising with low gradient disturbing moment

M _z , the driver can respond independently.

The parameters of the real drive train may not be known exactly or change over the lifetime. To create realistic conditions, the delay (PT1 time constant) of the modeled drive dynamics is the

Electric machine 1 (block 18) compared to the real driving dynamics of the

Electric machine 1 (block 10) reduced by 25%. The spring stiffness and

Damping in the model of the torsion spring / damper element (block 19) are compared to the real torsion spring / damper element / (relative block 14) reduced by 10%. In operation, an adaptation of the model parameters to the parameters of the real drive makes sense, a deviation of 10% therefore seems realistic.

FIG. 4 shows simulation results for a starting process of the vehicle with a setpoint torque M _Des ramp-shaped by the driver. The

_{Disturbance observer} correction _torque AM _0bS falls due to the above

described parameter deviation initially something. At time t = 0.2 s, the disturbance torque M _z jumps from 0 Nm to 150 nos. A corresponding disturbance can occur on the real drive, eg due to an error in the data communication or a component failure arise. The reaction of the disturbance observer correction torque AMo _b s, which represents an estimated value for the disturbance torque M _z , can be clearly seen. For example, the electric machine 1 can be switched off as soon as the disturbance observer correction torque AM ₀ b _S exceeds a torque threshold of 40 Nm. The shutdown command can be less than 5 ms after the

Fault be set, ie in a period in which the angular velocity cc e _h and thus the vehicle movement has not yet responded to the fault. FIG. 4 shows the curves of the angular velocities CUEIM, ÃŸiMObs and o _eh .

The angular velocity o _eh acts equally on the real drive and on the dynamic model. This results in a high degree of robustness of the method compared to vibration excitations due to road bumps, in icy conditions, in braking, ABS or ESP interventions. As shown above, the torque thresholds are compared with the disturbance observer

Correction torque AMo _b s modified depending on the operating state of the drive. In addition, the integral component of the disturbance observer correction torque AMo _b s can be modified or initialized depending on the operating state of the drive.

The disturbance observer correction torque AM ₀ bs represents an estimated value for the disturbance torque M _z , thus describes a deviation of the actual torque of the electric machine 1 from the target torque. The model can also be used to detect a magnitude too low torque on a drive unit and take action against it, for example by another

Drive unit compensates for the missing torque component. Too much retarding torque can also be detected and e.g. the electric machine 1 are switched off, if there is a risk that the drive wheels 5, 6 too slow or block.

Figure 5 shows a possible embodiment of an apparatus for performing the method, wherein in a first control unit 30, which is designed as a vehicle control unit, the target torque M _Des determined and transmitted via a bus system 32 to a second control unit 31, wherein the second control unit 31 as Electric machine control unit is designed. The first controller 30 receives the

Angular velocities CUEIM and ccv _eh , which are also reported via the bus system 32. For safety reasons, the first control unit 30 may require the encoder signals required to determine the angular velocities CUEIM and cc _eh also read directly. It can also own, the first controller 30th

assigned encoder be installed.

Signal propagation times of the data transmission are simulated in blocks 33, 34 and 37. The dynamic model is calculated in the first controller 30. The

Signal delays are compensated in the first controller 30. This is done by delaying the desired torque M _Des with the aid of a block 35 such that the response of the model in the form of the angular velocity coeiMObs coincides with the delayed values of the angular velocities CUEIM and cc _eh available in the control unit 30. In the case of a detected unwanted torque delivery of

Electric machine 1, the electric machine 1 and / or a power supply 36 are turned off. This shutdown is preferably carried out by means of a

Hardware signal and is configured redundant for security reasons. For this purpose, the first control unit 30 sends to the second control unit 31 a switch-off signal for the electric machine 1. Alternatively or additionally, a corresponding switch-off signal is output to the energy supply 36 by the first control unit 30.

In a further embodiment, a first dynamic model is calculated in the first control unit 30 and a second dynamic model is simultaneously calculated in the second control unit 31. Upon detection of an unwanted torque delivery of

Electric machine 1 with the aid of the first dynamic model, the first control unit 30 switches off the power supply 36. Will with the help of the second

Dynamic model detects an unwanted torque output of the electric machine 1, the second control unit 31 switches off the electric machine. With these two separate dynamic models and two shutdown paths, high redundancy and security against e.g. a software or

Hardware error in one of the two control units 30 or 31, which would prevent a required shutdown. The redundancy can be increased again if 31 different donors are assigned to the first control unit 30 and the second control unit. Thus, the acquisition of input and

Output variables of the drive train or the motor vehicle and / or the electric machine 1 redundant, which are supplied to the two dynamic models. The dynamic model or partial model of the drive train can drive units such as electric machines and hydraulic and internal combustion engines or units such as clutches, torque converter, lock-up clutches,

Transmission and propeller shafts, the chassis with brake system and the tires as well as related controls. From power units, aggregates and chassis or their suspensions, inertias, elasticities, mechanical play, friction, losses, dynamic behavior, limits and the drive dynamics can be modeled. An adaptation of the

Model parameter to the parameters of the real drive train makes sense.

The illustrated method can be used in hybrid, electric or hydraulic vehicles but also vehicles with conventional combustion engine drive.

Claims

claims

1 . A method for detecting unwanted driveline reactions of a

Motor vehicle with at least one drive unit, wherein at least one input variable (M _Des ) of the motor vehicle and / or the drive unit (1) is input to the drive train and at least one output variable (OJ _E IM) measured on the motor vehicle and / or the drive unit (1) characterized in that the at least one input variable (M _Des ) is supplied to a dynamic model at least partially replicating the drive train of the motor vehicle, wherein the dynamic model based on the at least one input variable (M _Des ) at least one model output variable (oo _E iMObs) determined, which is compared with the at least one measured output variable (O _E IM), wherein an unwanted driveline reaction is inferred in a difference between the measured output (OJ _E IM) and the model output (oo _E iMObs).

2. The method according to claim 1, characterized in that

Presence of several measured outputs (OJ _E IM) and several model outputs (oo _E iMObs) one of the measured

Output variables (ω _Ε | _Μ ) with an associated model output variable (oo _E | _M obs) is compared.

3. The method according to claim 1 or 2, characterized in that a

Disturbance (M _z ) of the drive train of the motor vehicle is estimated in the dynamic model.

4. The method according to claim 3, characterized in that a value (AM _0bS ) of the disturbance variable (M _z ) of the drive train of the motor vehicle in the dynamic model is estimated by methods of Störgrößenbeobachtung.

5. The method according to claim 3 or 4, characterized in that the value

(AMobs) of the disturbance variable (M _z ) of the drive train of the motor vehicle in the dynamic model is compared with at least one threshold, wherein when exceeding a first threshold value or _{falling below} a second threshold value by the value (AM _0bS ) of the disturbance variable (M _z ) at least one drive unit (1) of the motor vehicle is influenced or switched off.

A method according to claim 3, characterized in that the threshold depends on an operating condition of the drive train.

A method according to claim 6, characterized in that the estimated value (AMo _b s) of the disturbance variable (M _z ) on the, the drive train of the motor vehicle at least partially replicating part of the dynamic model

is returned.

Method according to at least one of the preceding claims, characterized in that a torque and / or a power of

Drive unit (1) used as an input variable and / or estimated as a disturbance variable (M _z ), wherein the dynamic model is a target torque, a target power, a calculated actual torque and / or a calculated

Actual performance of the drive unit (1) uses and / or appraises.

9. The method according to at least one of the preceding claims, characterized in that the dynamic model simulates a drive dynamics of at least one drive unit (1) and / or other aggregates.

10. The method according to at least one of the preceding claims, characterized in that the dynamic model simulates torques and / or performance of mechanically driven ancillaries. 1 1. Method according to at least one of the preceding claims, characterized in that the dynamic model elasticities and / or mechanical play in the suspensions of the drive unit (1) and / or in the suspensions of other units and / or in a chassis and / or in the drive train at least partially replicates.

12. The method according to at least one of the preceding claims, characterized in that the dynamic model or the parameters of the dynamic model based on a time course of at least one Input variable (M _Des ) and / or at least one measured output variable (OÜEIM) are adapted.

13. A device for detecting unwanted driveline reactions of a

Motor vehicle with at least one drive unit, wherein at least one input variable (M _Des ) of the motor vehicle and / or the drive unit (1) is input to the drive train and at least one output variable (OJ _E IM) measured on the motor vehicle and / or the drive unit (1) , characterized in that means are present which supply the at least one input variable (M _Des ) to a dynamic model at least partially replicating the drive train of the motor vehicle, wherein the dynamic model based on the at least one input variable (M _Des ) has at least one model output variable (M _Des ). oo _E iMObs), which is compared with the at least one measured output variable (OJ _E IM), wherein an unwanted driveline reaction is inferred for a difference between the measured output (OO _E IM) and the model output (oo _E iMObs).

14. The device according to claim 13, characterized in that a first, the dynamic model comprehensive control device (30) via a, at least one input variable (MDes) and / or at least one measured output variable (ωΕΙΜ, ooVeh) transmitting communication device (32) with a second control unit (31), wherein the first control unit (30) comprises a first dynamic model and the second control unit (31) comprises a second dynamic model, wherein the first control unit (30) is connected to a power supply (36) the second control unit (31) is connected to the drive unit (1).

15. The apparatus of claim 19 or 20, characterized in that the first control device (30) as a vehicle control unit and the second control unit (31) as

Drive unit control unit is formed.