DE102009058089A1 - Determining the linear relationship of two signals determined by means of NOx sensors in an SCR exhaust aftertreatment system - Google Patents

Abstract

A first aspect of the invention relates to a measurement method for an SCR exhaust aftertreatment system of a vehicle, wherein the SCR exhaust aftertreatment system includes a first NOx sensor disposed downstream of the SCR catalyst and the urea introduction device, and a second NOx sensor disposed in the SCR catalyst or downstream of the SCR catalytic converter is arranged comprises.
According to the method, a first signal is determined by means of the first NOx sensor. This first signal may also be a delayed signal. In addition, a second signal is determined by means of a second NOx sensor. Based on the first and the second signal, a linearity indication (for example, a correlation coefficient) is determined, which is a measure of the linear relationship between the two signals.
The linearity specification can be used to differentiate between NOx slip or NH3 slip.

Description

The invention relates to SCR exhaust gas aftertreatment systems (SCR - selective catalytic reduction).

One method for reducing nitrogen oxide emissions in automotive diesel engines is the so-called SCR process, ie the selective catalytic reduction of stock oxides. Ammonia (NH3) is used to complete the reaction. The products of the reaction are water (H ₂ O) and nitrogen (N ₂ ). The ammonia used for the SCR reaction is introduced in the form of an aqueous urea solution (typically 32.5 percent urea) before the SCR catalyst in the exhaust line, for example by means of metering pump or injector injected. From the urea-water solution formed by a hydrolysis reaction ammonia and CO ₂ . The ammonia reacts with the nitrogen oxides in the exhaust gas in a special SCR catalytic converter.

Depending on their size, SCR catalysts can only store a certain amount of NH3. The urea dosage should correspond on average to the urea necessary to reduce nitrogen oxide emissions. It should be noted that the engine nitrogen oxide emission is dependent on the respective speed and the respective torque of the engine, so that the urea dosage should be adjusted accordingly. If the urea dosage is too low, the efficiency of the stock oxide reduction deteriorates. This is also referred to as NOx slippage, i. H. the SCR catalyst lets too much nitrogen oxide pass through. However, if the urea dosage is too high, the resulting ammonia may not react with the nitrogen oxides in sufficient quantity due to the ammonia over-supply. In this case, ammonia can enter the environment, which can lead to an odor nuisance. In this case one also speaks of NH3-slip.

1 shows essential components of a typical SCR exhaust aftertreatment system 5 for use in a motor vehicle. The SCR exhaust aftertreatment system 5 includes a first NOx sensor 1 Abgasstromaufwarts a urea dosing device 3 and an SCR catalyst 4 , The first NOx sensor 1 is used to measure engine-side nitrogen oxide emissions. Downstream exhaust gas of the SCR catalyst 4 is a second NOx sensor 2 arranged. This can also be done in the SCR catalyst 4 himself (not shown). The second NOx sensor 2 measures both NOx slippage and NH3 slippage.

NOx sensors currently used in vehicle applications can not differentiate between NOx and NH3; Namely, NOx sensors have a so-called NH3 cross-sensitivity. Therefore it is not possible from the behind the SCR catalyst 4 measured sensor signal to distinguish directly between NH3 and NOx slip. This disadvantage represents a significant deficit for the correct regulation of the urea dosage.

From the publication WO 2009/036780 A1 is an SCR catalyst with a NH3 level monitoring known that detects the NH3 level using two NOx sensors in two different ways, with the respective errors, for example, due to the ammonia cross-sensitivity of the second sensor, at least partially compensate.

The object of the invention is to provide a measuring method which, in spite of the NH3 cross-sensitivity of the second NOx sensor, permits a differentiation between NOx slip and NH3 slip, that is to say differentiation between NOx slip and NH3 slip. H. which makes it possible to differentiate whether the substance detected by the second NOx sensor is NOx or NH3. Furthermore, the invention aims at specifying a corresponding device. Furthermore, the invention is directed to the disclosure of a control method for controlling the urea dosage using this measuring method.

The object is solved by the features of the independent claims.

• – einen SCR-Katalysator,
• – eine Harnstoff-Einbringungsvorrichtung,
• – einen ersten NOx-Sensor, welcher abgasstromaufwärts des SCR-Katalysators und der Harnstoff-Einbringungsvorrichtung angeordnet ist, und
• – einen zweiten NOx-Sensor, welcher in dem SCR-Katalysator oder abgasstromabwärts des SCR-Katalysators angeordnet ist.

A first aspect of the invention relates to a measurement method for an SCR exhaust aftertreatment system of a vehicle. Such an SCR exhaust aftertreatment system includes:

• An SCR catalyst,
• A urea introduction device,
• A first NOx sensor arranged upstream of the exhaust gas upstream of the SCR catalyst and the urea introduction device, and
• A second NOx sensor disposed in the SCR catalyst or downstream of the SCR catalyst.

According to the method, a first signal is determined by means of the first NOx sensor. This first signal may also be a delayed signal; this will be discussed in more detail later in the description. In addition, a second signal is determined by means of a second NOx sensor. Based on the first and the second signal, a linearity indication is determined, which is a measure of the linear relationship between the two signals.

The linearity specification can be used to differentiate between NOx slip or NH3 slip. This is explained in more detail below: The first signal determined by means of the first NOx sensor indicates the engine NOx emissions before the urea dosing. The second signal determined by the second NOx sensor indicates both NOx emissions after the SCR catalyst and NH3 emissions after the SCR catalyst. For the NOx emissions measured by the second NOx sensor, there is a highly linear relationship with the first signal measured by the first NOx sensor, i. H. In this case, there is a high correlation between the first and the second signal. The correlation can be increased if the first and the second signal are synchronized with one another by a corresponding delay of the sensor signal of the first sensor. On the other hand, if NH3 is mainly present at the second NOx sensor, the correlation between the two signals is low.

Preferably, a correlation coefficient between the two signals is calculated as a linearity specification. By determining the correlation coefficient of the two sensor signals, a dimensionless measure of the degree of linear relationship between the two signals can be obtained. In the following, it is assumed that the magnitude of the correlation coefficient can become at most +1, but the correlation coefficient does not necessarily have to be normalized to a maximum of 1 in the sense of the invention. At a value of the correlation coefficient of +1, there is a completely positive linear relationship between the two signals. If the correlation coefficient is 0, the two features do not depend linearly on each other.

Therefore, with a correlation coefficient close to +1, NOx slip and a correlation coefficient close to 0 can be judged as NH3-shudder. Thus, in spite of the NH3 cross-sensitivity of the second NOx sensor, the method makes it possible to distinguish between NOx slip and NH3 slip.

Preferably, to obtain the first signal, the sensor signal of the first NOx sensor is delayed in time by the time lag of the exhaust gas flow (ie, NOx emissions) between the position of the first NOx sensor and the position of the second NOx sensor at least partially compensate. This time delay corresponds approximately to a dead time and would falsify the result if the time delay is not taken into account in the calculation. Due to the temporal synchronization, the degree of linearity is therefore further increased in the case of NOx slippage. In order to substantially completely compensate for the influence of the time delay, the time delay of the sensor signal of the first NOx sensor is preferably selected such that it approximately corresponds to the time delay of the exhaust gas flow between the position of the first NOx sensor and the position of the second NOx sensor equivalent.

The time delay can also be inherently realized by taking into account a time offset in the determination of the correlation coefficient.

The method preferably provides that this time delay is calculated as a function of the respective exhaust gas volume flow. For example, it can be provided that the method calculates the time delay of the exhaust gas flow (ie the NOx emissions) between the position of the first NOx sensor and the position of the second NOx sensor from the exhaust gas volume flow and the volume of the exhaust system between the two NOx sensors , In accordance with the respectively calculated time delay, the sensor signal of the first NOx sensor can be delayed in time.

Another aspect of the invention relates to a control method for controlling urea introduction into an SCR exhaust aftertreatment system. Here, the urea introduction is controlled by utilizing the above-described linearity. For example, the linearity specification can be used as the controlled variable. However, the regulation is preferably carried out on a sensor signal of the second sensor that is plausibilized or evaluated with the linearity specification (eg the correlation factor).

Another aspect of the invention relates to an apparatus for determining a linearity indication for a previously discussed SCR exhaust aftertreatment system having two NOx sensors in a vehicle. In this case, the device comprises first means for determining a linearity indication (for example of a correlation coefficient) based on a first signal and a second signal which is a measure for the linear relationship between both signals. In this case, the first signal has been determined by means of the first NOx sensor and the second signal by means of the first NOx sensor.

The above explanations of the measuring method according to the invention can also be transferred to the apparatus for determining the linearity specification in the same way.

Preferably, the device comprises a delay element for delaying a sensor signal of the first NOx sensor, wherein the time-delayed signal is used as the first signal from the means for determining the linearity indication. The time delay preferably corresponds approximately to the time delay of the exhaust gas flow between the position of the first NOx sensor and the position of the second NOx sensor.

Furthermore, the device preferably comprises means for determining the time delay, for example as a function of the respective exhaust gas flow, as has been discussed above with reference to the method according to the invention.

Another aspect of the invention relates to an SCR exhaust aftertreatment system for a vehicle. This includes not only the above-discussed components of a conventional SCR exhaust aftertreatment system with two NOx sensors but also the above-described device for determining a linearity specification, and in particular the control described above.

An embodiment of the invention will be described below with reference to the drawings. In these show:

1 a conventional SCR exhaust aftertreatment system 5 ;

2 an embodiment for determining the correlation coefficient;

3 a first exemplary course of the time-shifted sensor signal 18 (in ppm) of the first NOx sensor 1 and the sensor signal 12 (in ppm) of the second NOx sensor 2 and the course of the resulting correlation coefficient signal 24 ;

4 a second exemplary course of the time-shifted sensor signal 18 (in ppm) of the first NOx sensor 1 and the sensor signal 12 (in ppm) of the second NOx sensor 2 and the course of the resulting correlation coefficient signal 24 ; and

5 an embodiment of NH3 level control.

1 shows a conventional SCR exhaust aftertreatment system 5 and has already been discussed above in the introduction to the description. Such an SCR exhaust aftertreatment system 5 has the disadvantage that due to the NH3 cross-sensitivity of the second NOx sensor 2 can not be easily distinguished between the presence of NH3 slip or NOx slippage. As discussed above, in the case of NOx slippage, there is a linear relationship between the measured NOx emissions at the second NOx sensor 2 and the time synchronized signal of the first NOx sensor 1 , On the other hand, in the case of NH3 slip, NH3 on the second NOx sensor 2 measured, the correlation between the two signals is low.

The invention exploits this linear relationship between the signals. Preferably, this is the correlation coefficient of the signals of the two NOx sensors 1 and 2 calculated. This indicates the degree of linear relationship between the two signals. A value of +1 gives a completely positive linear relationship between the signals. If the correlation coefficient is 0, the two signals are not linearly dependent on each other. Therefore, with a correlation coefficient close to 1, NOx trapping and a correlation coefficient close to 0 can be judged to be NH 3 -shoot. The disadvantage of the NH3 cross-sensitivity of the second NOx sensor 2 is cleared out.

2 shows an exemplary embodiment of a block 10 for determining the correlation coefficient. To determine the correlation coefficient in the block 10 becomes the sensor signal 11 of the first sensor 3 and the sensor signal 12 of the second sensor 2 used. In the block 13 is derived from an exhaust gas volume flow signal (in m ³ / h) and the volume of the exhaust system between the two NOx sensors 1 and 2 the time delay 15 NOx emissions from the position of the NOx sensor 3 to the position of the NOx sensor 2 calculated. According to this calculated time delay 15 becomes the sensor signal 11 in the block 16 delayed in time. In the block 17 is determined by status signals 21 and 22 of the first NOx sensor 1 or the second NOx sensor 2 as well as the delayed sensor signal 18 of the first NOx sensor 1 and the sensor signal 12 of the second NOx sensor 2 determines if the NOx sensors 1 and 2 deliver valid readings. The determined status becomes a validity signal 19 output.

In the block 20 the actual calculation of a correlation coefficient signal takes place 23 between the delayed sensor signal 18 of the first NOx sensor 1 and the sensor signal 12 of the second NOx sensor 2 , Further, an averaged correlation coefficient signal 24 and a status signal 25 determines which indicates whether the correlation coefficient signal 23 is valid.

The following MatLab program code describes the one in the block 20 running algorithm in detail.

The one in block 20 Calculation algorithm performed calculates for a freely selectable number (referred to as num_correlation in the source code) on measured values of the input signals 18 (referred to in the source code as x) and 12 (referred to in the source code as y) "online", ie without storing the input signals between the correlation coefficients 23 (referred to in the source code as fac_correlation).

In the first partial block initialization of the calculation, all calculation variables are assigned the initial values at the start of the calculation.

In the sub-block Restart of the calculation, a reset of the calculation quantities takes place analogously to the initialization. H. a restart of the calculation if a defined number (num_reset) of successive measured values is invalid.

In the sub-block calculation, the valid measured values of the input signals are calculated 18 and 12 (in the source code x or y) the running average values and bn, the variances sn and rn and the covariance cn calculated. From this, in turn, the correlation coefficient fac_correlation can be determined according to a freely selectable number of measured values num_correlation. Then the calculation starts again. From the individual values of the correlation coefficient, a sliding average value fac_correlation_avrg (corresponding to the mean value signal 24 in 2 ) about the last num_avrg calculation results.

Measured values of the input signals 18 (x in source code) and 12 (y in source code) with an associated status signal 19 (valid in the source code) with the value "not valid" (ie valid = 0) are not taken into account for the calculation in the sub-block calculation.

For valid values for the signal 23 (in the source code fac_correlation) and 24 (in the source code fac_correlation_avrg) becomes the status signal 25 (in source code fac_correlation_valid) set to 1.

With the concept described above, it is possible to continuously distinguish between NH3 and NOx slip without additional measurement technology.

3 shows an exemplary course 30 the time shifted sensor signal 18 (in ppm - parts per million) of the first NOx sensor 1 before the SCR catalyst and the course 31 of the sensor signal 12 (in ppm) of the second NOx sensor 2 after the SCR catalyst and the course 33 of the resulting correlation coefficient (shown here is the moving average waveform 24 the correlation coefficient multiplied by the factor 1000). In the example of 3 For example, the averaged correlation coefficient has a small amount close to 0 (up to a maximum of 0.35), ie, the correlation between both sensor signals 18 and 12 is low. This small value of the correlation coefficient indicates the presence of NH3 slip.

4 shows a further exemplary course 30 ' the time shifted sensor signal 18 (in ppm) of the first NOx sensor 1 before the SCR catalyst and the course 31 ' of the sensor signal 12 (in ppm) of the second NOx sensor 2 after the SCR catalyst and the course 33 ' of the resulting correlation coefficient (shown here is the moving average waveform 24 the correlation coefficient multiplied by the factor 1000). How out 4 can be seen, corresponds to the course 30 ' approximately the course 31 ' multiplied by a linearity factor. Therefore, here the correlation coefficient has a high amount close to 1, ie, the correlation between both sensor signals 18 and 12 is high. This high value of the correlation coefficient indicates the presence of NOx slip.

With the aid of the above-described concept, it is possible on the sensor signal of the second NOx sensor 2 based control functions can be improved, since by means of the inventive concept can be reliably distinguished between NOx-slip and NH3-slip. Therefore it is possible to prevent on the one hand NH3-reduced dosage and thus increased NOx-emissions and on the other hand to prevent NH3-overdosing and thus to minimize the urea consumption as well as the ammonia slip.

• – die Berechnung des Wirkungsgrads des SCR-Katalysators,
• – die Diagnose von NOx-Schlupf,
• – die Steuerung und Regelung der Harnstoff-Dosierung,
• – die Modellierung des SCR-Katalysators,
• – die Adaption der Harnstoff-Dosierung, beispielsweise infolge von Streuungen, Alterung und Fehlern im Abgasnachbehandlungssystem.

The distinction between NH3 and NOx through the correlation makes it possible to improve various applications in exhaust aftertreatment. These applications include, for example:

• The calculation of the efficiency of the SCR catalyst,
• - the diagnosis of NOx slip,
• The control and regulation of the urea dosage,
• The modeling of the SCR catalyst,
• - The adaptation of the urea dosage, for example due to scattering, aging and errors in the exhaust aftertreatment system.

5 shows an embodiment of an NH3 level control (or in other words: control of urea dosing) with NOx slip diagnosis and SCR efficiency determination. Here, the in 5 illustrated arrangement the correlation coefficients according to the invention for different applications. In the block 10 will - as already in connection with 2 explained - based on the sensor signal 11 of the first NOx sensor 1 and the sensor signal 12 of the second NOx sensor 2 determining a correlation coefficient signal as a measure of the linear relationship between the time-synchronized sensor signals; for example, the correlation coefficient signal 23 or the correlation coefficient signal 24 out 2 , It should be noted that the number of moving average 24 used values can also be selected to 1.

The correlation coefficient signal 23 or 24 will be in a block 40 which uses the SCR catalyst 4 modeled to the actual NH3 level 42 of the SCR catalyst 4 to determine. The block 40 also takes an actual dosing signal 41 contrary, which indicates the actual dosage with urea. In addition, the block takes 40 the sensor signals 11 and 12 opposite. For determining the actual NH3 level 42 becomes the sensor signal 12 via the correlation coefficient signal 23 or 24 rated. When determining the actual NH3 level 42 namely, on the basis of the correlation coefficient signal 23 or 24 distinguished whether the sensor signal 21 NH3 or NOx indicates. If the correlation coefficient signal 23 or 24 indicates that the sensor signal 12 indicates a certain amount of NH3 (in the case of NH3 slip), which becomes the sensor signal 12 corresponding NH3 amount subtracted from the current NH3 level.

The determined NH3 actual level 42 comes with a nominal NH3 level 43 compared and the difference 44 between NH3 actual level 42 and NH3 target level 43 in the block 45 evaluated to a target dosage 46 to determine urea. Depending on the target dosage 46 Urea is via the urea dosing device 3 (S. 1 ) into the SCR catalyst 4 brought in. The actual actual dosage 41 urea may vary from the target dosage 46 differ; For example, because the metering valve currently promoted high actual dosage 41 can not inject. block 47 sets the controlled system between nominal dosage 46 and determined or estimated actual dosage 41 wherein the dosing device 3 (S. 1 ) Part of block 47 is.

The control loop with the blocks 45 . 47 and 40 is used to quickly control the NH3 filling level. The slow adaptation of the NH3 filling level takes place via the adaptation of the nominal NH3 filling level 43 , The correlation coefficient signal according to the invention 23 or 24 will be in block 48 for determining the nominal NH3 level 43 used. Basically, block determines 48 the nominal NH3 level 43 based on one or more state variables of the exhaust system 49 (For example, the temperature of the exhaust gas and the size of the exhaust gas mass flow) and based on the NOx signal 11 of the first NOx sensor 1 , Thus, at a low temperature (eg 150 ° C), the nominal NH3 level is 43 high, because at low temperature, the NH3 storage capacity of the SCR catalyst 4 is high. Conversely, at high temperature (eg 400 ° C) the nominal NH3 level is 43 low, since in this case the storage capacity is low. In addition, in block 49 the NH3 target level to the currently existing engine emission, namely the NOx signal 11 of the first NOx sensor 1 , adapted: At low emission the target level is 43 less than at high emission. The nominal NH3 level 43 also becomes dependent on the correlation coefficient signal 23 or 24 adjusted: If the correlation coefficient signal 23 or 24 indicates that NH3 slip prevails, should the NH3 target level 43 be reduced, and if the correlation coefficient signal 23 or 24 indicates that NOx slip prevails, the target NH3 level should be 43 increase.

Besides the dosing control described above, the correlation coefficient signal becomes 23 or 24 still in block 50 used for NOx slip diagnosis. The NOx slip diagnosis 50 evaluated with the correlation coefficient signal 23 or 24 the sensor signal 12 indicates whether this is the indication of NOx or NH3 and gives a NOx slip signal 51 out. For example, the NOx slip signal becomes 51 the NOx slip diagnosis 51 then marked valid (by an additional unillustrated bit) when the correlation coefficient signal 23 or 24 indicates that there is actually NOx slippage. Based on the NOx slip signal 51 If a certain threshold value is exceeded, a warning can be output to the driver.

Further, the correlation coefficient signal becomes 23 or 24 in block 52 for determining the efficiency 53 of the SCR catalyst 4 used. The efficiency 53 For example, as the ratio between the NOx signal 11 of the first NOx sensor 1 and the NOx signal 12 of the second NOx sensor 2 (in the case of NOx slippage). This so determined efficiency is dependent on the correlation coefficient signal 23 or 24 marked valid (by an additional bit not shown) when the correlation coefficient signal 23 or 24 indicates that there is actually NOx slippage.

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

• WO 2009/036780 A1 [0006]

Claims (15)

Measurement method for an SCR exhaust aftertreatment system ( 5 ) of a vehicle, wherein the SCR exhaust aftertreatment system - an SCR catalyst ( 4 ), - a urea introduction device ( 3 ), - a first NOx sensor ( 1 ), which downstream of the SCR catalyst ( 4 ) and the urea introduction device ( 3 ), and - a second NOx sensor ( 2 ), which in the SCR catalyst ( 4 ) or downstream of the SCR catalyst ( 4 ), comprising the steps of: - determining a first signal ( 18 ) by means of the first NOx sensor ( 1 ); Determining a second signal ( 12 ) by means of the second NOx sensor ( 2 ); and - determining a linearity indication ( 23 . 24 ) based on the first signal ( 18 ) and the second signal ( 12 ), which is a measure of the linear relationship between the two signals. Method according to claim 1, wherein as linearity indication a correlation coefficient ( 23 . 24 ) of the two signals is determined. Method according to one of the preceding claims, wherein the first signal ( 18 ) to a sensor signal ( 11 ) of the first NOx sensor ( 1 ) is delayed signal. The method of claim 3, wherein the time delay is approximately equal to the time delay of the exhaust gas flow between the position of the first NOx sensor (16). 1 ) and the position of the second NOx sensor ( 2 ) corresponds. The method of claim 4, further comprising the step of: - Determining the time delay in dependence of the exhaust gas flow rate. Method according to one of the preceding claims, wherein the sensor signal of the second sensor ( 2 ) is evaluated as a function of the linearity specification as essentially NOx specification or as essentially NH3 specification. The method of claim 2, wherein between NOx-slip or NH3-slip depending on the correlation coefficient ( 23 . 24 ), where for a correlation coefficient close to 1 NOx slip is assumed and for a correlation coefficient close to 0 NH3 slip is assumed. Control method for regulating urea introduction in an SCR exhaust aftertreatment system ( 5 ), the urea introduction taking advantage of the Linearitätsangabe determined according to any one of the preceding claims ( 23 . 24 ) is regulated. A control method according to claim 8, wherein the sensor signal of the second sensor ( 2 ) by means of the linearity indication ( 23 . 24 ) Is evaluated. A control method according to any one of claims 8-9, wherein the linearity indication ( 23 . 24 ) for determining the actual NH3 level ( 42 ) and / or for determining the nominal NH3 level ( 48 ) is used. Device for determining a linearity specification for an SCR exhaust aftertreatment system ( 5 ) of a vehicle, the exhaust aftertreatment system comprising - an SCR catalyst ( 4 ), - a urea introduction device ( 3 ), - a first NOx sensor ( 1 ), which downstream of the SCR catalyst ( 4 ) and the urea introduction device ( 3 ), and - a second NOx sensor ( 2 ), which in the SCR catalyst ( 4 ) or downstream of the SCR catalyst ( 4 ), the device comprising: - first means ( 20 ) for determining a linearity specification ( 23 . 24 ) based on a first signal ( 18 ) and a second signal ( 12 ), which is a measure of the linear relationship between the two signals, wherein the first signal by means of the first NOx sensor ( 1 ) and the second signal by means of the second NOx sensor ( 2 ) has been determined. Apparatus according to claim 11, wherein the first means ( 20 ) for determining a correlation coefficient ( 23 . 24 ) of the two signals are configured. Device according to one of claims 11 to 12, further comprising - a delay element ( 15 ) for determining the first signal ( 18 ) by delaying a sensor signal ( 11 ) of the first NOx sensor ( 1 ), wherein the time delay approximately the time lag of the exhaust gas flow between the position of the first NOx sensor ( 1 ) and the position of the second NOx sensor ( 2 ) corresponds. SCR exhaust aftertreatment system ( 5 ) for a vehicle, comprising - an SCR catalyst ( 4 ), - a urea introduction device ( 3 ), - a first NOx sensor ( 1 ), which is arranged downstream of the SCR catalytic converter and the urea introduction device, - a second NOx sensor ( 2 ), which is arranged in the SCR catalytic converter or downstream of the SCR catalytic converter, and 20 ) for determining a linearity specification according to one of claims 11 to 13. SCR exhaust aftertreatment system ( 5 ) according to claim 14, comprising a control circuit for controlling urea introduction, wherein - the control circuit comprises the device for determining the linearity indication and - the control circuit is arranged, the urea introduction using the linearity indication ( 23 . 24 ).

Images