GB2440236A

GB2440236A - Method for the regulation of the boost pressure of an internal combustion engine

Info

Publication number: GB2440236A
Application number: GB0713473A
Authority: GB
Inventors: Joerg Ballaue; Siegmar Lampe; Daniel Beese; Holger Braun
Original assignee: Audi AG
Current assignee: Audi AG
Priority date: 2006-07-14
Filing date: 2007-07-11
Publication date: 2008-01-23
Anticipated expiration: 2027-07-11
Also published as: FR2904369B1; ITMI20071378A1; GB2440236B; GB0713473D0; CN101105151A; DE102006032836A1; FR2904369A1; DE102006032836B4; CN101105151B

Abstract

The boost pressure of an internal combustion engine is regulated, whereby the boost pressure regulating device has a bypass device, for preference in the exhaust gas flow of the internal combustion engine. The bypass device exhibits an actuator which clears the bypass device when a specified actuator reference value is reached. According to the invention, as a function of specified and/or detected combustion engine parameters ('engine filling', rlsol and engine speed, nmot). An adaptation amount is determined as a correction value (pwgad), by means of which the operating point-dependent specified reference value of the actuator is adapted for a specified regulation deviation (dpvdk) between a boost pressure actual value and a boost pressure reference value.

Description

1 P6065

Description

Method for the regulation of the boost pressure of an internal combustion engine for motor vehicles The invention relates to a method for the regulation of the boost pressure of an internal combustion engine in accordance with the preamble to Patent Claim 1.

Such a boost pressure regulating system is known, for example, from DE 195 02 150 Cl. A boost pressure regulating device in that case comprises an integral regulator, of which the integration of the regulation deviation is limited to a specified limit value in order to avoid high overshoot. In addition to this, various different limit values are specified for stationary and dynamic operational states, whereby the dynamic limit value is provided with corrections dependent on operational characteristic values and with an adaptive correction, as well as being additionally increased by a safety margin. The intention of this is that, even in the event of dirtying or wear of the * components involved iii the regulating system, a sustained good regulating quality will be assured. A similar boost pressure regulating system is also known from DE 198 12 843 A I, with which, in order for the boost pressure regulation to operate over a wide working range vitliout the regulating procedure becoming too slow, or overshoot occurring during the regulation, a manipulated variable for an actuator is transformed in such a way that, afer the transformation of the values, a linear relationship pertains between the manipulated variable and the regulating value. A correction value for a limit value, which is specified for an integral fraction of an integral regulator i1 connection with a boost pressure regulating system is also known from DE 197 2 861 Al.

As a result of series scatter incurred by the method of manufacture, for example with the spring preliminary tension in the waste gate area, as a specific example in the area of a spring-loaded pressure capsule which takes effect on a bypass valve, a different regulating behaviour may be incurred for different vehicles, which is undesirable. The same also applies to ageing manifestations of the components in the area of the boost pressure regulating valve, which can likewise vary from vehicle to vehicle.

The object of the present invention is therefore to provide a method for the boost pressure regulation of an internal combustion engine, by means of which, in particular in the event of series scatter and ageing manifestations of the components of the controlled system, a high and reproducible regulating quality can be provided, with the least possible regulation deviation between the reference boost pressure and the actual boost pressure. S. * . * ..*

This object is resolved with the features of Claim I. * . **.

**:* According to Claim I, as a function of specified and/or detected combustion engine :. parameters, an adaptation amount is determined as a correction value, by means of which the reference or threshold value of the actuator of a bypass device, dependent *:* on the operating point, is adapted for a specified regulation deviation between a boost pressure actual value and a boost pressure reference value, dependent on the operating point.

With such a method, a system behaviour which remains consistent over the series in respect of regulating quality and reproducibility can be guaranteed, since, due to the determination of the correction value for the operating point-dependent reference value of the actuator, even in the event of series scatter and ageing manifestations of the components of he controlled system of the boost pressure regulation, such an adaptation is possible which takes account of these manifestations In other words, with such an operating point-dependent adaptation routine for a boost pressure regulating system, which takes effect omi the reference value ofan actuator of ihe bypass device, regulation of the boost pressure can take place with the least possible regulation deviation.

In this situation, for preference, the correction value is determined as a function of a reference engine filling and/or an engine revolution speed of the internal combustion engine. Specifically, in this Situation the actuator reference value is a reference pressure of a spring-loaded pressure capsule, which in turn is a component part of a bypass valve forming a boost pressure regulating valve, which for preference can be subjected to pressure by a control device of the boost pressure regulating system. The correction value in this situation is a pressure value which for preference is subtracted from the actuator reference value.

The correction value is for preference determined by means of an adaptation device in :. an adaptation routine, which is subdivided into an adaptation clearance, an index determination in the engine revolution speed direction, an index determination in the reference engine filling direction, an interpolation of an adaptation group of *:*::* characteristics, an intermediate storage of the adaptation values, clearance for the writing of the adaptation group of characteristics, a quality check of the learned values, and a copying of the learned values into a memory array. The correction value is in this situation deposited in a non-volatile memory array (adaptation array), of which the addressing variables are for preference the engine revolution speed and the reference engine filling.

The invention is described in greater detail hereinafter on the basis of figures.

These show Fig I A sketch in diagrammatic form showing the principles of an electronic boost regulating system, Fig. 2 A block circuit diagram in diagrammatic form of the reference value adaptation according to the invention of a pressure capsule of the boost pressure regulating device, Fig. 3 The writing of an intermediate or buffer memory, Fig. 4 The writing of a work adaptation array, and Fig. 5 By way of example, the manipulation of the work adaptation array in one direction.

Fig. I shows in diagrammatic form a sketch of the principle of an electronic or electro-pneumatic boost pressure regulating system for an internal combustion engine I, of which the optimum boost pressure is regulated by means of a boost pressure : * control device 2 of the boost pressure regulating system. A pressure sensor, not shown *..... here, detects the boost pressure and passes this information on to the boost pressure * *, control device 2, which in turn activates a pulse valve 3, of which the manipulated variable controls the opening cross-section of the pulse valve in an inherently known manner. *..*

As an further be derived from Fig. I, located in the suction pipe 4 is a compressor S.....

* wheel 5 of an exhaust gas turbocharger 6, of which the turbine wheel 7 is located in the exhaust pipe 8 From the exhaust pipe 8, a bypass line 9 branches off as a waste gate, by means of which at least a part of the exhaust gas can be conducted into the exhaust gas channel, bypassing the turbine wheel 7. l'o clear the bypass line 9, a boost pressure regulating valve or bypass valve 10 is provided, which exhibits a pressure-loaded pressure capsule I), which, as a function of the preferentially electrical or electro-pnetimatic pulse valve 3, is subjected in a specified manner to a specified boost pressure as an actuator reference value, dependent on the operating point.

S

The basic function in this situation is as follows: The engine control device calculates a reference boost pressure at every point in time, from the driver's wishes. On the basis of this reference boost pressure and the boost pressure measured downstream of the conipressor, an actuating signal is calculated for the pulse valve 3, which is located between the control line 12 to the waste gate pressure capsule 11 and the Suction pipe 4 and the pressure line downstream of the compressor (see Fig. 1). If it is intended that the boost pressure should be increased in accordance with the reference boost pressure, by actuating the pulse valve 3 the valve cross-section between the control line 1 2 and the suction pipe 4 is steadily increased This causes the pressure to drop in the waste gate capsule II, and the bypass valve 10 begins to close due to the spring preliminary tension of the waste gate pressure capsule II. This causes the turbine mass flow to increase steadily, the implemented power which is made available to the compressor, and also the boost pressure.

: .. If it is intended that the boost pressure should be reduced in accordance with the S...

reference boost pressure, by actuating the pulse valve 3 the valve cross-section * * between the control line 12 and the suction pipe 4 is steadily reduced. This causes the pressure to rise in the waste gate capsule II up to a maximum of the value of the boost pressure, and the bypass valve 10 begins to open against the spring preliminary tension of the waste gale pressure capsule II. This causes the turbine mass flow to *5** decrease steadily, the implemented power which is made available to the compressor, * and also the boost pressure.

Fig 2 shows in diagrammatic form the method according to the invention for the boost pressure regulation of the internal combustion engine I, with which the adaptation device 14 according to the invention is incorporated into the model-based pilot device 13. Specifically, the pilot device 12 consists of a first control device 13a, in which, as a function of the reference mass flow through the bypass line 9, i.e. what is referred to as the waste gate throughflow, and of a waste gate side or bypass side liii pressure characteristic curve a reference pressure is specified at the pressure capsule II, which iii turn, in a second control device I 3b, in conjunction with a boost pressure by way of a boost pressure pulse valve group of characteristics, leads to a reference pulse duty factor in the bypass area.

In order to adapt the reference pressure at the pressure capsule II in respect of the series scatter and ageing manifestations of the components of the controlled system of the boost pressure regulation which might arise, by means of the adaptation device 14 as a function of the engine reference filling, the engine revolution speed, and an operating point-dependent specified regulation deviation (dpvdk) between a boost pressure actual value and a boost pressure reference value, an adaptation amount is determined as a correction value (pwgad), which represents a pressure value which is subtracted from the calculated capsule reference pressure value.

Specifically, in this situation, with the adaptation device 14 an adaptation routine is run through, which is subdivided into a large number of hierarchies, the first of which is a clearance for adaptation. In this situation, clearance is given for the take-up of data into an intermediate memory of the adaptation device 14, still to be described in (I.s greater detail, when the boost pressure build-up has been concluded, and a D-fraction * * of the load pressure regulation has been formed during the transient event. Clearance for reading a memory array takes place, by contrast, in an interpolation routine, which is likewise still to be explained in greater detail. **.*

As an additional condition, clearance of the data take-up first takes place when the 6***** * waste gate regulation is active and the eiigine temperature has exceeded a threshold which can be applied. As well as this, the pulse duty factor for the pulse valve 3 must lie between the minimum and maximum limits which can be applied. Specifically, for the situation in which (lie pulse duty factor is located at the lower or upper limit, the continuation of the adaptation has no sense, since the setting range of the actuator or pulse valve 3 is fully exhausted.

Moieover, (lie reference load demand which is written by the quotient from the reference engine filling (rlsol) and the maximum engine filling (rlmax) is intended to trigger the adaptation clearance, which can be applied for the full and part load opelating Phases. Under part load, for the evaluation raiige of the adaptation routine, the gradient of the reference boost pressure should lie within a band which can be applied, whereby clearance for adaptation should be prevented during the dynamic pressure build-up.

Provided that all the individual conditions are present, clearance for writing of the correction value takes place after an additional debounce time. In principle, the selection of fewer conditions is also possible.

A further adaptation routine of the adaptation routine 14 is represented by the index setting in the direction of the engine revolution speed (nmot) and in the direction of the reference engine filling (rlsol). For the assigning and reading out of values from the memory array, whole figure indices are needed for the addressing of the individual *.:: array elements. In this case, by way of example, two indices are formed in the nmot direction and two indices in the rlsol direction. Together, they describe the adjacent array elements of the current working and operating points respectively.

The indices for the array axes are constantly determined from the current values of the engine revolution speed (nmot) and the reference engine filling (risol) Which indices are allocated to which value ranges of the engine revolution speed and the reference S.....

* engine filling respectively is represented by the table shown in Fig. 3.

Within the index calculation, bits are formed which display the moment of an index change In order to prevent a "toggling" of these values in the case of a stationary working point, an index change is first displayed when it has occurred twice in the same direction. The index change bits are required for the point of time of the evatuation of the system status and the subsequent writing of the intermediate memory.

Next, the interpolation of the adaptation group of characteristics takes place, whereby, for the generation of the correction value (pwgad) for the current working point, the memory array (PWGADAP) is read out.

Once the interpolation clearance has been given, the array can be linear or bilinear interpolated or read out respectively. If the current working point lies within the limits of the memory array, the bilinear interpolation is started. If the working point is outside the limits, linear interpolation takes place. For the linear interpolation, fixed engine revolution speed and reference engine filling values are drawn on, which have likewise been calculated during the indexing calculation in accordance with the current index.

The bilinear interpolation is carried out in accordance with the equation: g(x1,y) = U00 x [((x-xi)(y-yi)I(xo-xi)(yo-yi))] + U01 x [((x-xi)(y1-yo)/(xo-xi)(Y -yo))] + U10 x [((x-x)(y-y1)/(x i-xo)(yo-yi))] + U1 x [((x-xo)(y1-yo)/(xj-xo)(yi-yo))} The factors U, U01 U10 and U are the adjacent points of the current working point which are generated from the memory array. The value x stands for the engine revolution speed nmot and the value y for the reference engine filling rlsol. The index * i represents the current working point, the index 0 the left or lower intermediary point respectively starting from the working point, and the index I the right or upper * intermediary point *.* * The linear interpolatioii is carried Out when one of the two array inputs (engine revolution speed nmot or reference engine filling rlsol) falls short of or exceeds the minimum or maximum intermediary point. In this case, interpolation takes place into the other direction in each case.

As can further be derived from Fig. 3, the intermediate point indices and the amendment amount (pwgreg) belonging to them in each case are transferred into an intermediate memory, whereby the value (pwgreg) of the application amount is the boost pressure regulation adaptation which is calculated, by way of an amplification characteristic curve which can be applied, from the regulation deviation (which is read out by the amount of the system constant after an index change) To do this, the temporal sequence of the index change bit is delayed via a ring memory The delay time can be applied. The amendment amount (pwgreg) is added later to the current adaptation value at the individual working point of the memory array, in order to determine the final correction value (pwgadp).

In addition, the adaptation value of the first working point can be determined directly and proportionally after the transient event of the boost pressure. This is necessary, because at the time of the transient event it is not assured that an index change is taking place. This proportional adaptation value is then assigned to the element of the memory array which was determined by the index before the point of time of the adaptation clearaiice.

In addition to this, the multiple entry of the same working points into the intermediate *.:::: memory can be prevented. To do this, every new working point during a filling cycle of the intermediate memory is identified in a bit array. If this working point occurs again in (lie same cycle, the entry of this working point into the intermediate memory * is prevented.

After the end of an operational phase of the boost pressure regulation, clearance is **** issued for the transfer of the values contained in the intermediate memory. The values ****** * 1 deposited in the intermediate memory are transferred into a work adaptation array, as is represented diagrammatically in Fig. 4.

The amendment amount (pwgreg) is added to the value of a work adaptation array to which it pertains. Every working point contained in the intermediate memory is unitiall)' identified for this purpose in a work bit array, in order that this cannot be manipulated by the subsequent continuity check, which will be referred to in greater detail hereinafter. The values of the intermediate memory cannot be transferred directly into the memory array and into the bit array, since, during the subsequent manipulation of the matrix elements, the boust pressure regulation could again become active and therefore, when the current correction value (pwgad) is read out, discontinuities could occur. With the transfer into a work adaptation array and the l0 copying of the array values (with boost pressure regulation not active) into the adaptation array after the continuity check, a consistent read out of the correction value (pwgad) can be assured at any time.

After the transfer of the intermediate memory contents into the work adaptation array, in the next hierarchical step the continuity manipulation of the array takes place. To do this, first the lines (nmot direction) are manipulated in the work adaptation array individually for each working point transferred out of the intermediate memory. In other words, this means that adjacent working points which are not started in the current learning cycle or, respectively, have not been identified as adequately adapted in the bit array, will be raised or lowered in relation to the current working point, in order to avoid discontinuities. If an adjacent working point is represented in the learning cycle which is already adapted (entry in the bit array), the manipulation will be stopped for this and all following working points in the line concerned of the work *:*::* adaptation array. After the manipulation of the line for the current working point of * the learning cycle, the gap is then manipulated in the same manner.

Specifically, for a clearance of the manipulation of the line direction, a check is made on whether the elements to the left of the element currently under consideration *s*.* * which was transferred out of the intermediate memory into the array, have been identified or not For the left-hand element, the element at the bottom left, and the element top left, a check is made in each case as to whether a marking has been set.

The same check also takes place in the right-hand direction. If no marking is found, clearance is given for the manipulation for the left, the right, or both sides.

If clearance is given for the matrix manipulation in the line direction, a mean value is formed from the left and right values of the working point which is to be manipulated (p = (p+ p1, )12). This is either carried out as far as the left or right edge of the work adaptation arlay, or only as far as a point which is already marked as "learned".

This procedure is represented for one direction by way of example in Fig 5. In this situation, the new value for the index I is a learned value transferred out of the intermediate memory 11.

After the line of the current working point has first been checked and manipulated if necessary, a check is now conducted on the gaps. This relates to the gap of the current working point and the gaps of the elements of the line which may have been manipulated. A check is carried out in each case for the upper and lower element of the work bit array, as to whether a manipulation is permissible, and a clearance is issued into the upper direction, lower direction, or both.

If a clearance for the matrix manipulation is effected in the gap direction, then, by analogy with the manipulation in the line direction, a mean value is formed from the upper and lower values of the working point which is to be manipulated. This is either carried out one after another as far as the upper or lower edge respectively of the work adaptation array, or only as far as a point already marked as "learned".

Because all the values transferred out of the intermediate memory into the work adaptation array were initially marked in the work bit array, after the ending of the continuity check a check of these values is carried out, with regard to their quality.

Only a value to be marked as "learned" will be identified in the work bit array. * * ****

After (lie continuity iianipulation has been carried out for all the values of the * intermediate memory, the values of (lie work bit array and the work adaptation array are transferred into the bit array or the memory array respectively, from which the current correction values (pwgad) are read out as a function of the operating point and are subtracted from the calculated capsule reference pressure. The memory array is a non-volatile adaptation array (PWGADAP), the addressing variables of which, here in the example and as most preferred, are the engine revolution speed (nmot) and the reference engine filling (risol). P6065

Claims

Claims I. Method for the boost pressure regulation of an internal

combustion engine with a boost pressure regulating device, by means of which the boost pressure of an internal combustion engine is regulated, whereby the boost pressure regulating device exhibits a bypass device, for preference in the exhaust gas flow of the internal combustion engine, and whereby the bypass device exhibit an actuator which clears the bypass device when a predetermined actuator reference value is reached, characterised in that * . * *.* S..' as a function of the specified and/or acquired combustion engine parameters (rlsol, nmot) an adaptation amount is determined as a correction value (pwgad), by means of which the operating point-dependent reference value specified for the actuator for a specified regulation deviation (dpvdk) is adapted between a boost pressure actual value and a boost pressure reference S...

* value.

* ..*.* * . 2 Method according to Claim I, characterised in that the correction value (pwgad) is determined as a function of a reference engine filling (risol) and/or an engine revolution speed (nmot) of the combustion engine. - 3. Method according to Claim I or 2, characterised in that the actuator reference value is a reference value of a spring-loaded pressure capsule, which is a constituent part of a bypass valve forming a boost pressure regulating valve, which for preference can be subjected to pressure via a pulse valve controlled b a control device of the boost pressure regulating system, and that the correction value is a pressure value which is subtracted from the actuator reference value.

4. Method according to one of Claims I to 3, characterised in that the correction value is determined by means of an adaptation device, in which the correction value determined is deposited in a non-volatile memory array (PWGADAP), the addressing variables of which are the specified and/or determined combustion engine parameters, in particular an engine revolution speed (nmot) and/or a reference engine filling (rlsol).

5. Method according to Claim 4, characterised in that the adaptation device :. exhibits a data take-up clearance device, by means of which clearance is give for the data take-up if at least one, and for preference all, of the following conditions is/are fulfilled: * .* * S * * a) The boost pressure build-up has been concluded, b) The D-fraction of the boost pressure regulating device has died out, c) The bypass device is active or cleared for operation, ***.

*: d) The engine temperature corresponds to a specified minimum temperature or has exceeded it.

e) The pulse duty factor or control signal respectively for a boost pressure pulse valve of the boost pressure regulating device is located within a predetermined lower and upper pulse duty factor limit, which is specified as a function of a pulse duty factor minimum value and a pulse duty maximum factor respectively, fl The reference load requirement reaches or falls short of a specified reference load value, whereby the reference load requirement is determined for preference by the quotient from the reference engine filling (ilsol) and the maximum engine filling (rlmax), g) In the event of a part load operation, the gradient of the reference boost pressure lies within a predetermined range 6 Method according to Claim 5, characteriscd in that Ihe data take-up is cleared when the minimum of one condition is present after an additional debounce time.

7. Method according to one of Claims 4 to 6, characterised in that for the assignment and reading out of values of the memory array (PWGADAP) at least one index, and for preference two indices, is/are formed in the direction of at least one combustion engine parameter, which describe in the memory array a current operating or working point respectively of the combustion engine.

8. Method according to Claim 7, characterised in that at least one index, and for preference two indices, is/are formed in the engine revolution speed direction and at least one index, and for preference two indices, are formed in the reference engine filling direction, which together describe the adjacent array * elements of a current working point. ***

9 Method according to Claim 7 or 8, characterised in that the index or indices ***.

are constantly determined from the current determined or calculated values of * ****S * the specified combustion engine parameters, in particular the engine speed and/or the reference engine filling, to which specific index values are allocated in a group of characteristics.

10. Method according to one of Claims 7 to 9, characterised in that, during the index calculation, bits are Formed which indicate the point of time of an index change IC) I I. Method according to Claim 10, characterised in that an index change is only indicated if it occurs at least twice in the same direction.

12. Method according to one of Claims 4 to II, characterised in that, for the determination of the correction value (pwgad) for the current working point, the memory array (PWGADAP) is read out and interpolated.

13. Method according to Claim 12, characterised in that, in the event of a current working point lying within the limits of the memory array (PWGADAP), an interpolation is carried out in accordance with the following equation: g(x,y1) = Uoo x [((x,-xt)(y1-yJ)/(xo-x,)(yo-yi))] + U01 x [((x-x,)(y,-yo)/(xo-xI)(yi-yo))J : ... + U)0 x + U1 x [((xrxo)(yi-yo)/(xi-xo)(yi-yo))] * *. whereby U00, U01 U10 and U11 are the adjacent points of the current working point which are generated from the memory array, whereby x stands for a first combustion engine parameter, in particular for the engine revolution speed (nrnot) and y for a second combustion engine parameter, in particular for the *SS* reference engine filling (risol), whereby the index i represents the current S.**.

* working point, the index 0 represents the intermediate point going from this to the lefi or lower intermediate point respectively, and the index I represents the intermediate point going from this to the right or upper intermediate point respectively.

14. Method according to one of Claims 12 or 13, characterised in that, for the silliation of a current working point lying outside the limits of the memory array (PWGADAP), a linear interpolation is carried out, drawing on specified combustion engine parameters, in particular the engine revolution speed (nmol) and the reference engine filling (risol), for which, during an index calculation for the memory array (PWGADAP), in each case a fixed value is determined according to the current index. jt

15. Method according to one of Claims 7 to 14, characterised in that the indices and an amendment amount (pwgreg), which enters into the correction value (pwgad), are transferred into an intermediate memory, whereby the amendment amount (pwgreg) is a value which is determined as a function of a regulation deviation, which is read out by the amount of the system constants after an index change, for preference by way of an amplification characteristic curve which can be applied, for which the temporal sequence of the index change bit displaying the point of time of an index change is delayed, and delayed for preference via a ring memory.

16. Method according to Claim 15, characterised in that the amendment amount (pwgreg) is added at a specified later point of time to a current adaptation :..::: value, entering into the correction value (pwgad), at the individual working point of the memory array (PWGADAP) in each case. * ** * * S * S*

* 1 7. Method according to Claim 16, characterised in that the adaptation value of a first working point is proportionally determined directly after a transient event of the boost pressure, whereby this proportional adaptation value is assigned to that element of the memory array (PWGADAP) which was determined before the point of time of the clearance of the data take-up of the adaptation device by the index.

I 8. Method according to one of Claims 1 5 to 1 7, characterised iii that, in order to avoid a multiple entry of same working points into the intermediate memory, each new working point is identified during a filling cycle of the intermediate memory in a bit array, so that, in the event of a working point arising again in the same cycle, the entry of this working point into the intermediate memory will be prevented. f7

19. Method according to one of Claims IS to 18, characterised in that, after the end of an operational phase of the boost pressure regulation. a transfer is undertaken of the values stored in the intermediate memory into a working adaptation array, whereby an amendment amount (pwgreg) is added to the value of the working adaptation array which pertains to it.

20. Method according to Claim 19, characterised in that each working point contained in the intermediate memory is initially identified in a working bit array.

21. Method according to Claim 19 or 20, characterised in that, in the boost pressure regulation system is not active, and/or after the ending of a continuity I..::: test of the array values, the array values of the working adaptation array are transferred into the memory array (PWGADAP) * 5. * * *

* 22. Method according to Claim 21, characterised in that, during the continuity test 5** of the array values of the working adaptation array, a manipulation is carried out in the line and column direction, making recourse to the working points *I..

transferred from the intermediate memory.

**..** * S 23. Device for carrying out the method according to one of Claims I to 22.