Method and device for detecting the closing time of a metering valve

Abstract

Method for detecting a closing time (205) of a metering valve (71), which has a solenoid coil and a movable armature mechanically coupled to a valve needle, and which is part of a metering unit (80) with which a liquid for nitrogen oxide reduction is metered in the direction of flow of the exhaust gas upstream of a catalytic assembly of the exhaust gas aftertreatment system of an internal combustion engine (1) to the exhaust gas stream (20) via an injection unit (70), characterized in that, in order to detect the closing time (205), a current profile (203) of the coil current (201) flowing through the solenoid coil of the metering valve (71) is recorded in a first stage, wherein the metering valve (71) is actuated with a standard quick-extinguishing duration at the switching-off moment or with a quick-extinguishing duration varied compared to the standard quick-extinguishing duration, and a measuring window is positioned in time such thatthat in a second stage the closing time (205) is determined based on the current profile (203) recorded in the first stage within the time-positioned measuring window and subsequently the determined closing time (205) is calculated back for a standard rapid extinguishing duration of the metering valve (71).

Description

State of the art

[0001] The invention relates to a method for detecting a closing time of a metering valve, which has a magnetic coil and a movable armature which is mechanically coupled to a valve needle and which is a component of a metering unit with which, in an exhaust aftertreatment system of an internal combustion engine, a liquid for reducing nitrogen oxides is metered into the exhaust gas flow via an injection unit in the flow direction of the exhaust gas upstream of a catalytic assembly of the exhaust aftertreatment system.

[0002] The invention further relates to a device, in particular a control unit, for carrying out the method according to the invention.

[0003] To comply with current emissions regulations, a nitrogen oxide storage catalyst (NSC) is predominantly used in small diesel engines. This reduces nitrogen oxides (NO, NO2) in the exhaust tract. The nitrogen oxides are initially stored in the catalyst. When the catalyst's absorption capacity is exhausted, an engine control unit adjusts to a rich exhaust gas mixture. This reduces the nitrogen oxides temporarily stored in the catalyst to nitrogen.

[0004] This method works well in the low and medium load and temperature ranges. To comply with future emissions regulations, the high-load range must also be monitored for limit values.

[0005] To ensure that the nitrogen oxide storage catalyst (NSC) reduces nitrogen oxides even during high-load operation and at high temperatures, it must be operated in the so-called DiAir mode. DiAir stands for "Diesel NOx Aftertreatment by Adsorbed Intermediate Reductants." This mode involves injecting hydrocarbons (hydrocarbons) upstream of the nitrogen oxide storage catalyst, a process also known as HCI (hydrocarbons injection). Typically, additional diesel fuel is injected into the exhaust system for this purpose.

[0006] This requires a special DiAir dosing system, which consists of the assemblies - Injection unit, essentially consisting of a metering valve and a heat sink, - Dosing unit, essentially consisting of a pressure stage (e.g. a feed pump) and a pressure sensor, and - Control unit, e.g. implemented as hardware and / or software in the engine control.

[0007] As emissions-relevant components or subsystems, DiAir components are subject to the strictest requirements of emissions legislation. All subcomponents must be monitored for correct functioning and malfunctions as part of on-board diagnostics (OBD).

[0008] Of particular importance in the DiAir dosing system is the monitoring of the dosing valve in the injection unit. These dosing valves have a solenoid coil. Actuating the solenoid coil moves an armature that is mechanically connected to a valve needle or is designed as such. This armature opens the valve against a spring force, allowing a fluid under high pressure to be injected into the exhaust duct. Monitoring the armature movement and thus the timing of valve opening and closing is of particular interest.

[0009] Deviations in the opening time are detected or determined by the so-called BIP algorithm and compensated by adjusting the current application duration. BIP refers to the time of armature movement after activation. Typically, the second time derivative of the current is evaluated, and the corresponding minima and maxima in the current curve are assessed. This method is known from the applicant's DENOXTRONIC systems and is already in series production.

[0010] The document DE 37 30 523 C2 describes a method and a device for controlling the mechanical movement of a solenoid valve armature, for the actuation of which a magnetic winding is provided, wherein, after switching off the movement current through the magnetic winding, induction voltages on the magnetic winding caused by the movement of the solenoid valve armature, which are assigned to mechanical switching times (BOP, EIP), are monitored and the signals thus generated are each raised to a detectable signal level by means of an external energy source.

[0011] DE 10 2007 003 211 A1 describes a method and a device for controlling an electromagnetic valve, wherein a valve needle of an electromagnetic valve assumes a first position when de-energized and a second position when energized, wherein, upon transition from the second position to the first position of the valve needle, a subsequent energization occurs for a specific period of time at a predeterminable time. Preferably, provision is made for the current to drop slowly rather than being rapidly extinguished. The duration of the subsequent energization can be selected such that a braking force acts on the valve needle, causing it to strike the valve seat at a moderate speed.

[0012] DE 10 2012 209 967 A1 describes a method, a control and / or regulating device and a computer program for operating a solenoid valve, wherein a valve element of the solenoid valve can be brought into a first position by a bias voltage and into a second position by magnetic force. Provision is made for determining, during a period without current supply, a time integral using a current flowing through a solenoid coil of the solenoid valve or a variable characterizing the current, and for comparing the time integral with at least one threshold value, and for inferring a state of the solenoid valve from the comparison with the at least one threshold value. Among other things, the method can be used to monitor the injection of a reducing agent into an exhaust gas aftertreatment system of an internal combustion engine.

[0013] Further algorithms for determining a launch time point (BMP) and the time of anchor strike (MSP) are also described in the applicant's not yet published patent applications with the internal file numbers R.348498 and R.346347.

[0014] Closing processes in injection valves are supported by a quenching voltage to enable a rapid dissipation of the energy stored in the solenoid coil. This is referred to as fast quenching. Fast quenching guarantees a fast closing process through a rapid dissipation of the magnetic forces. Although deviations in the closing point (EIP) have previously been determined using the same algorithm as used to determine the BIP, the accuracy and robustness in vehicle operation are considered too low due to various disturbances and a lack of flexibility. Another problem with previous methods is a fixed measurement window for recording the closing point (EIP). Due to drift, the closing point can quickly fall outside the measurement window and thus go undetected.Therefore, according to the current state of the art, the correct closing time (EIP) cannot be determined over the entire service life of the valve.

[0015] It is therefore an object of the invention to provide a method with which a drift of the closing time (EIP) can be determined over the entire service life of the valve and by means of a resulting adjustment of the control duration, this drift can be compensated.

[0016] It is a further object of the invention to provide a corresponding device, in particular a control unit, for carrying out the method. Disclosure of the invention

[0017] The problem concerning the method is solved by the features of claims 1 to 15.

[0018] According to the invention, in order to detect the closing time (EIP), in a first stage a current profile of the coil current flowing through the solenoid coil of the metering valve is recorded, wherein the control of the metering valve takes place with a standard quick extinguishing duration at the switch-off moment or with a quick extinguishing duration that varies compared to the standard quick extinguishing duration, and a measuring window is positioned in time in such a way that in a second stage the determination of the closing time is determined on the basis of the current profile recorded in the first stage within the temporally positioned measuring window and then the determined closing time is calculated back for a standard quick extinguishing duration of the metering valve.This two-stage process allows the dosing valve's closing time to be precisely determined regardless of its age, and allows multiple control methods and detection algorithms, some of which were already mentioned at the beginning, to be used together. This allows for a high degree of flexibility and also offers expansion options. Depending on the aging state of the dosing valve, a suitable control method and / or detection algorithm can be selected.

[0019] In a preferred method variant, it is provided that in the first stage of the detection method, the closing time of the metering valve is delayed in time by means of a quick extinguishing duration that is shorter than the standard quick extinguishing duration at the moment the metering valve is switched off, so that a clear detection of the closing time is enabled in the second stage of the detection method. This makes it possible, particularly with new metering valves, to shift features in the current waveform for a closing time back in time in the recorded current waveform, so that a clear evaluation of these features and thus an exact determination of the closing time of the metering valve is enabled. Especially with new metering valves, when the standard quick extinguishing duration is used, a closing time cannot be clearly detected due to its rapid occurrence.

[0020] On the other hand, the method also provides that in the first stage of the detection process, the dosing valve is controlled with the standard quick extinguishing duration if, due to age, the dosing valve already has a delayed closing time or if this has been clearly detected.

[0021] In a further advantageous method variant, the temporal position of the measurement window for a measurement period is adjusted such that the time at which the armature started moving in the previous measurement period is used as the center for the next measurement window to determine the closing time. This ensures that the analysis period is always optimally adjusted for evaluating the closing time or for detecting a feature in the current curve for a closing time, enabling a clear and reproducible determination of the closing time.

[0022] In this case, the measuring window can advantageously be shifted step by step from one measuring period to the next until a closing time can be detected within the new measuring window or the end of a search area is reached.

[0023] A particularly advantageous method variant provides for the current waveform to be recorded within the measurement window with a constant number of measurement points and at a constant sampling frequency. This saves memory space and reduces the load on the μprocessor, since only the relevant part of the current waveform during the valve closing phase is recorded and stored for further processing.

[0024] This further processing to determine the closing time can be advantageously carried out if, in the second stage of the detection method, the closing time is determined by calculating the first time derivative of the current waveform from the current waveform or from a filtered current waveform and detecting a first and a second zero crossing (N1, N2), whereby the closing time of the metering valve is defined as the time for the second zero crossing of the first time derivative of the current waveform. This makes use of the characteristic current waveform as it occurs when a valve closes. This is characterized by the initial occurrence of a minimum, which corresponds to the first zero crossing N1 in the time-differentiated current waveform, and subsequently of a maximum, which corresponds to the second zero crossing N2 in the time-differentiated current waveform. This typical curve is also referred to as an EIP characteristic.To verify the plausibility of the minima and maxima, a second time derivative of the coil current waveform or the filtered current waveform can be used. Another detection method is the evaluation of the inductance waveform of the solenoid coil. Filtering has the advantage that short-term fluctuations can have only a minor impact on the closing time determination. In addition to this method, other algorithms are conceivable, such as those mentioned at the beginning.

[0025] When varying the quick-quenching duration to clearly determine the closing time, it is necessary to determine a closing time for the standard quick-quenching duration from the closing time determined in the second stage of the detection process by multiplying it by a correction factor. The correction factor is calculated as a function of the applied quick-quenching duration and a holding current, both of which are linked via an energy balance of the solenoid coil. This allows the closing delay to be calculated during normal dosing operation of the valve. The correction factor, based on the applied quick-quenching duration and the holding current, can be stored, for example, in a characteristic map unit for different values, allowing for a quick calculation.

[0026] A particularly advantageous method variant provides for the implementation of the detection process in a two-mode process, whereby the closing time is determined in two different modes over the service life of the dosing valve. First, in a normal dosing mode, it is checked whether a closing time can be determined by applying the standard quick-extinguishing duration and gradually shifting the measurement window. If this is not the case, a so-called EIP detection mode is applied, in which the quick-extinguishing duration is gradually reduced. The advantage here is that the closing time or EIP time can be determined directly without back-calculation if the EIP characteristic is shifted significantly backwards in time during normal dosing operation with the standard quick-extinguishing duration.

[0027] It can be provided that the number of closing times not found in the normal dosing mode is incremented and the EIP discovery mode is activated when an applicable limit is exceeded.

[0028] In order to detect a possible malfunction of the dosing valve within the scope of an on-board diagnosis (OBD), which is required to meet legal requirements, a process variant provides that when a minimum permissible quick extinguishing time is reached in the EIP discovery mode, the dosing valve is classified as faulty because no closing time could be determined, and a corresponding error entry is made, e.g. in a higher-level engine control system.

[0029] A further advantageous method variant provides for the implementation of the detection process in a single-mode process, whereby the closing time is only implemented in an EIP detection mode. Initially, an expected measurement window position is specified for a feasible quick-extinguishing duration for new metering valves, and the measurement window for detecting the closing time is gradually shifted back in time until it is detected. The advantage of the single-mode process is the simplicity and stability of the function, since no switching between modes is required. The process can be implemented once, for example, during suitable operating conditions or after a specific operating time.

[0030] It can be provided that if the closing time is not detected and a predefined time limit for shifting the measuring window is reached, the quick extinguishing duration is varied step by step, starting from the initial value for new dosing valves.

[0031] A preferred method variant provides that, based on a determined opening time, which can be detected, for example, using methods as mentioned above, and the closing time, a correction of the control duration of the metering valve is carried out in order to minimize metering tolerances. This can be particularly advantageous if a drift in the closing time occurs due to wear and / or aging, which can then be monitored and its effects corrected. Depending on the aging state of the metering valve, the metered quantity or the metering strategy can be adjusted. For example, for an old valve, a long metering duration is selected instead of several short metering cycles.

[0032] In addition, the method according to the invention and its variants can also be used to determine the shortest possible dosing time and to use it specifically to fine-tune the dosing.

[0033] A preferred use of the method described above and its variants provides for use in a dosing system with which, during high-load operation of an internal combustion engine designed as a diesel engine, hydrocarbons in the form of diesel fuel are metered into the exhaust duct upstream of a nitrogen oxide storage catalyst in the flow direction of the exhaust gas to reduce nitrogen oxides. The dosing system comprises the following assemblies: injection unit, essentially consisting of a metering valve and a cooling body; dosing unit, essentially consisting of a feed pump, which can be designed as a reciprocating piston pump, and a pressure sensor; and control unit. This dosing system is also known as a DiAir system and is used in particular for nitrogen oxide reduction in small diesel internal combustion engines, such as those used in cars.In DiAir mode, additional diesel fuel is injected, especially during high-load operation of the internal combustion engine, as already described above. This fuel injection can also be used to increase the exhaust gas temperature during diesel particulate filter (DPF) regeneration.

[0034] In principle, the method can also be advantageously used in exhaust gas aftertreatment systems in which an ammonia-releasing fluid, e.g., an aqueous urea solution, is metered into the exhaust duct upstream of an SCR catalyst to reduce nitrogen oxides. Even in such systems, the reliable functioning of the feed pump is crucial. Furthermore, there are corresponding legal requirements regarding on-board diagnostics (OBD), which can be met using the method described above. Such systems are known as DENOXTRONIC 5.x from the applicant, which can determine the actual opening duration and, from this, the actual metered amount of fluid from the opening time (BIP) and closing time (EIP) of the metering valve. The design of the DENOXTRONIC metering valve is similar to the metering valve used in the DiAir system, so that the use of the method described above, with its variants, offers similar advantages as with the DiAir system.

[0035] The object of the device is achieved in that the control unit comprises devices for implementing the previously described method with its variants, and in particular, an ADC unit (analog-to-digital converter) for recording a current profile of the coil current through the solenoid coil, calculation units for temporal differentiation of the current profile or a filtered current profile, determination of minima and maxima, and correction of the determined closing time, as well as devices for influencing a quick extinguishing duration at the moment the metering valve is switched off. The implementation can be at least partially software-based, whereby the control unit can be designed as a separate unit or as an integral component of a higher-level engine control system.Typically, no hardware changes are required, as these devices are already part of valve control units, which simplifies the implementation of the method according to the invention.

[0036] The invention is explained in more detail below using an exemplary embodiment illustrated in the figures. It shows: Fig. 1 an example of a technical environment for the invention, Fig. 2 shows a schematic representation of a dosing system, Fig. 3 in a first flow diagram schematically shows different current profiles and closing times of different dosing valves when operating with a standard quick extinguisher, Fig. 4 in a second diagram schematically shows different current profiles and closing times of different metering valves when operating with a shortened quick extinguishing and Fig. 5 in a third curve diagram a current curve, a filtered current curve and the first time derivative of the current curve for a dosing valve.

[0037] Fig. Figure 1 shows an example of a technical environment in which the method according to the invention can be applied. The illustration is limited to the components necessary for explaining the invention.

[0038] In the Fig. 1 shows, by way of example, an internal combustion engine 1 designed as a diesel engine, comprising an engine block 10 and an exhaust duct 30 through which an exhaust gas stream 20 is conducted. The exhaust duct 30 has an exhaust gas purification system which, in the example shown, is a catalytically coated component arranged in the flow direction of the exhaust gas and initially has a nitrogen oxide storage catalytic converter 40 (NSC) and a diesel particulate filter 50 (DPF). Upstream of the nitrogen oxide storage catalytic converter 40 (NSC), an injection unit 70 is attached to the exhaust duct 30, with which hydrocarbons (HC), for example in the form of fuel, can be injected during a DiAir phase in order to be able to reduce nitrogen oxides during high-load operation at high temperatures. In addition, this fuel injection can also be used to increase the temperature of the exhaust gas during regeneration of the diesel particulate filter 50 (DPF).

[0039] The injection unit 70, together with a dosing unit 80, belongs to a dosing system 60, which can be controlled by a control unit 101. The functionality of the control unit 101 can be implemented in a software- and / or hardware-based manner in a higher-level engine control unit 100, e.g., an ECU (Electronic Control Unit), as is common in diesel engines.

[0040] The metering unit 80, which is also referred to as PSU-HCI (Power Supply Unit for Hydro Carbon Injection), is fed with diesel fuel via an inlet 81 from the return line of a high-pressure pump (not shown here) for the fuel injection of the internal combustion engine 1. Excess fuel can be fed into a tank 90 (see Fig. 2) flow back. The injection unit 70 has a water cooling system 72 to protect the components installed therein from the high temperatures in the exhaust duct 30.

[0041] Fig. 2 schematically shows further details of the dosing system 60. In the dosing unit 80, the pressure of the diesel fuel is increased from the inlet 81 via a feed line 82 by means of a feed pump designed as a reciprocating piston pump 86 to an injection pressure, typically 10 bar. When a quantity is requested from the dosing system 60, a dosing valve 71, which is a component of the injection unit 70, is opened. The hydraulic diagnosis as well as the control and monitoring of the dosing system 60 is carried out via a pressure sensor 87. In the flow direction of the diesel fuel, valves 85 are provided upstream and downstream of the reciprocating piston pump 86 to prevent backflow or backflow. By means of a return device 88, excess fuel, for example from leaks, can be conveyed via the outlet 83 into the tank 90 in a return line 84.

[0042] The reciprocating piston pump 86, the pressure sensor 87 and the metering valve 71 are controlled by the control unit 101 or receive feedback from these components, e.g. about temperatures and the current flowing through a solenoid coil of the metering valve 71.

[0043] The method according to the invention is based on a two-stage detection method for determining the valve closing time (EIP). In a first stage, the current consumption through the solenoid coil of the metering valve 71 is determined by positioning a measurement window such that an EIP is detected with high reliability. For this purpose, it may be necessary to adjust the control of the metering valve 71 so that the EIP time is delayed or shifted backward. A suitable measure for shifting the EIP time is to shorten the so-called quick extinguishing duration at the moment of shutdown of an output stage in the control unit 101.

[0044] In the second stage of the procedure, the EIP time is determined based on the current waveform recorded in the first stage. The current waveform is analyzed for corresponding EIP characteristics and the EIP time is then calculated back for normal operation.

[0045] The measures in the individual procedural stages are described in detail below.

[0046] In the first stage, the current waveform containing the EIP characteristic is recorded and displayed as clearly as possible. The EIP characteristic manifests itself as a local maximum in the current waveform. To achieve this, it may be necessary, particularly with new metering valves 71, which still have a fast closing characteristic, to delay the EIP time by reducing the fast extinguishing duration so that there is sufficient time between the end of the fast extinguishing and the EIP time. Since the EIP time is continually pushed back over time as a result of aging, the reduction in the fast extinguishing duration can be continued. From a certain EIP time, the fast extinguishing for EIP recording no longer needs to be reduced and can remain at the same value as during normal operation.Since changing the quick extinguishment affects the closing behavior, and continuous recording of the EIP time is not necessary, EIP recording should only be performed occasionally. This operating state is referred to as EIP discovery mode.

[0047] Fig. 3 shows, in a first waveform diagram 200, various current waveforms 206, 207, 208 for metering valves 71 of different ages, which are operated with a standard quick-quenching duration, wherein a coil current 201 is plotted as a function of time 202. A first and a second current waveform 206, 207 show the waveform for a first and a second metering valve 71, which are relatively new and each have an early EIP time (first closing time 209 EIP1 and second closing time 210 EIP2), the differentiation and thus recognition of which is difficult. A third current waveform 208 shows the waveform for a third metering valve 71, which is relatively old and has a later EIP time (third closing time 211 EIP3), but whose EIP feature is clearly recognizable.

[0048] Fig. 4 shows in a second trend diagram 200 according to the Fig. 3 shown curve diagram 200 different current curves 206, 207, 208 for the different old metering valves 71 according Fig. 3, which are operated with a reduced quick extinguishing time. The first closing time 209 (EIP1) and the second closing time 210 (EIP2) of the two new metering valves 71 are Fig. 3 are delayed and are shifted back in time and are clearly distinguishable. The third closing point 211 (EIP3) of the old metering valve 71 has also been shifted back, but is still clearly recognizable.

[0049] In order to save memory space and to relieve a µprocessor in the control unit 101, only a part of the current waveform 203 (compare Fig. 5) recorded during the valve closing time. The method uses an EIP measuring window with a fixed number of measuring points and a constant sampling frequency, i.e. only the measuring points within the measuring window are recorded and then evaluated in the second process stage. To achieve the best result and to make it easier to determine the EIP time in the second stage, the time at which the valve armature started to move (BMP) in the previous measurement is used as the center for the next EIP measuring window. If a valid EIP time cannot be determined, the EIP measuring window is shifted step by step for the next measuring window until the EIP time is found. If no EIP is found and the end of the search range is reached, the search continues at the beginning of the search window.

[0050] In the second stage, the EIP time is determined from the current waveform recorded in the first stage. This determination has no influence on the control of the valve. It is simply an algorithm and logic that calculates the EIP time.

[0051] During rapid quenching, the current is forced down to the quiescent current (close to zero). It is important to ensure that the residual magnetic flux linkage, fed by eddy currents, is greater than zero at the start of freewheeling operation or that residual remanence is ensured. However, since rapid quenching is only active for a short period of time, the magnetic field is partially retained due to magnetic hysteresis and the valve remains fully open. When rapid quenching is complete, the residual energy in the solenoid coil slowly decays via a freewheeling diode, generating a current curve in an e-function. As soon as the magnetic force has been reduced to such an extent that the force of the valve's return spring predominates, the valve needle or valve armature begins to move and the air gap in the magnetic circuit increases. This reduces the inductance and creates an induced voltage that is added to the circuit in the direction of current flow.The induced voltage generates an increase in current, which lasts until the needle or armature movement ends. Afterward, the current decays to zero in an e-function.

[0052] The EIP point can be identified when the current first decreases, then increases, and then decreases again. This characteristic can be identified using various mathematical methods, such as the first time derivative, the second time derivative of the current waveform, and the calculation of the inductance. The inductance can be determined using the following formula: L=(UBatt−IScoil*RSoil)*dt / d(IScoil)

[0053] Where L is the inductance of the solenoid coil of the metering valve 71, U Batt the battery voltage or supply voltage, I Spule the coil current 201 and R Spule the coil resistance.

[0054] The second time derivative is calculated by differentiating the filtered current waveform 204 twice and then examined for maxima and minima. This identification method has been shown to produce good and, above all, reproducible results for detecting needle or armature movement and thus also for detecting the EIP time.

[0055] Fig. 5 shows, by way of example, in a waveform diagram 200 in the upper part of the figure, a current waveform 203 of the coil current 201 as a function of time 202. In addition, a filtered current waveform 204 is shown, which is formed by means of low-pass filtering from the current waveform 203 or by means of a moving average value for discretely determined current values. In the lower part of the Fig. 5 shows the values ​​of the first time derivative of the current waveform 212 as a function of time 202. The first time derivative of the current waveform 212 is formed by simple differentiation of the filtered current waveform 204 and examined for zero crossings 213, 214 (N1, N2), whereby the second zero crossing 214 (N2) is used as the closing time 205 or EIP time.

[0056] The EIP time for the normal quick extinguishment duration is then multiplied by a factor k EIP calculated. The factor kEIP=f(tGAP,IHOLD) is a function of the quick extinguishing time t GAP and the holding current I HOLD , both of which are linked by an energy balance of the coil.

[0057] In a further embodiment of the method according to the invention, a distinction can be made between a two-mode method and a one-mode method for detecting the EIP time.

[0058] In the two-mode procedure, the EIP time is determined in two different modes over the lifetime of the valve. In normal dispensing mode, a check is carried out to determine whether the EIP time can be determined using the standard quick-clear function. If the valve is new, the EIP time is too early for reliable detection (see Fig. 3). If the number of undetected EIP times exceeds a predefined limit, the EIP discovery mode is activated, in which the quick deletion time is reduced and the EIP time is postponed (see Fig. 4). For older valves, the EIP time is also shifted back in time with the standard quick extinguishment, without the need to adjust the quick extinguish duration.

[0059] The following is a summary of the steps of the two-mode procedure: • Use the default quick delete duration first • Recording the current waveform 203 of the coil current 201 • Filtering the current waveform 201 to a filtered current waveform 204 • Calculation of the first time derivative of the current waveform 212 from the filtered current waveform 204 • Determination of the first zero crossing 213 (N1); a plausibility check can be carried out as follows: if all 3 values ​​before the first zero crossing 213 (N1) are less than zero, a falling current curve is detected and the search continues for the second zero crossing 214 (N2) • If the first and second zero crossings 213, 214 are not found, the search for the EIP feature can be aborted in this period for this measurement window, with the number of undetected EIP time points being incremented. In the next period, the measurement window is shifted back in time, and the detection process is repeated. • If the first and second zero crossings 213, 214 are found, the second zero crossing 214 is used as the EIP time as described above • At the next EIP time determination, the measurement window is centered on the first zero crossing 213 • However, if the total number of unrecognized EIP times exceeds a predefined limit, the EIP discovery mode is activated • For this purpose, the quick deletion time is gradually reduced until an EIP time is found • If the quick extinguishing duration has already been reduced to a minimum permitted value and no EIP feature has yet been found, the dosing valve 71 is classified as not functioning and corresponding error entries are generated, e.g. in the higher-level engine control 100. • Calculation of the closing delay in normal dosing mode with standard quick extinguishing duration based on the determined closing delay in EIP discovery mode using the relationship according to (2).

[0060] In the single-mode method, the EIP time is only determined in the EIP discovery mode, which can be performed once, for example, in suitable operating states or after a certain operating time.

[0061] The steps of the one-mode procedure are summarized below: • Start the detection procedure with an initial value for the quick extinguishing time, as is usual for new valves • Determination of the measurement window position starting with an expected value • Recording the measured values ​​for the coil current 201 within the previously defined measuring window • Search for the EIP feature and determine the EIP time according to the procedure described above • If no EIP feature corresponding to the zero crossings 213, 214 was found, the measurement window for the next measurement period is shifted back in time • If the measurement window has already passed through the entire preset time range and still no EIP feature has been found, the quick deletion duration is increased • If an EIP feature is found, the measurement window for the next EIP determination in the next period is centered at the first zero crossing 213.

[0062] According to the invention, the use of the previously described method with its method variants for detecting the closing time 205 (EIP) for metering valves 71 of a DiAir metering system is provided, as described at the beginning and in the Fig. 1 and Fig.2. For this purpose, current sampling with sufficient accuracy is provided on the DiAir metering valve controller. A digitized curve of the coil current 201 and the filtered current curve 204 derived from it can be determined, for example, using an ADC unit (analog-to-digital converter) and other calculation units within the control unit 101. By detecting the opening time (BIP) and the closing time 205 (EIP), the activation duration of the metering valve 71 of the DiAir system can then be corrected with the aim of minimizing the metering tolerance.

Claims

[1] Method for detecting a closing time (205) of a metering valve (71), which has a magnetic coil and a movable armature which is mechanically coupled to a valve needle, and which is a component of a metering unit (80) with which, in an exhaust gas aftertreatment system of an internal combustion engine (1), a liquid for reducing nitrogen oxides is metered into the exhaust gas flow (20) via an injection unit (70) in front of a catalytic assembly of the exhaust gas aftertreatment system, characterized byin that, in order to detect the closing time (205), a current profile (203) of the coil current (201) flowing through the magnetic coil of the metering valve (71) is recorded in a first stage, the metering valve (71) being controlled with a standard quick-extinguishing duration at the switch-off moment or with a quick-extinguishing duration which varies with respect to the standard quick-extinguishing duration, and a measuring window is positioned in time in such a way that, in a second stage, the determination of the closing time (205) is determined on the basis of the current profile (203) recorded in the first stage within the time-positioned measuring window, and the determined closing time (205) is then calculated back for a standard quick-extinguishing duration of the metering valve (71). [2] Method according to claim 1, characterized bythat in the first stage of the detection method, the closing time (205) of the metering valve (71) is delayed in time by means of a quick extinguishing time that is reduced compared to the standard quick extinguishing time at the moment of switch-off of the metering valve (71), so that a detection of the closing time (205) is clearly enabled in the second stage of the detection method. [3] Method according to claim 1, characterized by that in the first stage of the detection process, the dosing valve (71) is controlled with the standard quick extinguishing duration if, due to age, the dosing valve (71) already has a delayed closing time (205). [4] Method according to one of claims 1 to 3, characterized bythat the temporal position of the measuring window for a measuring period is adjusted in such a way that the time at which the armature started moving in the previous measuring period is used as the center for the next measuring window for determining the closing time (205). [5] Method according to one of claims 1 to 4, characterized by that the measuring window is shifted step by step from one measuring period to the next until a closing time (205) is detectable within the new measuring window or the end of a search area is reached. [6] Method according to one of claims 1 to 5, characterized by that the current waveform (203) is recorded within the measuring window with a constant number of measuring points and at a constant sampling frequency. [7] Method according to one of claims 1 to 6, characterized byin that in the second stage of the detection method, the closing time (205) is determined by calculating the first time derivative of the current curve (212) from the current curve (203) or from a filtered current curve (204) and detecting a first and a second zero crossing (213, 214), wherein the time for the second zero crossing (214) of the first time derivative of the current curve (212) is defined as the closing time (205) of the metering valve (71). [8] Method according to one of claims 1 to 7, characterized by that a closing time (205) for the standard quick extinguishing duration is determined from the closing time (205) determined in the second stage of the detection method by multiplying it by a correction factor, wherein the correction factor is formed as a function of the applied quick extinguishing duration and a holding current, both of which are linked via an energy balance of the magnetic coil. [9] Method according to one of claims 1 to 8, characterized by that the detection method is carried out in a two-mode method, wherein the closing time (205) is determined in two different modes over the service life of the dosing valve, wherein first in a normal dosing mode it is checked whether a closing time (205) can be determined when using the standard quick extinguishing time and gradually shifting the measuring window, and, if this is not the case, a so-called EIP discovery mode is applied, in which the quick extinguishing time is gradually reduced. [10] Method according to claim 9, characterized by that the number of closing times not found in normal dispensing mode is incremented and if an applicable limit is exceeded, the EIP discovery mode is activated. [11] Method according to claim 9 or 10, characterized bythat when a minimum permissible quick extinguishing time is reached in the EIP detection mode, the dosing valve (71) is classified as faulty and a corresponding error entry is made. [12] Method according to one of claims 1 to 8, characterized by that the detection method is carried out in a single-mode method, wherein the closing time (205) is only carried out in an EIP detection mode, wherein initially, with an applicable quick extinguishing duration for new dosing valves (71), an expected measuring window position is specified and the measuring window for the detection of the closing time (205) is gradually shifted backwards in time until it is detected. [13] Method according to claim 12, characterized bythat if the closing time is not detected and a preset time limit for shifting the measuring window is reached, the quick extinguishing duration is varied step by step, starting from the initial value for new dosing valves (71). [14] Method according to one of claims 1 to 13, characterized by that a correction of the control duration of the metering valve (71) is carried out on the basis of a determined opening time and the closing time (205). [15] Use of the method according to one of claims 1 to 14 in a metering system (60) with which, in high-load operation of an internal combustion engine (1) designed as a diesel engine, hydrocarbons in the form of diesel fuel are metered into the exhaust gas duct (30) in front of a nitrogen oxide storage catalyst (40) in order to reduce nitrogen oxides in the flow direction of the exhaust gas, wherein the metering system (60) comprises the assemblies of injection unit (70), essentially consisting of a metering valve (71) and a cooling body, metering unit (80), essentially consisting of a feed pump and a pressure sensor (87), and control unit (101). [16] Device, in particular a control unit (101), for detecting a closing time (205) of a metering valve (71), which has a magnetic coil and a movable armature which is mechanically coupled to a valve needle, and which is a component of a metering unit (80) with which, in an exhaust gas aftertreatment system of an internal combustion engine (1), a liquid for reducing nitrogen oxides can be metered into the exhaust gas flow (20) via an injection unit (70) in front of a catalytic assembly of the exhaust gas aftertreatment system, characterized bythat the control unit (101) has devices for carrying out the method according to claims 1 to 15 and in particular comprises an ADC unit for recording a current profile (203) of the coil current (201) through the magnetic coil, calculation units for the temporal differentiation of the current profile (203) or a filtered current profile (204), determination of minima and maxima and correction of the determined closing time (205), as well as devices for influencing a quick extinguishing duration at the switch-off moment of the metering valve (71).

Images