DE102005002385A1 - Electrohydraulic operating cylinder controlling method for actuation of charge-cycle valve, involves obtaining actual upstroke of cylinder based on control signals that depend on control factor and hydraulic oil operating parameter - Google Patents

Abstract

The method involves reading a reference value of a control factor, and determining an operating parameter e.g. temperature, pressure, and viscosity, of hydraulic oil. A control value is computed based on the reference value and the operating parameter by a control factor-transfer function. Control signals generated based on the computed value are adjusted to obtain an actual desired upstroke of an operating cylinder (3). Independent claims are also included for the following: (A) a computer program for executing a method for controlling an electrohydraulic operating cylinder (B) a control device for controlling one or more hydraulic operating cylinders.

Description

The The invention relates to a method for controlling a hydraulic Working cylinder or actuator, in particular a hydraulic Stellers for actuation a gas exchange valve of an internal combustion engine, according to the preamble of Claim 1, and a computer program and a control device according to the sibling claims 39 or 41 and 42.

Application Background Such cylinder or actuator and. the method according to the invention are electro-hydraulic adjustment systems, for example, the activity of valves, such as the gas exchange valves of an internal combustion engine or a compressor, or for actuating flaps, such as fast switching flaps in the intake manifold of a cylinder of an internal combustion engine, or to operate other mechanisms are used. Of particular interest are while hydraulic adjustment systems or working cylinder, a variable stroke, or a variable end position of a Allow adjustment.

In The rule is in such adjustment a fixed starting position present, from which an adjustment is carried out. In the example of the electro-hydraulic adjustment of a gas exchange valve is such a starting position by the concern of the gas exchange valve defined in a valve seat. For the inventive method However, this is not essential. There may be adjustment operations considered out of any rest position or starting position become.

object or object of the method according to the invention is generally the adjustment of the piston of a working cylinder from a given home position to a desired end position from Case by case may vary. It is important the end position or the adjustment or stroke sufficiently accurate adjust.

following This method is exemplified by an electro-hydraulic adjustment of gas exchange valves of an internal combustion engine or a compressor presented and formulated. With a direct transfer of the terms, for example the replacement of valve lift by travel or stroke, valve lift sensor by travel sensor or position sensor, speed by repetition frequency the operation an actuator, and gas exchange valve by piston of a working cylinder, this presentation also describes the more general applications.

In the case of electrohydraulic camshaft-free valve control units (EHVS), such as those described in the DE 101 27 205 and the DE 101 34 644 are known, stroke and timing of the gas exchange valves of an internal combustion engine can be freely programmed in principle. As a result, the operating behavior of the internal combustion engine as well as their specific fuel consumption and their emission behavior can be improved.

Indeed have electro-hydraulic camshaft-free valve controls yet an optimization potential with regard to the control of the valve lift on. The possible Improvements lie in the adjustment accuracy of the valve lift and in the time required for the representation of the stroke control technical effort. This also applies analogously for the control of adjustment or positioning operations in the other, the beginning mentioned systems, for the invention can be used.

From the DE 100 64 650 A1 A method for controlling gas exchange valves of an internal combustion engine with a variable opening function is already known, in which the control of the gas exchange valve takes place as a function of an evaluation of a pressure in the combustion chamber. It is provided that a valve lift sensor provides a position of the valve indicating stroke signal. The valve lift sensor can now be checked for correct function in which the course of the combustion chamber pressure expected for the valve position is compared with the measured combustion chamber pressure. For certain deviations, the valve lift sensor is automatically calibrated depending on this evaluation.

Advantages of invention

The inventive method with the characteristics of the independent Claim has in contrast the advantage that at least one hydraulic cylinder or Steller can be controlled without feedback on the travel or stroke of a Stellweg- or position sensor is needed.

The inventive method for controlling the stroke of a hydraulic actuator or working cylinder, hereinafter also referred to as a control function determined on the basis of existing information on the hydraulic operating parameters (state variables of the hydraulic oil) and optionally to other variables that influence the adjustment of the stroke, the control signals of hydraulic actuator so that a currently desired stroke of the actuator, or - in the example considered application - a currently desired stroke of a supplied hearing gas valve is adjusted within permissible tolerances.

With the state variables of Hydraulic oil are for example pressure and temperature in a pressure accumulator (rail) meant, the hydraulic auxiliary power for the operation of the working cylinder provides. information about the condition of the oil in the rail is cheaper by means of Standard technology (pressure and temperature sensors) easily recoverable. their Use in the control function of the valve lift thus allows a cost-effective realization the controller with regard to the required sensors.

alternative or additionally to the state variables pressure and temperature can also fabric sizes like the viscosity of the hydraulic oil considered, which influence the setting process. Advantageous is in particular the use of a corresponding viscosity sensor.

Of the indeed needed Scope of sensors for State variables of the oil generally depends from the characteristics of a concrete version of the hydraulic control system from. Key aspects include time variability or controllability of the state variables of the oil. For example, if in a simple system execution no pressure adjustment possible is because, for example, the rail pressure by means of pressure relief valve is fixed, then a pressure sensor can be omitted.

By the realization of the control function of the stroke as pure control deleted additional Effort to obtain information about the actually set Stroke of a respective setting operation (for example by means of time high-resolution Hubsensierung), as would be needed for example for a regulation.

A Realization of the control function according to the invention as a computer program in an electronic control unit shows about that There are also advantages in terms of memory and computing power requirements as well as the effort for the determination of system dependent parameters to be calibrated (application data).

One Another advantage of the method according to the invention is to see that it allows for various advancements that as demand-scalable, modular implementations of the control function can be implemented. This means that the simplest version of the Control function (basic function) in a simple way different Added extensions can be if this is to achieve the required positioning accuracy of Ventilhubs in a concrete system is necessary.

If, for example during the Hubstellvorgangs significant pressure oscillations occur at the high-pressure side inlet of the working cylinder and within the working cylinder and affect the achievable with the basic function adjustment accuracy of the valve lift too strong, then these pressure oscillations can be considered for example in a suitable modeling of the "stroke-determining oil pressure" (p _oil ) , The variable p _oil occurs in the control function of the stroke h according to the invention as an input variable of the conversion between a "setpoint stroke" (h _setpoint ) and the technical control variable of the hydraulic actuator, for example the activation time (t _m1 ) of a solenoid valve MV1.

For this exemplified realization of a pressure oscillation correction, there are equally advantageous alternatives, which can also be represented as modular additions to a basic function. For example, pressure oscillations can also be taken into account or compensated by a correction value which adds to the input variable "setpoint stroke" (h _setpoint ) additively or multiplicatively to the control function.

Of the primary Advantage of this for a pressure swing correction proposed solutions is that the effects of the mentioned pressure oscillations on the lifting process and on the resulting valve lift in the drive function so can be effectively compensated, that this solution even with high demands on the positioning accuracy and strong Pressure oscillations is applicable.

Further Advantages of this solution lie in their simplicity, intelligibility and modularity. These quality factors result significantly from the fact that the proposed implementation of a pressure swing compensation as independent, that is called modular supplement to a simple and efficient realizable stroke control function added to that for alone only in cases negligible Pressure oscillations is applicable.

In a further advantageous development of the method according to the invention become changes the actuator behavior during operation by means of an adaptation of the controller balanced, so that the required accuracy of the stroke control over the entire life of the internal combustion engine is guaranteed.

In another advantageous further development of the method according to the invention, deviations of the set valve lift from the desired value, ie the desired stroke h _desired , are determined by means of a Balancing, which is added to the pure control, for example, a basic function above, as a modular extension. Thus, the required accuracy of the stroke control can be ensured even if the pure control by itself is not accurate enough in each case, or should be waived to limit the effort on extensions such as said pressure oscillation correction.

Further Advantages and advantageous embodiments of the invention can the subsequent drawings, their description and the claims taken become.

drawings

It demonstrate:

1 an embodiment of a hydraulic actuator according to the invention;

2 the control of the first control valve (MV1) and the resulting stroke movement;

3 a possible definition of the operating parameter p _oil ;

4 and 5 Examples of hub transfer functions;

6 the functional module tm1_to_h;

7 the function module h_to_tm1;

8th a block diagram of the method according to the invention;

9 the function module ber_p_Oel;

10 the function module ber_T_Oel;

11 a stroke transfer function modeled on the scale approach;

12 an exemplary execution of the conversion h_to_tm1;

13 an alternative modeling of the block TW 12 ;

14 the functional relationship h (t _m1 ) for a case with or without large pressure oscillations;

15 the function module ber_p_Oel in the case of a pressure oscillation correction by means of the stroke-determining oil pressure;

16 an exemplary implementation of the function module ber_p_Oel_h_komp;

17 and 18 : Representations of the influence of pressure oscillations without and with compensation;

19a a representation of the average rail pressure (p _rail ) and the dependence of the stroke-determining oil pressure (p _oil ) from the desired stroke (h _setpoint ) in the case of a pressure oscillation compensation according to the invention;

19b a signal flow diagram of the functional module h_to_tm1 for a second realization of the pressure oscillation compensation according to the invention;

19c a representation of the correction factor of the target stroke for the second implementation of the pressure swing compensation;

20 the course of the transfer function h (t _m1 ) at three operating points of two identically constructed actuators, which show a copy dispersion of the setting properties;

21 an outline of the central Hubsteuerungsmodul h_to_tm1 for the adaptation according to the invention;

22 the addition of the valid for the current calculation pressure _offset p _offset to the pressure p _oil ;

23 the addition of the tm1 offset t _{m1_offset} valid for the current calculation to the value t _{m1_u} ;

24 a demonstration example of the effect of the adaptation according to the invention;

25 an outline of the central Hubsteuerungsmodul h_to_tm1 for the inventive control of the valve lift and

26 a Signalflußbild that an exemplary implementation of the function module stroke control 25 represents.

description the embodiments

In 1 is an embodiment of an inventive control of a hydraulic working cylinder 3 shown as a block diagram. The working cylinder is used here by way of example for actuating a gas exchange valve 1 an internal combustion engine or a compressor. However, the control according to the invention, as stated in the introduction, is not limited to this exemplary intended use of the working cylinder.

The gas exchange valve 1 can be designed both as an inlet valve and as an outlet valve. If the gas exchange valve 1 is closed, it lies on a valve seat 2 on.

The gas exchange valve is actuated 1 by a hydraulic working cylinder 3 , which is the central mechanical-hydraulic component of an electrohydraulic actuator 30 represents. Furthermore, the in 1 exemplified actuator 30 a first control valve MV1 and a second control valve MV2, hydraulic lines 11 such as 19a and 19b , a valve brake 29 and an optional check valve RV1. The components mentioned are in typical embodiments of a Stellers 30 integrated in a structural unit. In the description of the setting operations of the actuator 30 In addition to the properties of its components, the mass of one with the piston 5 connected gas exchange valve 1 and the friction conditions in a guide of the valve stem (not shown) included.

The working cylinder 3 is as a differential cylinder with a piston 5 formed with one-sided piston rod. The working cylinder 3 However, can also be performed with double-sided piston rod (not shown).

The piston 5 has a larger upper effective area A _ob and a smaller lower effective area A _unt . The upper effective area A _ob limits a first working space 7 of the working cylinder 3 , The lower effective area A _unt limits a second working space 9 , Both workrooms 7 and 9 be from a feedline 11 which are made up of the sections 11a . 11b and 11c supplied with pressurized hydraulic fluid, such as hydraulic oil supplied. For this purpose, the working cylinder 3 high pressure side via the feed line 11 and a first check valve RV1 installed therein with a high pressure accumulator 13 hydraulically connected, which provides the hydraulic energy for the adjustment process.

In the section 11b the feed line 11 which is the second working space 9 and the first workroom 7 connects, a first control valve MV1 is arranged. In the in 1 shown switching state, it is closed and de-energized.

The hydraulic fluid in the first working space 7 can have one in the section 19c unpressurized or with a low level pressure applied return line 19 which are made up of the sections 19a . 19b and 19c composed, be dissipated. In the return line 19 a second control valve MV2 is arranged, which in 1 is shown open. The second control valve MV2 is advantageously opened normally.

In the second workroom 9 can a closing spring 27 be provided, which is the gas exchange valve 1 with unpressurized working cylinder 3 in the closed position, that is in contact with the valve seat 2 , brings or holds in this position.

In a likewise advantageous alternative embodiment (in 1 not shown), the closing spring 27 be designed so that it applies the closing force alone and receives a correspondingly high amount of potential energy during the opening process. In this case, the hydraulic connection is eliminated 11c and the function of the second workspace 9 , that is, the application of the lower effective area A _under the piston 5 with the pressure of the high-pressure accumulator 13 , The hydraulic working cylinder 3 So in this case is designed as a single-acting cylinder. In an advantageous development of this embodiment, the spring 27 designed progressively, that is with one above the travel of the piston 5 increasing spring force.

It is also beneficial to the possibilities of hydraulic and mechanical power generation, as described above, to combine to the closing To provide force of the hydraulic actuator.

It is also possible and advantageous to the piston 5 to design, the effective area A that _whether changes over the path of the piston with at least one gradation, in particular smaller. For example, the piston 5 Two-part stepped run (not shown), wherein a first stage, which only on a first path section when opening the gas exchange valve 1 moved, an additional effective area for the pressure in the first working space 7 provides. In this embodiment of the hydraulic actuator so the upper effective area A _ob and thus the opening force of the actuator 30 during a first path section when opening the gas exchange valve 1 enlarged, with the advantage that the gas exchange valve 1 can be opened against higher gas forces and also faster.

Further, not mentioned in detail embodiments of the working cylinder 3 or Steller 30 are possible and equally suitable for the application of the control method according to the invention.

The high-pressure accumulator 13 is via a high pressure pump 17 supplied with high pressure hydraulic fluid (also referred to herein as oil). In the section 11a the feed line, which the high-pressure accumulator 13 with the second workspace 9 connects, a first check valve RV1 is provided so that a return flow of hydraulic fluid from the second Ar beitsraum 9 in the high-pressure accumulator 13 is prevented.

Between the first workspace 7 and the second control valve MV2 is a hydraulic brake device 29 intended. This hydraulic braking device 29 works as follows: when the piston 5 moved upwards and consequently the volume of the first working space 7 is reduced, the hydraulic fluid flows out of the first working space 7 through the section 19a the return line 19 off, until the top of the piston 5 the section 19a the return line 19 closes. Thereafter, the hydraulic fluid from the first working space 7 only via the hydraulic braking device 29 , which consists essentially of a throttle, drain, since the connection to the hydraulic braking device 29 at the top of the workroom 7 is arranged. By compared to the flow resistance of the section 19a the return line increased flow resistance of the hydraulic braking device 29 becomes the piston 5 braked before the gas exchange valve 1 on the valve seat 2 rests.

In high pressure storage 13 Temperature sensors T _rail and pressure sensors P _rail are arranged, which via signal lines with a control unit 31 are connected. The high pressure pump 17 , the first control valve MV1 and the second control valve MV2 are also via signal lines to the controller 31 connected and are controlled by this. The signal lines are in 1 shown as dashed lines.

The hydraulic brake device 29 can, as in 1 indicated by a signal line, be designed as an active braking device and the signal line from the control unit 31 be controlled if necessary. Also, the pressure P _{rail of} the high-pressure accumulator 13 depending on the desired adjusting movement or opening force of the gas exchange valve 1 by means of a corresponding control of the high-pressure pump 17 be managed.

If, as in 1 shown, the first control valve MV1 closed and the second control valve MV2 is opened, the pressure causes p _udr in the second working space 9 that is the gas exchange valve 1 against the direction of the arrow 15 moved and thus closed.

The force required for this is provided by the second working space 9 with high pressure hydraulic fluid from the feed line 11 is supplied while the pressure p _odr in the first working space 7 due to the hydraulic connection to the return line 19 drops rapidly and ultimately the very low pressure p _Rl in the section 19c the return line equalizes.

To open the gas exchange valve 1 the second control valve MV2 is closed and then the first control valve MV1 is opened. This results in a pressure equalization between the first working space 7 and the second workspace 9 instead of. As a result, the gas exchange valve opens 1 because of the first workroom 7 pressurized face A _ob the piston 5 is larger than that of the second workspace 9 pressurized annular surface A _{unt of} the piston 5 ,

For controlling the opening of the gas exchange valve 1 and especially of the resulting valve lift, the actuation of the first control valve MV1 is thus of great importance in two respects. First, with the opening of the first control valve MV1, the beginning of the opening movement of the gas exchange valve becomes 1 and, secondly, the duration of the activation-referred to below as the activation duration t _m1 -has a decisive influence on the stroke of the gas exchange valve 1 , The actuation duration t _m1 determines how long the first control valve MV1 remains open, from which the amount of the high-pressure accumulator 13 in the first workroom 7 flowing oil, which in turn determines the valve lift immediately. So by the first control valve 1 is closed again at the right time, the desired valve lift of the gas exchange valve 1 one.

If the gas exchange valve 1 is to be closed again, the second control valve MV2 is opened, so that the pressure p _odr in the first working space 7 collapses and that of the second workspace 9 on the piston 5 applied hydraulic force the gas exchange valve 1 closes.

The The inventive method described below is not on the previously described in the context of an example system execution limited. To the Example can Piezo valves instead of magnetic switching valves and / or proportional instead of switching valves are used. Also are multi-way valves instead of 2/2-way valves possible. For example, it is also possible the first control valve MV1 and the second control valve MV2 as functional parts a single control valve, this control valve can set more than two positions.

In a further possible embodiment, the first control valve MV1, or a combined control valve (MV1, MV2), are also actuated by means of hydraulic pressure force, wherein further control valves, such as electro-hydraulic servo valves, are used. In this case, the stroke control method according to the invention is used to control a servo valve, which serves to close the control valve MV1 and thus for metering the inflowing hydraulic fluid during the adjustment, so to determine that a desired adjustment of the piston 5 or a desired stroke of the gas exchange valve 1 is brought about.

The pressure supply can also be fixed instead of adjustable. The check valve RV1 may also be missing. In the hydraulic circuit also other components not shown here may be present, for example, a connection of the first working space 7 of the working cylinder 3 with the high-pressure accumulator 13 via another check valve.

The scope of the sensors can be compared to the in 1 example shown limited or extended. Thus, for example, there may be a plurality of pressure sensors, which are preferably distributed at different locations of the high-pressure accumulator 13 , but also directly at the beginning of each cylinder 3 , may be appropriate. A detection of the oil temperature can also - alternatively or in addition to the in 1 displayed location - at the high-pressure entrance or in the workrooms 7 and 9 single working cylinder 3 be provided.

Furthermore, additional temperature sensors may be provided on structural materials, such as cylinder head, cylinder, actuator or solenoid valve housings, or coil temperature of solenoid valves, and / or an oil viscosity sensing sensor. In particular, a stroke sensor may also be provided (in FIG 1 not shown), the travel or stroke of an electro-hydraulic actuator 30 or its associated piston 5 certainly. Corresponding information obtained with the mentioned sensors can be included in the method according to the invention for improving the control accuracy of the stroke control.

In typical and advantageous embodiments of a Steller 30 are its in 1 Components shown integrated in a structural unit. In extended embodiments, this integrated unit may also include a stroke sensor and / or other sensors and / or corresponding parts of the control electronics that are incorporated in 1 through the control unit 31 is represented.

In particular, so the controller 31 consist of several separate parts respectively electronic modules (not shown), which are interconnected by electrical lines or communication channels and the - in whole or in part - to individual writers 30 grown or on these steller 30 can be plugged.

The method described below for controlling a setting operation of a hydraulic working cylinder 3 or an electrohydraulic actuator 30 , which is based on the embodiment of 1 is in direct transmission in all of the above, as well as in other derivable by generalization system interpretations applicable. It is in particular independent of the intended use of the hydraulic working cylinder 3 ,

In the electro-hydraulic control of gas exchange valves of an internal combustion engine per cylinder at least one hydraulic actuator 30 for the inlet valve (s) and / or a hydraulic actuator 30 required for the exhaust valve (s). This means, among other things, that the high-pressure accumulator 13 in general, several stops 30 are connected, of which in 1 for the sake of clarity, only one, including the associated gas exchange valve 1 , is shown.

For each such system or for each individual hydraulic actuator 30 the basically similar task of a suitable determination of the respective stroke-determining control variable takes place in the case of a respectively preset nominal stroke and in each case currently available values of the essential influencing variables of the actuating process, such as, for example, the pressure and the temperature of the hydraulic fluid.

In 2 are the control of the control valves MV1 and MV2 over the time t together with the resulting movement of the gas exchange valve 1 and the piston 5 of the associated working cylinder 3 applied. The line 33 shows the valve movement over time until the end of the drive of the solenoid valve MV2. The resulting stroke h of the gas exchange valve 1 results from the activation of the first control valve MV1 with the activation time t _m1 .

The stationary stroke of the gas exchange valve after a short transient 1 (please refer 2 ) defines by way of example the control variable h considered in the stroke control. Other definitions, including, for example, characteristic values of the stroke during the opening process, such as the maximum stroke, are also possible. In the control of the stroke, a desired value or desired value h _{Soll is} specified for the control variable h, for example as a function of the power desired by the driver and the current rotational speed of the internal combustion engine.

The course of the movement and the resulting Stroke h of the gas exchange valve 1 among other things depend on the pressure and the temperature of the hydraulic power cylinder at fixed drive time t _m1 3 inflowing oil. This dependence is referred to as the "operating point dependence" of the relationship between activation time t _m1 and stroke h.

The functional relationship h (t _m1 ) between stroke h and activation duration t _m1 is hereinafter referred to as "stroke transfer function" and its inverse function t _m1 (h) as "inverse Hubtransferfunktion". Both depend on the operating point of the hydraulic working cylinder 3 , in particular of the hydraulic state variables pressure and temperature, from.

A suitable functional description of the dependence on the operating point for the inverse lift transfer function t _m1 (h), that is to say for the calculation of the drive time t _m1 , forms the core of a stroke control according to the invention. The analogous consideration for the Hubtransferfunktion itself leads to an inventive method for the calculation of the stroke h.

The desired benefits are achieved by a simple representation the operating point dependency achieved at the operating point by location-independent and time-independent (effective) influencing variables described becomes. Starting from time-dependent and possibly also location-dependent be physical factors these effective factors, for example as mean values or as characteristic individual values, such as Example initial values, defined.

According to the invention, for example, the dependence on the state variables of the oil is described by corresponding scalar influencing variables, that is to say individual values, _oil and _oil , which are to be defined appropriately in the above-mentioned manner. This is in 3 for the oil pressure p _oil illustrated.

The for an opening operation of a working cylinder 3 relevant effective oil pressure p _oil is in this example locally to that point on the high-pressure accumulator 13 related to the feedline 11 is connected, and time determined as the "mean" of the pressure at this point over a suitable time interval. As an alternative to the local determination of the pressure p _Oel can also be the supply point to the second working space 9 be used, which is also in 3 is displayed. As a simple alternative to the time determination can also be an instantaneous value of the pressure, for example, just before the start of the lifting motion t _if used (in 3 Not shown).

For approximate calculation of _oil , the example of a pressure sensor in the high-pressure accumulator 13 measured pressure p _rail used. If p _rail at one to the relevant working cylinder 3 , or to the junction of the relevant feed line 11 , sufficiently close location in high-pressure storage 13 If necessary, the local difference can be neglected 3 is accepted. This means that in this case p _{oil can} be represented by the time average, or alternatively also by the instantaneous value, of p _rail .

In the general case, a difference in the pressures p _rail and p _oil caused by the distance of the respective locations can be modeled on the basis of two effects which, depending on their size and the required accuracy of p _Oel , can be taken into account or neglected.

The first effect is a temporary average pressure drop, resulting from a medium, over the corresponding section of the high-pressure accumulator 13 (between the above places) results in flowing oil flow and thus from the adjusting operations, or adjusting strokes, of the considered actuator 30 and other (not shown) controller 30 can be calculated on the high-pressure accumulator 13 are connected.

The second effect is longitudinal vibrations of the pressure (pressure waves) caused by pulse-like withdrawal processes of hydraulic oil by individual actuators 30 , and / or by a pulsating delivery volume flow of a high-pressure pump 17 , in high pressure storage 13 be stimulated. Since such pressure waves are especially relevant when the actuating operations of individual controllers 30 repeat in an approximation periodically and with only slowly varying setting strokes, their contribution to the difference of p _rail and p _oil , that is a corresponding "pressure wave correction", can be described in a simple modeling, for example by empirically determined map data.

at Incidentally, the pressure waves mentioned are a different phenomenon than the pressure oscillations considered elsewhere for which one so-called pressure oscillation compensation is described.

For a time averaging in the determination of the effective oil pressure p _oil various variants are conceivable. In 3 For example, an averaging during the opening process is indicated. The opening process begins at time t _ob . The duration of the opening process corresponds to the activation time t _m1 .

The effective oil pressure p _oil averaged over an opening operation is usually smaller than the initial value of the pressure at the beginning of the movement, there the working cylinder 3 when opening the gas exchange valve 1 Oil from the high-pressure accumulator 13 extracts. Because of the compressibility of the oil, this will depend on the volume or geometry of the reservoir 13 , a more or less significant pressure drop. The mean pressure reduction depends in particular also on the stroke h of the open gas exchange valve 1 from. In addition to the removal of oil by the considered control process itself but also temporally obvious or overlapping adjusting operations of other hydraulic working cylinder to other, not shown gas exchange valves 1 belong, influence the considered control process by the pressure pulsations caused by them and / or by an overall greater reduction of the rail pressure p _rail . Such an influence can thus be taken into account in the case of a suitably chosen temporal averaging of the pressure p _rail .

With regard to the various influences or effects that are relevant or may be relevant for the calculation of the stroke-determining oil pressure p _oil , an algorithm for determining p _{oil can} advantageously be designed as follows: On the basis of measured values p _rail , the following is first calculates a base value, for example, a mean pressure in the high-pressure accumulator 13 , or in a part of this memory, at least approximately represented. To this basic value additively different correction values are added, as far as this is necessary to achieve the required accuracy of the stroke control.

The correction values are thereby determined specifically for individual influences which lead to deviations of the oil pressure p _oil from the stated basic value. The already mentioned influence of simultaneous or temporally overlapping positioning processes can thus be taken into account by a specific correction (overlay correction). Further correction values can be used, for example, to describe a pressure drop (flow loss) or to describe the effect of longitudinal vibrations (pressure waves) in the high-pressure accumulator 13 be formed, as already stated.

In principle, it is also conceivable to use only an instantaneous value, in particular a single measured value, of p _rail for a calculation of the pressure parameter p _oil and to dispense with averaging.

Analogous to p _Oel a suitable quantity T _{Oel is} defined and determined to describe the influence of the oil temperature in the Hubtransferfunktion. In the simplest case, a current measured value T _rail can already be found at a measuring point on the high-pressure accumulator 13 is used to serve as an estimate for T _oil . More elaborate and potentially better definitions and calculation methods for _oil are conceivable, for example with the inclusion of further measured temperature values such as the engine temperature T _{Mot of} the internal combustion engine.

Alternatively or additionally, a variable η _Oel can be determined, which determines the viscosity of the working cylinder 3 describes incoming oil. In typical cases, the viscosity η _{Oel is} responsible for most of the influence of the temperature T _oil , or it can even replace T _Oel as an influencing variable in the stroke control.

In the following, the size T _{Oel is used} as an influencing variable, and this also applies to the case. should stand that the viscosity η _{Oel is used} to describe the temperature influence. Basically, it is possible and permitted to describe all explicitly mentioned quantities by any ensemble of equivalent dimensions.

In accordance with selected definitions of the variables p _oil and T _oil results in a dependence of the Hubtransferfunktion h (t _m1 ) of these sizes as in the 4 and 5 is shown by way of example. In 4 is the dependence of the Hubtransferfunktion of the pressure p _oil at a constant temperature T _oil and in 5 depicting the dependence on the influencing variable T _Oel at a constant pressure p _Oel , wherein in each case three interpolation points for the varied state variable are selected.

With arbitrary values for p _Oel and T _Oel , a continuous functional set of lift transfer functions h (t _m1 ) generally results as a function of the influencing variables.

This relationship is written as follows: h = f stroke (t m1 ; T Oil , p Oil , E x ) (1)

In addition to the operating parameters T _Oel and P _Oel , further influencing _variables are optionally taken into account, which are represented in equation (1) by an additional argument E _x . For the general context _hub f is also the term "Hubtransferfunktion" is used.

Examples of further influences include additional temperature parameters, for example the temperature of solenoid valves, or gas forces acting on a gas exchange valve 1 act during the stroke-determining adjustment process. Such gas forces result for example from a high internal cylinder pressure p _cyl in the internal combustion engine at the time of opening the gas exchange _valve 1 , In this case, for example, p _zyl can be used as another scalar influencing _{parameter in} a _{method according to the} invention Description of the Hubtransferfunktion h (t _m1 ) can be used according to equation (1).

An algorithmic representation of the relationship (1) as described below can be realized, for example, as a function module "tm1_to_h" in the control unit. Such a module is in 6 shown. It represents a calculation function according to the invention for determining the given values of the control variable t _m1 , the influencing variables p _Oel and T _Oel and optionally further influencing variables E _x resulting hubs or "Isthubs" h = h _Ist .

The symbol E _x is here, as already used in equation (1) and in subsequent cases, to designate further influencing variables, these variables may vary from case to case, even with multiple use of the symbol within a diagram.

The Calculation function mentioned, for example, for monitoring purposes of the valve be used.

For the control of the stroke h, that is to say for setting a desired stroke h _{desired in} each case for an actuating operation, the calculation of the (1) inverse functional relationship becomes t m1 = f tm1 (h; T; Oil , p Oil , E x ) (2) With
h = h _nominal needed. This function is the "inverse stroke transfer function" or "manipulated variable transfer function" f _{tm 1.} Once again, E _x stands for any other influencing variables that may have been taken into account.

After a realization of the stroke control according to the invention is thus in the control unit 31 an algorithmic representation of the context (2), for example, as a computer program available. Such an inventive computer program or functional module "h_to_tm1" of the control unit is in 7 shown. It determines the resulting activation time t _m1 from given values of the desired stroke h _Soll , the influencing variables p _Oel and T _Oel and, if appropriate, further influencing variables E _x .

Before each actuation of a working cylinder 3 the function module tm1_to_h carries out a calculation in which the currently desired setpoint stroke h _setpoint and estimated values for the effective influencing variables present at the time of the setting process, such as, for example, T _oil and p _oil , are used as input variables. These influencing factors directly or indirectly take into account all the essential influences of the stroke-determining actuating operation or the forces which become effective in this actuating process.

According to the invention, hydraulic forces and friction forces of the actuating system which depend on the oil state, as well as the forces acting on hydraulic control valves, such as the solenoid valve MV1, are taken into account in that the conversion is suitably dependent on at least one operating parameter p _oil (effective oil pressure) and / or T _Oil (effective oil temperature) or equivalent information takes place. Instead of or in addition to T _Oel , as already mentioned, a measure η _{Oel can be used} for the viscosity of the hydraulic oil. If required, further temperature parameters and other factors are considered.

For the determination of current values or estimated values of the influencing variables, functional modules are in the control device in the exemplary embodiment of the invention 31 including a module "ber_p_Oel" and / or a module "ber_T_Oel", as in 8th shown.

In the 8th The interconnection of the calculation functions ber_p_Oel and ber_T_Oel of the influencing variables p _Oel and T _Oel with the central conversion function h_to_tm1 of the stroke control, which relates to the signal flow, is shown as a block diagram. According to equation (2), the desired value h _{Soll of} the stroke, the influencing variables p _Oel and T _Oel and, if appropriate, further influencing variables E _{x are introduced} as input variables into the calculation h_to_tm1. In addition, a valve index v is transmitted to the central conversion function, whose value indicates which gas exchange valve 1 or for which controller 30 the respective calculation is carried out. The valve index v can also occur under the further input variables E _{x of} the modules ber_p_Oel and ber_T_Oel. Furthermore, the optional connection of a function module "adaptation" is shown, which is an adaptation of the central conversion function to changes in the properties of the actuator 30 in operation serves.

The according to 8th for a respective adjustment calculated value of the drive time t _m1 is output to a responsible output unit, in the required manner, the electrical control of the associated first control valve MV1 of the respective working cylinder 3 performs.

In the 8th occurring modules ber_p_Oel, ber_T_Oel, h_to_tm1 and adaptation are explained in detail below.

In 9 is a first embodiment of a function module ber_p_Oel shown.

In doing so, use the one for the size p _oil de value is an estimated value for an actual later, namely during the setting process, actually existing effective pressure p _{oil of} the oil. This effective pressure p _Oel has a significant influence on the control process considered, since it the opening speed and thus also for a given control duration t _m1 the stroke of the gas exchange valve 1 significantly influenced. The accuracy requirements for calculating _oil are correspondingly high.

The module ber_p_Oel estimates this stroke-determining pressure p _oil based on existing information about the pressure in the high-pressure accumulator 13 , for example from measured values p _rail of the high-pressure accumulator 13 installed pressure sensor, and optionally further information about the change over time of the rail pressure between the time of calculation of p _oil or the last detection of p _rail and the timing of the adjustment of the gas exchange valve 1 for which this calculation is performed. The further information is represented by the input E _x . In order to represent such a temporal prediction or even a temporal averaging of measured values in the calculation of p _oil , for example, as in 9 indicated, corresponding times t _mess , at which the measured values used p _rail are detected, or alternatively corresponding crank angle φ _{KW, mess occur} as further input variables of the module ber_ p_Oel, as well as, for example, the time t _ob or the crank angle φ _{KW, whether} for the beginning of the lifting movement. The times can also be understood as time differences with respect to a reference time, such as the time of a current calculation of the control function of the hub. If crank angles φ _{KW, mess} and / or φ _KW, are used, the further input variables denoted by E _x usually also include the engine speed n _mot at which the actuator 30 is installed. The same applies to the use of the actuator 30 on a compressor or for other possible uses. Furthermore, in the case of averaging or filtering of quantities within the module ber_p_Oel and / or using time-dependent physical models, memory variables M _x may occur, which are typically read and updated every time the calculation ber_p_Oel is performed, as also in FIG 9 is displayed.

Among the other input variables E _{x of} the module ber_ p_Oel can, for example, information about the control or the current setting of an adjustable high-pressure pump 17 and / or information about temporally overlapping with the control process considered Stellvorgänge of another, at a high-pressure accumulator 13 connected working cylinders 3 be included in order to be able to calculate a good estimate for p _Oel in a correspondingly complex algorithm.

In a simple, exemplary implementation, the module processes ber_p_Oel in an incremental manner with a time-equidistant sampling of the signal of a pressure transducer on the high-pressure accumulator 13 the pressure measured values p _rail = p _rail <i>, i = 1, 2,... detected thereby at the times t = t _mess <i>, by means of a low-pass filtering, continuously receive a filter value p Oil = low pass (p rail <i>) (3) determined from the measured values. The filter function can be selected, for example, as a control PT1 element with a suitable time constant.

In order to determine the relevant value of the influencing variable T _Oel with regard to a lifting operation to be carried out in each case, another functional module ber_T_Oel is in the control unit 31 present, exemplary in 10 is shown.

In the embodiment shown, the module calculates the model size T _oil based on a detection quantity T _rail , which is the temperature of the high-pressure accumulator 13 Characterized, and optionally further information such as the engine temperature T _mot . Further input variables are again represented by an input E _x . Analogous to 9 Memory variables M _x can also occur.

For example, a simplest version of the function ber_T_Oel specifies a current measured value T _rail T Oil = T rail (4) directly to the subsequent conversion function on, wherein the detection of the sensor values T _{rail is} often enough to account for the - comparatively slow - change the temperature T _rail . In typical embodiments, the individual detection value can still be checked for plausibility and switched to a substitute value in the case of "outliers" caused by disturbances, which should not be further elaborated here.

Finally, as an important part of the stroke control according to the invention, an exemplary algorithm for the conversion of the target stroke h _Soll into the activation time t _m1 according to equation (2) is to be specified, which is done below. A corresponding algorithm is used, for example, as a computer program in the control unit 31 implemented as such and an execution of the module h_to_tm1 (see 7 ).

As advantageous algorithmic representations of the inverse Hubtransferfunktion (manipulated variable transfer function) according to equation (2) or the Hubtrans According to equation (1), generally suitable are so-called parametric representations or parameter functions, that is to say calculation formulas based on a number of predetermined (elementary) function terms and a number of parameters which occur as scaling parameters or coefficients. Such a description is hereinafter for the first Hubtransferfunktion f _hub according to equation (1) above.

For example, for the function f _stroke , a parametric representation using a scale _approach may be used in suitable cases. Such a description approach can be derived, for example, as a solution of an equation of motion that the lifting process of the gas exchange valve 1 describes sufficiently well, if possible elastic oscillations are weak. Further requirements of the derivation concern the behavior of the solenoid valve MV1. The 4 and 5 show examples in which the requirements are met. For other cases, such a basic approach can be extended, for example by means of appropriate vibration corrections or an alternative approach can be selected. This will be explained in more detail below, in describing advantageous developments of the invention.

In the case of modeling by means of a scale approach, the transfer function f _stroke according to equation (1) assumes, for example, the following form: h = f stroke (t m1 ) = HWA · fa (STP / HWA · (t m1 - TW)), (5) using a suitable scale function fa and scale parameters HWA, STP and TW. In general, another parameter H0 can be additively added. However, such a stroke offset is not considered here.

The scale parameters HWA, STP and TW depend on the influencing variables p _Oel , T _Oel ,... _Occurring or to be considered in equation (1): HWA = HWA (p Oil , T Oil , ...) TW = TW (p Oil , T Oil , ...) STP = STP (p Oil , T Oil , ...) (6)

A direct dependence of the function fa on the influencing variables can also be used if required. This is for example by means of a one-parameter function set fa (x) = f (a; x) (7) representable, wherein the so-called shape parameter a then analogous to the equations (6) in the form a = a (p Oil , T Oil , ...) (8th) can be recognized. The variable x in equation (7) stands for the argument of the function fa in equation (5), or more generally for a function term which depends on the activation time t _m1 and optionally on one or more of the parameters p _Oel , T _Oel , ... depends.

In 11 the course of a function f _stroke calculated according to equations (5) and (7) is shown by way of example. The scale parameter STP denotes the asymptotic slope, HWA a curvature parameter. The parameter TW corresponds to the point on the t _m1 axis, where the function, when extrapolated, intersects the t _m1 axis in the range t _m1 <T ₀ . Since the values t _m1 <T ₀ and in particular negative t _m1 values do not represent permissible activation times, this extrapolated curve is in 11 Not shown.

By an appropriate choice of the parameters HWA, STP and TW as a function of p _Oel and T _Oel can in the 4 and 5 Hubtransferfunktionen shown are very well approximated. The tabular presentation of the result values of such an adaptation, plotted over the given interpolation points p _oil and _oil on the pressure and temperature axes, is referred to as a characteristic diagram. Such a map is in appropriate form in the data memory of the electronic control unit 31 can be stored as a field of values.

The embodiment considered here should be limited to the simple case of the function (5) without variable shape parameters a. By way of example, let the scale function be obtained from a one-parameter set of solutions according to equation (7) fa = f0 = ln coshyp (9) specified for the parameter a = 0.

In a representation according to the invention of the module tm1_to_h used to calculate an actual value h = h _actual , for example, the function f0 in accordance with equation (9) is represented in tabular form as a characteristic curve in the data memory of the control unit 31 and evaluated by a program piece which calculates the algorithm according to equation (5) by means of interpolation for the value of the argument STP / HWA * (t _m1 -TW) to be calculated first. Instead of a characteristic curve, a polynomial or a rational function, or a piecewise composition of such functions, can be used to represent the function f0. The values of the parameters HWA, STP and TW are determined in a first step by the data memory of the control unit 31 stored map values of HWA, STP and TW are interpolated for the given values of the variables p _Oel and T _Oel .

Analogous to the parametric representation according to equation (5) for the Hubtransferfunktion (see equation (1)) can also be a parametric representation of the inverse Hubtransferfunktion f _{tm1 found} according to equation (2) and an inventive representation of the stroke control are based.

In particularly advantageous embodiments, the analogous representation of the calculation rule in accordance with equation (1) is at the same time associated with an analog representation of the inverse context (2) and vice versa. For the scale function according to equation (5), this is satisfied, as can be easily seen. The inverted function fa _inv given to the function fa results directly t m1 = f tm1 (h) = HWA / STP · fa inv (h / HWA) + TW (10)

For the scale function f0 given in equation (9), the inverse function f0 _{inv can be} explicitly stated: f0 inv = arcoshyp exp (11)

By inverting the scale approach ( 5 The representation of the stroke control function f _tm1 (h) obtained according to equation (10) is again a scale approach, the scale parameters being expressed directly by means of the parameters already known according to equations (6). For the stroke control according to the invention, the equation (10) with h = h _{Soll is} evaluated.

The particular advantage of an invertibility of the scale approach is the fact that both the Hubtransferfunktion and the inverse Hubtransferfunktion based on the same parameters, in the example so STP, HWA and TW, can be displayed, which is a particularly efficient implementation in the control unit 31 For both functions together and a saving of effort for the application of the data allows. In addition, there are advantages in that both representations are numerically inversely related to each other practically without any approximation error, which would be difficult to achieve in the case of different approaches for stroke calculation and stroke control.

For the algorithm given in equation (10) 12 the signal flow of a program representation, which for the realization in the control unit 31 is particularly suitable. The parameters HWA, TW and STP are implemented as characteristic diagrams as a function of p _Oel and T _Oel and the function fa _{inv by way of} example as a characteristic fa_inv, wherein a representation by elementary functions would also be possible here, as indicated above for the case of f0 , Otherwise occur in the block diagram in 12 only elementary arithmetic operations on.

The Calculation of the scale parameters can also be done in other ways as in the exemplary embodiment specified. Instead of or in addition for example, tables and maps can be more general Calculation rules are used, as they are, for example from an analytical, especially a physically model-based representation the parameter in dependence of the influencing variables can.

The concrete form of the dependencies in the equations (6) can be obtained, for example, by deriving equation (5) as a solution to an above-mentioned approximate equation of motion of the gas exchange valve 1 or piston 5 - At least approximately - be determined. The scale parameters are first represented as a function of the hydraulic forces, frictional forces and other forces occurring in the equation of motion. These forces, in turn, depend, inter alia, on oil substance quantities such as viscosity and density, which can be represented as a function of _oil and _oil .

Such relationships can be used in whole or in part in the calculation of the parameters in the context of a realization of the module h_to_tm1 according to the invention. Advantages result for example from a representation of the stroke control and / or the Hubberechnungsfunktion, in which in a first step, the substance sizes of the oil as a function of _oil and _oil and then the parameters, for example the scale parameters, a parametric description of the used Transfer function can be calculated depending on the substance sizes and optionally other factors.

For a concrete representation of individual of the mentioned dependencies, in turn tabular representations (maps) can be used. The advantage of such an embodiment of the calculation modules according to the invention, in which the dependence of substance sizes of the oil is shown explicitly, is the fact that it may be sufficient for a change in the hydraulic medium used, only in the control unit 31 to store stored information on the material properties, for example map data for viscosity and density, to the changed conditions.

Other embodiments of the stroke control or stroke calculation according to the invention may include, for example, additional calculation parts to take into account further influencing variables. 13 shows an example of such a more complex modeling of the time scale parameter TW in a representation as a signal flow-oriented block diagram. The time parameter TW is an additive of one of the oil temperature and a from the temperature of the first control valve MV1 dependent share together. Both parts also depend on p _oil . The modeling of the solenoid valve temperature uses, in addition to the oil temperature, the engine temperature and optionally further information.

Finally, limit themselves in an implementation according to the invention the stroke calculation according to Equation (1) or the stroke control according to the equation (2) used descriptive approaches not even in the embodiment mentioned approaches. So are next to the scale approaches for example, also polynomial approximations, rational functions as well as piecewise Compositions of these approaches usable. For polynomial function terms, for example, in the denominator a rational function, according to the invention are the polynomial coefficients or to equivalent Parameters and piecewise approaches if necessary, the transition points in the manner described as a function of the influencing variables.

Among the polynomials, those of the maximum degree 2, that is to say straight lines and parabolas, as well as their piecewise compositions are particularly suitable for representing the stroke transfer function. They are comparatively easily invertible, which opens up the advantages already mentioned in the case of the scale approach. For example, the in 4 and 5 Lift transfer functions shown by way of example are approximated by means of a parabola for the area of small strokes (acceleration area) and an attached straight line section for the area of large strokes (stationary area). The inverse function then directly provides a representation of the inverse hub transfer function which can be used for the stroke control. The occurring root function (square root) can be used for a conversion of the algorithm as a computer program in the control unit 31 For example, in tabular form, that is to say as a characteristic. Alternatively, functional approximations, for example by a rational function, are also suitable for the representation of the square root function in the control unit 31 suitable and advantageous.

Particularly suitable are the said piecewise functional representations of the Hubtransferfunktion in the above-described specific embodiments of a working cylinder 3 in which the upper effective area A _ob the piston 5 above the hub with at least one gradation changes. In these cases, it is often advantageous if stroke areas with different effective area of the piston 5 be treated separately in a piecewise description of the Hubtransferfunktion.

In the case of the 1 and 2 explained hydraulic cylinder 3 Depending on the geometric design and the moving masses, more or less pronounced pressure oscillations occur during a lifting operation (see also FIG 3 ). These vibrations affect both the high-pressure accumulator and the work spaces 7 and 9 of the working cylinder 3 as well as the corresponding hydraulic connections, for example the feed line 11 , As a result, shows the movement of the gas exchange valve 1 also over time correspondingly stronger or weaker pronounced vibrations, especially at the beginning of the movement (see 2 ).

As a consequence, the dependence of the resulting valve lift h on the actuation time t _{m1 of} the first control valve MV1, which is to be considered for the control of the stroke, also shows system-dependent more or less pronounced oscillations, especially in the range of small to medium strokes h. This is exemplified in 14 shown. Shown is a Hubtransferfunktion h (t _m1 ) in a case in which the course of this curve is characterized by vibrations, especially in the lower region. For comparison, a calculated course is shown in which these vibrations are "averaged". This calculated curve also exemplifies the case of concrete systems with only minor or negligible vibrations.

The causes of such vibrations of the Hubtransferfunktion are typically in strong pressure oscillations between the lower and upper working space 9 and 7 , which begin with the beginning of the opening of the first control valve MV1 and with the in 3 illustrated pressure oscillations in the high-pressure accumulator 13 are coupled. The pressure oscillations depend on the one hand on the dynamics of the control valve MV1 and the piston 5 on the other hand, however, also strongly on constructive parameters, as well as on the elasticity and density of the oil. In addition, but also non-linear friction effects at the beginning or during the movement of the working cylinder or the gas exchange valve 1 contribute to vibrations in the course of the lifting movement or the Hubtransferfunktion, wherein such friction effects also react on the pressure oscillations and reinforce them.

Around the required accuracy of the stroke control even in the case of strong To achieve pressure oscillations, the vibrations must be in one concrete solution For this Case considered appropriate and described.

This can and should - as stated in the introduction - be achieved on the basis of the already described simple solution of a stroke control by this solution by a suitable "Druckschwin The expansion is aimed at the far-reaching compensation of the entire described effect, that is, at the sufficiently precise modeling and compensation of the vibrations caused by the totality of all influences during the stroke transfer function h (t _m1 ) or in the stroke control used inverse lift transfer function t _m1 (h).

This object can be achieved according to the invention that during the Hubstellvorgangs on the high-pressure side inlet and within the working cylinder 3 occurring pressure _{oscillations are} taken into account in a suitable modeling of the size "p _oil ". The size "p _oil " is yes, as already related to 3 explained, defined as "stroke-determining oil pressure" and occurs in the control function of the stroke h according to the invention as input of the conversion between the size "Sollhub" and the corresponding technical control variable of the electro-hydraulic actuator 30 - In the example, the drive time t _{m1 of} the first control valve MV1 - on.

The functional and realization advantages of such Solution - like the high achievable positioning accuracy, intelligibility, simplicity of Application and resource efficiency, as well as modularity and scalability - are already introductory been described.

Below is a realization of this can be used if necessary, modular pressure swing compensation based on 15 ff explained.

The starting point is the basis of 8th ff described inventive realization of the stroke control. According to this block diagram, the function module h_to_tm1 executes the conversion of a setpoint stroke h _Soll into the activation time t _m1 required for the stroke control. From further function modules ber_p_Oel, ber_T _oil ... the current values of the influencing variables of the lifting process, ie p _oil , _oil , ..., taken into consideration in the module h_to_tm1, are calculated in advance and transferred to the module h_to_tm1.

If now the other module ber_p_Oel, as already described, would be executed in a simple form without taking account of occurring pressure oscillations, it would be in this case in the stroke control to unacceptable deviations of the actually set valve lift h _is the predetermined setpoint h _Soll .

In an advantageous development of the method according to the invention, therefore, the function module ber_p_Oel is designed such that it avoids or compensates for the abovementioned deviations by calculating a value of p _oil dependent on the _desired stroke h _desired value.

15 exemplifies this general solution concept in a form in which the function module ber_p_Oel contains two submodules or subfunctions ber_p_Oel mittel and ber_p_Oel_h_komp whose respective calculation results are additively linked and thus yield the output value p _{Oel of} the overall module.

The subfunction ber_p_Oel medium performs in this embodiment, a simple calculation without vibration _corrections , resulting in the result _size p _{Oel_m} results. The further module ber_p_Oel_h_komp, which is responsible for the pressure oscillation compensation, determines a correction _value p _{Oel_h_komp which is} added to p _{Oel_m} after the presence of both partial results as a function of the desired stroke h _setpoint and further parameters: p Oil = p Oel_m + p Oel_h_komp (12)

The sub-module ber_p_Oel medium can be carried out as described above in connection with the 8th has been described for the function module ber_p_Oel as a whole. The size p _{Oel_m} can, for example, as an initial pressure at the beginning of the opening of the control valve MV1 of the relevant actuator 30 To be defined. In this case, p _{Oel_m} not depend on the stroke h _target. An average pressure drop as a function of the quantity of oil taken off per unit of time, as well as longitudinal oscillations of the pressure (pressure waves) in the high-pressure accumulator induced by conveying and removal processes 13 , can be taken into account in the calculation of p _{Oel_m} , as described above.

An exemplary realization of the subfunction module ber_p_Oel_h_komp is in 16 shown as a block diagram. The block diagram represents a program- _technical execution of an algorithm which determines the correction _variable p _{Oel_h_komp} for a given operating point and desired stroke h _Soll on the basis of an operating-point-independent oscillation form. This is represented by a function f _swinging in 16 is implemented as a characteristic f_schwing example.

Amplitude, phase and frequency of the vibration shown as a function of the desired stroke are calculated specifically for the respective operating point, this at least by a state variable for the high pressure side effective oil pressure, for example, a pressure value p _rail , if necessary, by an oil temperature T _oil , and optionally further Influencing variables is described.

Example becomes in 16 the oil pressure represented by a - possibly still suitably processed - detection variable p _rail . Alternatively, you can For this purpose also the calculation result p _{Oel_m of} the submodule ber_p_Oel medium or another suitable model size for the effective oil pressure can be used.

In the 16 implemented solution represents a practical specialization of the following general approach: p Oel_h_komp (h) = A · f vibratory (B + C * h + D * h 2 + ...), (13) which is evaluated with h = h _target .

The parameters amplitude A, phase B, "frequency" C, etc. are generally dependent on p _rail or optionally p _{Oel_m} , T _Oel and possibly other influencing variables: A = A (p rail , T Oil , ...) B = B (p rail , T Oil , ...) C = C (p rail , T Oil , ...) etc. (14)

In the in 16 used specialization accounts for non-linear terms in h, that is, D = 0, etc. The function f _swing , which describes the shape of the vibration is realized there as an example as a characteristic. But it can also be approximated by analytic function terms.

The parameters A, B and C are in 16 illustrated by simply calculable rational terms depending on the oil pressure, the coefficients are designed in terms of temperature dependence as a characteristic curves: A = p + C 0 B = (C 1 + C 2 · P) / (C 3 + p) C = 1 / (C 3 + p) (15) With
p = p _rail ,
C ₀ = C ₀ (T)
C ₁ = C ₁ (T)
C ₂ = C ₂ (T)
C ₃ = C ₃ (T) and
T = T _oil .

Also practical is a representation of the temperature dependence by series developments in T, typically limited to Terms of maximum degree 2.

The Parameters A, B, C according to equation (14) can alternatively also be used as maps be realized, which is also a practical solution. Furthermore, it may be advantageous to set the parameters A, B, C depending on of fabric sizes like viscosity and elasticity represent the hydraulic oil respectively to calculate what is already above for the analog Case of the parameters of a functional description of the Hubtransferfunktion accomplished and justified has been.

The numerical values scalar parameters, characteristics and maps are obtained by an adaptation of the inverse Hubtransferfunktion calculated according to the selected algorithm, for example equations (10) to (15) to real, for example determined by measurements gradients and in a suitable form in the data memory of the controller 31 stored.

In this adaptation, it is advantageous to determine the values of the characteristic _maps used by way of example for the scale parameters HWA, STP and TW such that the gradients of the transfer _function calculated without the correction term p _{Oel_h_komp} average or interpolate the oscillations occurring in real, for example, measured progressions.

With the addition of the correction term, which the function module ber_p_Oel_h_komp calculated, then the deviations of the real, vibrational courses from the "uncompensated" interpolating Course through a suitable determination of the parameters of this function module best possible reduced.

To illustrate this fact is in 17 a set of measured lift transfer functions is shown along with the calculated calculated runs without pressure swing compensation.

18 shows the change with the addition of the pressure swing compensation. In this case, calculated and measured curves are almost identical except for small residual errors. In both cases, inverse stroke transfer functions t _m1 (h) were calculated and represented in the curve diagrams in the form of the transfer functions h (t _m1 ), which means only a permutation of the axes.

In 19a are out for the bottom of the curves 18 the average rail pressure, which is 50 bar, as well as the stroke-determining oil pressure p _oil with pressure swing compensation shown. The rail pressure corresponds approximately to the value of p _oil calculated without vibration compensation. The difference of the curves thus representing up to scaling _oscillating f the course of the particular for this example waveform.

A further, equally advantageous realization of a pressure oscillation compensation has already been mentioned above, which can also be added as a modular supplement to an existing simple stroke control function. In this solution, the Druckschwin Compensation compensated by a correction value to the input h _add the existing control function additively or multiplicatively.

An advantage of this solution is that it allows for an approximately simultaneous control of several gas exchange valves of a cylinder, such as two intake valves, essential parts of the calculations for controlling the valve strokes must be performed only once. This includes, in particular, the evaluation of characteristic diagrams, such as the maps HWA and TW occurring in the above example, which only have to be done once if both the maps and the current values of the influencing variables p _Oel and T _{Oel are} equal. In the first solution of the pressure oscillation compensation, this requirement would no longer be met in the case of different valve strokes.

19b shows a block diagram exemplifying the program execution of such a solution for the function module h_to_tm1 according to 8th describes. The existing simple control function (without pressure oscillation compensation) is shown therein as a function module h_to_tm1_basis. This functional module is realized by way of example in the form, as has been described above for the first embodiment of a stroke control function according to the invention.

The input variables p _Oel , T _Oel , v (valve index) and E _x (further sizes) of the function module h_to_tm1 are passed on directly to the module h_to_tm1_basis. On the other hand, the input quantity h _{setpoint is} first divided by a correction factor f _corr and only then the "pressure oscillation compensated" setpoint value of the valve lift calculated in this way is introduced into the corresponding input of the function module h_to_tm1_basis.

The correction factor f _corr is determined according to f corr = 1 + Δ hkomp (H Should , p Oil , T Oil ) (16) is formed, where the value Δ _hkomp ber_delta_hkomp calculated from a part of calculation module and provided. For the realization of this calculation can be largely based on the in 16 referenced algorithm belonging to the first described above solution of a modular pressure oscillation compensation. It is merely an adaptation of the amplitude, that is, of the multiplied by the calculated value of f_schwing factor in the sense that there is an additional dependence on the desired stroke h _Soll addition. This dependence can be approximated by a simple rational function term.

The quality of this second exemplary embodiment of a pressure oscillation compensation corresponds to the quality of the first specified solution, for which in 18 a comparison between calculation and measurement is shown. This looks practically the same in the case of the second solution. The course of the correction factor f _korr determined in this example for the lower of the curves is shown in _FIG 19c shown. It corresponds in shape to the course of the pressure-oscillation-compensated _oil pressure p _{oil of} the first solution.

In a further alternative realization of the pressure oscillation compensation, it is also possible to determine a correction term which, in addition or multiplicatively to the calculation result t _m1 , adds to an existing control function.

In general, it is possible to compensate for any, not too great a shift in the value of the input variable p _{Oel of} a conversion function (manipulated variable transfer function) f _tm1 by a correction term, which is either for example to the further input variable h _{setpoint of} the function f _tm1 , or to its calculation result t _m1 , additive or multiplicative addition. This general possibility, which is not limited to the influence variable p _oil alone, can be carried out in the case of several calculations of the stroke control function to be carried out at the same time of calculation (for several actuators) 30 ) can be used advantageously to shift existing differences of the individual influencing variables of these controllers, such as p _Oel , into corresponding correction terms at the desired values h _Soll or the result values t _m1 , with the advantage _arising in typical applications that then essential parts of the calculations for the concerned operators 30 are identical and therefore only need to be performed once. This applies in particular to the evaluation of non-controller-specific maps.

In principle, parameters such as the quantities HWA, STP and TW which occur in the exemplary representation of the manipulated variable transfer function f _tm1 according to equation (10) can be determined on the basis of data, for example maps, which are individual for individual hydraulic actuators 30 or even uniformly - in only one copy each - for groups of identical or similar behaving hydraulic actuator 30 in the data memory of the controller 31 be filed.

The individual-individual calculation of the parameters of the manipulated variable transfer function, or the use of individual individual characteristic maps, in particular represents a possible solution to the quite practice-relevant problem that also identically constructed electro-hydraulic actuators 30 due to unavoidable tolerances in the production do not behave completely identical to each other, that is, more or less significant - and possibly in view of the allowable Hubfehler may have too large - Exemplary scattering.

However, this does not solve the problem of a drift in the properties of hydraulic actuators which likewise occurs in practice 30 or working cylinder 3 that is, the problem of changes in performance over time.

In a further advantageous further development of the method according to the invention described above, it is therefore provided that by means of an adaptation of the controller, despite possibly occurring changes in the operating behavior of a controller 30 the required quality or accuracy of the stroke control of the actuator is guaranteed.

This is done according to the invention by means of suitable adaptation parameters in the algorithms for calculating the manipulated variable transfer function f _tm1 and its influencing variables. In concrete terms, this means that one or more adaptation parameters are introduced into these algorithms, or that parameters already present in them are configured appropriately as adaptation parameters.

It is advantageous if a general dependency of operating point-specific influencing variables, especially the oil pressure p _oil and the oil temperature T _oil , is considered for these adaptation parameters, and if this dependence is realized in tabular form in the form of characteristic maps, at least with respect to a part of the influencing variables. On the basis of a continuous operating point-related recognition respectively identification of behavior changes of a controller 30 If necessary, corresponding changes in the operating point-related stored values of the adaptation parameters or the operating-point-related characteristic map values are made in relation to the setting of the valve lift in order to minimize or minimize stroke errors as a result of the drift.

In an advantageous extension of this method, it is also possible to consider a dependency on the desired value h _Soll , or of an equivalent quantity, for an adaptation parameter. In this case, it is basically sufficient to use only a single adaptation parameter, which is generally determined as a function of the operating point and of the setpoint stroke.

The variable values of the adaptation parameters or adaptation maps are thereby individually for individual controllers 30 filed to account for a generally individually pronounced drift.

Of the primary Advantage of such an adaptation of the stroke control function exists in that the effects of changes in operating behavior or a drift on the Hubstellvorgang and the resulting Valve lift in the control function can be compensated so effectively can, that this solution even with high demands on the positioning accuracy and major changes of the operating behavior during operation or over time.

In many cases, another advantage opens up in the fact that the often existing production-related copies scatters of identical actuator 30 can be compensated by means of an adaptation of the individual adaptation parameters or adaptation maps. This can already happen at the end of the assembly, as soon as one out of modifiers 30 , Control unit 31 and possibly other components existing overall system is assembled or assembled. Such a compensation of specimen spreads by means of the adaptation parameters is advantageous, since in this case the other parameters used in the stroke control function, and exemplarily realized as characteristic diagrams, are not individual for individual actuators 30 must be executed. This drastically reduces the need for data storage and / or computation time.

20 shows an example of two actuators 30 ₁ and 30 ₂ the course of the transfer function h (t _m1 ) at three operating points (rail pressures) at T _Oel = 30 ° C, these two basically identical actuator 30 visibly distinguish, that is, have a clear specimen distribution.

According to the statements made above, the hub transfer function or its inverse are written as follows: h = f stroke (t m1 ; T Oil , p Oil , E x see (1) t m1 = f tm1 (h; T; Oil , p oil , E x see (2)

As operating point-specific parameters, the pressure p _oil and a temperature T _Oel and optionally present further influencing variables E _{x occur} , for example, on the gas exchange valve 1 acting gas forces, on.

Using a suitable algorithmic representation of the inverse Hubtransferfunktion or manipulated variable transfer function is for a respective pending adjustment of a gas exchange valve 1 from the desired target stroke h = h _setpoint the driving time required t _m1 of the magnetic valve MV1 is calculated. In the control unit 3l is this algorithm realized as a computer program, for example.

The exemplary implementation of an inventive adaptation of the stroke control described below is generally based on the assumption that a module h_to_tm1 as well as further modules required for calculating the influencing variables of this module (see for example FIG 8th ) available. These modules can be realized according to one of the embodiments described above. In these cases, parameterized representations of the manipulated variable transfer function occur, wherein the parameters are defined as functions as a function of p _Oel and T _Oel and optionally further influencing variables.

However, this type of execution of the module h_to_tm1 is not necessarily required to carry out the adaptation. Rather, the adaptation according to the invention generally assumes that at least one generally operating-point-dependent parameter, if appropriate also a plurality of such parameters, is present or added in a form suitable for the adaptation, whereby this suitable expression is based on a data record which is for each to be adapted hydraulic actuator 30 is present in an individual copy, the values of these data - at least partially - during operation can be changed.

For an exemplary realization, these valve-individual data sets are designed as adaptive, that is to say variable characteristics or as adaptive characteristic diagrams as a function of at least one of the influencing variables, that is to say _oil , _oil , etc. In an advantageous form in practice, see below, for example, adaptive maps are used over p _oil and _oil .

Such adaptive characteristics or maps can optionally For example, as so-called grid maps with defined nodes the input variables and / or as so-called associative maps with freely distributed (also changeable or delete and add) support points, which are stored together with the map values, trained be.

The for the Adaption to be used parameters may under certain circumstances - in whole or in part - under the already existing parameters of a basic approach of the manipulated variable transfer function, or a basic version "h_to_tm1_basis" of the module h_to_tm1, found and in for the adaptation purposes suitable manner, for example as adaptable Functions or maps, be trained. alternative or additionally can Parameters for added the purposes of adaptation become.

• a) ein Druck-Offsetparameter poffset = poffset (pOel, TOel, ...), (17)der additiv zur Eingangsgröße p_Oel von h_to_tm1 hinzutritt,
• b) ein t_m1-Offsetparameter tm1_offset = tm1_offset (pOel, TOel, ... ), (18)der zur Ausgangsgröße von h_to_tm1 addiert wird, woraus sich die benötigte Ansteuergröße t_m1 ergibt,
• c) eine Hub-Skala hskala = hskala (pOel, TOel, ...), (19)die multiplikativ zum Hub h, speziell dem Sollhub h = h_Soll vor der Berechnung der inversen Transferfunktion gemäß Gleichung (2) hinzutritt,
• d) ein Hub-Offset hoffset = hoffset (pOel, TOel, ...), (20)der additiv zum Hub h, speziell dem Sollhub h = h_Soll vor der Berechnung der Stellgrößen-Transferfunktion gemäß Gleichung (2) hinzutritt.

Examples of such suitable and advantageous adaptation parameters, which can optionally be added as required to any execution of the module h_to_tm1 (that is, to any base approach of the stroke control function), are:

• a) a pressure offset parameter p offset = p offset (p Oil , T Oil , ...), (17) the additively added to the input variable p _oil of h_to_tm1,
• b) a t _m1 offset parameter t m1_offset = t m1_offset (p Oil , T Oil , ...), (18) which is added to the output of h_to_tm1, resulting in the required control variable t _m1 ,
• c) a stroke scale H scale = h scale (p Oil , T Oil , ...), (19) multiplicative to the stroke h, specifically the target stroke h = h, to _{be added} prior to the calculation of the inverse transfer function according to equation (2),
• d) a stroke offset H offset = h offset (p Oil , T Oil , ...), (20) the additive-h to the stroke, especially the target stroke h = h _set before calculating the manipulated variable transfer function according to equation (2) should occur.

In generalization of the uniform scaling of the stroke (case c)), independent scaling of the stroke in separate regions of an area-wise subdivision of the stroke range, or of the t _m1 range, can also be considered. Advantageously, for example, a separate scaling for small or large strokes, with two independent adaptation _parameters h _{scale_low} and h _{scale_high} . The transition point can thereby be oriented at the transition from an acceleration range into an asymptotic range (at approximately constant speed) of the stroke movement, or at a transitional stroke, in which, in the case of a two-part stepped embodiment of a piston 5 as described earlier - whose upper effective area A _ob changed.

Incidentally, this generalization represents a specific example of the above-mentioned general possibility that an adaptation parameter, in this case h _scale , is also considered or determined as a function of the desired value h _Soll .

In all cases, these parameters can easily and advantageously in a modular manner to an already existing realization of a Basi sentence as module h_to_tm1.

Alternatively or additionally, multiplicative or additive adaptation parameters can be added to individual parameters which occur in the basic approach of stroke control, if such parameters are not used and executed directly as already indicated as a possibility as adaptive parameters. In a piecewise description of the Hubtransferfunktion, for example, based on two areas, can be introduced in this way, the above-mentioned scales h _{scale_low} and h _{scale_high} in an elegant way.

In the case of a multiplicative adaptation parameter, for example h _scale , a "perturbation" representation is often advantageous in the following way: H scale = 1 + Δh scale , (21) in this case Δh _scale serves as an adaptation parameter and in the manner mentioned is generally represented as a function of influencing variables, in particular p _Oel and T _Oel .

In practice, the use of the printing and the t _m1 -Offsets proves to be a very good choice, which can be compensated by the appropriate tracking of these two adaptation sizes already typical, even pronounced changes in the actuator behavior in the stroke control to the extent necessary.

In the case of high demands on the accuracy of the stroke control and strong actuator drift, the addition of a stroke scale - especially in the expression according to equation (21) - is recommended as the third parameter, which results in a generally noticeable and practically in all cases sufficient improvement of the adaptation quality. In this case, the adaptive scaling of the stroke can also be limited to a partial region of the stroke, for example in the sense of the abovementioned parameter h _{skala_high} .

Alternatively, it is also advantageous to use combinations of the parameter t _{m1_offset} with the parameter h _scale or with the two parameters h _{scale_low} and h _{scale_high} , the stroke scales being in the form ( 21 ) may be formed.

Furthermore, as already stated, it is also possible to consider, in addition to the operating-point-specific influencing variables, a dependency on the nominal stroke h _desired or an equivalent variable for an adaptation parameter. Basically, then a single such parameter is sufficient for the purpose of adaptation.

For example, an adaptation _value t _{m1_offset} can be _calculated as a function of influencing variables, such as p _Oel and T _Oel , and an "uncorrected" drive time t _{m1_u} , the variable t _{m1_u} being calculated by means of a calculation function f _{tm1_basis} for a non-drifted actuator 30 (Reference plate) is determined: t m1_u = f tm1_basis (H Should ; p Oil , T Oil , ...) and t m1 = t m1_u + t m1_offset (t m1_u ; p Oil , T Oil , ...)

The function t _{m1_offset} can be described in an already known manner advantageously by a parameter function as a function of t _{m1_u} . The parameters or coefficients of this function, which depend on the influencing variables p _Oel , T _Oel ,..., Can be represented by way of example by means of characteristic diagrams via p _Oel and T _Oel , the values of these characteristic maps being adaptable.

In a particularly preferred embodiment, said parameter function is approximated by a piecewise polynomial having fixed t _{m1_u} support points. The optimization task described below - see equation (29) - for the determination of the adaptation values, in this case the coefficients of the piecewise polynomial, can then be attributed to a problem of linear regression known from the prior art. For this purpose, efficient non-iterative solution methods are known, which are very well suited for implementing the adaptation according to the invention in a control unit ( 31 ) are suitable.

All these with regard to the choice of the adaptation parameters mentioned Concepts are also advantageous in that they - besides the compensation of a stellar drift - also a compensation of allow production variations resulting in production, as described above.

The core of the adaptation according to the invention is that in the case of a drift, that is to say a gradual change of properties of an actuator 30 during operation, the present invention adaptation variables are selectively changed to avoid otherwise occurring all too large errors of the stroke control.

For this purpose, the invention assumes that there is feedback about the actuator behavior that makes it possible for a single actuator to be customized 30 its positioning accuracy or the error of stroke adjustment, respectively, the change in this positioning accuracy or the control error over time to track and evaluate.

These measurements and reviews can depending on the requirements in the regular operation of a controller 30 be carried out, or in specially performed measuring cycles, for example, under the creation or exploitation of special operating conditions, in which, for example, a single actuator 30 undisturbed by other, otherwise usually superimposed settlements other Steller 30 is examined.

The feedback regarding the stroke can be done by way of example by a direct detection of the stroke by means of a stroke sensor. Alternatively, measurement signals are evaluable, in which the stroke of the piston 5 , or the valve lift of an associated gas exchange valve 1 , and / or the course of the opening or closing of a gas exchange valve (in which the stroke also comes in) significantly.

• – die Erfassung des Schließzeitpunktes eines Gaswechselventils 1 zum Beispiel durch Detektion des beim Auftreffen in den Sitz ausgelösten Körperschalls (zum Beispiel mit Klopfsensor), woraus die Rücklaufzeit beim Schließen und daraus wiederum der Ventilhub h bestimmbar sind;
• – die schnelle Erfassung des Signals eines Differenzdrucksensors (nicht dargestellt), das – bei einer Erfassung des Druckabfalls über dem Magnetventil MV1 – ein nicht-lineares Maß für den Durchfluss am Magnetventil MV1 liefert. Daraus ist über ein entsprechendes Durchflussintegral der Hubverlauf und damit der resultierende Hub H des Gaswechselventils 1 rekonstruierbar. Auch bei einer anderen geeigneten Positionierung eines Differenzdrucksensors, und sogar auf Basis eines Absolutdrucksensors, ist eine analoge, indirekte Bestimmung des Ventilhubs möglich.
• – die Erfassung der Zylinderfüllung zum Beispiel mittels eines Sensors für den angesaugten Luftmassenstrom, wobei aus der sich in Abhängigkeit des Hubes h eines Einlassventils einstellenden Zylinderfüllung – gegebenenfalls auch durch einen Vergleich dieser Füllungen bei einer Variation des Schließzeitpunktes des Einlassventils – auf den Hub zurückgeschlossen werden kann.

As possibilities of such an indirect determination of the stroke or lifting error may be mentioned here by way of example:

• - The detection of the closing time of a gas exchange valve 1 for example, by detecting the structure-borne sound (for example, with a knock sensor) triggered when it hits the seat, from which the return time during closing and, in turn, the valve stroke h can be determined;
• - The rapid detection of the signal of a differential pressure sensor (not shown), which - upon detection of the pressure drop across the solenoid valve MV1 - provides a non-linear measure of the flow at the solenoid valve MV1. This results in the stroke profile and thus the resulting stroke H of the gas exchange valve via a corresponding flow integral 1 reconstructed. Even with another suitable positioning of a differential pressure sensor, and even based on an absolute pressure sensor, an analog, indirect determination of the valve lift is possible.
• - The detection of the cylinder charge, for example by means of a sensor for the intake air mass flow, which can be deduced from the depending on the stroke h of an intake valve cylinder filling - possibly also by comparing these fillings in a variation of the closing time of the intake valve - the stroke ,

Further examples are conceivable. For the purposes of the present invention, it suffices if an optionally non-linear, monotonous measure w (h) (22) of the actual stroke h = h _{Is determined} from suitable measurement signals and the required for the evaluation of Hubfehlern reference value w (h _setpoint ) can be calculated. With direct stroke sensing or calculation, the function w is the identity.

Alternatively, for the purpose under consideration, a non-linear measure may also be used d (h is - H Should ) with d (0) = 0 (23) can be used, provided that this can be determined from an indirect feedback on the Arbeitzylinderhub.

The goal of the adaptation consists of occurring lifting errors or a direct or indirect measure available therefor according to (22) or (23) F = w (h is ) - w (h Should ) respectively F = d (h is - H Should ) (24) to minimize or minimize by suitably adjusting the adaptation parameters.

To carry out this method, a function module "adaptation" is provided, which is exemplified in 8th is shown. His method of working, as well as the corresponding measures or adjustments to other modules, are described in more detail below.

The module "Adaption" evaluates error measures according to (24), which are to Stellabgängen a gas exchange valve 1 at known operating points, for example characterized by p _Oel and T _Oel belong. If the evaluation leads to the fact that detected deviations of the strokes are to be reduced from the setpoint, the module "Adaption" adjusts one or more of the map values of exemplary - for representation of operating point dependent adaptation parameters - used adaptive maps, these changed values in each case to the currently considered operating points p _oil, with T _oil to be reduced Hubabweichungen belong. in the specific case of associative maps "learning" improved or new map values are usually classified as the addition of new and possibly cancellation is realized of old values.

The Module "Adaption" checks if the Adjustment leads to improvements and in any case ensures that a single adjustment not inadmissible Can have an impact, for example by means of a change in each case enough small steps.

Possibly will be an unfavorable Reversed adjustment or if so, a preliminary check of a possible Adjustment arithmetically, for example as part of a state of the art The technique known optimization method for finding "optimal" values to be adjusted Parameter, done - too refrain from the beginning.

at Such an exemplary realization may be the mode of operation the adaptation according to the invention be compared with a "slowed down" regulation in its basic scheme, in which not a single setting process, but whole ensembles of for example 10 or 100 or more actuations for the Characterization of a "regulatory error" and the rating a (desired) temporal change of this error by the "control intervention" are used.

By way of example, an adaptation by means of print offsets and actuation duration offsets is implemented below, which are implemented as adaptive characteristic diagrams as a function of p _Oel and T _Oel . Instead of T _Oil could also be the viscosity η _oil are used as described above.

These adaptation maps, which are available on a valve-specific basis, can be realized within the central stroke control module h_to_tm1. Accordingly, the adjustment intervention on the part of the module "adaptation" in 8th indicated by a double arrow leading to the h_to_tm1 module.

Alternatively or additionally, parameter values in the calculation modules ber_p_Oel, ber_T_Oel or other such modules can also be used as adaptation values in the sense of the invention and adjusted by the module "adaptation" 8th indicated by additional double arrows (not shown). For example, it would also be possible to consider an above-mentioned adaptation parameter p _offset in the module ber_p_Oel and add it to the calculation result p _oil .

For the exemplary embodiment of an adaptive stroke control according to the invention, the central stroke control module h_to_tm1 is adjusted according to FIG 21 divided into the submodules ber_p_Oel_a, h_to_tm1_basis and ber_tm1.

The task of in 22 By way of example, the module ber_p_Oel_a is used to add the pressure _offset p _offset valid for the current calculation to the pressure p _oil , which results in the output value p _{Oel_a} .

The respectively considered valve setting process is designated by the valve index v, which serves as a selector for the evaluation of the "correct" p _{offset characteristic} field.

The module h_to_tm1_basis implemented by way of example according to equations (10) and (11) a "base approach" f _{tm1_basis} the manipulated variable transfer function f _tm1 in the embodiment as a scale approach with the scale function fa _inv = arcoshyp exp.

In this case, scale parameters HWA, TW and STP occur which, as already mentioned, depend on p _Oel and T _Oel and optionally also on further influencing variables.

The block diagram in 12 represents an exemplary implementation of this algorithm, thus also the manipulated variable transfer function f _{tm1_basis} or the module h_to_tm1_basis according to 21 , wherein the scale parameters are implemented as maps in dependence on p _oil and T _oil .

The maps HWA, TW and STP are not executed valve-individual in this example. Therefore, it is assumed that the assumption is quite practical that the adaptation by means of the pressure and t _m1 offsets is also able to compensate for the production-related differences in copy of hydraulic working cylinders. The data applied in the maps HWA, TW and STP go for this purpose by a representative "average" behavior of a hydraulic actuator 30 out.

According to 21 the offset-corrected (adapted) pressure parameter p _{Oel_a} , which was previously calculated by the module ber_p_Oel_a, is used instead of p _Oel in the calculation of the module h_to_tm1_basis (in the implementation according to FIG 12 ). The output _{size of} this module is called t _{m1_u} . In this case, the index "u" stands for "uncorrected", since a drive duration offset correction follows.

The subsequent offset or adaptation _correction of t _{m1_u} is the task of the modul ber_tm1, which in 23 is shown. Again, the selection of the correct adaptation characteristic field for the evaluation of the parameter t _{m1_offset is} effected by means of the current value of the valve selector v. This is added to t _{m1_u} , which provides the final desired control variable t _{m1 of} the stroke control.

For the final version of the module "Adaption" according to 8th Let us assume by way of example that on the basis of the available feedbacks or acquisition variables of the system, the actually set stroke h _{actual can be} calculated and calculated.

For the evaluation of the quality of the adaptation is exemplified the quadratic error F 2 = (h is - H Should ) 2 (25) calculated individual actuations of a working cylinder.

Alternatively, the analogically defined quadratic error for the drive time t _m1 can be considered: F 2 = (t m1_ber - t m1 ) 2 , (26) where t _{m1 is} the actual driving time of a lifting operation. The size t _{m1_ber} however, is the (theoretical) drive time, which _is h the measured stroke or Isthub would have resulted. It is recalculated from the actual stroke h _{ist of} the lifting operation by means of the calculation algorithm for the inverse stroke transfer function (see equation (2)) present in the stroke control: t m1_ber = f tm1 (H is ; p Oil , T Oil , E x ). (27)

The average error of the stroke control, which according to (25) or (26) at a given operating point, here exemplified by p _Oel and T _Oel , calculated or determined, should - so the goal of adaptation - be as small as possible and possibly (if in principle possible) do not exceed specified error thresholds.

For the evaluation of such a mean error, which the module "adaptation" performs at regular or irregular time intervals again and again, an ensemble of setting events (measurements) is considered, which are assigned to an operating point p _oil , T _oil , which ensemble in each case also comprises a certain number of different activation times t _m1 or strokes h.

In this case, a sum of the error squares according to (25) or (26), possibly weighted according to criteria such as, for example, the stroke or the "age" of a measurement, is formed. Here is an example of the root of the mean square error introduced as a measure of goodness: wmqf = root (ensemble mean of F 2 ) (28)

By way of example, the module "adaptation" can be carried out such that on the basis of an existing ensemble of measurements at a given operating point, it recalculates the parameter values p _offset and t _{m1_offset} belonging to this operating _{point in such a} way that the measure ( 28 ) is minimized at least approximately: wmqf → Min. (29)

With a number of two parameters to be determined here (p _offset and t _{m1_offset} ), this task can be _{solved in} principle on the basis of only two measurements with strokes not too close to one another, if measuring errors and / or errors of the stroke modeling can be kept small enough.

In this case, under fairly general conditions, which are generally satisfied in the applications to be considered here, the minimum according to the problem (29) is assumed to be 0. This means that the adapted calculated Hubtransferfunktion as "2-point shape" is determined by the two measuring points, that is, through these points. Accordingly, the target values p _offset and t _{m1_offset of} the adaptation can be determined from the following equation system: H is (1) - h Should (1) = 0 h is (2) - h Should (2) = 0 (30) the measurement points being represented by the numbers 1 and 2 in brackets. The drive times associated with these points are correspondingly designated t _m1 (1) and t _m1 (2).

From (30) results with the inverse Hubtransferfunktion f _{tm1_basis of} the base approach of the stroke control, in the module h_to_tm1_basis according to 21 is implemented, according to equation (2), the transformed system of equations: t m1_offset + f tm1_basis (H is (1); p Oil + p offset , T Oil , E x ) = t m1 (1) t m1_offset + f tm1_basis (H is (2); p Oil + p offset , T Oil , E x ) = t m1 (2), (31) from which an implicit equation for the determination of p _offset results by the elimination of t _{m1_offset} : f tm1_basis (1) - f tm1_basis (2) - t m1 (1) + t m1 (2) = 0 (32)

The argument _lists of the inverse transfer function f _{tm1 of} the basic approach for the two measurement points are abbreviated, as indicated above.

The sought target value p _offset is thus to be determined as the zero of the left-hand side of equation (32), which is considered to be a function of p _offset . This is feasible approximately with iterative methods known in the art. A gradient method (Newton) may be mentioned by way of example, whereby one can often limit oneself to one or two iterations.

With the thus found approximate solution for p _offset , an associated target value of t _{m1_offset} is then _calculated from one of the equations from (31). The above-mentioned calculation steps are by means of algorithms which can be implemented simply and efficiently as computer programs with a corresponding realization of the module "adaptation" in the control unit 31 feasible.

This exemplified solution for the module "adaptation" is advantageously carried out for the practical application in such a way that the two measuring points considered in each case are already obtained by averaging the strokes from a plurality of individual measurements to the same operating point and with the same activation time be. This suppresses statistical errors of the measurement, which could otherwise disturb the procedure.

As well it is advantageous if the target values of the adaptation parameters are not entered directly into the adaptive maps, but over the time Sequence of reassessments and reassessments of adaptation be time-low-pass filtered. In each case, the current certain filtered values in the maps. In this way will ensure that changes the controller only slowly and errors from calculations and / or associated Measurements of the adaptation process in the stroke control can be suppressed.

As an alternative to the embodiment mentioned by way of example, it is also possible to use a larger ensemble in the respective recalculation of p _offset and t _{m1_offset} , that is to say the solution of the optimization _task according to (29), with advantage with regard to the quality of the adaptation. For this too, a system of equations can be derived and solved approximately. The system of equations results from the stationarity conditions equivalent to condition (29) d (RMSE) / d (p offset ) = 0 d (wmqf) / d (t m1_offset ) = 0 (33)

Using the error measure according to equation (26), this _results in an explicit equation for t _{m1_offset} and an implicit equation for p _offset , from which an approximate solution p _offset can be determined by means of known (iterative) methods. This solution method leads to a fast (computationally efficient) algorithm, for example, in the goal of a potentially very fast adaptation for the realization of the module "adaptation" in the control unit 31 can be selected.

Alternatively, a method known from the prior art for the iterative solution of the optimization task (29) can also be used. Examples include cyclic descent of coordinates, the steepest descent method and conjugate gradient methods. The higher computing time requirement, for example, a program piece, such a method example in the control unit 31 does not adversely affect the adaptation of gradual changes (as is generally the case here). The program piece can be executed in a practice-typical implementation, in which various control functions are executed as a multitasking program on the basis of a single computer unit, for example in a fast time or high-priority interruptible slow time frame or in a so-called background program.

As already described are the new target values the adaptation variables in one advantageous embodiment temporally filtered, with the filtered values in the adaptive Maps are entered. Unfavorable effects of possible respectively existing approximation errors and Measuring errors of the adaptation process on the stroke control become this way suppressed.

there allows it's enough often new made rating of quality the stroke control that drifts due to unfavorable adjustment of parameters recognized in time and prevented.

In some of the exemplary embodiments of the adaptation module, a selected evaluation measure, for example the quality measure (28) with the definition (25), is computationally determined for given values of the adaptation parameters, for which resulting strokes are calculated to known values of the influencing variables and t _m1 times Need to become.

In such cases proves to be very beneficial if the chosen mathematical Description of the inverse Hubtransferfunktion according to equation (2), which is based on the module h_to_tm1_basis, is invertible. In this case, immediately is also a mathematical description the one needed for the stroke calculation Transfer function according to equation (1) present and based on the already existing parameters the inverse Hubtransferfunktion easily and efficiently displayed in the control unit. In addition to this advantage, the adaptation process benefits greatly from the fact that Model error under these conditions practically not applicable come, as otherwise, that is, with different control and stroke calculation algorithms, as a rule would be given.

In the above embodiment, the invertibility of the stroke control algorithm is satisfied. For the stroke calculation used in the module "adaptation", the equation (5), supplemented by a drive-time offset t _{m1_offset} , is thus advantageously obtained. h = f stroke (t m1 ) = HWA · fa (STP / HWA · (t m1 - TW + t m1_offset )) (34) where fa is the inverse function to fa _inv and the maps HWA, TW and STP are evaluated for the inputs P _Oel + p _offset and T _Oel . The adaptation _parameters t _{m1_offset} and p _offset, in turn, are determined to the values p _oil , T _{oil of} the given operating point.

By the way, equation (34) can be used recognize that the parameter t _m1 offset represents an example of an adaptation parameter which adapts a basic parameter, in this case the parameter TW of the basic approach, in an additive (respectively subtractive) manner. In a functionally equivalent implementation, the map TW itself could instead be adapted, in which case it would be in an individual copy for each actuator 30 should be present.

The potential of the map adaptation according to the invention using the exemplified embodiment of the invention is shown in FIG 24 presented for a realistic designed test scenario. It depends on the behavior of two real controllers 30 gone out, showing a clear copy spread. This is in 20 for some Hubtransferfunktionen individual operating points shown. Based on this, a drift of the first actuator 30 ₁ assumed that his behavior in 6000 operating hours so drifts that he behaves in the end as Steller 0 ₂ . The drift was computationally generated by means of interpolation. At intervals of 300 operating hours, a map adaptation was carried out according to the method described by way of example for three values of p _Oel and T _Oel , that is, at 9 operating points of the first actuator, on the basis of approximately 11 measurements at different strokes. The change in the quality standard wmqf, which is calculated here over the entire ensemble of 100 points (with all operating points and activation times respectively strokes), is in 24 shown.

The Goodness of Lift control shows over the no significant changes during the entire operating period. The realistic Threshold set at 0.08 mm for the required quality of the stroke control is maintained at a considerable distance. The achieved thereby adjustment quality essentially from the base approach used and not from the adaptation.

To the Comparison also shows how the drift of the actuator behavior on the goodness the stroke control affects if no adaptation is used. In In this case, the threshold set for the required quality of the stroke control exceeded after more than 1500 operating hours.

The test example demonstrates the significantly different transfer functions of the identical controllers 30 , which was assumed in this example, at the same time also the possible and advantageous application of the adaptation according to the invention for the compensation of specimen scattering.

It is possible and occasionally advantageous to carry out the adaptation method described during the operation of a controller 30 or to limit an internal combustion engine equipped therewith to certain operating conditions and / or different procedures - for example, fast or slow adaptation - to use depending on the existing operating conditions. Also, if appropriate, a redundant design of adaptable parameters is advantageous, not all parameters being considered simultaneously. For example, depending on the operating state, different parameter sets, possibly also based on different methods, can be adapted.

The choice or the extent of the influencing variables characterizing the operating point is likewise not restricted. Advantages arise in an embodiment in which the changes in material properties of the oil, such as viscosity, are included in the control function of the stroke. This is possible, for example, on the basis of the detection values of a viscosity sensor (not shown) which is located in the oil circuit of the in 1 described system is installed. In such a case, the changes in viscosity caused, for example, by aging and / or contamination of the oil during operation are already taken into account in the control (without adaptation) so that no lifting errors can arise therefrom. Otherwise, that is without a viscosity sensor, an adaptation according to the invention is necessary in order to reduce or avoid the lifting errors which otherwise result from viscosity changes.

In a further advantageous embodiment of the invention it is provided that - for example, by the module adaptation on the basis of a weighted by this quality measure, as the measure wmqf according to equation (28) - regularly and / or checked at certain times, whether a required quality or accuracy the stroke control is still guaranteed. Alternatively or additionally, it can be checked whether adaptation parameters used remain within predefined permissible ranges. If it is determined that one or the other is no longer the case, for example, because an excessive change in the properties of a writer 30 and / or of the hydraulic oil used to operate the actuator, then or should an information or message be generated which indicates this fact. This information can be stored, for example, as diagnostic information or error message in an error memory and / or displayed by information means, such as a warning lamp. If such a situation is classified as critical, if necessary, on further examination, a protective measure may be taken, for example by a restricted operation of the operator concerned 30 , a decommissioning of this author, or even a shutdown of the entire Sys tems.

In the exemplary embodiments described so far, advantageous methods of the stroke control according to the invention of a hydraulic working cylinder 3 which is based on simple and efficiently realizable solutions without feedback from the actual hub, that is, from pure controls. The addition of an adaptation according to the invention furthermore also solves the problem of actuator drift that occurs in practice, ie the problem of changes in the actuator behavior over time.

on the other hand has a pure control the disadvantage that for typical applications of the Method, such as the control of gas exchange valves an internal combustion engine, needed a lot high positioning accuracy of the hub only with considerable effort in terms of Modeling various influences (Pressure, gas power and others), or if not reachable (reliably) is. This problem is not solved by the naturally slow adaptation.

In a further advantageous further development of the invention, it is therefore provided that deviations of the set stroke of an actuator 30 from the desired value, ie the desired stroke h _desired , be compensated by means of a fast control, which is added to an already described pure control, for example, a basic function above, as a modular extension. Also in this development, a supplementary adaptation can again be provided or added. On the one hand, the adaptation can serve to ensure the positioning accuracy of the basic function in the manner already specified, even in the case of actuator drift, and / or to compensate for specimen scattering. On the other hand, the regulation provided in the further development can also be supplemented by an adaptation or designed as an adaptive control. For the realization of corresponding developments of the invention, recourse may be had to solutions known in the art.

The primary advantage of a development of the method according to the invention in the sense of a regulation is that a detected error of the stroke control of a hydraulic working cylinder 3 is reduced or compensated very quickly. Thus, the required accuracy of the stroke control can be ensured even if the pure control by itself is not accurate enough in each case, or should be waived to limit the effort on extensions such as the above-mentioned wave correction and / or said pressure swing compensation.

Furthermore, it is advantageous that a controller under the control - or "precontrol" - the hub used and in particular carried out so that it already has a high degree of accuracy on its own. In this way, the demands on the positioning accuracy can be maintained even in highly dynamic operating points, that is, for example, in a very rapid change in the pressure p _rail in the high-pressure accumulator 13 and / or upon a sudden shutdown or engagement of the operation of a portion of the high pressure accumulator 13 connected actuator 30 or working cylinder 3 ,

Particularly advantageous is the combination with an already described method for adapting the precontrol. On the one hand, in this case, even if there is a change in the properties of a controller which are relevant for the setting process 30 and / or a hydraulic oil used, the accuracy of the precontrol substantially and thus the condition that the scheme continues - despite the above changes - can ensure the requirements of the positioning accuracy. On the other hand, as an option, the information about the positioning errors of the stroke obtained by the control in stationary operating points can be advantageously introduced into the process of adaptation of the precontrol.

In many cases, the control can respect _(Is h) of the required response to a stroke actually presented a working cylinder 3 from anyway, for example, for the purpose of monitoring or even the adaptation mentioned above, go out existing sensors. It is important that a corresponding sensor provides a fast and sufficiently accurate feedback at least in certain operating conditions or for certain areas of control variables such as the stroke h _Soll , so that the scheme can be used at least under these conditions. Suitable examples are the direct detection of the travel or stroke by Stellwegbeziehungsweise position sensor, or the reconstruction of the stroke from the signal of a differential pressure sensor, as mentioned above.

For the realization of a control, in the present case, it is important that before or at a calculation time of the stroke control for an imminent setting operation of a working cylinder 3 from the evaluation of the corresponding sensor signal for a past, ideally the last, setting process of this working cylinder whose actual stroke can be determined and compared with the desired stroke of this past positioning process.

For an exemplary embodiment of the inventions According to the invention control is assumed that at a calculation time of the stroke control of Isthub h _{Ist_1} the last setting operation, for example, from the signal of a stroke sensor, is determined. Thus, the deviation from the desired stroke h _{Soll_1} this last _adjustment process as a "control error" Δ h_1 = h Soll_1 - H Ist_1 (35) educated. On the basis of the control error Δ _{h_1} , a control algorithm calculates a correction _{variable Δtm1} that is used for an additive correction of the output _variable t _{m1_steu of} a " _feedforward control." This solution is described below with reference to _FIG 25 and 26 continue running.

25 shows an example of a corresponding implementation of the central Hubsteuerungsmodul h_to_tm1 according to 8th The feedforward control is present as a further submodule h_to_tm1_basis, which realizes one of the solutions according to the invention described above. 2) using the scale approach given in equation (10), see also 12 , The output quantity of the module h_to_tm1_basis is referred to as the pre-control _value t _{m1_steu of} the control _variable t _m1 . The control variable t _m1 finally results by adding the correction _value Δ _tm1 determined by the "control" module according to _FIG t m1 = t m1_steu + Δ tm1 , (36)

The module h_to_tm1_basis in this example is opposite to the one in 12 The realization has been expanded to the effect that it _outputs the last desired value h _{Soll_1} as well as the values tw ₁ , hwa ₁ and stp ₁ resulting from the last previous calculation of the manipulated variable transfer function of the parameters of this transfer function (in the selected exemplary representation). In addition to the actual value h _{Ist_1 of} the last _setting process, determined as specified, and optionally further input variables E _x , these variables are introduced into the "control" module.

An exemplary realization of the module "control" is in 26 shown as a block diagram, wherein a known simple control algorithm (PI controller) is used. In this case, the setting error Δ _{h_1} corresponding to equation (35) is formed directly at the input of the controller and multiplied by the output _{variable rising tm1_1 of} a submodule ber_steig_tm1. The result is a transformed into the Ansteuergrößenbereich control error Δ _{t_1} is integrated over an integrator member. The subsequent multiplication with a preferably calibratable amplification factor Ki yields the "I component" of the controller correction. A "P portion" is formed as a product of the transformed control error Δ _{t_1} and another preferably calibratable gain Kp.

The addition of the I component and the P component supplies the correction _value (controller _intervention ) _Δtm1 , which, as described _above , is added to the pilot control _{variable Δtm1_steu} in the higher-level module h_to_tm1.

Using the parameters tw ₁ , hwa ₁ and stp _1, the partial calculation module ber_steig_tm1 determines the slope of the manipulated variable transfer function tm1 (h) for the operating point, which is characterized by the setpoint h _{Soll_1} and the operating _{variables of} the last setting process. In an alternative embodiment of the submodules h_to_tm1_basis and "Regelung" of the module h_to_tm1 according to 25 The sub-calculation ber_steig_tm1 can also be moved to the module h_to_tm1_basis, which then transfers the slope _{value tm1_1} to the module "control." This functionally equivalent embodiment has the program advantage that the "control" module is then independent of the calculation algorithm used in the h_to_tm1_basis module is.

The example based on the 25 and 26 executed solution for the inventive control of the stroke of a hydraulic working cylinder 3 can be modified and / or expanded in many ways. Thus, any other solutions for the realization of the basis function h_to_tm1_basis can be used, which also includes the possible further extensions described above such as pressure compensation, wave correction and others. Corresponding alternatives to the linear PI controller used are known in the art, and non-linear and / or adaptive control algorithms may be used in particular This means that the control can be switched off in the presence of certain exclusion conditions and the control function according to the invention operates in this case as a pure control system 8th an adaptation according to the invention with the possible embodiments of the module h_to_tm1 be combined by means of which in the manner described a control of the stroke of a working cylinder 3 is realized.

In special implementations of the control function according to the invention within a control device 31 can do one or the other calculation module - in whole or in part - instead of running as a computer program in dedicated digital hardware.

Furthermore, in such a control device 31 in addition to the stroke control function also further, for the operation of an electro-hydraulic actuator 30 required respectively provided modules or functions, such as functions for controlling the closing operation and / or monitoring, be realized.

The control unit 31 can be performed - in whole or in part - as a so-called on-site electronic module, for example, to a controller 30 , or at a pressure accumulator 13 , or - in the case of an aforementioned electrohydraulic valve control - is mounted on a corresponding cylinder head.

The control unit 31 can be used to control multiple controllers 30 be responsible, in particular such controller, connected to a shared high-pressure accumulator 13 are connected. In a possible and advantageous embodiment of the control device 31 can do this for every single person 30 a vendor-specific electronic module may be present in which some or all of the functional modules, one for the controller 30 perform specific calculation or the signal of the actuator 30 associated sensor, for example, a stroke sensor, detect or process, and optionally some or all memory cells, the individual data, such as adaptation values, the actuator 30 included, are integrated.

Especially advantageous is an integration of some or all of these functions in an application specific circuit (ASIC), which is in the vendor specific Electronic module is installed.

It is also advantageous if in such an electronic module and corresponding components are integrated, for the electrical control of a controller 30 or its electrical components, for example a solenoid valve MV1, are required. These include, among other things, a drive signal generation unit and an output stage.

In a particularly preferred embodiment, such a vendor-specific electronic module, for example in the form of a Aufsetzteils, at a controller 30 grown or mounted. Such an embodiment has, inter alia, the advantage that the electrical components of a controller 30 how solenoid valves MV1 and MV2 and / or actuator-specific sensors can be directly connected by means of simple connecting elements with corresponding electrical connections of the attachment part. As a result, lines (as exemplified in 1 shown) or harnesses, which otherwise to a controller 30 , or its electrical components, would be introduced.

All contained in the drawings, the description and the claims Features can both individually and in any combination with each other invention essential be.

Claims (42)

Method for controlling a control device (MV1, MV2) of a hydraulic working cylinder ( 3 ) or Stellers ( 30 ), characterized by the following method steps: - reading in a setpoint value (h _setpoint ) of a control variable (h), - determining at least one operating parameter (T _oil , p _oil ), - calculating a control variable (t _m1 ) as a function of the setpoint value (h _setpoint ) and the at least one operating parameter (T _Oel , p _Oel ) by means of a manipulated variable transfer function (t _m1 = f _tm1 (h _Soll ; T _Oel , p _Oel , T _mot , ...)) with: f _tm1 : manipulated variable transfer function h _If : desired value of the control variable (setpoint stroke) T _oil , p _oil : hydraulic operating parameters T _mot , ...: further operating parameters and output of the control variable (t _m1 ) to an output unit for generating a corresponding control signal. A method according to claim 1, characterized in that the operating parameters (T _Oel , p _Oel ) and / or the further operating parameters are defined as scalar quantities or by individual values, such as an instantaneous value, an average, and / or a momentary or average time change , being represented. A method according to claim 1 or 2, characterized in that a hydraulic operating parameter (p _oil ) as an effective, the stroke of a working cylinder ( 3 ) Defining oil pressure is defined. A method according to claim 1 or 2, characterized in that a hydraulic operating parameter (T _oil ), the temperature of a working cylinder ( 3 ) characterized in a lifting operation of inflowing oil. A method according to claim 1 or 2, characterized in that a current value of a substance size, in particular the viscosity, one for operating the working cylinder ( 3 ) is used and used as a hydraulic operating parameter. Method according to one of the preceding claims, characterized in that the one or more hydraulic operating parameters (T _oil , _oil ) are measured directly. Method according to one of the preceding claims, characterized in that the one or more hydraulic operating parameters (T _Oel , P _Oel ) are estimated from one or more measured values (T _rail , P _rail ). A method according to claim 7, characterized in that the one or more hydraulic operating parameters (T _Oel , p _Oel ) stored in a respective characteristic field or a characteristic depending on one or more measured values (T _rail , p _rail ), or values calculated therefrom are. A method according to claim 7 or 8, characterized in that in the estimation of the hydraulic operating parameters (T _Oel , p _Oel ) from measured values (T _rail , P _rail ) further parameters (n _mot , t _mess , t _ob , h _Soll ). Method according to one of the preceding claims, characterized in that a mean pressure loss (flow loss) along a line to a working cylinder ( 3 ) and / or in a part of a high pressure accumulator ( 13 ) is determined as a function of a mean withdrawal volume flow and taken into account in the calculation of a stroke-determining oil pressure (p _oil ). Method according to one of the preceding claims, characterized in that a local-temporal course of a pressure wave or a longitudinal vibration in a high-pressure accumulator ( 13 ) and is taken into account in the calculation of a stroke-determining oil pressure (p _oil ) (wave correction). Method according to one of the preceding claims, characterized in that an influence of simultaneous or temporally overlapping setting operations is determined and taken into account in the calculation of a stroke-determining oil pressure (p _Oel ) (overlay correction). Method according to one of the preceding claims, characterized in that high-frequency pressure oscillations, which at a high-pressure side connection of the working cylinder ( 3 ) and / or within the working cylinder ( 3 ) occur during a setting operation of the working cylinder, and which lead to oscillations in the course of the manipulated variable transfer function (f _tm1 ) in dependence on the target stroke (h _Soll ), by an oscillating component dependent on the _desired stroke (h _Soll ) in the calculation of a stroke-determining oil pressure ( p _oil ) are taken into account (pressure oscillation compensation). Method according to one of the preceding claims, characterized in that a pressure loss and / or a wave correction and / or an overlay correction and / or a pressure oscillation compensation not in the calculation of a stroke-determining oil pressure (p _oil ), but in other ways, for example as additive or multiplicative Correction of a desired value (h _setpoint ) before the calculation of a manipulated variable transfer function (f _tm1 ), or as an additive or multiplicative correction of a drive variable (t _m1 ), are taken into account. Method according to one of the preceding claims, characterized in that the estimation of the operating parameter or parameters (T _Oel , p _Oel ) of the determination of the operating parameters (T _Oel , P _Oel ) at a future time of the control of the actuator (MV1, MV2 ) serves. Method according to one of the preceding claims, characterized in that the estimation of the operating parameter (T _Oel , p _Oel ) by means of averages, filter functions, in particular a low-pass filter, and / or physical models takes place. Method according to claim 16, characterized in that that the filter functions or physical models with memory variables work, and that the memory variables at every estimation re-imported and updated. Method according to one of the preceding claims, characterized in that the operating parameter or parameters (T _oil , p _oil ) are formed by averaging from measured values (T _rail , p _rail ) or calculated on the basis of such average values. Method according to one of the preceding claims, characterized in that disturbances, in particular hydraulic forces or frictional forces, depending on the or the operating parameters (T _oil , p _oil ) determined and taken into account in the control of the actuator (MV1, MV2). Method according to one of the preceding claims, characterized in that as a further operating parameter of the manipulated variable transfer function (f _tm1) a temperature of an internal combustion engine (t _mot), engine speed (n _mot), and / or on a piston ( 5 ) or to a gas exchange valve ( 1 ) acting gas forces are taken into account. Method according to one of the preceding claims, characterized in that the manipulated variable transfer function (t _m1 = f _tm1 (h _Soll ; T _Oel , p _Oel , T _mot , ...)) is represented by characteristics and / or maps. Method according to one of the preceding claims, characterized in that the manipulated variable transfer function (t _m1 = f _tm1 (h _Soll ; T _oil , p _oil , T _mot , ...)) is designed as a parameter function, or on the basis at least one parameter function is calculated. A method according to claim 22, characterized in that the parameter function has one or more, in particular elementary, function terms and corresponding scaling coefficients, and that the scaling coefficients from the hydraulic operating parameters (T _Oel , p _Oel ) and / or the other Operating parameters (T _mot ) are determined. Method according to claim 23, characterized that an elementary function term in the parametric representation the manipulated variable transfer function is performed as a characteristic, as a polynomial or as a rational function. Method according to one of the preceding claims, characterized in that the range of values of the setpoint stroke (h _setpoint ) is subdivided into a plurality of sections, and in that the manipulated variable transfer function (t _m1 = f _tm1 (h _setpoint ; T _oil , p _oil , T _mot , ...)) for each of these sections is represented separately by a map, a characteristic or a parameter function. Method according to one of the preceding claims, characterized in that the manipulated variable transfer function (t _m1 = f _tm1 (h _Soll ; T _Oel , p _Oel , T _mot , ...)) with respect to the relationship between t _m1 and h _Soll on simple Mode is invertible, and that the inverse function (stroke transfer function) can be calculated with the maps used to represent the manipulated variable transfer function, characteristics and / or function terms and scaling coefficients. Method according to one of the preceding claims, characterized by the following manipulated variable transfer function: t m1 = f tm1 (h) = HWA / STP fa inv (h / HWA) + TW with: HWA (T _Oel , p _Oel , ...): Curvature parameter STP (T _Oel , p _Oel , ...): slope parameter TW (T _Oel , p _Oel , ...): delay time parameter fa _inv : inverted scale function. Method according to Claim 27, characterized in that the inverted scale function fa _{inv has the} following form: fa inv = arcoshyp exp. Method according to one of the preceding claims, characterized in that in the manipulated variable transfer function (f _tm1 ) at least one multiplicative or additive adaptation parameter (p _offset , t _{m1_offset} , h _offset , h _scale ) is provided, which is generally or as needed Function of operating parameters (T _oil , p _oil ), and possibly as a function of the desired stroke (h _Soll ) or an equivalent size, is displayed; with: p _offset : pressure offset parameter t _{m1_offset} : control _duration offset parameter h _offset : stroke offset parameter h _scale : stroke scaling parameter A method according to claim 29, characterized in that at least one scaling parameter (h _scale ) in the form H scale = 1 + Δ hskala is specified. Method according to Claim 29 or 30, characterized in that at least one multiplicative or additive adaptation parameter (p _offset , t _{m1_offset} , h _offset , h _scale , Δ h _scale ) is calculated as a function of operating _variables (T _oil , p _oil ), and that the calculation rule or other data used in the calculation is variable for the purpose of adaptation. Method according to Claim 27, 28 or 29, characterized in that at least one multiplicative or additive adaptation parameter (p _offset , t _{m1_offset} , h _offset , h _scale , Δ h _scale ) is used as a characteristic field as a function of operating _variables (T _oil , p _oil ) is calculated or calculated using at least one such map, and that the data of the map for the purpose of adaptation are variable. Method according to one of the preceding claims, characterized in that an actual value (h _actual ) of the control variable (h) is detected directly or indirectly. Method according to one of the preceding claims, characterized in that a measure (F) for the deviation of the actual value (h _actual ) from the desired value (h _setpoint ) of the control variable (h) of an individual working cylinder ( 3 ) is formed. A method according to claim 34, characterized in that depending on one or more values of the deviation (F), which for respective actuating operations of a working cylinder ( 3 ), and possibly, depending on the values of the operating parameters (T _Oel , p _Oel ) and / or the further parameters (T _mot ) and / or the desired strokes (h _setpoint ), one or more adaptation parameters ( p _offset , t _{m1_offset} , h _offset , h _scale ) for this working cylinder ( 3 ) are adapted individually. A method according to claim 34, characterized in that a special operating state and / or a special mode of operation of a working cylinder ( 3 ) or a machine equipped therewith, to produce deviations (F) for given values of the hydraulic operating parameters (T _Oel , P _Oel ) and / or the further operating parameters (T _mot ) and / or the desired stroke (h _Soll ) of the working cylinder ( 3 ) in the best possible way. Method according to one of the preceding claims, characterized in that the respective current for a setting operation of a working cylinder ( 3 ) carried out a calculation of the control variable (t _m1 ) _comprises a partial calculation, which operates on the basis of the deviations (F) of one or more previous adjusting operations control of the control variable (h) of the working cylinder ( 3 ) realized. Method according to one of the preceding claims, characterized in that on the basis of the deviations (F) of one or more preceding actuating operations, and / or on the basis of the changes of adaptation parameters (p _offset , t _{m1_offset} , h _offset , h _scale ) , a diagnostic or error information is generated. Computer program, characterized in that to carry it out one of the above methods is suitable. Computer program according to claim 39, characterized that it can be stored on a storage medium. Control unit ( 31 ), in particular for controlling one or more hydraulic power cylinders ( 3 ) or Steller ( 30 ), characterized in that it is suitable for carrying out one of the above methods. Control unit ( 31 ), in particular for controlling one or more hydraulic power cylinders ( 3 ) or Steller ( 30 ), characterized in that it consists of one or more parts, at least one of which is in close proximity to a hydraulic actuator ( 30 ), or together with a hydraulic cylinder ( 3 ) or Steller ( 30 ), and / or with a high-pressure accumulator ( 13 ), and / or is combined with other parts of an electro-hydraulic actuating system to form an integrated mechatronic unit (mechatronic module).