FR2736093A1 - METHOD AND DEVICE FOR CONTROLLING THE FUEL SUPPLY OF AN ENGINE - Google Patents

Abstract

Method and device for regulating the fuel supply of an engine for accelerations and decelerations. The differences in the actual values of rotational speed and temperature are applied to a PID controller controlling a turbine burner fuel supply valve. The temperature actual value (TI) is the inlet temperature of the high pressure turbine and the rotation speed actual value (NI) is its rotation speed. The maximum allowed value (TM) is weighted before being applied to the PID controller. The weighting is done by a temperature reduction (DELTAT) from the maximum value (TM) to a weighted value (Tpond = TM - DELTAT) depending on the differences in temperature and rotational speed and the heating or cooling behavior of the motor housing.

Description

The invention relates to a method and a device for regulating

  the fuel supply of an engine during acceleration and deceleration operations, using

  maximum values set for the speed of rotation of the tur-

  bine and for the temperatures according to the different flight situations, the differences of which are applied as actual values to a PID regulator controlling a supply valve of a fuel supply line to the turbine burner.

  Such a method and such a device are known.

  DE-38 30 804. The process described in this document

  The purpose of this is to avoid over-oscillations in temperature, which are unacceptable when the engine is accelerated; for that

  it uses a PID (Proportional Integral Differen-

  tiel) using mainly input variables

  setpoint differences, weighted and differences

  the actual value of the speed of rotation and the

  temperature.; the difference in setpoint is the difference

  reference to a measured temperature of the high pressure turbine and a maximum authorized temperature or the difference between a measured rotation speed of the high pressure turbine and a maximum authorized rotation speed of the high pressure turbine. The maximum temperature allowed depends on

  the flight status of the aircraft or the operating states of the

  tor. The actual measured value is detected in a pyrometric way.

stick.

  The disadvantage of this process is that it does not

  take into account the differences in warm-up behavior or

  cooling of the crankcase and the blades at the time of access

  leration or deceleration of the engine. This difference regularly leads to errors when entering the actual temperature value at the blades of an engine; these errors result in a real measured temperature, too low during the acceleration phase. The input stage of the turbine is therefore subjected to excessive thermal demand and results in the maximum authorized temperature value being exceeded. The disadvantage is that this reduces not only the life of the turbine stage but also an excess of the speed of rotation at the time of acceleration which results in an excess of the annoying thrust for the control of the maneuvers. flight. In the case of deceleration, the current regulation does not take into account the thermal state of the motor housing so that during a

  further acceleration, the starting point, is so

  nante, a fictitious unchanged thermal state of the casing; that is

  reflected by considerable adaptation faults in the

  fuel in the new acceleration phase and

  may result in thermal overload of the turbine blades.

  The object of the present invention is to create a process

  and a device for regulating the fuel supply

  engine rant taking into account cooling errors or

  temperature for entering the actual temperature value

  engine erasures and avoiding any temperature overshoot or

  thrust, to achieve a longer life of the mo-

  tor for a maximum authorized thermal load, and guaranteed

  flying and flight maneuvers.

  To this end, the invention relates to a method of the type defined above, characterized in that the real value

  of temperature (TI) is the inlet temperature of the tur-

  high pressure and the actual value of the rotational speed

  tion (NI) is the rotation speed of the high pressure turbine or high pressure compressor and the maximum value

  temperature (TM) allowed before entering the tank

  PID regulator is weighted, the weighting being done by a

  temperature reduction (AT) of the maximum time value

  temperature (TM) at a weighted temperature value (Tpond = TM - AT) as a function of temperature and speed differences

  of rotation and of the heating and cooling behavior of the motor for acceleration or deceleration of the mo-

  tor. The invention also relates to a device charac-

  terized in that the temperature measurement installation and the speed detector take the actual values of the high pressure turbine or the high pressure compressor and a first summation point (S1) decreases by a reduction in temperature (AT ), a maximum temperature (TM) fixed according to the flight situation, and the temperature value (Tpond), weighted, resulting is applied to a downstream PID regulator which controls a regulator of the fuel supply of the motor a first order filter preceding the first summation point (S1) to delay the temperature reduction and a time constant either for heating or cooling (Te, Ta) depending on the position of an adjusting element upstream of the filter is implemented according to the temperature difference between the maximum value of

  the temperature (TM) and the actual value of the temperature (TI).

  Thus, the maximum permitted temperature of the tur-

  high pressure tank is weighted for heating behavior

  engine cranking or cooling and the resulting differences in temperature and speed are

applied to a PID regulator.

  This solution offers the advantage that the behavior

  in regulation is not defined by differences in temperature

  ideal, abstract, especially high pressure turbine

  but that the behavior actually existing for heating and cooling the

  turbine housing. The thermal radiation from the motor housing

  distorting the entry of the temperature of the turbine inlet stage is low for a low outlet rotation speed, low and develops more slowly at acceleration

  as the temperature rise of the turbine stage.

  The heat radiation values at the base of the detection

  tion of the temperature and which thus form the actual values of the temperature, consist of the values of radiation of temperature of the stage of entry of the high pressure turbine and the values dispersed coming from the casing, so that the entry of the temperature in acceleration phase results in measured values, actual too low and thus predetermines a temperature difference higher than the turbine inlet temperature, maximum allowed, than this is the case

  in reality. The same remark applies with the algebraic signal.

  that reversed for the deceleration phase; there is however no direct negative influence on the regulation cavity at the level of decelerations, since this is not defined

  analogously by a minimum temperature limit value.

  Furthermore, the present invention guarantees that in the event of deceleration, advantageously at any time, the evolution of the thermal state over time of the turbine casing is monitored; and thus the starting point at the time of a new acceleration is adapted to the thermal behavior of the motor housing. This is why the present invention avoids the deviation of the known methods and control devices from the

real situation.

  According to a preferred implementation of the method, to weight the maximum authorized temperature, a temperature reduction (AT) is delayed by a filter preferably a

  first order filter; the filter copies the ther-

  motor housing mismatch with different time constants

  annuities (Te, Ta) when the motor housing receives heat

  during a heating or acceleration phase or during

  that it restores heat during a cooling phase

  ment or deceleration, between the different engine operating states, during the different maneuvers

aircraft rolling or flying.

  This has the advantage of not only detecting the thermal radiation from the motor housing and its effect on

  the actual value measured by additional measuring installations

  corresponding and constantly detect them by corresponding measurement means but also thanks to fixed time constants, determined only once, we hold

  account of the influence of the motor housing.

  These time constants are preferably defined

  during the engine test for all the transitions

  sageable in the state of the engine to another; this information is stored and is called automatically when necessary. This has the advantage that the engine in flight does not have heavy, additional measuring devices for the

  entering the increasing or decreasing temperature of the casing but only a small plate fitted with semi-

  conductors, light in relation to the weight of a measuring device

  sure, and from which we call the constants of

pre-determined times.

  In the simplest version of the regulation process

  lation according to the invention, it is determined and recorded pre-

  reference two time constants as a function of the difference between the actual value (TI) of the temperature and the set value (TM) of the temperature, namely a time constant (Te) for heating the crankcase at the time of

  acceleration and a time constant (Ta) for cooling

  smoothing of the motor housing during deceleration. An overheating of the motor housing corresponds to a direct contribution

  until the actual value (TI) of the temperature

  ture has reached the maximum value (TM) of the temperature up to

  that, a threshold (TK). For this we first use a

  test circuit, then during engine tests,

  kills a fine adaptation of the time constants Te and Ta to

using engine behavior.

  According to the other preferential implementation of the pro-

  yielded from the invention, the temperature reduction (AT) presents

  an upper limit (ATmax) and it increases proportionally-

  to a difference (ND) between the actual value (NI) of the

  rotational speed and a rotational speed threshold value

  tion (NS) which is 5 to 20% less than the rotational speed

  maximum tion (NM). Thus the temperature reduction depends on the interval between the speed of rotation of the high pressure turbine and a constant threshold. This threshold value ensures

  advantageously entering zero the weighting of the temperature

  maximum ture authorized when the speed threshold is reached, ie; the weighting no longer depends on the speed of rotation. Beyond this threshold, the regulation behavior of the PID regulator no longer requires a reduction in

  the maximum allowable temperature.

  The rotation speed of the high pressure turbine equal to the rotation speed of the high pressure compressor is used, since this rotation speed is advantageously the fastest dynamic reaction signal in response to a variation in the fuel supply. The distance between the rotation speed of the high pressure turbine and the constant threshold defines the starting value of the maximum authorized temperature, weighted, for an acceleration and thus represents

  a measure of the initial warm-up state of the engine.

  Preferably, the temperature reduction (AT) is also weighted by a coefficient proportional to the temperature difference between the maximum temperature value

  (TM) and the actual measured temperature value (TI) and

  be the smallest temperature reduction (ATmin) and the largest temperature reduction (ATmax), taking a value

normalized between 0 and 1.

  This has the advantage that the weighting depends

  the speed of rotation can be influenced by

  their normalized between 0 and 1 for the temperature difference. As a further reduction in temperature

  (ATmax) for the temperature difference, we choose a value

  their equal to the upper limit (ATmax) of the temperature reduction for the speed difference, so that advantageously, in case of faults or errors in the input values of the speed difference , as for the adaptations of threshold errors or errors of entry of real values of the speed of rotation, the logic depending on the temperature difference is ensured by the corresponding proportional element, thus the reduction of temperature is directed in all cases towards zero and the load

  maximum temperature for the turbine stages is completely exhausted without exceeding.

  According to another preferential implementation of the process of the invention, the increase in the reduction in temperature

  ture (AT) on the basis of the speed difference30 (ND) is multiplied by the normalized values resulting from the temperature difference (TD) at a multiplication point (M) to be applied to the filter whose output signal

  (AT) is subtracted at a summation point (S1) from the maximum temperature value (TM) to form the pon-

  derated (Tpond). This multiplication means that the rise propor-

  tional between the smallest temperature reduction (ATmin) and the largest temperature reduction (ATmax) is combined with the rise in temperature reduction based on the difference in rotational speed. It further results that the temperature reduction (AT) when the smallest temperature reduction (ATmin) is reached and thus when the actual temperature (TI) approaches the maximum authorized temperature (TM) is set to 0 by the above multiplication and

  the PID regulator regulates the maximum temperature

male allowed.

  According to the invention, this method is implemented by

  a regulatory device mentioned at the beginning.

  For this, the motor includes a temperature measurement installation preferably a pyrometer to define the actual value of the inlet temperature of the turbine as well as a speed sensor to define the actual value of the speed of rotation of the turbine. The installation

  for measuring the temperature and the speed sensor

  tion take the actual values of the high pressure turbine or the high pressure compressor. A first summation point (S1) decreases a fixed maximum temperature (TM), depending on the

  vol, a temperature reduction (AT) to balance the

  their maximum before this signal is applied to a regulation

  PID author. The output signal from the PID regulator controls an element for adjusting the fuel supply to the engine. A

  first order filter to delay the temperature reduction

  erection is located upstream of the first summation point (Sl); the delay is defined by a heating time constant or

  cooling (Te, Ta) which is used depending on the po-

  sition of an adjustment element upstream of the filter.

  These measures; The advantage of regulation is that it does not require the development of apparatus and therefore does not entail additional weight compared to known regulation methods and devices, but only to use

  the existing calculation and regulation capacities, and

  so much closer to reality,

  so that the permissible thermal load of the motor is com-

  completely exhausted for the maximum duration of the components submitted

  at the highest thermal load; the engine thus offers the maximum maneuvering possibilities for the aircraft.

  According to a preferred embodiment of the in-

  vention, at a second summation point (S2), we form the dif-

  speed of rotation (ND) between a threshold value of the speed of rotation (NS) and the actual measured value of the speed of rotation (NI); this summation point is followed

  by a proportional element whose output signal is

  plicated at a multiplication point (M). At a third summation point (S3), a temperature difference (TD) is formed between the maximum temperature value (TM) and the actual temperature value (TI). This temperature difference is applied to a proportional element but by limiting the

  minimum value and the maximum value of the time difference

  temperature. The output signal of the proportional element is coupled to the multiplication point (M). The exit from the point of

  multiplication (M) is applied to the first order filter.

  This preferred embodiment offers the advantage

  tage that the maximum permitted temperature reduction is set to zero before a rotational speed is reached

  maximum when you reach the threshold and before reaching-

  dre the maximum temperature allowed; so for example in

  the final phase of an acceleration, we have the maximum temperature

  authorized male which represents the control variable to be

  nir for the PID regulator; this makes it possible to adjust a corresponding maximum thrust for a maximum speed of rotation without the motor being overloaded by temperature overruns or by overruns of the rotational speed,

  interfering with the execution of correct flight maneuvers.

  The present invention will be described below in detail.

  more detailed with the aid of a preferred embodiment.

  rentiel shown in the accompanying drawings, in which: - Figure 1 shows a regulating device for setting

  implementation of the regulation process according to the present invention

  tion. - Figure 2 shows a block diagram of an engine equipped with lines for the measurement and regulation signals used to control the fuel supply using the

  regulation method according to the invention.

  - Figure 3 shows a timing diagram of the speed of rotation

  of the high pressure turbine and the temperature of the turbine

  high pressure tank applying the regulation process according to the invention. Figure 1 shows a control device 1

  for implementing the regulatory process according to the invention

  tion in order to regulate the fuel supply of an engine during acceleration and deceleration operations, at

  using maximum values set for rotational speeds

  tion of the turbine and the temperatures, for different situa-

  flight or rollover maneuvers, and whose differences from the actual values are applied to a regulator

  PID 10 (Proportional-Integral-Differential regulator). The re-

  DIP 10 regulator controlled via a supply line

  15, a fuel supply valve shown in Figure 2 under reference 2; it defines food

  in fuel from the burner shown in FIG. 2 under the reference

  rence 14. The fuel supply valve is mounted in a fuel supply line with the reference 3 in Figure 2

  The actual value of the temperature TI is the temperature

  erasure of the high pressure turbine which, in this

  example is determined pyrometrically at the level of

  bes of the fixed grid of the high pressure turbine bearing the reference 16 in FIG. 2. The actual value of the rotation speed NI is the rotation speed of the high pressure turbine measured in the high pressure compressor 11 according to the

figure 2.

  Maximum allowable temperature values

  TM for the different operating states of the motor 17 se-

  lon Figure 2 are weighted for the different taxiing or flight maneuvers of the aircraft before being applied to the PID regulator 10; for this, the maximum values TM are reduced by a reduction in temperature AT at the level of a

  first summation point S1. This reduction in temperature AT is modified as a function of a difference in speed of ro-

  tation, of a temperature difference and of the behavior on heating and cooling of the motor casing as shown under the reference 12 in FIG. 2, by

  several function levels 6, 7, 8, between a maximum value

male ATmax and zero.

  In the exemplary embodiment of FIG. 1, a threshold is first set for the speed of rotation Ns according to different operating states of the motor, and which is 5 to 20% below the speed of NM maximum rotation of the high pressure turbine, which gives the formula:

NS = NM (5 to 20%). NM

  A second summation point S2 forms the difference in speed of rotation ND between the threshold NS and the real value NI of the speed of rotation according to the formula:

ND = NS - NI;

  the signal is applied to a logical combination with an element

  proportional 6. When the actual value NI is greater than

  lower than this threshold NS, there will be a negative value for ND, so that function 8 sets the temperature reduction AT to zero; thus depending on the speed of rotation, the motor can be accelerated to the maximum fixed value TM of the temperature. If the actual speed NI is below this threshold NS, the temperature reduction AT increases until it reaches

  maximum reduction in temperature ATmax; in this example-

  ple, this maximum reduction is equal to 100 C; this reduction

  tion is made in proportion to the difference in rotational speed ND between the threshold NS and the actual value NI. In this example, the gradient corresponds to 4 C for 1% variation in the speed of rotation. The maximum reduction of the temperature ATmax, for example equal to 100 C, is obtained in this embodiment for a difference in rotation speed of

25% compared to the NS threshold.

  The logic combination output signal in

  proportional element 6 is applied to a point of multipli-

  cation M which multiplies the reduction in temperature set by logic combination 6 as a function of the speed of rotation

  with normalized values between 0 and 1. These values

  their normalized values depend on the temperature difference TD between the actual value TI and a maximum authorized value TM of the temperature of the high pressure turbine according to:

TD - TM - TI

  This temperature difference TD is formed in

  a third summation point S3. In a logical combination

  that with proportional element 7, we attribute to this difference

  temperature rence TD, normalized values between 0 and 1. As long as the temperature difference

  TD exceeds the value ATmax the output of the logistic combination

  that 7 is level 1; so the multiplication to the point

  multiplication M gives overall a reduction in temperature

  ture defined by the proportional element 6 only from the difference in rotational speed. When the temperature difference TD exceeds a minimum value ATmin, the output

  of the logical combination 7 is set to the value 0; the multi-

  plication at the multiplication point M thus overall gives a constant temperature reduction of 0 C and the weighted temperature Tpond at the input of the PID regulator 10 is set to TM; the maximum temperature value on the blades of the turbine inlet stage can thus be expected without thermal overshoot. Between the multiplication point M and the summation point S1, there is a first order filter 4 which delays the temperature reduction AT to simulate the behavior at

  heating or cooling the motor housing. The cons-

  filter time aunt T is equal to Te for heating and Ta for cooling. These constants are taken from a memory 9 via an adjustment element 7 with the switching positions 0 for cooling and 1 for heating to be applied to the filter 4. The choice between the heating time constant and the cooling time constant depends on a logic function 8 whose

  the input is linked to the summation point S3 so that the difference

  TD temperature rence forms the input signal. As long-

  time that a temperature difference threshold TK is not exceeded, that is to say that the inlet temperature of the

  turbine is close to the maximum temperature, the motor housing

  heaters and the output of logic function 8 is set to state 1; the heating time constant Te of the motor housing defines the time constant t of the filter 4 When the temperature difference TD exceeds the temperature difference threshold TK, then the input stage of the turbine is at a higher temperature low and the casing gives off heat to the environment; this is done with a cooling time constant Ta. In this case, the logic function 8 fixes its start at zero and the adjusting element 5 is put in the position 0; so for the time constant T of filter 4, we have the

  constant of cooling time Ta of the casing.

  The temperature difference threshold is preferably set equal to the maximum temperature reduction ATmax;

  this means that any decrease in temperature reduction

  ture is done by multiplication with a weighting coefficient

  depending on the temperature difference and

heating time constants.

  Figure 2 shows a block diagram of a mo-

  sensor 17 equipped with measurement lines 18, 19 and feedback signals

  gulation 15 for controlling the fuel supply implementing the method according to the invention. A regulating device 1 like that shown in detail in FIG. 1, receives signals for measuring the real values of the speed of rotation NI and of the temperature TI as well as the maximum value of the temperature TM and the thresholds of the speed of NS rotation. The output signal from the regulating device 1 is transmitted by the signal line 15 to an adjustment valve 2 equipping the fuel supply 3 to the burner 14 of the engine 17. The inlet temperature of the high pressure turbine TI is measured on a blade 16 of the fixed grid of the first stage of the turbine. The rotation speed of the NI high pressure turbine is determined in zone 20 of the high pressure compressor

at the measurement point 11. FIG. 3 shows a chronogram of the speed of rotation NI of the

  high pressure turbine and high pressure TI temperature using the regulation process according to

  the invention. At time to, there is a state of thermal equilibrium

  that in the engine running at the speed of rotation NIl by

  example in slow motion. At time tl, we predetermine a new

  vale speed of rotation NI2 at which the motor must be accelerated. A maximum value TM of the temperature of the blades of

  the high pressure turbine which cannot be exceeded is as-

  associated with this predetermined rotation speed. Like the car-

  engine ter is still relatively cold, in the initial phase, due to the dispersing effect of the engine casing, a temperature of the blades is detected below the actual value, so that in the case of usual regulation, there is too great a difference in temperature between the value

  actual temperature TI and maximum value TM predeter-

  mined by a PID regulator; this annoyingly leads to

  the excessive temperature curve Tu shown in broken lines

  terrompus; the blades are thus damaged. The overshoot is coupled with an overspeed

  NU which hinders the execution of correct flight maneuvers.

* This is why, according to the invention, the predetermined maximum value TM of the inlet temperature of the high pressure turbine at time t1 is reduced by the value ATmax.

  to arrive at the weighted temperature Tpond.

  As soon as the speed of rotation NI at time t2 exceeds the initial speed of rotation NIM corresponding to a

  maximum temperature reduction ATmax, the reduction in tem-

  TF temperature at the filter inlet is pro-

  portable until at time t3 a threshold speed of rotation NS is reached. Then, the temperature reduction TF at the filter input is zeroed. During the first phase of acceleration, between instants t1 and t4, the temperature reduction AT follows the filter inlet temperature TF with cooling time constants Ta

  bigger. As soon as at time t4, the temperature difference

  ture TD exceeds the temperature difference threshold TK, we switch to smaller heating time constants Te and the temperature reduction T is reduced correspondingly to the heating time constant Te to zero.

  This shows the effect of the delay filter according to the

  vention using different time constants to heat and cool the motor housing; for the real curve of the temperature TI and the speed of rotation NI, it is clear that one now reaches exactly the predetermined maximum values without exceeding. The integrated safety logic according to FIG. 1 has no influence on the embodiment shown in FIG. 3 thanks to the reduction in temperature depending on the speed of rotation of the functional stage 6, obtained by the

  weighting coefficient depending on the temperature difference of the functional stage 7.

Claims (7)

R E V E N D I C A T IONS   1) Method for regulating the fuel supply of a   engine during acceleration and deceleration   tion, using maximum values set for the turbine rotation speed and for the temperatures according to the different   different flight situations, the differences of which are applied   as actual values to a PID regulator controlling a supply valve for a fuel supply line to the turbine burner, characterized in that the actual temperature value (TI) is the inlet temperature of the high pressure turbine and the actual value of the   rotational speed (NI) is the rotational speed of the tur-   high pressure or high pressure compressor and the   their maximum allowable temperature (TM) before   introduction into the PID regulator (10) is weighted, the weight   ventilation by a reduction in temperature (AT) of   the maximum temperature value (TM) at a time value   weighted temperature (Tpond = TM - AT) according to the differences   those of temperature and rotational speed and behavior   to the engine heating and cooling (12) for access   leration or deceleration of the engine (17).   2) Method according to claim 1, characterized in that   temperature reduction (AT) is delayed by   diary of a filter (4) and this filter (4) copies the thermal behavior of the motor casing (12) with different time constants (Te, Ta) for heating and cooling between different operating states in situations of different theft.   3) Method according to claim 1 or 2, characterized in that to memorize the behavior on heating of the casing (12) of the engine under the effect of the acceleration of the engine (17) or to record the behavior on cooling of the casing (12) during the deceleration of the engine (17), the thermal time constants (Te, Ta) during the supply of heat or the evacuation of heat from the casing (12) of the engine, at the passage of an operating state to the other of the engine are determined during the test for different operating states of the engine and this information is stored in memory.   4) Method according to any one of claims 1 to 3, characterized in that   temperature reduction (AT) has an upper limit   greater (ATmax) and increases in proportion to a difference   rence (ND) between the actual value (NI) of the rotation speed and a rotation speed threshold (NS) this threshold being 5 to 20% lower than the maximum rotation speed (NM), the temperature reduction ( AT) being zero for the negative differences.   ) Method according to any one of claims 1 to 4,   characterized in that the temperature reduction (AT) enters a minimum reduction   temperature (ATmin) and a maximum reduction in temperature   ture (ATmax) increases in proportion to the temperature difference between the maximum value and the temperature (TM) and   the actual measured temperature value (T) and at this increase   mentation we associate the normalized values between 0 and 1, the greatest reduction in temperature (ATmax) being equal to the upper limit of the reduction in temperature   (ATmax) due to the difference in rotational speed.   6) Method according to claim 5, characterized in that the rise in temperature reduction (AT) on the basis of the difference in rotational speed (ND) is multiplied by   normalized values resulting from the difference in temperature   ture multiplied at a multiplication point (M) and applied to the filter (4) whose output signal (AT) is subtracted at a first summation point (S1) of the maximum value of the   temperature (TM) to form the weighted temperature (Tpond).   7) Device for regulating the fuel supply to an engine, comprising a temperature measurement installation, preferably a pyrometer to define the actual value (TI) of   the inlet temperature of the turbine and a speed sensor   rotation speed to define the actual value of the turbine speed (NI), characterized in that the temperature measuring system (16) and the speed sensor (11) take the actual values of the high pressure turbine (13) or high pressure compressor   (20) and a first summation point (S1) decreases by one reduction   temperature (AT), a maximum temperature (TM) fixed according to the flight situation, and the temperature value (Tpond), weighted, resulting is applied to a PID regulator   (10) downstream which controls an adjusting member (2) of the feed   fuel (3) from the engine (17), a filter (4) from the   mier order preceding the first summation point (S1) to delay the temperature reduction and a time constant either for heating or cooling (Te, Ta) depending on the position of an adjustment element (5) upstream of the filter (4) is   implementation depending on the temperature difference   between the maximum temperature value (TM) and the value   their actual temperature (TI).   8) Device according to claim 7, characterized by a second summation point (S2) forming the difference between   a speed threshold (NS) and the actual value me-   after the rotation speed (NI), a proportional element (6) is provided downstream whose output is applied to a multiplication point (M), a temperature difference between the   maximum temperature value (TM) and the actual value of.   the temperature (TI) being formed at a third point of som-   mation (S3) connected to a proportional element (7) with a diff-   minimum (ATmin) and maximum (ATmax) temperature reference, element whose output is coupled to a multiplication point (M), the output of this multiplication point (M) cooperating with the first order filter (4)