DE102013209070A1 - Method and device for detecting a start of movement of electromechanical actuators - Google Patents

Abstract

The invention relates to a method for detecting the start of movement of an electromechanical actuator which is driven by means of at least one magnetic coil, the actuator being part of a hydraulic component, such as a magnetic pump or a solenoid valve, the point in time when the actuator begins to move within a tightening phase by evaluating a Current Idv (t), which flows through the magnet coil, the calculation of a temporal inductance curve L (t) of the magnet coil is determined from the current curve and the evaluation of an inductance change. The invention further relates to a device, in particular a control and evaluation unit, for carrying out the method. According to the invention, it is provided that, after the start of the tightening phase, a secant is calculated from the inductance curve L (t) before the expected time of the start of movement and the time of the start of movement is defined when the inductance of the inductance curve L (t) defines a certain difference from the calculated inductance values on the secant exceeds. The method and the device enable an increase in the detection accuracy of a start of the stroke or movement of the actuator or the armature of a diaphragm pump or a metering valve. In a preferred use of the method, dosing systems for dosing reducing agents in an SCR system can achieve an improved diagnosis with regard to the start of an injection period (GDP), which among other things Incorrect dosing can be prevented.

Description

State of the art

The invention relates to a method for detecting a start of movement of an electromechanical actuator, which is driven by at least one magnetic coil, wherein the actuator is part of a hydraulic component, such as a magnetic pump or a solenoid valve, wherein the time of starting movement of the actuator within a suit phase by means of evaluation of a Current I _{dv (t)} , which flows through the magnetic coil, the calculation of a time course of inductance L _{(t) of} the magnetic coil is determined from the current waveform and the evaluation of an inductance change.

The invention further relates to a device, in particular a control and evaluation unit, for carrying out the method according to the invention.

Hydraulic, non-pressure compensating actuators are driven by electromagnets in various applications. Typical examples are solenoid pumps and solenoid valves. These actuators typically need to be moved against a spring force and a hydraulic force. In this case, a magnetic coil of the electromagnet is energized. The magnetic force increases and at the moment when there is an equilibrium of forces between the opening magnetic force and the closing pressure forces, the actuator starts to move.

For example, such pumps and / or metering valves, which are driven by means of an electromagnet, are used in nitrogen oxide reduction units for exhaust gas purification.

In connection with future legal requirements regarding the nitrogen oxide emission of motor vehicles, a corresponding exhaust aftertreatment is required. Selective catalyst reduction (SCR) can be used to reduce the NO _x emissions (denitrification) of internal combustion engines, especially diesel engines, with temporally predominantly lean, ie oxygen-rich combustion. In this case, a defined amount of a selectively acting reducing agent is added to the exhaust gas. This can be, for example, in the form of ammonia, which is added directly in gaseous form, or else from a precursor substance in the form of urea-water solution (HWL). Such HWL-SCR systems have been used for the first time in the commercial vehicle segment.

Corresponding metering devices are for example from the DE 196 07 073 A1 known. A urea-water solution (HWL), eg under the name AdBlue as 32 , 5% solution known, is promoted by a line from a tank to a metering valve and metered into an exhaust tract of an internal combustion engine, in particular in the lean-burn engines or diesel engines to reduce the nitrogen oxide concentration of such engines. The mechanism of action has been adequately described in the specialist literature (cf., for example WEISSWELLER in CIT (72), pages 441-449, 2000 ).

In current metering systems, as known, for example, under the name DENOXTRONIC 5.1, a membrane pump sucks the AdBlue solution from the reagent tank in a delivery module and compresses it to the system pressure of 4.5 to 8.5 bar required for atomization. The dosing module doses the amount of AdBlue required for the NO _x reduction by atomization into the exhaust gas flow upstream of the SCR catalytic converter. Control of metering and heating strategy as well as for on-board diagnostics (OBD) can be performed by a higher-level motor control or by a dosing control unit. With the processing of the current engine operating data and all required sensor data, the amount of reducing agent is matched exactly to the engine operating point and to the catalyst-specific properties for maximum nitrogen oxide reduction.

The present invention particularly relates to a monitoring function of the metering valve in the SCR system. This is an actuator diagnosis, which constantly monitors whether the metering valve reacts to an activation and accordingly the reducing agent is injected into the exhaust gas line. The opening and closing of the valve are controlled by the current. The current flowing through the coil is measured by the dosing control unit and is ready for diagnosis. The start of the injection period (BIP - Start of Injection Period) is detected on the basis of the current profile.

According to the prior art, the metering valve is monitored by the course of the current in the tightening phase. The monitoring is based on the second derivative of the current flow of the metering valve, which is received by the controller at each tightening phase. The current kink, which is detected as an armature movement, is detected by the difference between the local maximum and the local minimum of the derivative of the current profile, if this difference (as so-called quality number) exceeds a threshold.

The applicant's application, which has not yet been published, with the internal file number R.346122, discloses a method and a device for detecting a release time of an electromechanical actuator described. In this case, after a control of a solenoid driving the actuator, the release time of the actuator is determined by means of evaluation of a temporal actual current flow through the magnetic coil. It is provided that a modeled current waveform is determined by the solenoid under the assumption of a non-moving actuator, that the actual current waveform during the control of the solenoid is determined and compared with the modeled current waveform and that upon reaching a predetermined deviation of the two current histories of each other, the release time is recognized. Method and apparatus are intended to improve the accuracy of the determination of the release time of the actuator, for example an anchor of a diaphragm pump, and thus to improve the accuracy of a pressure determination in a system for metering a urea-water solution for reducing nitrogen oxide in the exhaust gas of an internal combustion engine.

The not yet published application of the applicant with the internal file number R.346123 also describes a method and a control and evaluation unit for detecting a movement start of an electromechanical actuator, which is driven by at least one solenoid, wherein the actuator part of a hydraulic component, such as a magnetic pump or a solenoid valve is, wherein a movement start of the actuator by means of evaluation of a current flowing through the magnetic coil, and its temporal derivatives determined and using tabular values or curves as a function of further characteristics with the determined start of movement, a pressure is determined. It is provided that for determining the start of movement, a relative inductance determined from a temporal Induktivitätsverlauf and a time course of the relative inductance is evaluated. Method and apparatus to allow an increase in the detection accuracy of a stroke or movement start of the actuator or the armature of a diaphragm pump in a preferred use of the method and thus improve the accuracy of a pressure detection.

From the DE 10 2011 002 630 A1 For example, a method for detecting a change of the activation state of an electromagnetic actuator and a circuit arrangement with an electromagnetic actuator is known. The electromagnetic actuator has an electromagnet with an inductance and a mechanically controlled by the solenoid armature. The method includes evaluating an inductance value of the inductance over time.

A not yet published application of the applicant with the title "Method and device for detecting an anchor stop of an electromechanical actuator" (internal reference R.346347) relates to a method for detecting an armature stop of an electromechanical actuator after driving a solenoid driving the actuator, wherein the armature stop of the actuator is determined by evaluating a temporal actual current waveform through the magnetic coil. In this case, it is provided that an inductance profile is determined from the actual current profile and that the armature stop is detected when a predetermined inductance threshold is reached.

The recognition results with the current methods, as described above, show in practice, however, that the start of movement of the actuator (Begin Motion Point - BMP) and thus the exact time of the beginning of the injection phase (BIP) can not always be determined exactly. Reasons for this are u.a. Fluctuations in the supply voltage, mechanical tolerances on the metering valve, e.g. different spring constants, and PWM clock switching in 24V electrical systems.

It is therefore an object of the invention to provide an optimized method for monitoring the functionality of the metering valve, with which a reliable and robust detection of a start of movement (BMP) of the actuator or the beginning of the injection phase (BIP) is detected and, above all, blockage of the Valve can be detected in the suit phase.

It is a further object of the invention to provide a device, in particular a control and evaluation unit for carrying out the method.

DISCLOSURE OF THE INVENTION The object of the invention relating to the method is achieved by calculating a secant from the inductance curve L _(t) before the expected start of the movement and by defining the time of the start of movement when the inductance of the inductance curve L is defined _(t) exceeds a certain difference from the calculated inductance values on the secant. Compared to the aforementioned prior art, the proposed method makes it possible to increase the detection accuracy of a start of stroke or movement of the actuator or the armature of a diaphragm pump or a metering valve. Interference, such as tolerances in the spring constant of return springs, temperature influences and a different system pressure in such hydraulic systems can be significantly reduced, whereby a reliable functional diagnosis of such hydraulic components, such as magnetic pumps or solenoid valves, can be realized. The method can also be implemented quite easily by means of appropriate software and hardware.

The detection method envisages that the determination of the secant comprises a multi-stage algorithm in which first the current profile is digitized within an evaluation period into individual current values I _{dv (n)} and from the individual current values I _{dv (n)} and their change ΔI _{dv (n )} the inductance curve L _{(t) is} calculated, on the basis of the calculated inductance values a plurality of inductance gradients G _{(m) are} formed, from the inductance gradients G _(m) a middle gradient G is formed and with this middle gradient G the secant and the inductance values on the secant are determined become. This can be accomplished by comparatively simple mathematical operation and can be implemented by means of a software extension into existing diagnostic methods. In addition, a high degree of robustness against interference can be realized.

A movement diagnosis of the actuator that is robust with respect to interference influences and system tolerances can be achieved, in particular, if, as provided by a preferred method variant, the time of the start of movement is defined for the time at which the difference of the inductance values between the inductance curve and inductance values on the secant ΔL _{(1 )} ... ΔL _{(N) reach} a maximum maximum (ΔL) and / or exceed an applicable threshold ΔL _max . Compared to a pure evaluation of the current flow of the coil current can hereby be achieved a reliable motion diagnosis for the actuator.

A variant of the method provides that the evaluation period for determining the secant within the tightening phase before the expected start of movement of the actuator, defined by a start time t _b and an end time t _c , and the time t _d until the end of the evaluation of the inductance difference values ΔL ₍₁₎ ... ΔL _{(N) can be} adapted depending on the system. Thus, the evaluation method can be adapted to different systems with different time behavior or reaction behavior. In addition, by determining the evaluation ranges for the secant calculation and / or the evaluation period of the inductance difference values ΔL ₍₁₎ ... ΔL _(N), the accuracy or quality of the method can be determined as a function of the available computing or memory resources be determined. Due to this flexibility, the evaluation process can also be optimized under different operating conditions of the system.

With regard to a cost-effective and efficient implementation of the evaluation, it has proved to be advantageous if the current through the solenoid I _dv detected by means of an analog-to-digital converter with a sampling frequency f _a and the individual current _values I _{dv (n)} stored in a microcontroller and discrete inductance L _(n) and the Induktivitätsdifferenzwerte .DELTA.L ₍₁₎ ... .DELTA.L _(N) are calculated on the inductance of the secant, wherein the number of values from the analysis period to determine the secant and out of the analysis period of the Induktivitätsverlaufs and is determined from the sampling frequency f _a .

If, as is provided by a preferred _{variant of the method,} the supply voltage U _{batt is also} detected and stored by means of an analog-to-digital converter with a sampling frequency f _a to calculate the inductance values L _(n) , errors in the inductance calculation due to short-term voltage fluctuations of the supply voltage, here the on-board battery, or be avoided as a result of a clocking of the supply voltage.

If the sampling frequency f _{a is} adjusted depending on the system and its time constants, the computational effort can be reduced to a necessary level, which is particularly advantageous with regard to the memory resources and computation hardware. For slow systems with a larger time constant, a small sampling frequency f _a is sufficient. For systems with a small time constant, on the other hand, the sampling frequency f _a must be increased in order to take account of subtle changes in the calculation and evaluation. The stored current values must in any case cover the expected time range in which the movement start of the actuator is expected.

A variant of the method provides that the current values I _{dv (t)} are filtered before their detection and digitization by means of a low-pass filter. The filtering can be done analogously in the current determination within the hardware, for example by means of an RC element, or can be realized software-based by forming a moving average from a certain number of individual values. As a result, briefly occurring glitches can be eliminated, which would otherwise lead to misinterpretation.

An advantageous use of the method according to the invention with its variants described above provides the use for monitoring a metering valve in a metering unit as part of a metering system, with a urea-water solution can be introduced as a reducing agent in the exhaust passage of an internal combustion engine, which in the flow direction of the exhaust gas after the introduction of the urea-water solution comprises an SCR catalyst, wherein the actuator is designed as an armature of the metering valve. In particular, lean-burn engines or diesel engines can thus be brought about a nitrogen oxide reduction, so that future legal requirements for pollutant emissions can be met. Here, the monitoring of the functionality of the metering valve is of great importance. In particular, the method can be used to reliably detect the beginning of the injection period (BIP). A blocking of the metering valve can thus be detected within the tightening phase and thus an undesirable addition of reducing agents can be prevented. In addition, it can thus be detected and displayed as part of the on-board diagnostic (OBD) inoperability of the metering valve.

The object of the invention relating to the device is achieved in that the control and evaluation unit comprises at least one analog-to-digital converter, calculation units for inductance calculation, gradient calculation, mean calculation, secant calculation, difference calculation and maximum value formation, comparators within checking units and memory units for carrying out the previously described Having method with its variants. The functionality of the method can be implemented in this case at least partially software-based. In this case, the implementation is cost-effective by an appropriate software extension in the control and evaluation, or if this is designed as part of a higher-level engine control, in the parent engine control. In a metering system, as provided for by a preferred use of the method, the control and evaluation unit can also be implemented in a corresponding metering control unit of the metering system.

In a preferred control and evaluation unit, the analog-to-digital converters are designed as FADC (Fast Analog-Digital Converter), so that with high sampling frequency f _a, the individual current values I _{dv (n)} and possibly the individual supply voltage _values U _{batt (n) can be} digitized and stored for further calculation in one or more memories (so-called buffer) of a microcontroller. The processor load can be reduced to a minimum.

The invention will be explained in more detail below with reference to an embodiment shown in FIGS. Show it:

1 a schematic representation of a metering system as an example of a technical environment in which the method according to the invention can be used,

2 an equivalent circuit diagram for the control of a metering valve,

3 and 4 each a history diagram for the current flow and the switching states of the drive switch,

5 and 6 further course diagrams for the current flow, the course of the time derivative of the current and an inductance curve and

7 and 8th each a flowchart for two different evaluation algorithms.

1 shows an application example of the method according to the invention in a simplified schematic overview of a dosing 1 using a urea-water solution (HWL), also known as AdBlue, from a storage container 10 in an exhaust duct 40 an internal combustion engine can be injected. The injection takes place in the flow direction of the exhaust gas flow 41 in front of an SCR catalyst 50 , Such an arrangement is known, for example, under the name DENOX-TRONIC 5.1 as a product of the Applicant.

In a conveyor line, the HWL is a filter 11 by means of a diaphragm pump 20 to a dosing unit 30 promoted, with a constant amount per stroke in a pressure line 25 between diaphragm pump 20 and dosing unit 30 is encouraged. The dosing unit 30 doses the HWL into the exhaust duct as needed 40 , In the flow direction of the HWL in front of and behind the diaphragm pump 20 there is an inlet valve 22 and a pressure valve 23 , To avoid pressure surges is between diaphragm pump 20 and the pressure line 25 a pressure shock absorber 24 arranged. An ice pressure damper 12 Prevents damage to the device on the input side if the HWL should freeze in extremely cold temperatures.

In a return line arranged in parallel to the conveyor line can by means of a return pump 60 the conveyor system be emptied, with the HWL back into the reservoir 11 can be pumped. In the flow direction of the HWL is on the input side of the feed pump another inlet valve 61 and an outlet valve on the output side 62 , In addition, in this purge string another ice pressure damper 12 intended.

For example, the system is designed for a nominal pressure of 6.5 bar. This pressure is via the diaphragm pump 20 constructed and must be monitored. A separate pressure sensor is not provided in newer systems for cost reasons. As the drive of the diaphragm pump 20 by means of a magnetic coil 21 can be done based on the current through the solenoid coil 21 a print model be determined and thus a pressure indexing possible, as described for example in the above-mentioned document R.346123.

For controlling the diaphragm pump 20 , the return pump 60 and the dosing unit 30 serves a control and evaluation unit 70 which may also be designed as part of a dosing control unit or higher-level engine control unit. The functionality can be software-based. In the control and evaluation unit 70 can also the temporal current consumption of the solenoid 21 the diaphragm pump 20 and a metering valve 103 (please refer 2 ) be evaluated.

The evaluation method according to the invention relates in particular to the metering valve 103 the dosing unit 30 and provides that on the basis of the time course of the calculated inductance of the solenoid of the metering valve 103 an armature or needle movement of the metering valve 103 is detected and thus an opening of the valve is detected, at which moment the injection phase begins (BIP).

In principle, the evaluation method described below can also be transferred to other systems where reliable detection of an armature movement of a lifting magnet has to be detected.

2 shows in an equivalent circuit diagram 100 greatly simplifies the circuit for controlling the metering valve 103 the dosing unit 30 out 1 , By switching on and off the so-called highside switch 101 (HS) and lowside switch 102 (LS), the current I _dv through the solenoid of the metering valve 103 , which is shown as an inductance, to be controlled, the circuit _being supplied by an on-board power supply U _batt . To avoid induced voltage peaks is a freewheeling diode 104 (D) provided. High side and low side switch 101 . 102 can in the dosing unit 30 be integrated or as an integral part of the control and evaluation 70 be executed.

In 3 shows an example in a history diagram 110 a current course 115 for the current value 111 of the solenoid current as well as the low-side and high-side switching states 112 . 113 depending on the time 114 during a drive. Thereby one differentiates the suit phase 116 Wherein the time t _1, both the high-side switch 101 as well as the lowside switch 102 be turned on. In the so-called holding phase 117 , from the time t ₂ , the lowside switch 102 with closed highside switch 101 clocked. In the closing phase 118 will open after the highside and lowside switches 101 . 102 between the time t ₃ and t _{4 of} the highside switch 101 closed again for a short time. At time t ₀ can be based on a discontinuity in the current curve 115 be closed on a time for a start of movement of the anchor. The other versions therefore relate only to the suit phase 116 of the history diagram 110 ,

4 shows, as already in 3 shown, the current flow 115 for the current value 111 of the solenoid current as well as the low-side and high-side switching states 112 . 113 depending on the time 114 especially during the suit phase 116 ,

At time t ₁ , the highside switch turns 101 and the lowside switch 102 one. In the closed circuit thus applies: U _Batt = I _dv * R + L _(t) * dI _dv / dt (1)

U _{Batt is} the supply voltage or on-board voltage in the vehicle, I _dv the coil current through the solenoid of the metering valve 103 and R the ohmic resistance of the circuit (in the equivalent circuit diagram 100 in 2 not shown). The battery voltage U _Batt , the coil current I _dv and the resistor R can in the control and evaluation 70 be determined. L _(t) represents the inductance of the magnetic coil, which can change over time as a result of the armature movement.

The current flows at the beginning of the tightening phase 116 (Time t ₁ ) from the battery through the coil and increases in accordance with a PT1 behavior with a time constant τ τ = L / R (2) where R is the resistance and L is the inductance of the circuit. The inductance of the circuit mainly consists of the solenoid of the metering valve 103 , Lines and other components can be neglected in this regard.

During this time, the magnetic field in the coil builds up simultaneously. At time t _mb the current reaches the first local maximum and at time t _mc the first local minimum. The maximum and minimum are next to each other and form a kind of "Stromknick". Approximately before the time t _mb , the magnetic force exceeds the holding force of the armature due to counter-pressure of the closing spring and friction. The anchor starts to move. The movement of the armature induces a voltage against the applied battery voltage. As a result, the current decreases briefly and then increases again, wherein the current increase then according to the PT1 behavior with a new time constant τ * τ * = L * / R (3) increases, where L * represents the inductance of the coil with the attracted anchor. Because of the changed air gap, L *> L, where according to (1): L _(t) = (U _{Batt (t)} - _dv I _(t) · mR) / dI _{dv (t)} / dt t ₁ <t <t _mb (4) L * _(t) = (U _{Batt (t)} - I _{dv (t)} · R) / dI _{dv (t)} / dt t _mb <t (5)

Between the time t _mb and t _mc , the current I _{dv increases} more slowly and the value for the time derivative dI _{dv (t)} / dt decreases. The consequence is an increase in the inductance, which can be used as a characteristic for an armature movement. The corresponding course is in 5 exemplified.

It should be noted that this is not a differential inductance, but a static inductance, which corresponds only to the inductance of the circuit in steady state operation. The dynamic behavior, as is usually the case during the tightening phase 116 occurs is not considered here.

However, it is sufficient to consider only the static inductance for detecting the movement of the armature (BIP).

5 show in a further progress diagram 110 during the suit phase 116 the current course 115 for the current value 111 of the coil current I _{dv (t)} , the value for the time derivative of the current value 119 dI _{dv (t)} / dt and its course 121 and the induction value 120 and the induction course 122 L _(t) as a function of time 114 , At the current course 115 It is very difficult to prove an anchor movement. But following the temporal Induktivitätsverlauf 122 , can be a beginning of the movement of the armature or in the case of the metering valve 103 a start of the injection period 123 (BIP) at a peak in the curve of the inductance curve 122 prove.

The evaluation algorithm according to the invention provides that, as shown in FIG 6 in a further progress diagram 110 exemplified, a secant 124 is determined, with the secant 124 after the start of the suit phase 116 (Time t ₁ or t _a ) by inductance values at times t _b and t _c before the beginning of the injection period 123 (BIP) is formed and the exact time of the beginning of the injection period 123 (BIP) at which a maximum inductance deviation is determined 125 max. (ΔL) in the inductance curve 122 from the previously calculated secant 124 occurs.

The evaluation algorithm is carried out in a multi-level process of the following in the following 7 and 8th flowcharts shown 130 is shown.

First, the metering valve current I _{dv (t)} via a shunt resistor by means of a so-called fast analog-digital converter (FADC) is detected with a sampling frequency f _a . Within the times t _b to t _d , the individual current values I _{dv (n) are} in a buffer 131 a microcontroller stored so that a digitized current waveform can be provided. It will be then N = (t _d -t _b ) f _a (6) Values saved. The same applies to the battery _voltage U _batt , the time course of which is likewise digitized in this way.

When sampling the current values by means of the sampling frequency f _a in the FADC, it is also necessary to ensure a sufficiently high sampling frequency f _a in order to take into account even subtle changes in the calculation and evaluation. The in the buffer 131 stored current values must in any case cover the expected GDP range.

It proves to be advantageous if the current values I _{dv (t) are} previously filtered by a low-pass filter. Before the inductance L is calculated.

This is then carried out in a second step within the inductance calculation unit 132 the determination of the corresponding discrete inductance values L _(n) L _(n) = (U _{Batt (n)} - I _{dv (n)} · R) / _dv .DELTA.I _(n) x f _a (7) where U _{Batt (n)} and I _{dv (n) are} the discrete ones in the buffer 131 stored battery voltage values and current values are. ΔI _{dv (n)} represents the difference between the current values and f _{a represents} the sampling frequency.

The next step will be in a gradient calculation unit 133 Gradient values between the times t _b and t _c determined. To do this, analogous to (6) M = (t _c - t _b ) · f _a (8) Single gradient values determined. In order to use each value in the calculation, two values each for the gradient are separated by k = ceil (M / 2) (9) calculated. It results from it P = floor (M / 2) (10) Value pairs with the following single gradient G _(m) G _(m) = (L _{(m + k)} - L _(m) ) / k, where m = 1 ... P (11)

L _{(m + k)} and L _(m) are temporally adjacent inductance values. Ceil (English ceiling function) and floor (English: floor function) are special rounding functions from mathematics, where ceil is the rounding to the next whole number and floor for the rounding to the next lower integer or corresponds to the integer part of the number.

A mean gradient G is then in a mean value calculation unit 134 is calculated by summing all the gradients G _(m ) calculated according to (11) and dividing the sum by the number k.

With the mean gradient G and the inductance value L _(k) becomes a secant 124 (S) formed. This is done in a secant calculation unit 135 , The values on the secant 124 can after S _(n) = L _(k) -G · (n-k) (12) be calculated. Wherein the value of L _(k) of the inductance-calculating unit 132 is made available.

The following is in a difference calculation unit 136 the difference between the value on the secant and the inductance value from the inductance calculation unit 132 educated.

From the difference values ΔL ₍₁₎ ... ΔL _(N) is in a process variant (see 7 ) in a maximum value calculation unit 137 the maximum difference max (ΔL) is determined and by means of a checking unit 138 compared with a threshold ΔL _max . If the maximum difference is max (ΔL) ≥ the threshold ΔL _max , the result "Anchor movement detected" 141 (GDP). If this is not the case, the result is "no armature movement" 142 output.

Usually the difference between secant 124 and inductance course 122 performed from the time t _b . Alternatively, the evaluation algorithm can be varied by setting the starting time of the calculation to the time t _c , which reduces the calculation steps. The calculation of the inductance values on the secant 124 and their comparison with the Induktivitätsverlauf 122 ends at time t _d .

In the in the 7 The methods shown are all values in the buffer 131 used for Induktivitätsdifferenz calculation. Although it requires more computation steps and thus more resources of the microcontroller, but has the advantage that with it an identification of the position of the maximum difference max (ΔL) can be determined. This could be used with appropriate plausibility checks.

The in the 8th illustrated method variant provides that after the secant determination in the secant calculation unit 135 the differences in the difference calculation unit 136 calculated one by one and using the verification unit 138 be compared with the threshold ΔL _max . If the difference ΔL _(j) ≥ the threshold ΔL _max , the result "armature movement detected" 141 (GDP). If not, it will be in a counter 139 the index around 1 elevated. J> n _d , the highest counter value at time t _d , which is in the checking unit 140 is queried, as a result, "no anchor motion" 142 output. If j is still ≤ n _d , the next difference value to the corresponding secant value is calculated and compared with the threshold ΔL _max . This process variant is designed a little resource-saving. The calculation of the secant 124 can do this with a default value 143 be initiated.

The time _tb , t _c , t _d and the threshold ΔL _max are application values and can be adapted depending on the system.

QUOTES INCLUDE IN THE DESCRIPTION

This list of the documents listed by the applicant has been generated automatically and is included solely for the better information of the reader. The list is not part of the German patent or utility model application. The DPMA assumes no liability for any errors or omissions.

Cited patent literature

• DE 19607073 A1 [0006]
• DE 102011002630 A1 [0012]

Cited non-patent literature

• WEISSWELLER in CIT (72), pages 441-449, 2000 [0006]

Claims (9)

A method for detecting a start of movement of an electromechanical actuator, which is driven by means of at least one magnetic coil, wherein the actuator is part of a hydraulic component, such as a magnetic pump or a solenoid valve, wherein the time of a movement start of the actuator within a suit phase ( 116 ) by evaluating a current I _{dv (t)} , which flows through the magnetic coil, the calculation of a temporal Induktivitätsverlaufs ( 122 ) L _{(t) of} the magnetic coil from the current ( 115 ) and the evaluation of an inductance change, characterized in that after the start of the tightening phase ( 116 ) from the inductance course ( 122 ) L _(t) before the expected time of the start of movement a secant ( 124 ) and the start of movement is defined when the inductance of the inductance curve ( 122 ₍₎ L _{t) having} a certain difference from the calculated inductance on the secant ( 124 ) exceeds. Method according to claim 1, characterized in that the determination of the secant ( 124 ) comprises a multi-stage algorithm, in which the current curve ( 115 ) are digitized within an evaluation period into individual current values I _{dv (n)} and from the individual current values I _{dv (n)} and their change ΔI _{dv (n)} the inductance curve ( 122 ) L _{(t) is} calculated on the basis of the calculated inductance values ( 120 ) a plurality of inductance gradients G _{(m) are} formed, from the inductance gradients G _(m) a middle gradient G is formed and with this middle gradient G the secant ( 124 ) and the inductance values on the secant ( 124 ). Method according to Claim 1 or 2, characterized in that the start of movement is defined for the time at which the difference in the inductance values between the course of the inductance ( 122 ) and inductance values on the secant ( 124 ) ΔL ₍₁₎ ... ΔL _{(N) reach} a maximum maximum (ΔL) and / or exceed an applicable threshold ΔL _max . Method according to one of claims 1 to 3, characterized in that the evaluation period for determining the secant ( 124 ) within the tightening phase ( 116 ) before the expected start of movement of the actuator, determined by a start time t _b and an end time t _c , and the time t _d until the end of the evaluation of the inductance difference values .DELTA.L ₍₁₎ ... .DELTA.L _(N) are adaptable depending on the system. Method according to one of Claims 1 to 4, characterized in that the current through the magnet coil I _{dv is} detected by means of an analog-to-digital converter with a sampling frequency f _a and the individual current _values I _{dv (n)} are stored in a microcontroller and discrete inductance values L _{( n)} and the inductance difference values ΔL ₍₁₎ ... ΔL _(N) to the inductance values of the secant ( 124 ), whereby the number of individual values from the evaluation period for determining the secant ( 124 ) and from the evaluation period of the inductance curve ( 122 ) as well as from the sampling frequency f _{a is} determined. A method according to claim 5, characterized in that for calculating the inductance _values L _(n) additionally the supply voltage U _batt by means of an analog-digital converter with a sampling frequency f _{a is} detected and stored. A method according to claim 5, characterized in that the sampling frequency f _{a is} adjusted depending on the system and its time constants. A method according to claim 5, characterized in that the current values I _{dv (t)} are filtered before their detection and digitization by means of a low-pass filter. Use of the method according to the aforementioned claims for monitoring a metering valve ( 103 ) in a dosing unit ( 30 ) as part of a dosing system ( 1 ), with a urea-water solution as a reducing agent in the exhaust gas channel ( 40 ) of an internal combustion engine can be introduced, which in the flow direction of the exhaust gas after the introduction of the urea-water solution an SCR catalyst ( 50 ), wherein the actuator as an anchor of the metering valve ( 103 ) is trained. Device for detecting a start of movement of an electromechanical actuator which is driven by means of at least one magnetic coil, wherein the actuator is part of a hydraulic component, such as a magnetic pump or a solenoid valve, wherein for evaluating a movement start of the actuator, a control and evaluation unit ( 70 ) is provided, with which a current I _{dv (t)} , which flows through the magnetic coil, a temporal Induktivitätsverlauf ( 122 ) L _(t) and inductance changes can be determined, characterized in that the control and evaluation unit ( 70 ) at least one analog-to-digital converter, calculation units for inductance calculation, gradient calculation, mean value calculation, secant calculation, difference calculation and maximum value formation ( 132 . 133 . 134 . 135 . 136 . 137 ), Comparators within review units ( 138 . 140 ) and memory units for carrying out the method according to claims 1 to 9. Apparatus according to claim 10, characterized in that the one or more analog-to-digital converters are designed as FADC and the digitized values for further calculation in one or several memories of a microcontroller can be stored.

Images