DE102016222732A1 - Method and device for performing a position control for a positioner unit - Google Patents

Abstract

The invention relates to a method for regulating the position of an encoder unit (3) in a drive system, comprising the following steps:
Performing a control function of a predetermined desired position (φ _s ) and an actual position of the actuator unit (3), wherein the control is performed on the basis of control parameters and provides a control variable (u) to the positioner unit (3);
Providing at least one correction parameter (â, b) for adapting the control;
Determining the at least one correction parameter (â, b) as a function of an error deviation (e), wherein the error deviation (e) is a deviation between a modeled dynamic quantity of the positioner unit (3) in accordance with a positioner model and a dynamic behavior of the actual position (φ _i ) Indicates encoder unit.

Description

Technical area

The invention relates to position controls for position encoders, in particular for position encoders in an internal combustion engine, such as a camshaft actuator. The invention further relates to measures for adapting position controls for position encoders.

Technical background

Many positioners in drive systems with an internal combustion engine are regulated to a desired position. An example of this is the camshaft actuator, which is mechanically coupled to a camshaft of the internal combustion engine and rotates the camshaft relative to a crankshaft by means of a hydraulic or electrical adjustment system. As a result, the phase position of valve opening times with respect to a crankshaft angle can be adjusted. The camshaft actuator is adjusted by specifying a manipulated variable, which can specify, for example, a control torque in the form of a duty cycle.

Conventional position controls are based on a PID controller to which a positional deviation is applied. From the integral portion of the controller, a hold duty ratio can be calculated which is applied to the system to compensate for quasi-stationary disturbances such as the spring moment of a return spring, frictional torque, leakage, external user disturbances, and the like. Positioners for positioners are currently being parameterized to adjust controller parameters and feedback gains of the PID controller portions to different operating points of the respective engine type, which may be predetermined for a hydraulic camshaft actuator by the temperature of the hydraulic oil, hydraulic pressure, engine speed and the like, for example to guarantee a uniform control quality over the entire operating range. Due to production series variations and aging effects during vehicle operation, significant changes in the control quality may occur due to changes in the physical behavior of the position indicator and the controlled system, which may result from overshooting in the case of setpoint jumps or stationary controller deviations between the actual and setpoint values that the properties of the entire drive system in terms of performance, consumption and emissions change adversely. Consideration of these effects by additional parameters requires time-consuming application, and it is therefore desirable to improve the control performance of positioners for position indicators and to increase their robustness against the effects of series scattering and aging.

Disclosure of the invention

According to the invention, a method for position control of a positioner unit in a drive system according to claim 1 and a device and the drive system according to the independent claims are provided.

Further embodiments are specified in the dependent claims.

• - Durchführen einer Regelung abhängig von einer vorgegebenen Sollposition und einer Istposition der Stellgebereinheit, wobei die Regelung anhand von Regelungsparametern durchgeführt wird und eine Stellgröße an die Stellgebereinheit bereitstellt;
• - Bereitstellen von mindestens einem Korrekturparameter, um die Regelung zu adaptieren;
• - Ermitteln des mindestens einen Korrekturparameters abhängig von einer Fehlerabweichung, wobei die Fehlerabweichung eine Abweichung zwischen einer modellierten dynamischen Größe der Stellgebereinheit gemäß einem Stellgebermodell und einem dynamischen Verhalten der Istposition der Stellgebereinheit angibt.

According to a first aspect, a method for position control of an actuator unit in a drive system is provided, comprising the following steps:

• - Performing a control function of a predetermined target position and an actual position of the actuator unit, wherein the control is performed based on control parameters and provides a control variable to the positioner unit;
• - providing at least one correction parameter to adapt the control;
• Determining the at least one correction parameter as a function of an error deviation, wherein the error deviation indicates a deviation between a modeled dynamic variable of the positioner unit according to a positioner model and a dynamic behavior of the actual position of the positioner unit.

The above method is based on a simplified mathematical modulator model, with which the dynamic behavior of the actuator unit is described. The positioner unit can affect the physical positioner (actuator) or the entire controlled system, including the position indicator. The modulator model forms a first derivative of the adjustment position (adjustment speed) in the event of a change in the manipulated variable, wherein it is shown that a change in the adjustment speed of the positioner unit in the event of a change the manipulated variable takes place with a time delay. The speed and the type of change in the adjustment speed at a change in the manipulated variable depends largely on the operating point of the actuator unit, the operating point by system parameters, as for the example of the camshaft position sensor, by the temperature of the hydraulic oil, the engine speed and the like is determined, and through effects of serial dispersion and aging. Alternative embodiments may provide a modulator model in which the manipulated variable is also mapped to higher derivatives of the position encoder.

The control can be designed for a nominal system behavior, and corresponding deviations of the real behavior of the physical position encoder unit detected by the adaptation device can be determined by the mathematical positioner model one or more correction parameters by which the control is adjusted accordingly.

The correction parameters are determined on the basis of a deviation between the behavior of the positioner unit and the behavior of the positioner model, which is defined with respect to the first or a higher derivative of the position, wherein the correction parameters correspond to model parameters of the positioner model. In other words, the correction parameters correspond to parameters of the positioner model, which make it possible to adapt the positioner model with respect to a dynamic behavior of the positioner unit.

In this way, dynamic properties of the position encoder unit can be taken into account or described in a position regulation by correction parameters. The correction parameters can now be used to adapt components of the control.

By providing the modulator model, which is adapted to a dynamic behavior of the actuator unit, it is possible to adapt various components of the controller, such as a PID controller, a dynamic feedforward control, a disturbance observer and the like to a deviating from a nominal behavior of the positioner unit behavior , Thereby, input disturbances such as a spring moment or a spring force of a return spring, a frictional torque, leaks, disturbance torques of external loads and the like can be compensated, and the steady state accuracy and transient performance can be improved. Through this automatic provision of the correction parameters, the behavior of the position control can be adapted to the actual system behavior, thereby improving the quality of control in case of series dispersion or aging effects. In addition, the provision of the above method makes it possible to avoid postdating of the corresponding position control even if the positioner unit is replaced.

Furthermore, the modulator model can be adapted by the at least one correction parameter and the correction parameter can be determined by minimizing the error deviation.

According to one embodiment, the control can be adapted by adjusting the control parameters as a function of the at least one correction parameter or by adapting a control input as a function of the at least one correction parameter.

It can be provided that the dynamic behavior of the actual position is indicated by the first time derivative or higher time derivatives of the actual position of the positioner unit, wherein the positioner model the manipulated variable to one of the first time derivative or one or more of the higher time derivatives of the actual position corresponding size maps.

Furthermore, a disturbance observer can be provided which adjusts the manipulated variable as a function of an inverse encoder model adapted by the at least one correction parameter in order to compensate for disturbances on the position encoder system by a stopper size.

In particular, the stop size can be determined as the difference between a filtered manipulated variable and the model value of the adapted inverse positioner model acted upon by the actual position.

Furthermore, the control can be acted upon by means of a dynamic precontrol which provides a predicted setpoint position and a pre-control manipulated variable component corresponding to the predicted setpoint position for acting on the manipulated variable, wherein the control is carried out as a function of the predicted setpoint position, the dynamic feedforward control depending on the precontrol manipulated variable component determined by the at least one correction parameter.

list of figures

• 1 eine schematische Darstellung einer Positionsregelung zur Regelung einer Stellgebereinheit in einem Antriebssystem eines Kraftfahrzeugs;
• 2a und 2b eine mögliche Realisierung der Ermittlung der Korrekturparameter für die Regelung mit Hilfe eines Gradientenabstiegverfahrens; und
• 3 ein erweitertes Regelungssystem mit zusätzlichem Störgrößenbeobachter und dynamischer Vorsteuerung.

Embodiments are explained below with reference to the accompanying drawings. Show it:

• 1 a schematic representation of a position control for controlling a positioner unit in a drive system of a motor vehicle;
• 2a and 2 B a possible realization of the determination of the correction parameters for the control by means of a gradient descent method; and
• 3 an extended control system with additional disturbance observer and dynamic precontrol.

Description of embodiments

In 1 is a simplified regulatory system 1 with a control unit 2 for controlling a positioner unit 3 intended. The control system 1 is for position control of the encoder unit 3 designed, for example, can be an adjuster for continuous phase adjustment of the camshaft in a drive system of a motor vehicle. Because the positioner unit 3 has different operating points as a function of operating states of the drive system and the operating points are also subject to series variations and aging effects, the position control is adaptively designed. This is an adaptation block 4 provided, which determines one or more correction parameters â, b, the control unit 2 be fed to adjust the control according to a changed behavior of the positioner system.

vorgesehen sein, wobei a_nom, b_nom vorgegebenen Streckenparametern für ein nominelles Systemverhalten des Stellgebersystems entsprechen.The control unit 2 has a simple control element 21 on, which may be formed for example as a PD controller. The PD controller may have a transfer function k _p + k _d · s and is for a nominal system behavior of the position encoder system 3 designed. To adapt the control to a changed behavior of the positioner system 3 not in the regulator 21 to intervene on the input side (ie as a control input) is a compensator 22 provided that a control deviation, in a differential element 23 is determined by subtraction between a predetermined desired position φ _s and an actual position φ _i , scaled accordingly. The transfer function of the compensator 22 can, for example, as a first-order filter with $$\frac{\widehat{a}s+\widehat{b}}{a_{nOm}s+b_{nOm}}$$

be provided, where a _nom , b _nom predetermined route parameters for a nominal system behavior of the positioner system correspond.

The control unit 2 provides a manipulated variable u, which usually indicates a control torque in the form of a duty cycle. The duty cycle can be defined by values between 0 and 1 or between -1 and 1. The duty cycle becomes the positioner system 3 predetermined, which responds according to the system behavior and by accelerating or decelerating a certain position φ starts.

The adaptation block 4 has a first filter 41 and an encoder model block 42 on. The first filter 41 serves to form the first derivative of the actual position φ _{i of} the position encoder system 3 and may, for example, have the following transfer function: $$\frac{s}{{\tau }_{2}s+1}\text{or discreet}\frac{{\phi }_k-{\phi }_{k-1}}{\Delta T}$$

wobei τ_d einer Konstante zur Berücksichtigung einer Totzeit entspricht.The encoder model block 42 includes a modulator model that corresponds to a simplified mathematical model describing dynamic system behavior. The modulator model forms a current manipulated variable, ie, in this embodiment, an indication of a duty cycle to a position change quantity, ie, for example, a first derivative of the manipulated position. The positioner function can be selected to reflect the actual behavior of the positioner unit 3 replicates. For example, the positioner model can have the following transfer function: $$\frac{1}{\widehat{a}s+\widehat{b}}{e}^{-{\tau }_{d}s}$$

where τ _d corresponds to a constant for taking into account a dead time.

In a second differential element 43 is an error deviation e between the dynamic behavior of the positioner system 3 , ie the first derivative of the positioning position at the output of the first filter 41 and the output of the mathematical positioner model. The error deviation e becomes a correction parameter block 44 which determines from the error deviation e the correction parameters â, b as corrected path parameters.

The modulator model is adjusted according to the correction parameters â, b, which are determined in the correction parameter block so that the error deviation e is minimized.

Der zweite Filter 51 kann beispielsweise folgende Übertragungsfunktion aufweisen: $$\frac{s}{{\left(\widehat{a}s+\widehat{b}\right)}^2}$$

In the correction parameter block 44 can the in 2a and 2 B are performed, which can determine the correction parameters by means of a Gradientenabstiegverfahrens. For this purpose, the error deviation e with an adaptation factor k _a and an output of a second filter 51 multiplied and the result correspondingly in an integration block 52 integrated. The output of the integration block corresponds to the first correction parameter â, which is in the second filter 51 is fed back. Inputs of the second filter 51 are still the second correction parameter b and the derivative $\dot{\widehat{\phi }}\text{,}$

The second filter 51 can, for example, have the following transfer function: $$\frac{s}{{\left(\widehat{a}s+\widehat{b}\right)}^2}$$

The output of the model is filtered and multiplied by the error from model and "measured" velocity (gradient descent method).

zugeführt werden. Weiterhin wird die Fehlerabweichung e mit dem Ausgang des dritten Filters 61 und einem entsprechenden Adaptionsfaktor k_b multipliziert und das Ergebnis einem zweiten Integrator 62 zugeführt. Der Ausgang des zweiten Integrators 62 entspricht dem zweiten Korrekturparameter b̂ und wird an den dritten Filter 61 zurückgeführt. Eine mögliche Übertragungsfunktion des dritten Filters kann sein: $$\frac{1}{{\left(\widehat{a}s+\widehat{b}\right)}^2}$$

Analogously, the second correction parameter b can be used correspondingly in the calculation scheme of FIG 2 B be determined. 2 B shows a third filter 61 , the first correction parameter â and the size $\dot{\widehat{\phi }}$

be supplied. Further, the error deviation e becomes the output of the third filter 61 and a corresponding adaptation factor k _b multiplied and the result a second integrator 62 fed. The output of the second integrator 62 corresponds to the second correction parameter b and is applied to the third filter 61 recycled. A possible transfer function of the third filter may be: $$\frac{1}{{\left(\widehat{a}s+\widehat{b}\right)}^2}$$

The calculation rules of 2a and 2 B are merely exemplary ways to track the correction parameters simultaneously and in real time according to the error deviation. Other options for determining the correction parameters from the error deviation can also be used here.

In addition to the present method by model matching to track the model parameters, a recursive parameter estimator (least squares method) or a Kalman filter could also be used. In the recursive parameter estimator, both the input (e.g., duty cycle) and the measurement (e.g., camshaft position) are used to estimate the model parameters. Recursive means changing the old estimate for the model parameter with a correction term that is the difference between the new measure and a prediction based on the old measure. Kalman filters are normally used to estimate non-measurable state quantities. However, parameters can also be defined as additional state variables. The new model parameter is also calculated from the old parameter and a correction term. The correction term again depends on the measured value and a predicted value based on the old measured value.

In 3 Another embodiment is shown in which the control system 1 of the 1 by a dynamic precontrol 7 and a disturbance observer 8th is extended. The dynamic feedforward control 7 is _designed to estimate a _precontrol manipulated variable _component u _fwd on the basis of the mathematical positioner model from the time profile of the predefined setpoint position φ _s and to explicitly take into account the manipulated variable _limitations . The reference position serves as output information for a trajectory calculated back by the positioner model, which is then to be realized by the precontrol. The trajectory can also correspond to the course of the desired position. This trajectory may include a manipulated variable limitation and possibly also a temporal filtering of the setpoint position φ _s .

This will be the actual control of the control unit 2 is relieved with the return of the predicted position φ _pred instead of the desired position φ _s . The _precontrol manipulated variable _component u _fwd becomes the manipulated variable u _R at the output of the control unit 2 added.

The dynamic feedforward control 7 can, for example, have the following transfer function: $$\frac{\widehat{a}{\mathrm{<figure-callout\; id="2"\; label="s"\; filenames="DE102016222732A1\_0012.png"\; state="\{\{state\}\}">s</figure-callout>}}^2+\widehat{b}s}{{\left({\tau }_{fwd}s+1\right)}^2}$$

This represents an inverse model for transferring the setpoint position φ _s to a manipulated variable u. Furthermore, the dynamic precontrol comprises a pre-filtering of the setpoint position φ _s with: $$\frac{1}{{\left({\tau }_{fwd}s+1\right)}^2}$$

The dynamic feedforward control 7 is thus adapted adaptively based on the correction parameters â, b.

It is also a disturbance observer 8th provided, a model block 81 for calculating an inverse encoder model and a third filter 82 having. The third filter 82 becomes the manipulated variable u of the positioner unit 3 or a size dependent thereon. The manipulated variable u of the encoder unit 3 must with a filter 82 be filtered, which has the same time constant as the filter 81 , Otherwise, the error (input fault) can not be determined correctly from the actual and the back-calculated control value. The inverse encoder model can have the following transfer function: $$\frac{\widehat{a}{\mathrm{<figure-callout\; id="2"\; label="s"\; filenames="DE102016222732A1\_0012.png"\; state="\{\{state\}\}">s</figure-callout>}}^2+\widehat{b}s}{{\left({\tau }_{2}s+1\right)}^2}$$

The third filter 82 can have the following transfer function: $$\frac{1}{{\left({\tau }_{2}s+1\right)}^2}{e}^{-{\tau }_{d}s}$$

The disturbance observer 8th is used to compensate for position deviations that may occur due to an input fault in the positioner system. If the position changes due to a fault, eg a spring moment of a return spring, a friction torque, leakage of the hydraulics or disturbance torques from external consumers, the disturbance can be caused by the disturbance observer 8th be compensated. The disturbance observer 8th can calculate the disturbance from the current manipulated variable u and the actual position φ _i . For this purpose, the actual position φ _{i is} converted into an observer manipulated variable value u _b via the inverse encoder model and compared with a filtered manipulated variable û. It is in a third difference element 83 formed a difference. The difference is added back to the manipulated variable u as the stop size u _H. The inverse encoder model also takes account of the correction parameters â, b, so that the inverse encoder model always to this current dynamic behavior of the positioner unit 3 adapted.

Because both the dynamic feedforward control 7 , the regulatory block 3 as well as the disturbance observer 8th Based on the mathematical positioner model, the corresponding model parameters must be tracked by adaptation with the aid of the correction parameters â, b, depending on the operating point or on series scattering and aging effects.

The proposed control system can take into account the dynamic properties of the positioner unit in the control system, in particular by the dynamic pilot control, whereby the control performance is improved. In particular, input disturbances can be compensated by the spring moment of the restoring spring, a frictional torque, leaks in the hydraulics, disturbance torques of the external consumers and the like, and thus the stationary accuracy and the transient behavior can be improved.

Adaptation and disturbance observers may be active at all times or only temporarily, and the last values of the stop size or the adapted model parameters are retained until the adaptation and the disturbance observer are reactivated. Alternatively, the adaptation may be performed depending on the magnitude of the error e between model size and measured velocity.

Claims (11)

Method for regulating the position of an encoder unit (3) in a drive system, comprising the following steps: - performing a control as a function of a predefined setpoint position (φ _s ) and an actual position of the positioner unit (3), wherein the control is carried out using control parameters and a manipulated variable ( u) to the positioner unit (3) provides; - providing at least one correction parameter (â, b) to adapt the control; Determining the at least one correction parameter (â, b) as a function of an error deviation (e), wherein the error deviation (e) is a deviation between a modeled dynamic variable of the positioner unit (3) in accordance with a positioner model and a dynamic behavior of the actual position (φ _i ) indicates the encoder unit. Method according to Claim 1 in which the modulator model is adapted by the at least one correction parameter (â, b) and the correction parameter (â, b) is determined by minimizing the error deviation (e). Method according to Claim 1 or 2 wherein the control is adapted by adjusting the control parameters depending on the at least one correction parameter (â, b) or by adjusting a control input depending on the at least one correction parameter (â, b). Method according to one of Claims 1 to 3 , wherein the dynamic behavior of the actual position (φ _i ) by the first time derivative or higher time derivatives of the actual position (φ _i ) of the positioner unit (3) is specified, the modulator model the manipulated variable (u) on one of the first time derivative or a the higher temporal derivative of the actual position (φ _i ) corresponding size maps. Method according to one of Claims 1 to 4 in which a disturbance-size observation is carried out, which adapts the manipulated variable as a function of an inverse encoder model adapted by the at least one correction parameter (â, b) in order to compensate for disturbances on the positioner system (3) by a stopper size. Method according to Claim 5 , wherein the stop size is determined as the difference between a filtered manipulated variable and the model value of the adapted by the actual position adapted inverse encoder model. Method according to one of Claims 1 to 6 , wherein the control of a dynamic precontrol (7) is applied, which provides a predicted set position (φ _pred ) and the predicted set position (φ _pred ) corresponding _{Vorsteuerungs} manipulated variable _component (u _fwd ) for acting on the manipulated variable (u), wherein the Regulation is performed depending on the predicted setpoint position (φ _pred ), wherein the dynamic _feedforward control (7) determines the pilot control manipulated variable _component (u _fwd ) as a function of the at least one correction parameter (â, b). Position control of an encoder unit (3) in a drive system, comprising: a control unit, which is designed to perform a control depending on a predetermined desired position (φ _s ) and an actual position of the actuator unit (3), wherein the control is performed on the basis of control parameters and provides a manipulated variable (u) to the positioner unit (3) ; - An adaptation block, which is designed to provide at least one correction parameter (â, b) to adapt the control; Determining the at least one correction parameter (â, b) as a function of an error deviation (e), wherein the error deviation (e) is a deviation between a modeled dynamic variable of the positioner unit (3) in accordance with a positioner model and a dynamic behavior of the actual position (φ _i ) indicates the encoder unit. Position control after Claim 8 in which a disturbance observer (8) is provided which is designed to adapt the manipulated variable as a function of an inverse encoder model adapted by the at least one correction parameter (â, b) in order to compensate for disturbances on the positioner system (3) by a stopper size. Computer program adapted to perform all steps of a procedure according to one of Claims 1 to 7 perform. A machine-readable storage medium having stored thereon a computer program according to claim 10.

Images