DE102005042490A1 - Operating process for internal combustion engine involves reporting difference in calculated masses of nitrogen oxides in time sequence - Google Patents

Abstract

The operating process is for an engine (10) whose exhaust region (13) contains at least one catalytic converter (16) with a temperature sensor (17) and a nitrogen oxide sensor (18), and with a reagent to convert nitrogen oxides. The difference in the calculated masses of nitrogen oxides are reported in time sequence. The masses for the differences are stored and evaluated in the production of a reagent signal dependent on the sequence of the results of the evaluation.

Description

The Invention is based on a method for operating an internal combustion engine, in the exhaust gas an SCR catalyst and downstream the SCR catalyst arranged a NOx sensor are and of a device for carrying out the method according to the Genus of independent Claims.

In the DE 199 03 439 A1 For example, a method and a device for operating an internal combustion engine are described, in the exhaust gas region of which an SCR catalytic converter (selective catalytic reduction) is arranged, which reduces the nitrogen oxides contained in the exhaust gas of the internal combustion engine to nitrogen with a reagent. The metering of the reagent preferably takes place as a function of operating parameters of the internal combustion engine, such as, for example, the rotational speed and the injected fuel quantity. Furthermore, the dosage is preferably carried out as a function of exhaust gas parameters, such as the exhaust gas temperature or the operating temperature of the SCR catalyst.

When Reagent is provided, for example, the reducing agent ammonia, that from a urea-water solution can be won. The dosage of the reagent or of The starting materials of the reagent must be carefully determined. A Too low a dosage has the consequence that nitrogen oxides in the SCR catalyst not complete anymore can be reduced. Too high a dosage leads to a reagent slip, on the one hand to an unnecessarily high Reagent consumption and on the other hand, depending on the nature of the reagent, resulting in an unpleasant odor nuisance can lead.

In the DE 10 2004 031 624 A1 (Not prepublished) is described a method for operating an SCR catalyst used for cleaning the exhaust gas of an internal combustion engine, in which a control or regulation of the reagent level in the SCR catalyst is provided to a predetermined setpoint. The specific specification of the desired value ensures, on the one hand, that a sufficient amount of reagent is available in transient states of the internal combustion engine for eliminating the NOx raw emissions of the internal combustion engine as completely as possible and, on the other hand, prevents reagent slippage.

Of the Reagent filling level of the SCR catalyst is determined on the basis of a catalyst model, that is the NOx mass flow entering the SCR catalyst, the the SCR catalyst leaving NOx mass flow, the catalyst temperature and optionally considered the reagent slip. The maximum possible Reagent filling level of the SCR catalyst depends in particular from the operating temperature of the SCR catalyst, where it is highest at low operating temperatures and with increasing Operating temperature drops to smaller values. The efficiency of the SCR catalyst depends on the catalytic activity which is low at low operating temperatures, with increasing Operating temperature goes through a maximum and with further increasing operating temperature drops again.

In the DE 199 60 731 A1 and DE 199 62 912 A1 In each case, NOx sensors are described which are provided for detecting the NOx concentration present in an exhaust gas flow. The described NOx sensors have a plurality of chambers, which communicate with each other via the diffusion barriers. The known multi-chamber NOx sensors have a cross-sensitivity to ammonia (NH3) due to the measuring principle. The ammonia contained in the exhaust gas as an example of a reagent leads via the reaction 4 NH3 + 5 O2 → 4 NO + 6 H2O to a falsification of the sensor signal. Accordingly, if an increase in the reagent dosage is carried out in the previously known procedures, the sensor signal will increase due to the reagent slippage occurring in the case of an existing overdose or correct dosage of the reagent and will decrease due to the increasing NOx conversion in the case of a previously existing underdosage. If, on the other hand, a lowering of the reagent dosage occurs, the sensor signal will slip due to the reduced reagent slip in a hitherto existing overdose of the reagent and increase at a previously existing correct dosage or underdosing due to the no longer complete NOx conversion.

In the DE 10 2004 046 640 A1 (not prepublished) a generic method for operating an internal combustion engine and an apparatus for carrying out the method are described, in which a NOx sensor is arranged with a cross-sensitivity to a reagent downstream of the SCR catalyst. In the exhaust region, at least one SCR catalyst is arranged, which is acted upon by a reagent, which contributes to the NOx conversion in the SCR catalyst. Provided is the calculation of at least one measure of the NOx concentration occurring downstream of the SCR catalyst, which allows for an increase in the accuracy of determining the dosage of the reagent.

One Reagent slippage may be calculated from the difference between the calculated Measure of the NOx concentration and the measured measure of the sum the NOx concentration and the reagent concentration determined become. Considered The fact is that both a reagent hatch and an insufficient NOx reduction reaction is a deviation between the calculated measure of the NOx concentration and the measured measure of the sum the NOx concentration and the reagent concentration in the same Effect direction. According to one Embodiment is at a high difference initially the dosage of the reagent is reduced. If there was a reagent slip, becomes the reduction of the dosage of the reagent to a Reduce reagent slippage. This will be the difference smaller. The reduction of the dosage of the reagent has in this case as a correct measure exposed. Was originally too low a dosage of the reagent before, so will the determined difference due to the lower NOx conversion on increase, so that it can be concluded that the reduction of Dosage of the reagent was wrong and instead an increase in the Dosage is to be made.

by virtue of the possible constant changes the quantity of dosing agent can oscillate around the optimum operating state, in which both a possible minimum NOx concentration downstream after the SCR catalyst as well as possible minimal reagent slippage should occur, not in all operating conditions be maintained.

Of the Invention is based on the object, a method for operating an internal combustion engine, in whose exhaust gas an SCR catalytic converter and a NOx sensor are arranged, and a device for carrying out the Specify a method that is as possible optimal exhaust gas cleaning result with minimal reagent slippage result.

The The object is achieved by those specified in the independent claims Features each solved.

Advantages of invention

The inventive method for operating an internal combustion engine, in the exhaust gas region at least an SCR catalyst is arranged, which is provided with a reagent is seen, which contributes to the NOx conversion in the SCR catalyst, sees that's at least a measure of the downstream after calculated NOx concentration occurring in the SCR catalyst and is measured with a NOx sensor that has a cross-sensitivity across from comprising the reagent. The difference between is determined the calculated measure of the NOx concentration and the measured sum of the NOx concentration and a reagent slip. The differences between the calculated measure and the Exhaust gas sensor signal. At least measures for the in chronologically determined differences. A rating is planned the crowd. A reagent signal indicating the dosage of the reagent determines is dependent of a given number and / or predetermined order of Evaluation results of the measures established.

The inventive approach avoids permanent misdosing through long-term adaptation. An essential advantage of the procedure according to the invention results in that the application effort is reduced. Especially is with the procedure according to the invention achieved a dosage of the reagent, which only a slight overshoot and undershoot.

advantageous Further developments and refinements of the procedure according to the invention arise from dependent Claims.

in principle Is it possible, immediately the differences of the further assessment. According to one embodiment, it is provided that the differences are made plausible and that the determined plausibility as dimensions for the Differences are saved and then evaluated. The plausibilities as results the plausibility can For example, reflect the finding whether an over- or Underdosing of the reagent is present.

A Embodiment provides that at a previous reduction the reagent signal a different number and / or order of valuation results of the measures for the differences as one increase is given. Another embodiment provides that a greater increase in the Reagent signal provided as a corresponding reduction of the reagent signal is. With these measures the emphasis will be on one as possible high NOx conversion laid. In a possibly present Underdose is as fast as possible the maximum possible NOx conversion reached again.

One embodiment provides that the dimensions for the differences with a low and / or high load of the internal combustion engine not stored or stored dimensions are deleted. Another According to an embodiment, the dimensions for the differences at a temperature of the SCR catalytic converter which is outside a predetermined temperature range and / or outside a predetermined range of the catalytic converter efficiency are not stored or stored dimensions are deleted. With these measures it is prevented that a change of the reagent signal due to special, not representative operating conditions of the internal combustion engine and / or characteristics of the exhaust gas is made.

The inventive device for operating an internal combustion engine initially applies Controller, that to carry out of the method is prepared. The control unit contains in particular a difference determination, which is the difference between that of one versus the other Reagent cross-sensitive NOx sensor provided exhaust gas sensor signal and the NOx concentration calculated by NOx concentration determination downstream determined after the SCR catalyst. Are still planned Difference memory for the determined differences as well as an evaluation arrangement for the evaluation the crowd for the Differences.

The control unit contains preferably at least one electrical storage in which the Procedural steps are stored as a computer program.

Further advantageous developments and refinements of the procedure according to the invention result from further dependent claims and from the description below.

drawing

1 shows a technical environment in which a method according to the invention runs and

2 shows a flowchart of a method according to the invention.

1 shows an internal combustion engine 10 , in their intake area 11 an air detection 12 and in their exhaust area 13 a reagent dosage 14 , a first NOx sensor 15 , an SCR catalyst 16 , an SCR catalyst 16 assigned temperature sensor 17 and a second NOx sensor 18 are arranged.

Downstream of the internal combustion engine 10 an exhaust gas flow ms_abg and a NOx raw concentration NOx_vK occur. Downstream of the SCR catalyst 16 occur an NOx concentration NOx_nK and a Reagenzmittelschlupf ms_Rea_nK on.

The air detection 12 represents a control unit 20 an air signal ms_L, the internal combustion engine 10 a rotation signal n, the first NOx sensor 15 a first NOx signal NOx_vK_mess, the temperature sensor 17 a temperature signal te_Kat and the second NOx sensor 18 an exhaust gas sensor signal S_nK available.

The control unit 20 represents one of the internal combustion engine 10 associated fuel metering device 25 a fuel signal m_K and the reagent dosage 14 a reagent signal S_Rea available.

The control unit 20 contains a torque determination 30 in which the air signal ms_L, the rotation signal n and a torque setpoint MFa are provided, and the torque Md of the internal combustion engine 10 determined.

The control unit 20 also contains a NOx raw concentration determination 31 in that the air signal ms_L, the rotation signal n and the fuel signal m_K are provided and which determines a calculated measure NOx_vK_mod of a NOx raw concentration NOx_vK.

The control unit 20 also includes a NOx concentration determination 32 , the NOx raw concentration NOx_vK, the temperature signal te_Kat, a space velocity RG and a reagent level ReaSp in the SCR catalyst 16 and the calculated amount of NOx_nK_mod for the NOx concentration NOx_nK downstream of the SCR catalyst 16 determined.

The calculated measure NOx_nK_mod for the NOx concentration NOx_nK and the exhaust gas sensor signal S_nK become a difference determination 33 provided which determines a difference D. In a result store 34 five differences D1-D5 can be stored, the time sequence being a time control 35 is given with a time signal t. The result store 34 Furthermore, the torque Md, the temperature signal te_Kat and a reset signal R are provided. Instead of the differences D1-D5 can in the result memory 34 that of a plausibility check 36 plausibilities P1-P5 determined from differences D are stored. The differences D1-D5 and the plausibilities P1-P5 are referred to below as measures for the differences D.

The dimensions D1-D5, P1-P2 for the differences D become an evaluation arrangement 37 provided, which contains several memory cells for evaluation results, wherein the evaluator results are entered as a 1 or a 0. The evaluation arrangement 37 provides a reagent signal determination 38 a reagent signal change dS_Rea available.

The reagent signal definition 38 continues to receive that from the first NOx sensor 15 provided NOx sensor signal NOx_vK_mess and that of the temperature sensor 17 provided temperature signal te_Kat provided and gives to the rating arrangement 37 a previously valid reagent signal S_Rea_alt from.

2 shows a flowchart with a first function block 41 begins, which provides a determination of the reagent signal S_Rea. In a second function block 42 a determination of the NOx raw concentration NOx_vK takes place. In a third functional block 43 is a determination of the temperature te_Kat of the SCR catalyst 16 intended. In a fourth function block 44 the calculation of an efficiency eta_Kat of the SCR catalyst takes place 16 , In a fifth functional block 45 the calculation of the NOx concentration NOx_nK_mod is provided. In a sixth function block 46 the detection of the gas sensor signal S_nK takes place. In a seventh function block 47 a determination of the difference D is provided. In an eighth function block 48 a determination of plausibilities P1-P5 is provided. In a ninth function block 49 the stored differences D1-D5 or the determined plausibilities P1-P5 are stored. In a tenth function block 50 an evaluation of the differences D1-D5 or the plausibilities P1-P5 is provided and in an eleventh function block 51 the determination of the reagent signal change dS_Rea takes place. Then it becomes the first function block 41 jumps back.

The method according to the invention works as follows:
The in the control 20 arranged torque determination 30 determined by the internal combustion engine 10 applied torque Md as a function of at least the predetermined torque setpoint MFa, which is provided for example by an accelerator pedal of a motor vehicle not shown in detail, in which the internal combustion engine 10 is arranged as a drive motor. The torque Md is at least approximately a measure of the load of the internal combustion engine 10 , In the determination of the torque Md may continue the rotation signal n and / or the air detection 12 provided air signal ms_L be considered.

The control 20 indicates the fuel signal m_K determined in particular on the basis of the torque Md to the fuel metering device 25 from. The fuel signal m_K sets, for example, a fuel injection timing and a fuel injection amount. The one in the internal combustion engine 10 burnt fuel leads to the exhaust gas flow ms_abg, which depends on the operating point of the internal combustion engine 10 may contain unwanted more or less high NOx raw concentration NOx_vK.

The largest possible elimination of the NOx raw concentration NOx_vK is at least the SCR catalyst 16 in the exhaust area 13 the internal combustion engine 10 arranged. Next to the SCR catalyst 16 For example, further catalysts and / or a particle filter can be provided. The SCR catalyst 16 supports the reduction reaction of NOx with a reagent that is either in the exhaust region 13 with the reagent dosage 14 metered or optionally provided within the engine. Instead of the reagent, a starting material may be provided. In the case of the reagent ammonia may be provided instead of the ammonia as the starting material, for example, a urea-water solution or, for example, ammonium carbamate. The dosage is determined with the reagent signal S_Rea, which in the illustrated embodiment of the reagent dosage 14 is made available. Alternatively, with an internal engine provision of the reagent, the fuel signal m_K can be modified with the reagent signal S_Rea in such a way that the required reagent is produced internally.

After the start of the method according to the invention takes place in the first functional block 41 the determination of the reagent signal S_Rea in the reagent determination 37 , When determining the dosage, the NOx raw concentration NOx_vK, which is used in the second functional block, is preferably taken into account 42 was determined. In the third function block 43 becomes at least a measure of the temperature te_Kat of the SCR catalyst 16 determined. For temperature detection, for example, the temperature sensor 17 upstream, immediately at or in or downstream of the SCR catalyst 16 can be arranged. It is essential here that the temperature signal te_Kat at least a measure of the operating temperature of the SCR catalyst 16 reflects. Instead of a temperature measurement, an estimate of the catalyst temperature te_Kat can also be provided.

In the fourth function block 44 the determination of at least one measure of the efficiency eta_Kat of the SCR catalyst takes place 16 in particular the operating temperature te_Kat of the SCR catalyst 16 depends. The determination can be made in the NOx concentration determination 32 be provided. The operating temperature essentially determines chen the catalytic activity, which is low at low temperatures, for example below 150 ° C, low, for example, 250 ° C has a maximum and then decreases again to higher operating temperatures. Furthermore, in determining the efficiency eta_Kat, the exhaust gas flow ms_abg can be taken into account, which can already be determined at least approximately from the air signal ms_L alone. Optionally, the fuel signal m_K can additionally be taken into account. From the variables mentioned can be the space velocity RG of the exhaust gas in the SCR catalyst 16 instead of the exhaust gas flow ms_abg in determining the efficiency eta_Kat of the SCR catalyst 16 can be taken into account. Furthermore, in particular the reagent level ReaSp is taken into account. The relationships given and in particular the determination of the reagent level ReaSp, which is the NOx concentration determination 32 is made available, the above-mentioned prior art according to DE 10 2004 031 624 be removed. This document is incorporated by reference.

In the fifth function block 45 becomes at least a measure of the NOx concentration NOx_nK_mod in the exhaust gas after the SCR catalyst 16 calculated. The calculated NOx concentration NOx_nK_mod after the SCR catalyst 16 can be calculated from the calculated raw NOx concentration NOx_vK multiplied by the term (1 - efficiency eta_Kat of the SCR catalyst 16 ).

In the following sixth function block 46 the exhaust gas sensor signal S_nK is detected. The exhaust gas sensor signal S_nK is from the second NOx sensor 18 provided downstream of the SCR catalyst 16 is arranged. In the absence of the first NOx sensor 15 is the second NOx sensor 18 the only NOx sensor. In principle, it is possible for the NOx concentration NOx_nK after the SCR catalyst 16 with a NOx sensor and to detect the reagent slip ms_Rea_nK with a reagent sensor. However, the utilization of a cross-sensitivity of the second NOx sensor is particularly advantageous 18 towards the reagent.

The cross-sensitivity occurs in particular when the measurement is based on the same physical processes. If ammonia is provided as the reagent, in both cases it is a reduction reaction within the second NOx sensor 18 , The exhaust gas sensor signal S_nK changes with a change in the NOx concentration NOx_nK after the SCR catalytic converter 18 and a change in reagent slip ms_Rea_nK. Therefore, the exhaust gas sensor signal S_nK of the second NOx sensor becomes 18 not denoted NOx signal, but generally exhaust gas sensor signal S_nK. The disadvantage here is that it is initially not possible to distinguish between the change in the NOx concentration NOx_nK and a change in the reagent slip ms_Rea_nK.

A reagent slippage ms_Rea_nK may occur when the reagent level is ReaSP in the SCR catalyst 16 exceeds the upper maximum allowable value. A reagent slip ms_Rea_nK can continue with load jumps of the internal combustion engine 10 occur even if at the same time depending on the direction of the load jump, the reagent signal S_Rea is completely reduced to zero or increased to a maximum value.

In the seventh function block 47 is in difference determination 33 the difference D between the exhaust gas sensor signal S_nK and the calculated NOx concentration NOx_nK_mod downstream decreases downstream of the SCR catalyst 16 determined. The determined differences D can be used in the plausibility check 36 be made plausible. The plausibility check 36 provides the plausibilities P, P1-P5. The plausibility check preferably provides a determination, based on the differences D, as to whether an overdose or underdose of the reagent is present. The plausibility check 36 determines, for example, the ratio of metered amount of reagent and in the SCR catalyst 16 reacted reagent amount. It is essential that at least one measure of the determined differences D in the result memory 34 is filed. The measure may accordingly be the determined difference D itself or, for example, the determined plausibility P. The differences D1-D5 or the determined plausibilities P1-P5, which are referred to below as dimensions D1-D5, P1-P5, are determined according to the ninth function block 49 in the result memory 34 saved. For example, it can be provided that the dimensions D1-D5, P1-P5 are respectively stored at specific times, the times being from the time control 35 are fixed, which provides the time signal t.

According to the tenth function block 50 Then the evaluation of the dimensions D1-D5, P1-P5 in the evaluation arrangement 37 which, depending on the evaluation result 0, 1, provides the reagent signal change dS_Rea, that of the reagent signal determination 38 is supplied. At this point, there is the advantage of storing the plausibilities P1-P5 over the storage of the differences D1-D5, since the plausibilities P1-P5 instead of an analog value as logical "0" or "1" can be stored. The reagent signal change dS_Rea changes, for example, pre-control values for the metering of the reagent, which in the exemplary embodiment shown differs from that of FIG NOx raw concentration NOx_vK and / or for example from the temperature signal te_Kat and / or the efficiency eta_Kat of the SCR catalyst 16 depend.

The evaluation arrangement 37 in particular carries out an evaluation of the dimensions D1-D5, P1-P5 on the basis of a preceding reagent signal change dS_Rea. When evaluating the dimensions D1-D5, P1-P5, the previously valid reagent signal ms_Rea_alt is taken into account. If the stored measures are more than logical values, the differences D1-D5 can be compared with a threshold, not shown in greater detail, to ensure that a response to a deviation is only made for larger differences D1-D5.

The following Cases can become result:

Increase of the reagent signal S_Rea:

at a present underdose of the reagent falls the Exhaust gas sensor signal S_nK due to the increased NOx conversion from. On the other hand, if there was already a correct or an overdose, this will increase Exhaust gas sensor signal S_nK due to the reagent slip ms_Rea_nK at.

Lowering the reagent signal S_Rea:

at a present overdose of the reagent drops the exhaust gas sensor signal S_nK due to the reduction of the reagent slip ms_Rea_nK from. On the other hand, if there was a correct or an underdosing, then it increases the exhaust gas sensor signal ms_Rea_nK due to the increasing NOx concentration NOx_nK on.

There both the reagent slip ms_Rea_nK and too low NOx conversion is a deviation between the calculated NOx concentration NOx-nK and the exhaust gas sensor signal In principle, S_nK can not work in the same direction the reagent slip ms_Rea_nK and too low NOx conversion be differentiated.

Accordingly, for the plausibility check in the evaluation order 37 different approaches are provided. First, it is assumed that in a previous step, the dimensions D1-D5, P1-P5 have led to a lowering of the reagent signal S_Rea. The lowering of the reagent signal S_Rea would have the consequence that the differences D become smaller. If, on the other hand, the differences D increase, it can be assumed that the second NOx sensor 18 no reagent slip ms_Rea_nK, but an increased NOx concentration NOx_nK has detected. The reagent signal S_Rea must therefore be increased. The evaluation arrangement 37 However, the reagent signal change dS_Rea is only available when a predetermined number and / or predetermined sequence of evaluation results 0, 1 of the dimensions D1-D5, P1-P5 has been established. The evaluation results 0, 1 are entered as a 1 or a 0 in the individual evaluation arrangement memories. A 1 is entered, for example, for a deviation of the dimensions D1-D5, P1-P5 in the wrong direction. Accordingly, a 0 is entered for a deviation of the dimensions D1-D5, P1-P5 in the right direction. For example, an increase of the reagent signal S_Rea may be provided if three successive evaluation results 0, 1 have the same value as shown in the first row of the evaluation arrangement memories. Furthermore, an increase in the reagent signal mRea may be provided if, for example, four out of five evaluation results 0, 1 have the same value as shown in the second line of the evaluation arrangement memories. Only then does the output of the reagent signal change dS_Rea take place, wherein the dosage is increased, for example, by 20 percent. The basis for decision-making is therefore the number of positive or negative evaluation results 0, 1 and / or the observance of a predetermined sequence of positive or negative evaluation results 0, 1.

Correspondingly, an increase in the reagent signal S_Rea would have the consequence that the differences D in the other direction likewise become smaller. If, on the other hand, the differences D increase, it can be assumed that the second NOx sensor 18 already detects the reagent slip ms_Rea_nK. The evaluation arrangement 37 in response to this, the reagent signal change dS_Rea is only issued when a predetermined number of evaluation results 0, 1 and / or a predetermined sequence of evaluation results 0, 1 have again been determined. It can again be provided that the reduction is not made until, for example, four out of five evaluation results 0, 1 have the same value or, for example, three consecutive evaluation results 0, 1 are the same. Only then does the supply of the reagent signal change dS_Rea take place, wherein the dosage can be lowered, for example, by 5 percent. The basis for the decision is also the number of positive or negative evaluation results 0, 1 and / or the observance of a predetermined order of positive or negative evaluation results 0, 1.

Depending on the previously valid reagent signal S_Rea old can in the evaluation arrangement 37 used given Number of evaluation results 0, 1 in the tenth function block 50 in the event of any necessary adaptation of the reagent signal change dS_Rea according to the eleventh function block 51 be taken into account. Furthermore, it can be provided that an increase of the reagent signal S_Rea resulting from the evaluation result 0, 1 is different from the amount of change in a lowering of the reagent signal S_Rea.

To further ensure proper long-term adaptation, the operating point of the SCR catalyst may also be added 16 and / or the duration of the evaluation. For example, the temperature signal te_Kat, which is at least an indication of the currently available reagent storage capacity of the SCR catalytic converter, can be taken into account 16 gives. Preferably, the efficiency eta_Kat of the SCR catalyst 16 considered.

Preferably, furthermore, the load of the internal combustion engine 10 considered. A suitable signal is for example the torque Md, which is the internal combustion engine 10 has to raise. In particular, it may be provided that at low loads, corresponding to the idling of the internal combustion engine 10 and / or at high loads, a rating of the non-representative in these operating conditions dimensions D1-D5, P1-P5 omitted.

Furthermore, it can be provided that, after a predetermined period of time within which the reagent signal S_Rea has not been changed, the dimensions D1-D5, P1-P5 are no longer used for the evaluation. This avoids a well-adapted reagent dosage 14 Over time, at least slightly unlearned over time.

The signals to be considered are, for example, the time signal t, the torque Md, the temperature signal te_Kat and the reset signal R, which in the exemplary embodiment shown is the result memory 34 to provide. In the simplest case, the signals are used to delete the dimensions D1-D5, P1-P5. Alternatively, the signals of the evaluation arrangement 37 to provide.

Instead of directly influencing the reagent signal S_Rea with the reagent signal change dS_Rea, it is possible to change the metering of the reagent by influencing the reagent fill level in the SCR catalytic converter 16 provided that a control or regulation of the reagent level in the SCR catalyst 16 is provided. Due to a change in the desired reagent level or the reagent actual level results in an indirect influence on the reagent signal S_Rea.

Claims (10)

Method for operating an internal combustion engine ( 10 ), in the exhaust area ( 13 ) at least one SCR catalyst ( 16 ), which is acted upon by a reagent, which is used for NOx conversion in the SCR catalyst ( 16 ) in which at least one measure (NOx_nK_mod) for the downstream of the SCR catalyst ( 16 ) NOx concentration (NOx_nK) and with a cross-sensitive to the reagent NOx sensor ( 18 ) and in which the difference (D) between the calculated measure (NOx_nK_mod) for the NOx concentration (NOx_nK) and the measured sum of the NOx concentration (NOx_nK) and a reagent slip (ms_Rea_nK) is determined, characterized in that the differences (D) are determined in chronological order, that measures (D1-D5, P1-P5) for the differences (D) are stored, that the measures (D1-D5, P1-P5) are evaluated and that a reagent signal ( S_Rea), which determines the dosage of the reagent, in dependence on a predetermined number and / or predetermined order of evaluation results (0, 1) of the dimensions (D1-D5, P1-P5) is determined. Method according to claim 1, characterized in that that the differences (D) are made plausible and that the determined plausibility (P1-P5) as dimensions get saved. Method according to claim 1, characterized in that that at a previous lowering of the dosage of the reagent a different number and / or a different order of evaluation results (0, 1) of the dimensions (D1-D5, P1-P5) for the differences (D) are provided as in a previous one increase the dosage. Method according to claim 1, characterized in that that a bigger increase in the Dosing of the reagent made as a lowering of the dosage becomes. Method according to claim 1, characterized in that that after a change of the reagent signal (S_Rea) the stored dimensions (D1-D5, P1-P5) for the differences (D) deleted become. Method according to Claim 1, characterized in that the dimensions (D1-D5, P1-P5) for the differences (D) at a low or high load (Md) of the internal combustion engine ( 10 ) or the saved dimensions (D1-D5, P1-P5) are deleted become. Method according to claim 1, characterized in that that the stored dimensions (D1-D5, P1-P5) for the differences (D) deleted if for a predetermined period of time no change in the reagent signal (S_Rea) was made. Method according to claim 1, characterized in that the dimensions (D1-D5, P1-P5) for the differences (D) at a temperature of the SCR catalytic converter lying outside a predetermined temperature range ( 16 ) and / or outside a predetermined range of catalyst efficiency (eta_Kat) or the stored measures (D1-D5, P1-P5) are cleared. Device for operating an internal combustion engine ( 10 ), in the exhaust area ( 13 ) at least one SCR catalyst ( 16 ), which is acted upon by a reagent, which is used for NOx conversion in the SCR catalyst ( 16 ), characterized in that at least one prepared for carrying out the method control device ( 20 ) is provided. Apparatus according to claim 9, characterized in that the control unit ( 20 ) a difference determination ( 33 ), which determines the difference (D) between the NOx sensor which is cross-sensitive to the reagent ( 18 ) provided exhaust gas sensor signal (S_nK) and of a NOx concentration determination ( 32 ) calculated downstream NOx concentration (NOx_nK_mod) downstream of the SCR catalyst ( 16 ) determines that a result store ( 34 ) for the dimensions (D1-D5, P1-P5) for the differences (D) and an evaluation arrangement ( 37 ) are provided for the stored dimensions (D1-D5, P1-P5).