FR2770465A1 - METHOD AND DEVICE FOR DETERMINING SKATING IN A CLUTCH - Google Patents

Abstract

The invention relates to a method and device for determining slip in a clutch between an engine and the transmission of a drive train using clutch input speed and wheel speed, whereby variations in speed are calculated using a mathematical model describing the dynamic performance of the drive train and are taken into account when slip is determined.

Description

 The present invention relates to a method for determining the slip in a clutch arranged in the transmission train of a motor vehicle, referred to simply as a vehicle in the following, as well as to a device for implementing the method.

 The clutch arranged in the transmission train of a vehicle between a drive motor and a gearbox is more and more often actuated automatically, a clutch actuation actuator being controlled by a control device according to the vehicle operating conditions. Such automatic clutches can also be arranged downstream of the gearbox. On the one hand, such automatic clutches considerably increase the comfort of using vehicles. On the other hand, they lead to a reduction in consumption because they operate more frequently at a speed more favorable to consumption, in particular in combination with automated gear change mechanisms. For reasons of low power consumption of the actuator, shortness of the time necessary for the operation, and comfort, the automated clutch is operated in such a way that it is only closed to the extent that this is necessary to avoid skating, or skating too high to be acceptable. Knowing the clutch slip is therefore necessary for many reasons.

 If the clutch output rotation speed, which is identical to the gearbox input rotation speed, is calculated by determining the average of the rotation speeds of the driven wheels and multiplying this value by the reduction ratio overall which occurs respectively between the input rotation speed of the gearbox and that of the wheels, the oscillations that appear in the transmission train are not taken into account, while the transmission train is a system sensitive to oscillations . It follows that a calculation difference which appears, as a result of the oscillations of the transmission train, between the measured speed of rotation of the engine, that is to say the speed of rotation of the clutch input, and the calculated input rotation speed of the gearbox, i.e. the clutch output rotation speed, is considered slippage, although no slippage is actually present. In order to achieve a certain degree of certainty with regard to such incorrect skid interpretations, it has hitherto been customary to introduce a fixed skid limit which must be exceeded in order for the previously explained difference between the speeds of rotation to be considered skating. In particular, in the case of longitudinal oscillations, called grazing, such as those which appear when driving at very low speeds of rotation or during a start-up where the torque is applied in the form of a torque, this limit of skating should be set to a very high value. This leads, during normal operation, to a slip that is not recognized as such, even though it is actually present, which can lead to unnecessarily high consumption and a reduction in service life. of the clutch.

 It is a first object of the present invention to provide a method for determining the slip in a clutch, arranged in the transmission train of the vehicle between an engine and a gearbox, which makes it possible to recognize a slip which appears in the clutch without requiring a measurement of the clutch output rotation speed, and in which oscillations of the transmission trains are taken into account.

 It is a second object of the present invention to provide a device for implementing such a method.

According to a first aspect, the invention provides a method of the type mentioned in the introduction in which the rotational speed nKi of the clutch input is measured and the rotational speed nKa of the clutch output is calculated from measuring the speed of rotation of at least one vehicle wheel and the overall reduction between the clutch output and the vehicle wheel,
characterized in that
variations Antn which appear as a result of variations in the operating parameters of the transmission train are calculated using a mathematical model which describes the dynamic behavior of the transmission train and in that
it is estimated that there is skating if Ini <i - flxal Anern> o.

 According to the invention, oscillations in the speed of rotation which appear due to the dynamic behavior of the transmission train are calculated in particular by taking into account the variation of the torque transmitted by the engine to the transmission train. To consider as slippage the difference between the measured input rotation speed of the clutch and the output clutch rotation speed, calculated from the measured rotation speed of the vehicle wheels and the reduction in gear together, this difference must exceed the variations in the dynamically calculated speeds of rotation.

According to a first embodiment, the Andean variation in rotational speed is determined using the following formula 60 #M
# ndyn ~ 1 =
4 # JM * fR
where AM = variation of the engine torque, JH = moment of inertia of the engine and fR = grazing frequency;
the variation AN of the engine torque can in particular be determined by comparing an engine torque signal and a filtered engine torque signal.

According to a second embodiment, the Andean variation in rotational speed is determined using the following formula: An (t + TR) dyn ~ 1 = #n (t) * e- # TR
where # = D 2 # / TR
where D = constant damping of the grazing oscillation and TR = duration of a grazing oscillation.

 According to a third embodiment, the variation #ndyn of rotational speed is the greater of the two values: # ndyn1 and # ndyn2, calculated above.

According to a second aspect, the invention relates to a method for determining the slip in a clutch arranged in the transmission train between an engine and a gearbox,
in which the clutch input rotation speed or the engine rotation speed nlci is measured and the clutch output rotation speed nKa is calculated using a measurement of at least one speed wheel rotation,
characterized in that
the total transmission train is represented by a mathematical model of the following form
x = Ax + Bu,
where x denotes the state variables, u the generating pairs, A the state matrix, and B the transmission train control matrix, and
the clutch output rotation speed is calculated using the mathematical model.

In this case, the resisting torque (ML) acting from outside on the transmission train can be calculated according to the following formula: ML = ME - JM * #M - JKF @ * # KF @
Where ME = engine torque, JH = engine inertia torque, OH = angular speed of the engine, J = Z = overall inertia torque on the vehicle wheels, rvz = angular speed of rotation of the vehicle wheels.

 According to a third aspect, the invention relates to a device for implementing a method according to the invention.

The objects, features and advantages of the present invention set out above, as well as others, will emerge more from reading the following description of preferred embodiments thereof, with reference to the drawings in which:
Fig. 1 represents a transmission train of a motor vehicle;
Fig. 2 represents a model of oscillations of the transmission train;
Fig. 3 represents oscillations which appear in the transmission train after a shortening of the torque;
Fig. 4 represents curves intended to explain the determination of the respectively applied engine torque;
Fig. 5 shows the damped oscillation of the input rotation speed of the gearbox;
Fig. 6 is a logic diagram intended to explain the calculation of the dynamic slip limit;
Fig. 7 shows a transmission train similar to that of FIG. 1, provided with additional sensors;
Fig. 8 is a representation intended to explain the determination of the resisting torque,
Fig. 9 is a logic diagram intended to explain the determination of the input rotational speed of a transmission;
Fig. 10 is an oscillation model of the transmission train; and
Fig. 11 is an oscillation model of the transmission train.

 According to FIG. 1, the transmission train of a vehicle comprises a combustion engine 2, which is connected by a clutch 4 to a gearbox 6 which is itself connected to the rear drive wheels 12 by a cardan shaft 8 and a differential 10 The front wheels 14 of the vehicle are not driven in the example shown.

 The clutch 4 is of a conventional structure and contains inter alia a clutch disc 16 which is fixedly connected in rotation to the crankshaft of the combustion engine 2, a pressure plate 18, which is connected fixedly to the input shaft of the gearbox 6 and which can, by means of an operating lever 20, be released from its friction engagement with the clutch disc 16, in opposition to the force d 'a disc spring.

 The gearbox 6 is a manually controlled gearbox, which can be controlled by a gear change lever 22.

 An actuator 24, for example an electric stepping motor, which is controlled by an electronic control device 26, is provided for operating the operating lever 20.

 The electronic control device contains, in a manner known per se, a microprocessor, memories, interfaces, etc. This device receives as input signals the signal from a rotation speed sensor 28, intended to detect the rotation speed of the clutch disc 16 or of the crankshaft of the combustion engine 2, the signal of a position sensor 30 intended to determine the position of the actuator 26 or of the operating lever 20, the signals from the sensors 32 and 34 of the speed of rotation of the wheels, as well as possibly signals relating to other operating parameters of the transmission train such as the position of a combustion engine throttle valve 2, etc. The rotational speeds of the non-driving front wheels 14 can also be brought to the control device 26.

 The structure and operating mode of the installation described above are known in themselves and are therefore not explained in more detail.

When the speed of rotation of the pressure plate 18 which is equal to the speed of rotation of the input shaft of the gearbox is calculated by calculating the average of the speeds of rotation of the rear wheels 22 and multiplying this value by the overall reduction of the gearbox 6 and of the differential 10, and taking the slip value the difference between this rotation speed of the pressure plate 18 thus calculated and the rotation speed of the clutch disc 6, the following difficulty appears:
the overall transmission train is a structure liable to oscillate, in which the engine, ie combustion engine 2, suspended in a flexible manner inside the vehicle oscillates relative to the vehicle which is essentially inert and which rests on the ground by the rear wheels 12, the transmission train acting as an elastic coupling element.

 The system liable to oscillate is shown diagrammatically in FIG. 2. In this figure, JH denotes the engine inertia torque, i the overall reduction of the transmission and c the elastic constant of the transmission train.

 If the inertia of the motor is stimulated by a sudden AM of torque, an oscillation of amplitude AN / c is established at the natural frequency of grazing O'bratage This oscillation is represented in FIG. 3 where the ordinate represents the speed of rotation n and the abscissa represents the time t.

The maximum angular speed resulting from this oscillation is equal to
d # MAX / dt = #M / C * #broutage
The calculations then give
#M
#Max = #broutage
VS
60 #M 60 #M
nMax = # # = = #
(2 #) JM ## grazing (2 #) # (2 #) JM # fbroutage
An = l is the maximum oscillation of speed or maximum difference in speed which results from the variation AN of the motor torque. This value is considered as clutch slippage only when the difference between the measured speed of the clutch disc 16 and the gearbox input rotation speed or clutch output rotation speed, calculated in multiplying the average of the rotational speeds of the rear wheels 12 by the overall reduction, is greater than # ndyn1.

The frequency of chatter chatter depends on the respective reduction of the gearbox. Chatter frequencies can be determined using measurements for each speed or can be determined using vehicle data. The frequencies for the other gear ratios can be determined from the grazing frequency for the first gear by:
fn = fi * speed 1
lvicess n
The variation of the engine torque AN is determined by comparing the engine torque signal and a filtered engine torque signal. The engine torque signal is for example measured, using a characteristic diagram of the engine speed and the throttle valve position or the engine speed and suction pressure, from which the engine torque can be read for given quantities. Likewise, the motor torque can be obtained directly from the motor control, for example by a data bus, such as a CAN bus. The filtered engine torque signal is derived from the engine torque signal by causing the engine torque signal to pass through a filter known per se, with a time constant
Filter TF. The filter time constant TF should not be chosen too small, as this would cause the filtered signal to follow the raw signal too quickly, and it would be impossible to carry out a precise determination of the variation AN of the motor torque. It is appropriate to use a filter time constant TF which corresponds to twice the period of the chattering oscillation. A determination of AN can be obtained from values of the motor torque memorized for this purpose at earlier times.

 Fig. 4 represents two graphs, among which the upper graph represents the difference between the signal Ms of engine torque and the filtered signal ME, F of engine torque. The curves of the lower graph are identical to those of the upper graph and represent the time constant TF of the filter; the smaller the filter time constant TF, the faster the filtered signal ME, F of motor torque follows the actual motor torque Ms. Determine the variation in torque AN by comparing the jerk signal ME and the filtered signal ME , F describes a damping, as a function of time, of the oscillation of the transmission train.

Due to the damping in the transmission train, the amplitude An of the chattering oscillation decreases as time passes. The method described above for determining the damping of the oscillation by the decrease in AN as a function of time is not sufficiently precise in general. For this reason, the amplitude for the next step is determined from the amplitude of the chatter oscillation, taking into account the damping of the oscillation by means of the damping constant.
D. For the new amplitude, we get:
Ani = K
or (Fig. 5) An (t + TR n (t) * e = n (t) e
In addition # = D # 2 # # f @) = D # 2 # / T @
During a period of the grazing oscillation, the command is called p times, i.e. the speed of rotation is read p times, so that the damping constant K is obtained , amortization constant per command interruption: K = ###
For a simplified calculation of K, the above term is subjected to a series development. For example, as a first approximation: (1 - = i - mx -
other elements of serial development can also be used when accuracy needs to be increased.

So for the dynamic limit of sliding according to the influence of the damping: # ndyn2 = #ndyn # K
In this way, two terms are available to determine the slip limit, namely, on the one hand the slip limit # ndyn1 calculated from the variation AN of the engine torque, and on the other hand the slip limit # ndyn2 calculated on the basis of the influence of damping, the greater of these two values being used as the slip limit:
AnJyn = MX (tndy, l "And, M-)
Unlike the state of the art which works by using a slip limit fixed at a very high value to eliminate the influence of oscillations in the transmission train, the invention makes it possible to work by using a slip limit realistic, adapted to the real oscillations of the transmission train.

 Fig. 6 shows in the form of a logic diagram the method described above for determining the slip.

 In step 100, the slip An is calculated in a conventional manner by subtracting from the measured speed of rotation of the engine the speed of rotation of the input of the gearbox determined by multiplying the average of the measured speeds of rotation of the wheels by the reduction ratio total.

 In step 102, the variation AN of engine torque is calculated, as explained with reference to FIG.

4.

 In step 104, the slip limit Anzn1 is calculated from the variation AN of engine torque according to formula (1).

 In step 106, the slip limit # ndyn2 is calculated based on the influence of the damping according to formula (2).

 In step 108, it is determined whether An1 is greater than An2. If the answer is positive, it is determined in step 110 that An1 is the value of the dynamic slip limit Anm. If the answer is negative, it is determined in step 112 that Anwn2 constitutes the dynamic skating limit Anm. It is then determined in step 114 if the slip An determined in a conventional manner is greater than Ante. If the answer is positive, it is estimated that the clutch is slipping. If the answer is negative, it is however estimated that there is no slip of the clutch.

 As an alternative to the method described above, there is also the possibility of reconstructing, at least approximately, the global state vector of the dynamic "train train" system, from measured quantities.

To this end, the input quantities motor torque Ms and load ML are introduced into the mathematical representation of the transmission train using its dynamic model, and a comparison of the measured quantities and of the corresponding quantities of the mathematical models is carried out. To this end, the difference between the measured quantities of the transmission train and the quantities determined from the mathematical model is assigned (observed) an appropriate weighting at the input of the mathematical model. The mathematical model is stimulated so that it oscillates in synchronization with the transmission train. In this way, non-measurable quantities can be deduced from the mathematical model. In the particular case of determining the slip, the non-measurable quantity of the input rotation speed of the gearbox or clutch output rotation speed is deduced from the mathematical model and is compared with the rotation speed. measured from the engine.

This comparison makes it possible to determine whether or not there is slippage.

 In Fig. 7 shows a transmission train which corresponds to that of FIG. 1, but which is provided with additional sensors, for example a throttle flap position sensor 36, a cardan shaft speed sensor 38, etc. These additional sensors are also connected to the electronic control device 26, inside which the mathematical model is installed.

 The problem posed by the calculation method described above by means of observers is that the resistive torque Mt, which results from the resistance to advancement, the slope, etc., is not known in the particular case of the vehicle. . It is therefore necessary to determine the resisting torque ML by means of an evaluation of the disturbing quantities. It is possible to use for this purpose several measurable quantities such as the speed of the vehicle, obtained from the rotational speeds of the wheels, and the rotational speed of the engine.

For the rotational speeds of the wheels, it is advantageous to also take into account the rotational speeds of the front wheels 14, which in general does not mean any additional cost since these rotational speeds are determined by means of ABS sensors, which are present anyway. Using the known inertias of the engine and the vehicle, it is possible to determine by means of a very simplified model, which is shown in FIG. 8, an estimated magnitude of the resistant torque ML according to the following formula: M @ = ME - JM * #M - JKF @ * # KF @
where Jx is the mass inertia torque of the engine, Jz is the mass mass inertia of the vehicle, reduced on the engine side, OM is the speed of rotation of the engine, is is the speed of rotation of the vehicle projected onto the motor side.

The dynamic mathematical model of the vehicle can be represented as a state as follows:
x = Ax + Bu
where the vector x represents the state quantities, that is to say the angle of rotation and the angular velocities, and where the vector u represents the couples which intervene: driving torque and resisting torque. The system is described by the state matrix A. The control matrix B makes it possible to project onto the individual state coordinates the individual moments which occur.

 Fig. 9 represents a logic diagram of the determination of skating, exposed above, from a complete mathematical model. The motor torque input quantity, which is measured, for example from the stress exerted on the housings where the motor is supported on the vehicle, or which is measured, for example from the speed of rotation and the angle of the throttle flap or from information concerning the engine control, intervenes on the vehicle 120. Measured quantities are determined, at 122, on the vehicle 120 using sensors. Furthermore, the resistive torque is determined at 124, as explained with reference to FIG. 8. The resistive torque and the engine torque are introduced at 126 as input quantities in a dynamic vehicle model. The measured values 122 are read mathematically at 128 in the dynamic model 125 of the vehicle. The difference between the measured values mathematically determined in 128 and the measured quantities, measured directly, is dynamically weighted at 130 and is introduced into the dynamic model 126. By an appropriate choice of dynamic weighting, the dynamic model is stimulated in a way as it oscillates in coincidence, or synchronization, with the vehicle, so that the input rotation speed of the gearbox or clutch output rotation speed can be calculated and compared with the speed of engine rotation or clutch input rotation speed. The clutch slip can therefore be calculated directly using this comparison.

 The following representation is used as an example of the construction of the model: in Figure 10, the following results are obtained for the slip clutch.

By using the known inertias of the engine 201, of the transmission 203 and of the vehicle 205 and of the clutch 202 with its transmissible torque as well as the rigidity of the transmission train 204, the dynamic model of the vehicle can be produced using the model of the
Figure 10:

By introducing the state vector x, the control vector u, the state matrix A and the control state matrix B, we obtain

D / dt x = Ax + Bu
According to another example according to Figure 11, it is possible to establish a model for the case of the clutch without slippage. The angles of rotation of the clutch input and output are then identical.

In this way, the inertia of rotation of the engine and the gearbox can be brought together. We obtain:

d x= Ax+Bu
dt
On a décrit essentiellement un procédé conforme à l'invention, et un dispositif de mise en oeuvre du procédé, au moyen desquels il est possible de distinguer, dans le patinage calculé à partir de la différence entre la vitesse de rotation du moteur et les vitesses de rotation de sortie de la boîte de vitesses ou tout au moins de roues individuelles de véhicule, entre un patinage réellement existant à l'embrayage et un patinage virtuel qui apparaît entre l'entrée de la boîte de vitesses et une roue, par suite de la dynamique de la chaîne de transmission: effet des oscillations de torsion de l'arbre de transmission entre la boîte de vitesse et les roues du véhicule.By introducing the state vector x, the control vector u, the state matrix A and the control state matrix B, we obtain

dx = Ax + Bu
dt
We have essentially described a method according to the invention, and a device for implementing the method, by means of which it is possible to distinguish, in the slip calculated from the difference between the rotational speed of the engine and the speeds of output rotation of the gearbox or at least of individual vehicle wheels, between a slippage actually existing at the clutch and a virtual slippage which appears between the input of the gearbox and a wheel, as a result of the dynamics of the transmission chain: effect of torsional oscillations of the transmission shaft between the gearbox and the vehicle wheels.

 If vehicle reactions or clutch actuations that are executed depending on the slip are induced, it is possible to induce the reaction of the vehicle or the clutch actuation in the event of actual slip on the clutch can and to avoid it in case of virtual skating. The command / regulation contains, for example, a part which causes the clutch to close, as soon as the slip is recognized.

 In appropriate vehicle running situations, there may be a clutch slip which therefore leads to the closure of the clutch under the effect of the command. In addition, load variations which are repeatedly induced can induce virtual slippage in the form of a difference in rotation speed, between the input rotation speed of the gearbox and the output rotation speed weighted by the overall reduction .

 As an additional example, we can cite the creation of a temperature model / stress model to determine the clutch temperature / clutch stress, in the control / regulation of the clutch with automatic actuation. The clutch temperature or the friction power applied in the clutch is calculated using the vehicle data. Depending on the order, reactions defined on the vehicle or when the clutch is actuated are carried out when a limit temperature / limit stress is exceeded. By means of the determination and the distinction between real and virtual skating, the control can command to induce or avoid modifications in the operation of the clutch.

 In a method for determining the slip in a clutch, the clutch input rotational speed or the engine rotational speed is measured and the clutch output rotational speed is calculated from the measurement of the speed of rotation of at least one vehicle wheel and the overall reduction between the clutch output and the vehicle wheel. According to a first embodiment of the method, Andean variations which appear as a result of variations in the operating parameters of the transmission train are calculated using a mathematical model which describes the dynamic behavior of the transmission train and it is estimated that there is skating if | nK1 - nia | Andyn> 0.

 The invention further relates to a device for determining a slip according to the method described above.

 The invention is also not limited to the example or examples of embodiment of the description. Rather, many alterations and modifications are possible within the framework of the invention, in particular variants, elements and combinations and / or materials which are for example inventive by combination or transformation of the particularities or elements or process steps described in the general description and the embodiments contained in the drawings, which lead by particulars to be combined with a new object or with new process steps or sequences of process steps, insofar as they relate to manufacturing processes, verification and machining.

Claims (8)

 1. Method for determining the slip in a clutch arranged between an engine and a gearbox in the transmission train of a vehicle,  wherein the rotational speed nKi of the clutch input is measured and the rotational speed nKa of the clutch output is calculated from the measurement of the rotational speed of at least one wheel of the vehicle and the overall reduction occurring between the clutch output and the vehicle wheel,  characterized in that  Andean variations which appear as a result of variations in the operating parameters of the transmission train are calculated using a mathematical model which describes the dynamic behavior of the transmission train and in that  it is estimated that there is skating if In: nka | #ndyn> 0.  2. Method according to claim 1, characterized in that  the Andean variation in speed is determined using the following formula  60 #M  # ndyn ~ 1 =  4 # JM * fR  where AN = variation of the engine torque, J.w = moment of inertia of the engine and fR = grazing frequency.  3. Method according to claim 2, characterized in that  the variation AN of the engine torque is determined by comparing an engine torque signal and an engine torque filter signal.  4. Method according to claim 1, characterized in that  the Andean variation of rotation speed is determined by the following formula: #n (@ + T @) dyn ~ 2 = #n (t) * e- # TR  where 2 #  # = D  TR  where D = constant damping of the grazing oscillation and TR = duration of a grazing oscillation.  5. Method according to claims 4 and 2, characterized in that the variation #ndyn of rotation speed is the greater of the two values: # ndyn1 and Anyn.  6. Method for determining the slip in a clutch arranged in the transmission train between an engine and a gearbox,  in which the clutch input rotational speed or the engine rotational speed nKi is measured and the clutch output rotational speed n # a is calculated using a measurement of at least a wheel rotation speed,  characterized in that  the total transmission train is represented by a mathematical model of the following form  x = Ax + Bu,  where x denotes the state variables, u the generating pairs, A the state matrix, and B the transmission train control matrix, and  the clutch output rotation speed is calculated using the mathematical model.  7. Method according to claim 6, characterized in that  the resisting torque ML acting on the transmission train from the outside is calculated according to the following formula: M @ = ME - JM * #M - JKFZ * # @@@  where ME = engine torque, JM = engine inertia torque, #M = engine angular speed, J: DZ = overall inertia torque on the vehicle wheels, z = angular speed of rotation of the vehicle wheels.  8. Device for implementing a method according to any one of the preceding claims.