FR2787856A1 - METHOD AND DEVICE FOR CONTROLLING AN AUTOMATIC GEARBOX SWITCHING MECHANISM - Google Patents

Abstract

A control method in which the switching mechanism (103--108) is actuated using an actuator (102); a drive element (109) travels a path following the actuation of the actuator (102), a stroke quantity (Sréel) representing the stroke is detected; a setpoint value (Scons) is formed for the stroke and the actuator is controlled during a first phase of the switching operation to set the detected stroke (Sréel) to the setpoint value (Scons). A reaction quantity is detected (Neréel, Sréel, I, Ig) which represents the reaction of the gearbox.

Description

State of the art.

  The invention relates to a method for controlling the

  switching of a switching mechanism of a gearbox

  automatic braids, according to which the switching mechanism is actuated using a controlled actuator, and a drive element travels a path according to

actuation of the actuator.

  The invention also relates to a device for

  the implementation of such a method.

  In vehicles fitted with a gearbox-

  its automatic, the automatic gear change is generally done by controlling the components of the transmission line, namely the engine, the clutch and the gearbox.

  speeds. These components must be ordered so that the chan-

  report management is done as quickly as possible but

  also in a comfortable way. The focal point of this operation

  ration is the synchronization of the gearbox. In

  known automatic gearboxes, this synchronization

  sation is generally an operation taking place in the

  time. This causes problems according to the different con-

  environment editions as well as the different rotational speeds, the different gear ratios (skipping gears; multiple gear changes, i.e. passing over one or more gear ratios), suspension, the temperature, the viscosity of the oil, etc.

  synchronization of the rotation speeds of the gearbox

  tresses cannot be done exactly by an operation

  mandated in a satisfactory time, according to the different

  situations for all rotational speed conditions

  different days. This can result in more wear

  gearbox and noisy operation from

  pleasant when changing the new gear,

  that the rotational speeds do not correspond exactly. The present invention aims to develop means for improving the gear change

  in an automatic gearbox.

  To this end, the invention relates to a method of the type defined above, characterized in that: - a quantity representing the stroke is detected, a reference quantity for the stroke is formed, - the actuator is controlled for at least a first

  phase of the switching operation so that the stroke

  tectée is set to the value of the target variable, - at least one reaction quantity is detected which represents the reaction of the gearbox and / or the

  switching and / or actuator, in response to an action-

  actuator, and - the setpoint variable is established during a current switching operation, at least as a function of the variable

  reaction detected during at least one operation

previous mutation.

Advantages of the invention.

  The method and the device for controlling a

  switching gearbox of an automatic gearbox according to the invention apply to a switching operation

  of the gearbox. The switching mechanism is activated.

  actuated by a controlled actuator. The switching mechanism

  tion comprises a drive element, generally a walkman, which travels a stroke according to the actuator control. We detect the magnitude that represents this race. According to the invention, a target quantity is formed.

  of the race and at least during a first phase of the operation

  switching ration, the actuator control is such that the magnitude of the detected stroke takes the value of the setpoint magnitude. Then at least one reaction quantity is detected representing the reaction of the gearbox and / or of the switching mechanism and / or of the actuator in response to the action of the actuator. The nominal value, during the actual switching operation, is established

  at least as a function of the reaction quantity entered during

  in at least one previous switching operation.

  Thanks to the adaptation of the actuator control according to the invention, the switching operations are carried out

  gearboxes optimally adapted to the conditions

existing switching conditions.

  According to the invention, during a first switching phase, it is first recognized when the synchronization of the gearbox begins. It is thus determined whether the time when the synchronization rings are in

  support and thus allow synchronization to be carried out. Se-

  lon the invention, we learn or adapt at the start of the syn-

  chronization, the position of the player.

  According to another development of the invention, the

  gearbox has an input shaft and as a large

  reaction quantity a rotation speed quantity is detected

  input representing the speed of rotation of the input shaft. In this embodiment, provision is made in particular for

  form the target variable for the switching operation

  current function as a function of the chronological evolution of the magnitude of the input rotational speed entered during at least one previous switching operation. It is then advantageous to compare the quantity representing the evolution over time of the speed of rotation of the input shaft with a predetermined set value. Then the input quantity of the current switching operation is formed as a function of the magnitude of the travel existing when the threshold is reached. This embodiment of the invention is based on the consideration that the coming into support

  synchronization rings is expressed by a character line

  gradient of the input speed of the gearbox

  speeds. We can in this way adapt the position of the

  the scale, when the synchronization rings come to bear

  tion, that is to say at the start of synchronization proper. According to another embodiment, a travel quantity representing the travel of the operating element is first detected as a reaction quantity

  mentioned above which is generally the player. It is no-

  much planned to form the target variable of the current switching operation as a function of

  the chronological evolution of the size of the race detected

  tee during at least one previous switching operation.

  This embodiment of the invention is based on the consideration

  ration that the coming into support of the synchronization rings results in a characteristic curve of the gradient of the player's stroke. The position of the player when the synchronization rings come to bear, that is to say at the start

  of synchronization proper, maybe this way

  adapted without knowing the input rotation speed

of the gearbox.

  It is also possible to provide electrical control

  the actuator and use, as a reaction quantity,

  tion, the intensity which represents the control current. he

  provision is made in particular to measure an intensity representing

  both the control current and form the con-

  sign of the current switching operation depending on

  the intensity measured during at least one switching operation

  anterior tation. This allows you to enter a quantity

  sensing the chronological evolution of the race, then the target quantity is formed as a function of a comparison of the measured intensity and of a threshold formed with the chronological evolution, during at least one previous switching operation. According to another advantageous development of the invention, as already indicated, the switching mechanism

  includes an operating element, usually the bala-

  deur, which travels a stroke according to the actuator command. In this case, a quantity representing the stroke is also detected. The actuator is controlled

  during at least a second switching phase, the syn-

  actual timing of the gearbox, according to a time interval. This time interval

  is then formed as a function of the reaction quantity detected

  tee. In this case it is in particular provided that the time interval for the current switching operation is formed as a function of the chronological evolution of the detected race.

  during at least one previous switching operation.

  The variant described above relates in particular to the synchronization proper, that is to say the adaptation of the input rotation speed of the gearbox to

  the speed of synchronization of the gear ratio

  braids to be passed, and in particular for emergency operation in the event of failure of the input speed sensor of the gearbox. In this case, it is necessary to switch to the usual synchronization with a speed regulator and a time control. This

  control over time is then adapted according to the invention.

  According to a particularly advantageous development

  geux of the invention, a speed of rotation value is entered

  tion representing the input rotation speed of the gearbox then, when the magnitude of the entered stroke is at the set value, in a second phase the actuator is controlled to adjust the input rotation speed of

  the gearbox, entered, on a predetermined value.

  In this realization, after reaching the point of syn-

  timing, switching from position or stroke regulation to speed regulation (synchronous speed regulation of the gear ratio of

destination).

  According to another development of the invention,

  when the gearbox input speed

  ses, entered, is at the expected value, in general the synchronization speed, then in a third phase the actuator is controlled so that the quantity entered from the

  stroke takes a second predetermined setpoint quantity.

  In this embodiment, after reaching the speed of ro-

  synchronization, to shift gears

  of destination, a speed regulation is switched

  tation (regulation on the synchronous speed of rotation of the

  destination speed ratio) on a po-

sition or race.

  According to another advantageous characteristic of the invention, during at least one previous switching operation, a quantity is formed representing the evolution over time of the speed of rotation of the input shaft and this quantity is compared with a threshold. predetermined and the target variable of the current switching operation is formed according to the magnitude of the stroke, existing

when you reach the threshold.

  The invention also relates to control device

  a switching mechanism of an automatic gearbox

  tick during a gearbox switching operation, the switching mechanism being actuated by means of a controlled actuator, characterized by means for - entering a quantity representing the stroke, - forming a reference quantity of the stroke, - command the actuator during at least a first phase of the switching operation so that the magnitude of the entered stroke takes the value of the setpoint, - enter at least one reaction quantity which represents the

  reaction of the gearbox and / or the

  mutation and / or actuator following a maneuver of

  the actuator, and form during a switching operation

  current value, the target quantity as a function of the reaction quantity entered during at least one operation

previous switching.

  More advantageously, this device is pre-

  seen moreover with means for: - entering a value of the input speed of the gearbox representing the input speed of the gearbox, and then - when the magnitude of the entered stroke has taken the value of the nominal variable, in a second phase, control the actuator so that the input speed of the

  gearbox, input, adjusts to a predefined value

  finished. Drawings The present invention will be described below from

  in more detail using the accompanying drawings in the-

which:

  - Figure 1 shows schematically the actuates-

  ment of an automatic gearbox,

  - Figure 2 shows the different functional blocks

  tional control device according to the invention, - Figures 3 and 4 each show in their part (a) and (b), two different variants of control of the switching sequence, - Figures 5, 6, 7 , 8 show different possibilities of adaptation,

  - Figures 9, 10, 11 show chronograms -

my.

Examples of realization.

  The present invention will be described below in

  with the help of different examples of implementation.

  Figure 1 shows a controller 101 receiving different input signals. The input signals are detailed in FIG. 2. The control device acts on a direct current motor 102 actuating a screw wheel 104 via a motor shaft equipped with a screw 103; using the screw transmission 105 and the gears 1 and 2,

  the pinion 106 of the switching cylinder is thus actuated.

  The switching cylinder has, for example, different paths for reverse R, neutral N and the

  gear ratios 1, 2, 3, 4. In addition, there is another path

  min cam for the fifth gear. To pass the different

  rents gear ratios, by rotation of the switching cylinder 107 and through the switching forks or their guide means 108, the

  player 109. The reference S indicates by way of example the di-

  rection of movement of the player.

  It does not matter for the present invention that the player 109 is actuated by a switching fork and a switching cylinder and an electric motor or by

  any control, for example hydraulic and / or tire-

  matic. The detection of the travel of the player, directly or indirectly, is of no importance for the invention. To achieve good, fast synchronization, we subdivide

  the gear shift operation in three ranges or pha-

  its different: 1. Detection of the start of synchronization. To detect the end of the first phase, the synchronization rings must be detected as precisely as possible.

  to allow synchronization by these.

  2. Synchronization. The problem of my-

  optimal way, on how to operate the player while

  during synchronization, to exert on it a force corresponding to the different speeds (rotational speeds). 3. End of synchronization or start of the passage of the new report. Here it is necessary to detect as precisely as

  possible from when you can pass the pro-

  prement said, that is to say the moment when the speeds of

  different rotation will be practically equal.

  To control the gearbox switches, the following will be described in more detail

  the control unit 101 using FIG. 2.

  The control unit receives a switching command B ve-

  from unit 201. This switching order B can be de-

  manually triggered by the driver of the vehicle but also

  automatically, for example depending on the load applied

  the engine or the accelerator pedal position

  or the trigger speed. The switching order

  tion B must contain information (start of start,

  destination) regarding the switch to execute operation

  ter. The rotational speed sensor 202 supplies the actual Nereal value of the input rotation speed of the gearbox to the control unit 101; the speed sensor 203 detects the actual value Nab of the speed of

  gearbox output rotation or vehicle speed.

  The travel sensor 1091 detects the path of the player 9 and applies this path as a Sreel stroke quantity to

the control unit 101.

  Optionally you can plan to measure, as

  their, the current Ig of the DC motor 102 using an intensity sensor 1022 to apply this quantity to

  the control unit 101. At the output of the control unit

  command 101, the electric motor 102 is controlled by the cou-

  control rant I; the clutch actuator 204 is controlled by the signal K and the engine control 205 by the signal M. The control unit 101 receives the described input signals from the central unit 1011 and that -ci 1011 receives the control current I. At the output of the unit 1011, there is the synchronous speed of rotation Ns for the speed controller NR 1012, the setpoint stroke Scons of the charge controller 1013 and the signal Ko coordination

  for switch 1014. The speed regulator

  tion 1012 receives the actual Nereal value of the speed of ro

  gearbox input and the LR 1013 position controller receives the actual Sréel value of the position of the player. Depending on the regulation deviation, regulators 1012 and 1013 supply values for the control current as an output. Optionally, the central unit 1011 can directly receive a Icons control value for the electric motor 102 (switching position at). The function of the control unit 101 will be described below

  using the following flowcharts.

  For the description of the flow charts, we use

  will conventionally be the letter Y to represent a positive answer to a question and the letter N to represent a

negative response.

  Figure 3 shows in its parts (a) and (b) a first variant of the control of the switching sequence. After the starting step 301, the switching order B is recorded in block 1011 (FIG. 2). In step 303, the switching order B is then used to

  find out whether a changeover is planned or not. If no commu-

  tation is not provided, we go directly to the final step 316 and we resume the sequence presented in Figure 3. If on the contrary we have a switching command, then in step 304 we first transmit the signal K to the clutch adjusting member 204 (Figure 2) and the signal M to the engine control unit 205 (Figure 2). The signals M and K open the clutch at this time while the engine control unit 205 receives a set value of the engine torque which

  generally corresponds to a reduction in the torque

  tor. As this part of the switching sequence control is not the subject of the invention, it will not be

  no more reference to this point here.

  In step 305a, a setpoint Scons is recorded as a function F1 of the switching order B

  and other quantities of the adaptation memory 1. As in-

  the switching order contains in particular information

  starting point and ending point.

  In step 305b, the value ac-

  Sréel stroke tuelle (supplied by the stroke sensor

  1091). Then the control unit 101 adjusts by the regulator

  position sensor LR 1013, the electric motor 102 leaving it supplied with power (step 306b) until in comparison step 306 it can be seen that the position of the Sréel player corresponds to the set point Scons Al, recorded in the 'step 305a. A1 is the right value

  to apply. If these values match, step 307 further

  sees the coordinate signal Ko from block 1011 at switch 1014. Then the position of the switch is modified so that now the control current I starts from the speed regulator 1012. The first phase of the switching operation during which one passes in the synchronization position of the player (come into contact with the rings of

synchronization) ends.

  Step 308 begins the second phase of switching

  tion during which the gearbox is synchronized-

  his. For this, in step 308 we take the speed of

  Nsyn synchronization rotation as function F of vi-

  gearbox rotation size and if applicable

other quantities are calculated.

  In step 310 the instantaneous value is recorded

  Nereal tane of the input speed of the gearbox by the sensor 202 to compare this speed in step 311 with the synchronization value Nsyn A2 calculated in

  step 308. The value A2 is thus the value which must be ap-

  plead. In step 311 it is determined that the speed of syn-

  timing has not yet been reached so that the speed controller 1012 continues to supply 1!

  the electric motor 102. When the rotational speed syn-

  chrone is reached, in step 312 we choose the signal

  Ko coordination to switch back from the speed regulator

  rotation tesse 1012 on the position regulator 1013. This completes the second switching phase, that is to say

the actual synchronization.

  In step 313, the third switching phase begins, during which the new gear ratio is passed. For this, in step 313, a Scons setpoint value for the position of the player is extracted from the memory. This is generally done according to the

  destination report and where appropriate other sizes.

  In step 314, the actual position of the Scons player is compared with the reference value extracted from the memory.

  Scons5 A3. The value A3 is the applicable value. The regulation

  position sensor 1013 continues to supply the electric motor

  than 102 until the condition shown in step 314 is satisfied. Then by sending the signals M and K, the clutch is closed and the control of the engine torque is ended by the control unit 101. After the final step 316,

  repeat the process shown in Figure 3.

  FIG. 4 shows in its parts a and b a variant of switching control. Steps 401 to 407 inclusive, i.e. the first switching phase thus corresponds to the sequence of operations already described in

  using blocks 301-307 in Figure 3.

  The second switching phase, i.e. the

  actual synchronization of the gearbox

  begins with step 408 during which a variable is extracted

  their set icons from memory for motor current. In step 409, a set duration is extracted

  Tcons as function F2 of the switching operation ac-

  tuelle B and other quantities such as for example the temperature

  operating mode of the adaptation memory 2. The

  Icons intensity value extracted in step 408 is ap-

  applied directly to motor 102, switch 1014 used

pant position a.

  Step 411 starts the clock and step 412 de-

  if the clock period is over. As long as

  that the clock period has not ended, the motor

  trique 102 receives the Icons setpoint intensity. If the

  the clock is finished, then by the coordination signal Ko, step 413 resets the switch 1014 to its

  position (a) on position controller 1013.

  The third phase begins with step 414.

  This phase consists of passing the new report; blocks 414, 415, 416, 417 correspond to blocks 313, 314, 315, 316

  already described using Figure 3.

  The switching sequence presented using Figure 4 does not require a gearbox input speed sensor unlike this

  which is necessary for the sequence of the se-

  lon figure 3. For this reason, the sequence shown

  using figure 4 is considered in particular as a va-

  emergency light in case of speed sensor failure

  gearbox input rotation.

  Figures 5, 6, 7, 8 show how actuali-

* enter the values entered in the adaptation memory 1,

  written (blocks 305a, 405a in Figures 3 and 4) and in the

  adaptation memory 2 (step 409 in FIG. 4). The opera-

  tions shown in figures 5, 6, 7, 8 take place in pa-

  parallel to the actual switching (Figures 3 and 4) and permanently adapt the content of the adaptation memories 1 and 2. Figure 5 shows a first variant with a

  gearbox input speed sensor.

  After the start-up step 501, in step 502 a threshold SW is formed as a function F4 of the current gear change B

  as well as other quantities, in particular specific quantities

  gearbox fiques; this threshold is either formed or extracted from a memory. In step 503 the

  current value of the gearbox input speed

  its Neréel and we form the derivative as a function of time Veréel

  of Neréel. In the comparison step 504, the graph is compared.

  from the Veréel speed of rotation corresponding to the speed

  input rotation speed of the gearbox at the threshold

  SW formed in step 502. This comparison continues until

  until the value of the rotation speed gradient reaches the threshold SW. When this is the case, step 505 records the stroke set at this time Sréel of the player; in step 506 as the new Scons value, function of F5,

  this value stored in the adaptation memory 1

  recorded in step 505, and quantities as a function of the operating temperature and the like. After the final step

  507, the sequence shown in FIG. 5 is restarted.

  The adaptation operation shown in the figure ensures permanent adaptation of the start of the synchronization operation to the current data. This is based on the fact that when the synchronizing rings are pressed, the gradient of the input rotation speed of the gearbox

  speeds typically increases. This is represented

seen in figure 11.

  Figure 6 shows a second variant

  adaptation not requiring a speed sensor

  gearbox input. After the start-up stage

  rage 601, the instantaneous value Sréel of the position of the climber is recorded and the instantaneous or current value of the control intensity I. From the value Srée1, the derivative in time Sréel 'is formed. In step 603 a Scons reference value is extracted from a memory, as a function F6 of the intensity of the control current I and of the recorded value Scons' for the speed of movement of the player. In step 604, the Sreel 'value is compared to the

  setpoint Scons' during the first communication phase

  tation. As soon as the condition stated in step 604 is fulfilled, the switching position of the Sréel player set at this time is recorded, and according to the value thus recorded and other quantities such as the gearbox temperature, we updates the setpoint variable Scons in the adaptation memory 1. After the final step 606, the sequence of operations shown at

Figure 6.

  The adaptation presented in Figure 6 is based on the consideration that thanks to the known characteristic of the motor can be determined the force acting on the player. In this way, a good adaptation of the position of the player is obtained even without using a speed sensor.

  gearbox input rotation size.

  A third variant of the adaptation will be de-

  described below using figure 7. This variant does not

  nor does it stop the speed of input rotation speed of the gearbox but supposes the measurement of the real intensity of the current of the electric motor 102. After the starting step 701 the instantaneous value Sréel of the position of the player and the instantaneous or current value of the intensity Ig (sensor 1022 in FIG. 2). From the Sréel value, we form the derivative in Sréel time '. In step 703 a value is extracted from a memory

  intensity setpoint Icons as function F6 of the

  their Serre 'of the speed of movement of the walkman and other sizes. In step 704 the Ig value is compared

  measured at the Icons setpoint formed in step 703.

  As soon as the condition stated in step 704 is fulfilled, we

  saves the switching position of the Sréel portable player

  frozen at this instant, and according to the value thus recorded

  and other quantities such as the temperature of the

  gearbox, we update the large-

  Setpoint deur Scons in the adaptation memory 1. After the final step 706, the operations are started again.

shown in Figure 7.

  According to FIG. 8, the updating of the adaptation memory 2 will be described (FIG. 4a, block 409). After the start-up step 801, in step 802, the current value Sreel of the position of the player is recorded, as well as the intensity of the control current I or, where appropriate, the current Ig actually measured in the electric motor 102. A starting from the Sréel position of the player, we form the derivative according to Sréel time '. In step 803, the intensity of the control current I or, where appropriate, the setpoint Scons', is extracted from a memory, from the intensity

Ig measured.

  During the second switching phase,

  Tee in Figure 4, that is to say the synchronization itself, in step 804 we compare the speed of movement

  actual Sréel 'of the player 109 at the set point value

  carried out in step 803. If the speed of movement reaches the

  setpoint, then in step 805 we take as

  their AT the time difference between the instant at which

  the clock (block 411, FIG. 4b) and the instant at which the condition represented in block 804 is satisfied. In

  block 806, this AT duration is recorded as a new du-

  Tcons setpoint, in adaptation memory 2. This new Tcons setpoint time is therefore generally more

  shorter than the old setpoint duration it replaces.

  If the speed of movement of the portable player does not reach the set value, one passes from interrogation 804 to interrogation 807; during the latter, it is checked whether the value Tcons formed in block 409 is finished (FIG. 4a). If not, we come back to

  the starting step 801. On the contrary, if the duration of consi-

  gne Tcons is finished, then in step 808 we take as new value Tcons the old value Tcons increased by

  value A, in adaptation memory 2. After step fi-

  nale 809, we redo the operations of the fi-

  gure 8. The new setpoint time Tcons is then longer

  longer than the old setpoint time that was replaced.

  The background to the flow of operations in FIG. 8 consists in adapting the duration of the synchronization operation proper to the conditions existing at

  any moment. For this, you can shorten the duration of consi-

  gene Tcons recorded in adaptation memory 2 in step-

  by the paths 804, 805, 806. But it is also possible to extend a too short synchronization duration Tcons, recorded in the adaptation memory 2 by passing

by paths 804, 807, 808.

  Figure 9 shows the end of the second phase, that is to say the end of the synchronization operation using a characteristic curve of the input rotation speed of the gearbox as a function time. The graph in FIG. 9 shows the evolution between blocks 309,

310, 311 of Figure 3b.

  Figure 10 shows the synchronization operation.

  command in time shown in Figure 4. It is recognized in the first phase that the player has arrived at

  the value Ssyn or Scons (interrogation 406a, FIG. 4a). Pen-

  during the synchronization time Tcons, the electric motor 0 is supplied with the setpoint current intensity Icons formed in block 408 (FIG. 4a). This can be done by

  a pulse width modulation or with an in-

  Setpoint density Icons, fixed (broken line in figure

  ). After the end of the synchronization period Tcons, com-

  begins the third phase during which we continue to

  move the player to pass the new speed report-

his.

Claims (11)

R E V E N D I C A T IONS   1) Method for controlling the switching of a switching mechanism of an automatic gearbox, according to which the switching mechanism (103-108) is actuated using a controlled actuator (102) and an element drive (109) travels a path as a function of actuation of the actuator (102), characterized in that - a stroke quantity (Sréel) representing the stroke is detected, - a reference quantity is formed (Scons ) for the race, - the actuator is controlled for at least a first   phase of the switching operation so that the stroke   tectée (Sréel) sets the value of the con-   sign (Scons), - we detect at least one reaction quantity (Neréel, Sréel, I, Ig) which represents the reaction of the gearbox and / or the switching mechanism (103-108) and / or actuator (102), in response to actuation of the actuator, and   - the target value (Scons) is established during an operation   current switching ration, at least as a function of the detected reaction quantity (Neréel, Sréel, I, Ig) during   at least one previous switching operation.   2) Method according to claim 1, characterized in that the gearbox comprises an input shaft and as   reaction quantity a speed velocity quantity is entered   input station (Neréel) representing the speed of rotation of the input tree.   3) Method according to claim 1, characterized in that   the switching mechanism includes a functional element   event (109) which travels through a stroke (S) as a function of actuation of the actuator (102) and as a quantity of   reaction we enter a stroke quantity (Sréel) representing both the race.   4) Method according to claim 1, characterized in that   the actuator (102) is controlled electrically and as a large   reaction deur we enter an intensity (I, Ig) representing the control current.   5) Method according to claim 2, characterized in that the target quantity (Scons) is formed for the current switching operation as a function of the time course (Veréel) of the quantity of the input rotation speed   (Neréel), entered during at least one switching operation previous tion.   6) Method according to claim 3, characterized in that the target quantity (Scons) is formed for the current switching operation as a function of the evolution over time (Sreel,) of the quantity of the stroke entered ( Sréel) during   minus a previous switching operation.   7) Method according to claim 4, characterized in that   we measure an intensity (Ig) representing the com-   command and the target quantity (Scons) of the current switching operation is formed as a function of the measured intensity (Ig) during at least one previous switching operation,   and we plan in particular to enter a quantity (Sréel,) re-   present the evolution over time of the race, and we   forms the target quantity (Scons) as a function of a   comparison of the measured current intensity (Ig) and of a threshold (Icons), formed as a function of the quantity (Sréel ') during at   minus a previous switching operation.   8) Method according to claim 1, characterized in that   the switching mechanism includes a functional element   ment (109) which travels a stroke as a function of actuation of the actuator (102), a stroke quantity (Sréel) representing the stroke is entered and the actuator is controlled during at least a second phase of the operation switching as a function of a time interval (Tcons), this interval (Tcons) being formed as a function of the detected reaction quantity (Sréel, I, Ig) and it is notably planned to form the interval (Tcons) of the current switching operation as a function of the evolution over time (Sréel ') of the detected stroke (Sréel) for at least   a previous switching operation.   9) Method according to claim 2, characterized in that during at least one previous switching operation a quantity (Veréel) is formed representing the evolution over time of the speed of rotation of the input shaft and this is compared. quantity (Veréel), at a predetermined threshold (SW) and the target quantity (Scons) of the current switching operation is formed as a function of the quantity of the stroke   (Sréel), existing when the threshold is reached.    ) Method according to claim 1, characterized in that a value for the input speed of the gearbox (Neréel) is entered representing the input rotation speed of the gearbox and when the entered stroke (Sréel) is at the value of the setpoint variable (Scons), in a second phase, the actuator is controlled to adjust the value of the speed of rotation   gearbox input, input (Neréel) at a value their predetermined (Nsyn) -   11) Method according to claim 10, characterized in that when the value of the input rotation speed of the   gearbox, input (Neréel) is at predeter value-   mined (Nsyn), in a third phase we order   the actuator so that the entered stroke (Sréel) takes the   their of a second predetermined nominal quantity (Scons) 12) Device for controlling a switching mechanism of an automatic gearbox during a gearbox switching operation, the switching mechanism (103-108) being actuated using a controlled actuator (102), characterized by means (101) for entering a quantity (Sréel) representing the stroke, - forming a reference variable (Scons) of the stroke, - controlling the actuator during at least a first phase of the switching operation so that the magnitude of the entered stroke (Sréel) takes the value of the reference magnitude (Scons), - enter at least one reaction magnitude (Neréel, Sréel, I , Ig) which represents the reaction of the gearbox and / or the switching mechanism (103-108) and / or the actuator (102) following an operation of   the actuator, and form during a switching operation   current value, the target quantity (Scons) as a function of the reaction quantity entered (Neréel, Sréel, I, Ig)   during at least one previous switching operation.   13) Device according to claim 12, characterized by means (101) for - entering a value for the speed of input rotation of the gearbox (Neréel) representing the speed of input rotation of the gearbox, and then   - when the size of the entered race (Sréel) has taken the   their of the target quantity (Scons), in a second   phase, control the actuator so that the rotational speed   input of the entered gearbox (Neréel),   rule on a predetermined value (Nsyn).