CN108374712B

CN108374712B - Method for fault detection in an SCR system by means of ammonia slip

Info

Publication number: CN108374712B
Application number: CN201810085539.XA
Authority: CN
Inventors: A.巴施托雷亚拉; T.普菲斯特
Original assignee: Robert Bosch GmbH
Current assignee: Robert Bosch GmbH
Priority date: 2017-01-30
Filing date: 2018-01-29
Publication date: 2021-08-27
Anticipated expiration: 2038-01-29
Also published as: KR20180089300A; CN108374712A; KR102406226B1; DE102017201400A1

Abstract

The invention relates to a method for fault detection in an SCR system of an internal combustion engine in a motor vehicle, having two SCR catalysts and at least two nitrogen oxide sensors, wherein a first nitrogen oxide sensor is arranged between the two SCR catalysts and a second nitrogen oxide sensor is arranged downstream of the two SCR catalysts. The method comprises the following steps: a continuous detection of the signal at the first nox sensor is carried out and the ammonia mass at the first nox sensor is determined therefrom. The overdosing is then carried out by increasing the mass of the reducing agent dosed, the converted ammonia mass of the first SCR catalyst being higher than the ammonia mass required for the reduction of nitrogen oxides, until the ammonia loading level on the first SCR catalyst is higher than the maximum ammonia loading level of the first SCR catalyst.

Description

Method for fault detection in an SCR system by means of ammonia slip

Technical Field

The invention relates to a method for fault detection in an SCR system having two SCR catalysts by means of ammonia slip (Ammoniak-Schlupf). Furthermore, the invention relates to a computer program which executes each step of the method when the computer program runs on a computer, and to a machine-readable storage medium which stores the computer program. Finally, the invention relates to an electronic control unit which is set up to carry out the method.

Background

The technology widely used today for reducing nitrogen oxides (NOx) in the exhaust gases of combustion motors in motor vehicles is selective catalytic reduction (Selective-Catalytic-RAcquisition) (SCR). In an SCR system, passing a dosing module as a reducing agent solution upstream of at least one SCR catalyst will also be commercially useful

Whereas the known urea-water solution is injected into the exhaust system. Produced from the urea-water solutionAmmonia reacts with the nitrogen oxides in the selective catalytic reduction over the SCR catalyst to produce elemental nitrogen.

Due to the adoption of stricter emission regulations, a plurality of SCR catalysts are used which have an influence on the same exhaust gas. In the event of an insufficient efficiency of the SCR catalytic converter for reducing the nitrogen oxide emissions in the exhaust system, it is provided that the fault detection is carried out by means of a vehicle-specific test method (usually implemented in an electronic control unit). For this reason, continuous monitoring is carried out during normal operation of the vehicle. For a common checking method, at least one nitrogen sensor is used upstream of the SCR catalyst and at least one nitrogen sensor is used downstream of the SCR catalyst. For a single SCR catalyst, two nitrogen oxide sensors are sufficient for calculating the efficiency of the SCR system and for monitoring the nitrogen oxide emissions at the same time.

In known monitoring strategies, the ammonia storage capacity of the SCR catalyst is known and taken into account as a characteristic for the SCR catalyst, for example, as a result of ageing or damage (fehlfank). In this case, the SCR catalytic converter is first filled by an over-stoichiometric dosing of the reducing agent solution, so-called overdosing, up to a maximum ammonia filling level. If the maximum ammonia loading level is reached, ammonia can no longer be stored by the SCR catalyst and ammonia slip occurs, for which pure ammonia flows out downstream of the SCR catalyst. The nitrogen oxide sensor has a lateral sensitivity (Querempfindlichkeit) to ammonia, so that the ammonia slip can be known indirectly and used as a defined starting point. The dosing of the reducing agent solution is then reduced (herunterfahren) to a small scale or cut off, so that the stored ammonia is reduced (abauen) by the selective catalytic reduction. The efficiency of the SCR catalyst can now be known, for example, for inferring its storage capacity therefrom. In principle, this method can be used with two SCR catalysts, but the dosing of the reducing agent solvent is very strongly suppressed by the SCR catalyst arranged upstream.

DE 102012202671 a1 discloses a method for diagnosing two SCR catalysts. The aging state of the second SCR catalyst is known from the difference between the sensor signal of the first sensor and the sensor signal of the second sensor during the purging of the two SCR catalysts.

DE 102012220151 a1 relates to a method for checking two SCR catalytic converters. Here, a first SCR catalyst arranged upstream is examined by: the nitrogen oxide concentration is influenced by a change in operating variables of the combustion motor and/or of the system. An insufficient storage capacity of the first SCR catalyst is identified if a signal of a sensor arranged between the SCR catalysts is subject to suppression. Similarly, a second SCR catalyst arranged downstream is examined by: the functional capability of the first SCR catalyst is first verified (verizizen) and the nox concentration is subsequently influenced again by a change in the operating variables of the combustion motor and/or the system. An insufficient storage capacity of the second SCR catalyst is identified if a signal of a sensor arranged downstream of the second SCR catalyst is subject to suppression.

Disclosure of Invention

The method relates to an SCR system of a combustion motor in a motor vehicle. The SCR system has two SCR catalysts arranged one behind the other in a common exhaust system. The exhaust gas first passes through the first SCR catalyst and is subsequently passed (weiterleiten) to the second SCR catalyst, so that both SCR catalysts have an influence on the exhaust gas. Furthermore, the SCR system has at least two nox sensors, which are likewise arranged in this exhaust system. A first nitrogen oxide sensor is arranged between the two SCR catalysts and is able to measure there the sum of the nitrogen oxide concentration and the ammonia concentration after exhaust gas aftertreatment by the first SCR catalyst. The second nitrogen oxide sensor is arranged downstream of the two SCR catalysts and is able to measure there the sum of the nitrogen oxide concentration and the ammonia concentration after exhaust gas aftertreatment by the two SCR catalysts. The ammonia concentration coincides with ammonia slip that occurs when the ammonia loading level exceeds a maximum ammonia loading level for the corresponding SCR catalyst. The components mentioned can be connected to a common electronic control unit, which controls these components. In this method, the signal of the first nitrogen oxide sensor is continuously detected and the ammonia mass on the first nitrogen oxide sensor is determined therefrom.

In the case of an active diagnosis, the quality of the reducing agent dosed is increased, in other words an overdosing is carried out. The ammonia mass of the first SCR catalyst converted from the dosed reducing agent mass exceeds the ammonia mass required for reduction on the first SCR catalyst. Thus, a mass of ammonia is provided on the first SCR catalyst that is greater than a mass of ammonia consumed by the SCR. As a result, the ammonia loading level on the first SCR catalyst increases. If the maximum ammonia loading level is exceeded, ammonia passes through the first SCR catalyst unutilized and ammonia slip occurs over the first SCR catalyst. The presence of an ammonia slip is detected from the signal of the first nitrogen oxide sensor when a first check is made as to whether an ammonia slip is present. If ammonia slip is finally detected at the first SCR catalyst, a first fault detection of the second SCR catalyst is carried out by means of the evaluation criterion explained below.

Optionally, after the signal of the first nox sensor is continuously detected and the ammonia mass at the first nox sensor is known therefrom, a passive diagnosis is carried out for a normal dosing before an overdosing, i.e. a dosing whose parameters are not changed by this method. In the passive diagnosis, a second check is also carried out as to whether ammonia slip is present on the basis of the signal of the first nitrogen oxide sensor. If ammonia slip on the first SCR catalyst is finally detected, a second fault detection of the second SCR catalyst is carried out by means of the evaluation criterion explained below. If a fault is output during the passive diagnosis when the second fault detection is performed, the overdosing is released and an active method (ablaufen) is carried out.

If, during the passive diagnosis, no fault is output during the second fault detection, the overdosing and the following steps of the active diagnosis described above are prevented. It is preferable in this case to be able to start Ghost counting (Ghost-Counter). The ghost count simulates the feasibility of the active diagnosis and determines the time at which a fault has been identified by the active diagnosis. If the active diagnosis has ended In the current driving cycle, the IUMPR (In-Use-Performance Ratio) is increased. If at least one of the operating parameters of the internal combustion engine does not indicate this, the ghost count is stopped and/or reset. The operating parameters studied in this respect include the temperature level and the temperature gradient of the second SCR catalyst, the mass flow of the reducing agent solution through the second SCR catalyst and the switch-on and switch-off conditions. The ghost count thus indicates with what probability a fault can be identified by: it places the travel period with the performed diagnosis in a proportional relationship with all travel periods. Thereby complying with the legal provisions associated therewith. In particular, it must not fall below the threshold (festlegen) specified here for the IUMPR.

As a possible evaluation criterion, it can be provided that the difference between the ammonia mass at the first nox sensor and the ammonia mass at the second nox sensor is determined. This difference is then integrated over a time interval. If the integrated difference is below a third threshold value at the decision time, a malfunction on the second SCR catalyst is identified, since sufficient ammonia cannot be stored by the second SCR catalyst. Otherwise no fault on the second SCR catalyst is identified. The integral of the difference can optionally be normalized to a maximum value that can be specified and is dependent on the temperature.

Alternatively, the amount of the difference can be used for evaluation instead of the difference itself. The fault detection takes place as described above. The evaluation of the poor quantity is advantageous in particular for the case in which the ammonia loading level of the second SCR catalyst is very close to the maximum ammonia loading level for the SCR catalyst or even has reached this maximum ammonia loading level. In this case, the smallest temperature change in the second SCR catalyst serves to be able to alternately release ammonia and then store it again. The differences thus detected have a changing sign which enables a reliable evaluation by the formation of the quantities.

As a further possible evaluation criterion, provision can be made for a correlation coefficient to be determined from the signal of the first nox sensor and the signal of the second nox sensor. If the correlation coefficient is above a fourth threshold, a fault on the second SCR catalyst is identified because the two signals are too similar to possibly store enough ammonia. Otherwise no fault on the second SCR catalyst is identified. The previously described phenomenon, in which ammonia can be released alternately and then stored again, can also be evaluated here. The signal of the first nox sensor and the signal of the second nox sensor are then differentiated from one another, so that the correlation coefficient decreases.

The two evaluation criteria accordingly generate an appropriate fault recognition for the second SCR catalyst per se.

Optionally, an additional third nitrogen oxide sensor can be arranged upstream of the first SCR catalyst, which third nitrogen oxide sensor can there measure the nitrogen oxide concentration before the exhaust gas treatment by the SCR catalyst. The third nitrogen oxide sensor is particularly advantageous when detecting ammonia slip at the first SCR catalytic converter.

In the presence of the third nitrogen oxide sensor upstream of the first SCR catalyst, it can according to one aspect be checked whether there is ammonia slip on the first SCR catalyst by: the signal of the nitrogen oxide sensor immediately downstream of the first SCR catalyst, and therefore of the first nitrogen oxide sensor, is compared with the signal of the nitrogen oxide sensor arranged upstream of the first SCR catalyst, and therefore of the third nitrogen oxide sensor. The ammonia slip is detected if the signal of the nitrogen oxide sensor immediately downstream of the first SCR catalyst is continuously greater than the signal of the nitrogen oxide sensor arranged upstream of the first SCR catalyst since a defined time. This means that an ammonia slip at the first SCR catalyst is detected if the signal of the first nitrogen oxide sensor is continuously greater than the signal of the third nitrogen oxide sensor since the specified time. In this checking method, no additional assumptions or calculations are required. Since the nitrogen oxides are only reduced between the third nitrogen oxide sensor and the first nitrogen oxide sensor, but it is not possible to add (hinzukommen) additional nitrogen oxides, an increase in the signal of the nitrogen oxide sensor immediately downstream of the catalyst can infer ammonia which escapes as a result of ammonia. According to another aspect, a check is made as to whether there is ammonia slip on the first SCR catalyst by: the signal of the nitrogen oxide sensor immediately downstream of the first SCR catalyst, and therefore of the first nitrogen oxide sensor, is compared with the expected signal for this nitrogen oxide sensor. The expected signal relates to the nitrogen oxide concentration at a nitrogen oxide sensor downstream of the first SCR catalyst, taking into account the occurrence of ammonia slip. It should thus be ensured that nitrogen oxides which are not converted into elemental nitrogen by the first SCR catalyst are falsely detected as ammonia slip. The ammonia slip is detected if the signal of the nitrogen oxide sensor immediately downstream of the first SCR catalyst is equally large or greater than the expected signal for this nitrogen oxide sensor. This means that if the signal of the first nitrogen oxide sensor is equally large or greater than the expected signal for the first nitrogen oxide sensor, ammonia slip at the first SCR catalyst is detected. In this inspection method, assumptions are made about the expected signal. In this way, it is already possible to reliably detect the ammonia slip if the signal of the nitrogen oxide sensor immediately downstream after the SCR catalyst does not correspond to the nitrogen oxide concentration after-treatment by the SCR catalyst.

The signal expected for the first nitrogen oxide sensor can preferably be calculated by a factor from the signal of the nitrogen oxide sensor arranged upstream of the SCR catalyst, and thus for the first SCR catalyst from the third nitrogen oxide sensor. The factor can be determined from the temperature of the exhaust gas mass flow upstream of the first SCR catalyst and the nitrogen oxide concentration and the ammonia filling level of the first SCR catalyst.

If, during the examination, it is determined that the temperature gradient has fallen below the first threshold value, the identification of the ammonia slip can preferably be prevented. If the temperature drops, a new storage location in the SCR catalytic converter is left empty, so that ammonia slip is not possible and therefore fault detection is falsified. Alternatively, this correlation can also be taken into account when identifying the ammonia slip, for example by: for learning the expected signal to adjust the factor.

The fault detection is advantageously only carried out if the integrated ammonia mass at the first nox sensor exceeds a second threshold value. This ensures that the ammonia mass available for reduction is sufficiently large for the second SCR catalyst for reliable conclusions to be drawn.

If there is ammonia slip at the first SCR catalyst and the integrated ammonia mass at the first nitrogen oxide sensor exceeds the second threshold value, further knowledge of the ammonia slip can be adjusted by the signal of the first nitrogen oxide sensor.

If there is ammonia slip at the second SCR catalyst, it is preferably possible to end the overdosing when an active diagnosis is carried out. The resulting ammonia slip is sufficient for a reliable evaluation. It is now sufficient to add the quantity of reducing agent necessary for the reduction of the nitrogen oxides, in order to avoid excessive consumption of the reducing agent. In this case, it can be provided that additional and/or variable parameters are used in particular for determining the factor when determining the ammonia slip. For example, the integrated quantity of reducing agent dosed in excess can be taken into account, since the probability of ammonia slip increases accordingly. Otherwise, the overdosing can also be ended if a defined reducing agent mass is dosed.

Furthermore, it can be provided that the overdosing is not carried out continuously but rather in pulses (pulseweise) when an active diagnosis is carried out. In the case of pulsed dosing, a smaller ammonia mass is available for further evaluation after the end of the overdosing than in the case of continuous dosing. As a result, the SCR system is less excited, resulting in less emissions therefrom.

The computer program is set up to carry out each step of the method, in particular when it is executed on a computer or a controller. The method can be implemented in conventional electronic controllers without structural modifications. To this end, the computer program is stored on the machine-readable storage medium.

The electronic control unit is obtained by loading the computer program onto a conventional electronic control unit, which is set up to carry out fault detection in the SCR system.

Drawings

Embodiments of the invention are illustrated in the drawings and are explained in detail in the following description.

Fig. 1 schematically shows an SCR system which comprises two SCR catalysts and three nitrogen oxide sensors and which can carry out fault detection by means of an exemplary embodiment of the method according to the invention.

Fig. 2a shows a flow chart of an exemplary embodiment of the method according to the present invention, in which an active diagnosis is shown.

Fig. 2b shows an additional part of the flow chart of a further exemplary embodiment of the method according to the present invention, in which a passive diagnosis is shown, which is arranged upstream of the active diagnosis of fig. 2 a.

Fig. 3a shows a graph of the signal of the nox sensor over time for detecting ammonia slip according to an exemplary embodiment of the method according to the present invention, in which there is no ammonia slip.

Fig. 3b shows a graph of the signal of the nox sensor over time for detecting ammonia slip, in which there is ammonia slip, according to an exemplary embodiment of the method according to the present invention.

Fig. 4a shows a diagram of the dosed mass of reducing agent, the signal of the nitrogen oxide sensor, the integrated ammonia mass at the first nitrogen oxide sensor and the integrated difference between the mass of ammonia at the first nitrogen oxide sensor and the mass of ammonia at the second nitrogen oxide sensor over time, in which a malfunction of the second SCR catalyst is not detected by an embodiment of the method according to the invention.

Fig. 4b shows a diagram of the dosed mass of reducing agent, the signal of the nitrogen oxide sensor, the integrated ammonia mass at the first nitrogen oxide sensor and the integrated difference between the mass of ammonia at the first nitrogen oxide sensor and the mass of ammonia at the second nitrogen oxide sensor over time, in which a malfunction of the second SCR catalyst is detected by an embodiment of the method according to the invention.

Detailed Description

Fig. 1 shows an SCR system 100 of a combustion motor, not shown, in a motor vehicle, having a first SCR catalyst 101 and a second SCR catalyst 102, in which a fault can be identified by means of an embodiment of the method according to the invention. The two

SCR catalysts

101 and 102 are arranged one after the other in the exhaust system 120, wherein the first SCR catalyst 101 is arranged closer to a dosing module 130 which injects a urea/water solution into the exhaust system 120 upstream of the two

SCR catalysts

101 and 102. Further, the SCR system 100 includes: a dosing module; a first nitrogen oxide sensor 111, which is arranged between the first SCR catalyst 101 and the second SCR catalyst 102 and is able to measure there the nitrogen oxide concentration after exhaust gas aftertreatment and the ammonia slip past the first SCR catalyst 101; a second nox sensor 112, which is arranged downstream of the second SCR catalyst 102 and is able to measure there the nox concentration after the exhaust gas aftertreatment and the ammonia slip over the two

SCR catalysts

101 and 102; and a third nitrogen oxide sensor 113, which is arranged upstream of the dosing module 130 and downstream of the first SCR catalyst 101 and is able to measure there the nitrogen oxide concentration of the exhaust gas before the exhaust gas aftertreatment by the

SCR catalysts

101 and 102. The three nox

sensors

111, 112 and 113 mentioned and the dosing module 130 are connected to and controlled by an electronic control unit 140.

Fig. 2a shows a flow chart of an exemplary embodiment of the method according to the present invention, in which an active diagnosis is shown. Initially, signal y of first nox sensor 111 was detected 200₁And from this, the ammonia mass NH at the first nox sensor 111 is determined 201₃z。

It is then checked in a query 202 whether the temperature gradient dT exceeds a first threshold value S₁. In further embodiments, further conditions for passive diagnosis can be checked in the query 202. Possible conditions are:

-preparation of the SCR system 100;

-preparation of said

nox sensors

111, 112 and 113;

signal y for the

nox sensors

111, 112 and 113₁、y₂And y₃In particular the signal y for the third nox sensor 113₃(ii) evaluation of (d);

-an evaluation of the exhaust gas mass flow;

-temperature level of the

SCR catalysts

101 and 102; and

-the modeled ammonia loading level of the

SCR catalysts

101 and 102.

If at least one of these conditions is not met, signal y continues for first NOx sensor 111₁Until all conditions are met 200.

If all the conditions checked in query 202 are met, then an active diagnosis is carried out, in which the mass m of reducing agent dosed into exhaust system 120 is increased_DosThereby overdosing 203 the SCR system 100. The overdosing 203 is set in such a way that the converted ammonia mass of the first SCR catalyst 101 exceeds the ammonia mass required for reducing nitrogen oxides. Depending on the embodiment and operating conditions, the overdosing 203 is carried out either continuously or in pulses. Thereby, the ammonia loading level on the first SCR catalyst 101 increases until it exceeds the maximum ammonia loading level of the first SCR catalyst 101. If this is the case, ammonia slip occurs at the first SCR catalyst 101.

Depending on the signal y of the first NOx sensor 111, the first check 204 is made as to whether there is ammonia slip₁To learn of ammonia slip over the first SCR catalyst. For this purpose, signal y of first nox sensor 111 is used₁With the expected signal y for the first nox sensor 111_eA comparison is made. Expected signal y_eSignal y from third nox sensor 113 by a factor F₃Is known and indicates the expected reduction of nitrogen by the first SCR catalyst. The factor F is in turn known from the temperature upstream of the first SCR catalyst 101, the exhaust gas mass flow, the nitrogen oxide concentration and the ammonia loading level of the first SCR catalyst 101. The integrated, overdosed mass of reducing agent is taken into account when determining the factor F.

A first check 204 is shown in fig. 3a and 3b as to whether there is ammonia slip at the first SCR catalyst. Both figures show the signal y of the first nox sensor 111 with respect to time t₁And stationSignal y of the third nox sensor 112₃And the expected signal y_eA graph of (a). In fig. 3a, signal y from third nox sensor 113 is measured by factor F₃Is known to the expected signal y_eSignal y of first nox sensor 111 is always exceeded₁. Thus, in this case, no ammonia slip is identified. Whereas in FIG. 3b, the expected signal y_eSignal y in region 300 below first nox sensor 111₁Thereby identifying ammonia slip for the region 300.

In another embodiment, a first check 204 is made as to whether there is ammonia slip on the first SCR catalyst 101 by: the signal y of the first NOx sensor 111 is measured₁Signal y directly connected to third nox sensor 113₃A comparison is made. If the signal y of the first NOx sensor 111₁Continuously exceeds signal y of third nox sensor 113 from a predetermined point in time₃The ammonia slip is identified.

If the ammonia slip is detected, the mass NH of ammonia at the first NOx sensor 111 is checked₃z is integrated 205. The mass of ammonia to be integrated

For fault recognition of the first SCR catalyst 101. With respect to the integrated ammonia mass on the first NOx sensor 111

Whether or not the second threshold value S is exceeded₂A first check 206 is performed. If this is the case, it can be assumed that there is sufficient ammonia available to the second SCR catalyst 102 for carrying out the fault identification. If the integrated ammonia mass

Is lower than the second threshold S₂Or it is equal to said second threshold valueS₂The method is repeated.

If a first check 204 is made as to whether ammonia slip is present, it is determined that ammonia slip is occurring at the first SCR catalyst 101, the overdosing 203 is terminated 207 and a dosed quantity m of reducing agent is added_DosDown to the necessary scale.

If a first check 204 is made not only as to whether ammonia is present, but also in this exemplary embodiment as to the integrated ammonia mass at the first nox sensor 111

Whether the second threshold S is exceeded₂The first check 206 is performed in the affirmative and the first failure recognition 209 is released.

A selection criterion is also selected according to the embodiment of the method according to the invention. In one embodiment of the method according to the invention, the ammonia mass NH at the first nox sensor 111 is determined from the mass of ammonia₃z and mass NH of ammonia on the second nitrogen oxide sensor 112₃n constituting the integrated difference

Calculate 208a according to equation 1:

(equation 1).

A first fault detection 209 of the second SCR catalyst 102 is then carried out by means of the selection criterion. If the difference is integrated

Is lower than a third threshold S₃Then a fault 210 is output because sufficient ammonia cannot be stored by the second SCR catalyst 102. If the integrated difference

AboveThe third threshold value S₃Or it is equal to said third threshold S₃The fault 211 is not output.

In a further embodiment of the method according to the invention, signal y from first nox sensor 112₁And the signal y of the second oxynitride sensor₂The correlation coefficient E is calculated 208b according to equation 2:

(equation 2).

The correlation coefficient indicates the signal y of the first nox sensor 111₁And the signal y of the second oxynitride sensor 112₂Until a decision time t_EThere are many similarities. A first fault detection 209 of the second SCR catalyst 102 is then carried out by means of the selection criterion. If the correlation coefficient is higher than a fourth threshold S₄The fault 210 is output because of the two signals y₁And y₂Too similar to allow sufficient ammonia to be stored. If the correlation coefficient is lower than the fourth threshold S₄Or it is equal to said fourth threshold S₄The fault 211 is not output.

It should be noted here that NH is the mass of ammonia in question₃z is integrated 205 and the integrated ammonia mass at the first NOx sensor 111 is used

Whether the second threshold S is exceeded₂The check 206 is performed while performing the

calculation

208a and 208b of the selection criteria.

The transition point 1 shown in fig. 2a by dashed lines shows the connection to the flow chart shown in fig. 2 as a further exemplary embodiment of a passive diagnosis arranged upstream of an active diagnosis.

In fig. 2b, initially, similarly to the active diagnosis, the signal y of the first nox sensor 111 is detected 200₁And from which is known 201Mass of ammonia NH at the first NOx sensor 111₃z. Subsequently, it is also checked in a query 202 whether the temperature gradient dT exceeds the first threshold value S₁. In a further embodiment, it is also possible to check further conditions for the passive diagnosis in the query 202. The conditions correspond to the aforementioned conditions for an active diagnosis. The same reference numerals as in fig. 2a mean that the steps correspond to each other. The three

steps

200, 201 and 202 listed above can therefore also be received from the active diagnosis and reference is made to the description thereof.

During the passive diagnosis, the operating parameters of the metering are not changed and, for this purpose, a normal metering is carried out. According to the signal y of the first NOx sensor 111₁A second check 220 is made as to whether there is ammonia slip. The second check is carried out in a manner corresponding to the first check 202 as to whether ammonia slip is present for the active diagnosis, so that reference is made to the description of fig. 2a and fig. 3a and 3b for the description of the second check.

If ammonia slip is detected, the ammonia mass NH at the first nox sensor 111 is likewise corrected₃z is integrated 221. Integrated mass of ammonia

Generally distinguished from the integrated ammonia mass when active diagnostics are performed

And is typically smaller. With respect to the integrated ammonia mass on the first NOx sensor 111

Whether the second threshold S is exceeded₂A second check 222 is performed, wherein the second threshold S is also checked₂And (6) adjusting. If the integrated ammonia mass

Is lower than the second threshold S₂Or it is equal to said second threshold S₂Then the method is repeated.

If a second check 221 is made not only as to whether ammonia is present, but also in this exemplary embodiment as to the integrated ammonia mass at the first nox sensor 111

Whether the second threshold S is exceeded₂The second check 222 is positive and the second fault identification 224 is released. A selection criterion is selected according to the embodiment of the method according to the invention. As in the case of active diagnostics, the ammonia mass NH at the first nox sensor 111 is calculated 223a, either according to equation 1₃z and mass NH of ammonia on the second nitrogen oxide sensor 112₃Integrated difference of z

Or from the signal y of the first nox sensor 111₁And the signal y of the second oxynitride sensor 112₂The correlation coefficient E is calculated 223b according to equation 2.

The second fault identification 224 is then performed in accordance with the selection criteria. Similar to the passive diagnosis, if the integrated difference is

Is lower than the third threshold S₃Or if the correlation coefficient E is higher than the fourth threshold S₄Then a fault 225 is output. If this is not the case, there is no output fault 226.

If a fault 225 is output during the passive diagnosis, the active diagnosis shown in fig. 2b is carried out starting from the transition point 1, i.e. the overdosing 203 is carried out. Otherwise, the execution of the overdosing 203 and the steps of the active diagnosis in fig. 2a, which follow it, are prevented.

In the embodiment shown in fig. 2b, ghost counting 227 is started after no output fault 226 has occurred in the passive diagnosis. The ghost count 227 is used for these situations to know the time at which the active diagnostics have identified the fault. If the active diagnosis has ended In the current driving cycle (ablaufen), the IUMPR (In-Use-Performance Ratio) is increased. If at least one of the operating parameters of the internal combustion engine does not indicate this, the ghost count 227 is stopped and/or reset. The ghost counter 227 associates the driving cycle with the diagnosis with all driving cycles and thus indicates the probability with which a fault is identified.

FIGS. 4a and 4b show the dosed mass m of reducing agent for the active diagnosis with respect to time t_DosSignal y of the nitrogen oxide sensor₁、y₂And y₃Integrated mass of ammonia on the first NOx sensor 111

And mass of ammonia NH at the first nox sensor 111₃z and mass NH of ammonia on the second nitrogen oxide sensor 112₃n integrated difference

A graph of (a). Increasing the mass m of reducing agent dosed in the event of an overdosing 203_Dos. Due to the resulting ammonia slip on the first SCR catalyst 101, the integrated ammonia mass on the first nitrogen oxide sensor 111

Is increased until the ammonia slip is determined when the first check 204 is made as to whether ammonia slip is present. In these figures the first check 204 is made as to whether there is ammonia slip by: determining signal y of first nox sensor 111₁Whether or when the third nitrogen oxide is exceededSignal y of sensor 113₃. The overdosing 203 is terminated 207 at this time.

In this case, only the integrated difference

Used as an evaluation criterion. In fig. 4a situation is shown in which no fault 211 is determined. The integrated difference

As depicted at decision time t_EAbove the third threshold S₃. In fig. 4b, a situation is shown in which a fault 210 is determined. The integrated difference

As depicted at decision time t_EIs lower than the third threshold S₃。

Claims

1. Method for fault detection in an SCR system (100) of an internal combustion engine in a motor vehicle for active diagnosis, having two SCR catalysts (101, 102) and at least two nitrogen oxide sensors (111, 112), wherein a first nitrogen oxide sensor (111) is arranged between the two SCR catalysts (101, 102) and a second nitrogen oxide sensor (112) is arranged downstream of the two SCR catalysts, comprising the following steps:

I. a signal (y) for the first NOx sensor (111)₁) Performing successive detections (200);

signal (y) from the first NOx sensor (111)₁) Is determined (201) by the mass of ammonia (NH) at the first NOx sensor (111)₃z）；

By mass of reducing agent dosed (m)_Dos) Is over-dosed (203), wherein the mass of converted ammonia of the first SCR catalyst (101) exceeds for nitrogen oxidation(ii) the mass of ammonia necessary for reduction of the substance until the ammonia loading level on the first SCR catalyst (101) exceeds the maximum ammonia loading level of the first SCR catalyst (101);

according to the signal (y) of the first nitrogen oxide sensor (111)₁) -a first check (204) as to whether there is ammonia slip on the first SCR catalyst (101); and is

V. if there is ammonia slip on the first SCR catalyst (101), performing a first fault identification (209) of at least the second SCR catalyst (102) by an evaluation criterion.

2. The method according to claim 1, characterized in that a passive diagnosis is carried out after method step II and before method step III, additionally with a normal dosing, said passive diagnosis comprising the following steps:

-depending on the signal (y) of the first nitrogen oxide sensor (111)₁) -performing a second check (220) as to whether there is ammonia slip on the first SCR catalyst (101);

-identifying (224) a second malfunction of at least the second SCR catalyst (102) by an evaluation criterion if there is ammonia slip on the first SCR catalyst (101);

-blocking the method steps III to V if no fault (226) is output while the second fault detection (224) is being performed.

3. Method according to claim 2, characterized in that a ghost counting (227) is started if no fault (226) is output during the passive diagnosis while the second fault detection (224) is being carried out.

4. Method according to one of claims 1 to 3, characterized in that the quality of ammonia (NH) at the first nitrogen oxide sensor (111) is known for the evaluation criterion₃z) and ammonia on the second oxynitride sensor (112)Amount (NH)₃n) of the difference (D) between n), integrating this difference (D) over a time interval (Δ t), if said integrated difference (D) is (D:)

) At the determination time (t)_E) Below a third threshold value (S)₃) -identifying a fault (210, 225) on the second SCR catalyst (102) and if the integrated difference (c/l:)

) At the determination time (t)_E) Above the third threshold value (S)₃) Or equal to said third threshold value (S)₃) No fault (211, 226) on the second SCR catalyst (102) is identified.

5. Method according to any one of claims 1 to 3, characterized in that the signal (y) from the first NOx sensor (111) is evaluated with respect to the evaluation criterion₁) And the signal (y) of the second oxynitride sensor (112)₂) If the correlation coefficient (E) is determined at the time (t)_E) Above a fourth threshold value (S)₄) -identifying a fault (210, 225) on the second SCR catalyst (102) and if the correlation coefficient (E) is at the decision time (t)_E) Is lower than the fourth threshold value (S)₄) Or equal to said fourth threshold value (S)₄) No fault (211, 226) on the second SCR catalyst (102) is identified.

6. Method according to any one of claims 1 to 3, characterized in that an additional third nitrogen oxide sensor (113) is arranged upstream of the first SCR catalyst.

7. The method of claim 6, wherein said first SCR catalyst is present in (i)101) If the signal (y) of a first nitrogen oxide sensor (111) immediately downstream after the first SCR catalyst is detected (204, 220)₁) Continuously greater than a signal (y) of a nitrogen oxide sensor (113) arranged upstream of the first SCR catalyst from a defined point in time₃) Then the ammonia slip is identified.

8. Method according to one of claims 1 to 3, characterized in that, when checking (204, 220) whether there is ammonia slip at the first SCR catalyst (101), the signal (y) of the first nitrogen oxide sensor (111) immediately downstream is detected if it is₁) Is equally large or larger than the expected signal (y) for the first NOx sensor (111)_e) Then the ammonia slip is identified.

9. Method according to claim 8, characterized in that the signal (y) from a nitrogen oxide sensor (113) arranged upstream of the first SCR catalyst (101) can be derived by a factor (F)₃) To calculate the expected signal (y)_e) Wherein the factor (F) is determined from the temperature upstream of the first SCR catalyst (101), the exhaust gas mass flow and the nitrogen oxide concentration, and the ammonia filling level of the first SCR catalyst (101).

10. A method according to any one of claims 1 to 3, characterized in that, in the checking (202), if the temperature gradient (dT) falls below the first threshold value (S)₁) That prevents the escape of the identification.

11. Method according to one of claims 1 to 3, characterized in that only the integrated ammonia mass (x, y) at the first nitrogen oxide sensor (111) ((x, y))

) Exceeds a second threshold value (S)₂) The fault detection is carried out (209, 224).

12. Method according to any of claims 1 to 3, characterized in that the overdosing (203) is ended (207) if there is ammonia slip on the second SCR catalyst (102) while the active diagnosis is being carried out.

13. A machine-readable storage medium on which a computer program is stored, the computer program being set up to carry out each step of the method according to any one of claims 1 to 12.

14. Electronic control unit (140) which is set up to carry out fault detection in SCR system (100) by means of the method according to one of claims 1 to 12.