DE10331334B4 - Method for operating an internal combustion engine - Google Patents

Abstract

Method for operating an internal combustion engine, in particular a motor vehicle, with an exhaust gas aftertreatment arrangement arranged in an exhaust tract of the internal combustion engine with at least one catalytic converter and a first lambda probe arranged upstream of the catalytic converter and a second lambda probe arranged downstream of the catalytic converter, with a first Step (a) a lambda value upstream of the catalytic converter is actively adjusted to a value less than 1 until an oxygen stored in the catalytic converter is completely discharged, and then in a second step (b) the lambda value upstream of the catalytic converter is actively adjusted to a value greater than 1 is until the catalyst is completely loaded with oxygen, the oxygen storage capacity of the catalyst being determined by means of an oxygen balance, with at least one Pa at the same time during the determination of the oxygen storage capacity parameters of the lambda probes are determined and a functionality of the lambda probes is determined from the parameter, the determination of at least one parameter of the lambda probes comprising a plausibility check, wherein in the first step (a) and / or ...

Description

The invention relates to a method for operating an internal combustion engine, in particular a motor vehicle, having an arranged in an exhaust tract of the internal combustion engine exhaust aftertreatment arrangement with at least one catalyst and upstream of the catalyst arranged first lambda probe and a downstream of the catalyst arranged second lambda probe, wherein for determining an oxygen storage capacity of Catalyst in a first step (a) a lambda value before the catalyst is actively adjusted to a value less than 1, until an oxygen stored in the catalyst is completely discharged, and then in a second step (b) the lambda value before the catalyst active on a Value greater than 1 is adjusted until the catalyst is completely laden with oxygen, wherein the oxygen storage capacity of the catalyst is determined by means of an oxygen balancing, at the same time during the determination of the oxygen at least one parameter of the lambda probes is determined and the functionality of the lambda probes is determined from the parameter, wherein the determination of at least one parameter of the lambda probes comprises a plausibility check, wherein in the first step (a) and / or in the second step (b) it is checked whether first and second lambda probe simultaneously display the expected rich or lean exhaust gas composition, according to the preamble of claim 1.

For the diagnosis of the state of a catalytic converter of an internal combustion engine, methods are known which measure an oxygen storage capacity (OSC) of the catalytic converter by means of active lambda adjustment. This storage capacity correlates with the hydrocarbon (HC) conversion in the catalyst. If the catalytic converter has good conversion properties, the lambda fluctuations upstream of the catalytic converter, which are actively generated by the lambda controller and detected by a first lambda probe, are smoothed by the oxygen storage capability of the catalytic converter. If the catalyst has no or only reduced conversion properties as a result of aging, poisoning by leaded fuel or by combustion misfires, the control oscillation present upstream of the catalytic converter strikes the lambda probe downstream of the catalytic converter. By comparing the signal amplitudes of the two lambda probes can be concluded that a functioning or defective catalyst, such as from DE 23 28 459 A1 known.

For diagnostic purposes, the OSC is additionally correlated with a light-off and conversion behavior of the catalyst. To measure the OSC, for example, by setting a rich lambda value of, for example, 0.95 to 0.98, an oxygen possibly stored in the catalyst is first removed and subsequently adjusted to a lean, oxygen-rich by setting a lambda value of, for example, 1.02 to 1.05 Exhaust gas filled with oxygen again. The OSC is measured by oxygen balancing. The emptying and filling of the catalyst with oxygen is determined and controlled by a jump of an output signal of a catalytic converter downstream of the lambda probe into the fats or lean.

From the DE 199 53 601 A1 a method for checking an exhaust gas catalytic converter of an internal combustion engine is known, wherein during a diagnostic time increases an oxygen loading of the catalytic converter and is concluded from the measured during several Lambdaregelschwingungen NO _x concentration on the conversion properties of the catalytic converter.

From the DE 198 01 626 A1 is a diagnosis of a catalyst in the exhaust gas of internal combustion engines is known, wherein the catalyst has both an oxygen and a nitrogen oxide storage capability and wherein the oxygen concentration in the exhaust gas before the catalyst is repeatedly increased and decreased so that the change in the signal before and after the Catalyst arranged exhaust probe maps. A first phase shift between the signals of both probes during the increase and a second phase shift during the lowering of the oxygen concentration is detected and the difference of the phase shifts determined. If this difference does not reach a predetermined threshold, an error signal is output.

From the DE 100 17 931 A1 For example, for the diagnosis of an exhaust gas purification system of a lambda-controlled internal combustion engine, it is known to apply a certain oxygen load to the catalyst, which is greater than the normal operating load. This is achieved by increasing the area of a desired value enclosed by the oscillation of the probe signal of the pre-catalyst lambda probe. By evaluating the oscillation of the signal of the Nachkat lambda probe, the emission control system can be checked. If this diagnosis results in a malfunction of the emission control system without the area of the oscillation of the probe signal of the pre-catalyst lambda probe having to be increased to or above a desired value, the pre-catalyst lambda probe is defective. Otherwise, a catalyst failure can be concluded.

From the DE 198 03 828 A1 and the DE 41 12 478 is a method of determining an oxygen storage capacity of a catalyst in which an oxygen content of the exhaust gas before and after the catalyst is determined, multiplied by the air or exhaust gas mass flow through the engine or catalytic converter and the product is integrated. The integral values are a measure of the amounts of oxygen that flow into and out of the catalyst. The difference in the integral values provides the change in the degree of oxygen filling of the catalyst in the integration period. In this case, it is ensured in the method that in the diagnosis of a complete filling of the catalyst with oxygen and then complete emptying or vice versa.

From the DE 100 26 213 A1 For example, a method for configuring a plurality of lambda probes of an internal combustion engine having two exhaust gas banks is known. For a given connection configuration of lambda probes and inputs of an engine control unit, it is detected to which inputs the lambda probes are respectively connected. On this basis, a correct assignment of the lambda probes is made to the exhaust banks.

From the EP 0 897 054 A1 a method for exchange testing of lambda probes is known. In a multi-cylinder internal combustion engine having at least two separate exhaust pipes, each with a catalyst and at least one lambda probe per exhaust pipe during a test period which corresponds to at least the reaction time or switching time of the lambda probe, air by means of an air source individually at least one exhaust pipe via at least one air termination before Lambda probe supplied. A thereby modified exhaust gas value is quickly and uniquely identified by the lambda probe associated with the exhaust gas system. If the expected signal does not originate from the lambda probe assigned to the exhaust gas line, an exchange of the lambda probes can be detected.

From the DE 101 17 244 A1 is a method for detecting interchanged connected oxygen sensors, which are arranged in the exhaust line of an internal combustion engine one behind the other and connected to a control device for the fuel injection, known. During a test period, the fuel injection is switched off or the air-fuel ratio changed and measured the reaction times until the occurrence of the resulting signal change in the O ₂ sensors and evaluated.

From the DE 100 15 330 A1 It is known to carry out NO _x regeneration measures in an internal combustion engine with NO _x storage catalytic converter in the exhaust gas line as a function of the state parameters of the NO _x storage catalytic converter. In the case of NO _x storage catalysts arranged in parallel in the exhaust gas system, a NO _x regeneration measure currently being carried out is terminated as a function of the latest point in time at which a state parameter value characteristic of the termination of an NO _x regeneration measure is measured on one of the NO _x storage catalytic converters.

From the generic EP 1 314 868 A1 For example, a method for determining a degradation of an exhaust gas sensor in a motor vehicle is known.

From the DE 198 44 994 A1 For example, a method is known for detecting a degradation in the response time of a lambda probe in front of a catalyst. In this case, a forced excitation causes small periodic changes in the fuel / air ratio. There is also provided a diagnosis of the catalyst, wherein a lambda probe signal is compared with the signal of another lambda probe after the catalyst. The compulsory stimulation described can also be used for carrying out the catalyst diagnosis. A diagnosis of lambda probe and catalyst simultaneously at one and the same forced excitation is not provided.

From the DE 197 06 382 A1 For example, a method for recognizing exchanged lambda probes is known. For this purpose, the fuel / air ratio is actively adjusted by the lambda controller and the reaction of the lambda probes observed. It is not intended to carry out a catalyst diagnosis at the same time. On the contrary, it should be irrelevant in the simultaneous execution of the commutation diagnosis with other diagnoses, whether the catalyst is warm or not.

The invention is based on the object, a method of o. G. Art regarding diagnosis times and diagnostic frequency for the exhaust aftertreatment device to improve.

This object is achieved by a method of o. G. Art solved with the features characterized in claim 1. Advantageous embodiments of the invention are specified in the dependent claims.

For this purpose, it is provided according to the invention that in the event that the plausibility check results in a plausibility violation, d. H. a different display of the exhaust gas composition before and after the catalyst by the two lambda probes, it is determined whether the first or second lambda probe is defective.

This has the advantage that diagnoses of the catalyst state and the state of the lambda probes run in parallel, so that diagnostic times are reduced and homogeneous phases are shortened for diagnoses. At the same time can be detected quickly and reliably, whether a lambda probe upstream or downstream of the catalyst is defective.

For example, the catalyst is a precatalyst or a main catalyst.

For example, the catalyst is a NOx storage catalytic converter or the exhaust aftertreatment device additionally has a downstream of the catalyst NOx storage catalytic converter, wherein under predetermined conditions a time-limited NO _x regeneration of the NO _x storage carried out catalyst, the NO _x regeneration as a first step (a ) and after completion of the NO _x regeneration, the second step (b) is performed. This has the advantage that the determination of the oxygen storage capacity of the catalyst is combined with a NO _x regeneration of the NO _x storage catalyst such that the regeneration phase of the NO _x regeneration is at the same time the oxygen purification phase for determining the oxygen storage capacity of the catalyst. This reduces and minimizes diagnostic times with deviation from the optimum operating mode of the internal combustion engine for a current operating state thereof.

For determining the faulty lambda probe, a predetermined operating mode of the internal combustion engine, in particular a homogeneous lambda 1 operation, is expediently set.

For an exhaust aftertreatment arrangement with two catalysts arranged in parallel in the exhaust gas system, the determination of at least one parameter of the lambda probes comprises, for example, a permutation of the respective first lambda probes upstream of the catalysts and / or the respective second lambda probes downstream of the catalysts, wherein step (a) for a second of the parallel Catalysts offset by a predetermined time later than for the corresponding other first of the two parallel catalysts, wherein a permutation of the terminal of the two first lambda probes of the two catalysts is determined when for a predetermined number of cycles of steps (a) and ( b) the first lambda probe of the second catalyst indicates a change in state of the exhaust gas before the first lambda probe of the first catalyst, and wherein a permutation of the connection of the two second lambda probes of the two catalysts is determined, if for a vorbestim mte number of cycles of steps (a) and (b) the second lambda probe of the second catalyst indicates a change in state of the exhaust gas before the second lambda probe of the first catalyst. The state change of the exhaust gas is, for example, a jump from lean exhaust gas composition to rich exhaust gas composition or a jump from rich exhaust gas composition to lean exhaust gas composition.

In a preferred development of the invention, the determination of at least one parameter of the lambda probes comprises a dynamic behavior of the lambda probes, wherein a gradient of the temporal change of an output signal of the lambda probes is determined and a faulty lambda probe is determined if the gradient is smaller than a predetermined value. The gradient of the change over time in the transition from rich exhaust gas composition to lean exhaust gas composition is determined, for example, at the end of step (b). A maximum gradient or an averaged gradient is expediently determined and compared with a corresponding, predetermined value.

In order to minimize the active intervention of diagnoses, which may have negative effects on both consumption and pollutant emission, further test cycles are requested depending on the result of the determination of the oxygen storage capacity.

For example, in the event that the oxygen storage capacity results in a value in the range of a fresh catalyst and the lambda probes have been recognized as functional, no further test cycles are requested for a predetermined time interval, in particular until the next restart of the internal combustion engine.

For example, in the event that the oxygen storage capacity gives a value in the range between a fresh catalyst and a defective catalyst, a predetermined first number, in particular 2 to 5, of further test cycles is required.

For example, in the event that the oxygen storage capacity gives a value in the range of a defective catalyst, a predetermined second number, in particular 6 to 10, of further test cycles is requested.

Conveniently, the first number of further test cycles is smaller than the second number of further test cycles.

Appropriately, in the first step (a) a lambda value in the range of 0.95 to 0.98 or less and in the second step (b) a lambda value in the range of 1.02 to 1.05, in particular 1.03, set.

In a preferred embodiment of the invention, the determination comprises at least one Parameters of the lambda probes a check of a lean stress and / or a fatigue voltage of the first and / or second lambda probe.

Further features, advantages and advantageous embodiments of the invention will become apparent from the dependent claims, and from the following description of the invention with reference to the accompanying drawings. These show in

1 a graphical illustration of an oxygen storage capacity (OSC) measurement of a precatalyst in transition from a NOx regeneration to a lean operation;

2 a graphical illustration of a combination of catalyst and lambda probe diagnosis;

3 a graphical illustration of a lambda probe Vertauschungserkennung for a two-bank concept with correctly connected lambda probes;

4 a graphical illustration of a Lambda probe interchangeability detection for a two-bank concept with reversed lambda probes after the catalysts of the exhaust banks;

5 a graphical illustration of a lambda probe Verschauschungserkennung for a two-bank concept with reversed lambda probes before the catalysts of the exhaust banks;

6 a schematic representation of an internal combustion engine with pre-catalyst and NOx storage catalyst;

7 a schematic representation of an internal combustion engine with two exhaust banks, each having a precatalyst with each upstream and downstream arranged lambda probes and

8th a schematic block diagram of a method with reduced diagnosis time for an active catalyst diagnosis.

The invention will be exemplified below for two exhaust aftertreatment arrangements as in 6 and 7 shown, described. In 6 has an internal combustion engine 10 an exhaust gas aftertreatment arrangement in an exhaust gas line, which has the following in the flow direction, a first lambda probe 12 , which is designed to output a steady-state lambda signal, a pre-catalyst 14 , a second lambda probe 16 , which is designed to deliver a jump signal, a NO _x storage catalytic converter 18 and a NO _x sensor 20 , In the execution of the internal combustion engine 10 according to 7 the exhaust aftertreatment arrangement comprises a first exhaust bank 100 and a second exhaust bank 200 that at 22 lead together and into a common NO _x storage catalytic converter 18 lead to a NO _x sensor 20 is subordinate. The first exhaust bank 100 includes a first lambda probe as seen in the flow direction 112 the first exhaust bank 100 , which is designed to output a steady-state lambda signal, a pre-catalyst 114 the first exhaust bank 100 and a second lambda probe 116 the first exhaust bank 100 , which is designed to deliver a jump signal. The second exhaust bank 200 includes a first lambda probe as seen in the flow direction 212 the second exhaust bank 200 , which is designed to output a steady-state lambda signal, a pre-catalyst 214 the second exhaust bank 200 and a second lambda probe 216 the second exhaust bank 200 , which is designed to deliver a jump signal.

1 and 2 illustrate oxygen oxygen storage capacity (OSC) measurements using lambda over time graphs utilizing an NO _x regeneration in an internal combustion engine having a configuration according to FIG 6 , In 1 is on the horizontal axes 24 the time t, on the vertical axis 26 a lambda value and on the vertical axis 28 applied a probe voltage. A graph 30 shows a time course of lambda before the pre-catalyst 14 , A graph 32 shows a probe voltage of the lambda probe 16 after the precatalyst 14 , A graph 34 shows a probe voltage of the NO _x sensor 20 after the NO _x storage catalyst 18 ,

In addition, the signals "Request the NO _x regeneration" 36 , "OSC measurement" 38 , "End of the measurement OSC" 40 and "release lean operation" 42 in their temporal correlation to the graphs 30 . 32 and 34 applied. At time t ₁ , a NO _x regeneration is requested, wherein previously the internal combustion engine was operated in a lean operation with lambda>> 1. In the time interval t ₁ -t ₂ becomes the NO _x regeneration before the pre-catalyst 14 a lambda in the rich range (rich exhaust gas composition) of, for example, 0.8 to 0.9 is set. This phase is ended by changing the lambda to the NO _x storage catalytic converter 18 from lean to rich, ie the probe voltage 34 rises abruptly. Even before the time t ₂ , the entire oxygen from the precatalyst 14 eliminated, which is due to the increase in probe voltage 32 the lambda probe 16 after the precatalyst 14 (Jump into fats) is visible. Once the whole in the NO _x storage catalytic converter 18 contained NO _{x is} implemented, the probe voltage also increases 34 because the regeneration exhaust gas is up to the NO _x storage catalyst 18 breaks through, what too an end of the NO _x regeneration at time t ₂ leads. Depending on the operating mode requirement, it would then be possible to switch to lambda 1 mode or lean mode. However, regardless of the NO _x regeneration in the time interval t ₁ -t ₂ subsequent operating mode after completion of the NO _x regeneration at t ₂ to a lean lambda 1 point (lambda 1.02 to 1.05) is switched and the oxygen introduced into the precatalyst accumulates until the probe voltage 32 the lambda probe 16 after the precatalyst 14 indicates the breakthrough of lean exhaust gas at time t ₃ . In the time interval 44 From time t ₂ to time t ₃ oxygen is in the precatalyst 14 registered until it is completely loaded with oxygen. Thus, the NO _x regeneration in the time interval t ₁ -t _{2 is used} as an OSC clearing phase, so that it does not have to run extra extra. As with arrow 46 in 2 indicated, the OSC measurement thus takes place in the time interval t ₁ -t ₃ and the NO _x regeneration in the time interval t ₁ -t ₂ . Thus, these two processes overlap in time or partially in parallel and take less time overall, as if both processes would run sequentially or separately.

Parallel to the OSC measurement 46 become the lambda signals 32 and 34 checked for their dynamic behavior and regarding the exchange of respective first and second lambda probes in a two-bank system, as in 7 shown, which will be described in more detail below.

As in 2 with arrow 48 indicated in the time domain before t _{1 is} a check of the lean stress of the probe signals 32 and 34 , As with arrow 50 indicated, takes place in the time interval between t ₁ and t ₂ , ie during the clearing phase of the OSC measurement or the NOx regeneration, a check of the fat tension of the probe signals 32 and 34 ,

At the end of the NO _x regeneration or the clearing step of the OSC measurement at time t ₂ , a jump from rich exhaust gas composition to lean exhaust gas composition takes place. This is used to develop a dynamic behavior of the lambda probe 12 in front of the precatalyst 14 to consider. In 2 is the dynamic measurement of the first lambda probe 12 With 52 characterized. As evaluation criteria for the dynamics, for example, the gradient is used as delta O ₂ / dt, delta voltage / dt or delta oxygen pump current / dt. For example, the gradient between the desired lambda values and the actually measured gradient of the lambda signal is compared. If the dynamic value falls below a predetermined threshold, errors are detected.

To evaluate the dynamics of the probe voltage 32 the second lambda probe 32 after the precatalyst 14 becomes the jump from rich exhaust gas composition to lean exhaust gas composition at time t ₃ at the end of the OSC measurement 46 used. In 2 is the dynamic measurement of the second lambda probe 16 With 54 characterized. The measurement and evaluation of the probe dynamics, for example, analogous to the first lambda probe 12 in front of the precatalyst 14 ,

For example, the maximum gradient which results when changing from rich exhaust gas composition to lean exhaust gas composition can also be used as a dynamic criterion. Alternatively, an averaged gradient is calculated, which results from a minimum fat lambda value to a minimum lean lambda.

A rough plausibility check of the second lambda probe 16 after the precatalyst 14 in relation to the first lambda probe 12 in front of the precatalyst 14 takes place in the respective measuring phase "lean" in the time interval t ₂ -t ₃ or "fat" in the time interval t ₁ -t ₂ . In the lean with Lambda> 1, for example, lambda = 1.03 to 1.05, after appropriate Entprellzeit, which takes into account gas runtimes and catalyst Ausläumzeiten, checked whether the second lambda probe 16 after the precatalyst 14 also shows lean. After setting a rich mixture before the pre-catalyst 14 is checked after appropriate debounce time, if the lambda after the pre-catalyst 14 also shows in bold.

Result in this rough plausibility between "lambda before the pre-catalyst 14 "And" lambda after the precatalyst 14 "Plausibility violations, because, for example, the first lambda probe 12 before the pre-catalyst "bold" indicates, whereas the second lambda probe 16 after the pre-catalyst indicates "lean" or vice versa, a suspected error is set. Subsequently, a precise check is triggered, which is to ensure the assignment of the plausibility error to the fault location, ie it is determined whether the first lambda probe 12 or the second lambda probe 16 is defective. For this review, an operating mode is advantageously requested, which is favorable for a quick fault finding, such as a homogeneous lambda 1 operation.

To ensure the simultaneous, simultaneous execution of the diagnosis, a release is defined, which results from the sub-approvals of the individual test functions, such as catalyst test, lambda test, etc. Alternatively, a global release is defined that takes into account the physical test conditions of the individual partial tests. Only when this release is given, the active measuring phase of the lambda adjustment is executed.

3 . 4 and 5 illustrate a flow of monitoring for probe interchange between the first lambda probes 112 . 212 before the pre-catalysts 114 . 214 and the second lambda probes 116 . 216 after the precatalysts 114 . 214 each of the two exhaust banks 100 and 200 , 3 shows a time course of lambda values in an exhaust aftertreatment device in the state "i. O.", 4 shows a time course of lambda values in an exhaust aftertreatment device in the state "reversed second lambda probes 116 . 216 after the precatalyst 114 . 214 " and 5 shows a time course of lambda values in an exhaust aftertreatment device in the state "reversed first lambda probes 112 . 212 in front of the precatalyst 114 . 214 ". This in 3 . 4 and 5 each upper axbox concerns the first exhaust bank 100 and the respective lower axbox relates to the second exhaust bank 200 , On the horizontal axes 10 is again the time t and on the vertical axes 26 a lambda value is plotted. graph 130 shows a time course of lambda according to the first lambda probe 112 in front of the precatalyst 114 for the first exhaust bank 100 and graph 132 shows a time course of lambda according to the second lambda probe 116 after the precatalyst 114 for the first exhaust bank 100 , graph 230 shows a time course of lambda according to the first lambda probe 212 in front of the precatalyst 214 for the second exhaust bank 200 and graph 232 shows a time course of lambda according to the second lambda probe 216 after the precatalyst 214 for the second exhaust bank 200 , In a system with two exhaust banks 100 and 200 , as in 7 illustrated, there is another possibility that the first lambda probes 112 . 212 or the second lambda probes 116 . 216 be installed reversed during installation or change. As a result, the lambda control parameters of an exhaust bank fit 100 . 200 not to the measured lambda signal. This results in incorrect control interventions that may result in misdiagnosis or emissions degradation.

An interchangeable detection becomes parallel to the active lambda intervention for the OSC measurement 46 ( 2 ), as in 3 graphically illustrated. For this purpose, the active lambda adjustment for the catalyst diagnosis for the two exhaust banks 100 and 200 timed request. The time offset T1 is selected such that initially the lambda adjustment for the first exhaust bank 100 takes place and the lambda adjustment for the second exhaust bank 200 will be carried out only if for the first bank 100 the test on the probe jump has already ended at t ₂ or, in the event of a fault, sufficient time for the diagnosis to expire in the first exhaust gas bank 100 was given. This necessary, time shift, for example, is fixed, in particular via a Delta-T, or the time shift is determined by models that the OSC of the precatalyst 114 . 214 depending on gas flow rate, take into account catalyst temperature. Evaluated is either the lean-fat jump or the fat-lean jump of the probe signals 132 . 232 respectively. 130 . 230 ,

4 illustrates how interchanged second lambda probes 116 . 216 after the precatalysts 114 . 214 affect and be recognized. After the lambda adjustment at t ₁ for the first exhaust bank 100 the second lambda probe reacts 116 after the precatalyst 114 for this exhaust bank 100 not, but it jumps the second lambda probe 216 after the precatalyst 214 the second exhaust bank 200 , as with arrow 58 indicated. In the subsequent adjustment for the second exhaust bank 200 at t ₁ 'there is the corresponding same fault response of the second lambda probe 116 after the precatalyst 114 the first exhaust bank 100 whereas the second lambda probe 216 after the precatalyst 214 the second exhaust bank 200 does not react, as with arrow 60 indicated. If this malfunction is detected in several measuring phases, then a fault "second lambda probes 116 . 216 after the precatalyst 114 . 214 exchanged ". arrow 56 denotes a maximum time to which a probe jump is expected.

5 illustrates how swapped first lambda probes 112 . 212 in front of the precatalyst 114 . 214 affect and be recognized. For monitoring here is the lambda setpoint 30a in front of the precatalyst 114 the first exhaust bank 100 and the lambda setpoint 30b in front of the precatalyst 214 the second exhaust bank 200 with the lambda actual value 130 . 230 the first lambda probe 112 . 212 in front of the precatalyst 114 . 214 compared with actively requested adjustment. First, in the time interval t ₁ -t _2, a lambda setpoint adjustment for the first exhaust gas bank 100 output (for example, bold: clear catalyst). In the event of a fault, the first lambda probe will not react 112 (Signal 130 ) in front of the precatalyst 114 the first exhaust bank 100 on the adjustment, but the first lambda probe 212 in front of the precatalyst 214 the second exhaust bank 200 (Signal 230 ). The fat adjustment is ended via the lambda signal 132 ( 3 ) after the precatalyst 114 the first exhaust bank 100 , In the subsequent adjustment to lean for the first exhaust bank 100 in the time interval t ₂ -t ₃ the signal becomes 230 the first lambda probe 212 in front of the precatalyst 214 the second exhaust bank 200 show lean while the signal 130 the first lambda probe 112 in front of the precatalyst 114 the first exhaust bank 100 not reacted. The exact opposite behavior results in the time staggered lambda adjustments for the second exhaust bank 200 in the time intervals t ₁ '-t ₂ ' and t ₂ '-t ₃ '. After a corresponding applicable number of Verstellintervallen with confirmed error, a fault "first lambda probes 112 . 212 in front of the precatalyst 114 . 214 exchanged ". A signal 62 for "suspected error" and a signal 64 for "error" is in 5 shown.

In order to keep the active intervention of diagnoses, which can be negative both for reasons of consumption and for reasons of emission, as low as possible, it is provided to evaluate, depending on the diagnosis result, how many test cycles are necessary for a reliable result. This is also in 8th illustrated by a schematic flow chart. Returns the test cycle in step 66 "OSC measurement performed (1st result)" with respect to the catalyst diagnosis (OSC measurement) a very large OSC value, which is in the range of a result for a fresh catalyst, so is in a decision step 68 the turnoff 70 "OSC >> moderately aged" to step 72 "Measuring end" is selected and the active diagnosis is inhibited for this trip, insofar as the lambda signals are plausible during the test and i. O. were. The test result for catalytic converter and lambda probes is set to "checked" for this case (cycle flag or Z_flag set). If the test result for the OSC value is from step 66 between the value for a very good catalyst and a defective catalyst (NO), so in the decision step 68 the turnoff 74 "Moderately aged>OSC> border cat" to step 76 For example, x is a number from 2 to 5, and it will request a number of checks corresponding to the number x before the catalyst is considered tested. If the test result for the OSC value is from step 66 in the range of a defective catalyst (NO O.), then in the decision step 68 the turnoff 78 "OSC <boundary cat" to step 80 "Number y tests for Z_flag" is selected, where y is, for example, a number from 6 to 10, and thus a number of tests corresponding to the number y is requested before the catalyst is considered checked in order to debounce the error.

Claims (17)

Method for operating an internal combustion engine, in particular a motor vehicle, having an exhaust gas aftertreatment arrangement arranged in an exhaust tract of the internal combustion engine with at least one catalytic converter and a first lambda probe arranged upstream of the catalytic converter and a second lambda probe arranged downstream of the catalytic converter, wherein for determining an oxygen storage capability of the catalytic converter in a first Step (a) a lambda value upstream of the catalyst is actively adjusted to a value less than 1 until an oxygen stored in the catalyst is completely discharged, and then in a second step (b) the lambda value before the catalyst actively adjusted to a value greater than 1 is until the catalyst is completely loaded with oxygen, wherein the oxygen storage capacity of the catalyst is determined by means of an oxygen balance, while at the same time during the determination of the oxygen storage capacity least a parameter of the lambda probes is determined and the functionality of the lambda probes is determined from the parameter, wherein the determination of at least one parameter of the lambda probes comprises a plausibility check, wherein it is checked in the first step (a) and / or in the second step (b) whether first and the second lambda probe simultaneously display the expected rich or lean exhaust gas composition, characterized in that in the event that the plausibility check results in a plausibility violation, ie a different display of the exhaust gas composition before and after the catalyst by the two lambda probes, it is determined whether the first or second lambda probe is defective. A method according to claim 1, characterized in that the catalyst is a precatalyst or a main catalyst. A method according to claim 1, characterized in that the catalyst is a NOx storage catalytic converter or the exhaust aftertreatment device additionally comprises a catalytic converter downstream of the NOx storage catalytic converter, wherein under predetermined conditions a time-limited NO _x regeneration of the NO _x storage catalytic carried out, the NO _x Regeneration as the first step (a) performed and after completion of the NO _x regeneration of the second step (b) is performed. Method according to at least one of the preceding claims, characterized in that for determining the faulty lambda probe, a predetermined operating mode of the internal combustion engine, in particular a homogeneous lambda 1 operation, is set. Method according to at least one of the preceding claims, characterized in that for an exhaust gas aftertreatment arrangement with two catalysts arranged in parallel in the exhaust gas system, the determination of at least one parameter of the lambda probes comprises an exchange of the respective first lambda probes upstream of the catalysts and / or the respective second lambda probes downstream of the catalysts, wherein the step (a) for a second of the parallel catalysts offset by a predetermined time is executed later than for the corresponding other of the first two parallel catalysts, wherein an exchange of the connection of the two first lambda probes of the two catalysts is determined, if for predetermined number of cycles of steps (a) and (b) the first lambda probe of the second catalytic converter indicates a state change of the exhaust gas upstream of the first lambda probe of the first catalytic converter, and wherein an exchange of the connection of the two second lambda sensors of the two catalytic converters is determined if, for a predetermined number of cycles of steps (a) and (b ) the second lambda probe of the second catalyst indicates a state change of the exhaust gas before the second lambda probe of the first catalyst. A method according to claim 5, characterized in that the state change of the exhaust gas is a jump from lean exhaust gas composition to rich exhaust gas composition or a jump from rich exhaust gas composition to lean exhaust gas composition. Method according to at least one of the preceding claims, characterized in that the determination of at least one parameter of the lambda probes comprises a dynamic behavior of the lambda probes, wherein a gradient of the temporal change of an output signal of the lambda probes is determined and a faulty lambda probe is determined if the gradient is smaller than a is predetermined value. A method according to claim 7, characterized in that the gradient of the time change in the transition from rich exhaust gas composition to lean exhaust gas composition at the end of step (b) is determined. A method according to claim 7 or 8, characterized in that a maximum gradient or an averaged gradient is determined and compared with a corresponding, predetermined value. Method according to at least one of the preceding claims, characterized in that further test cycles are requested as a function of the result of the determination of the oxygen storage capacity. A method according to claim 10, characterized in that in the event that the oxygen storage capacity gives a value in the range of a fresh catalyst and the lambda probes have been recognized as functional, for a predetermined time interval, in particular until the next restart of the internal combustion engine, no further test cycles requested. A method according to claim 10 or 11, characterized in that in the event that the oxygen storage capacity results in a value in the range between a fresh catalyst and a defective catalyst, a predetermined first number, in particular 2 to 5, is requested by further test cycles. Method according to one of claims 10 to 12, characterized in that in the event that the oxygen storage capacity results in a value in the range of a defective catalyst, a predetermined second number, in particular 6 to 10, is requested by further test cycles. A method according to claim 12 and 13, characterized in that the first number of further test cycles is smaller than the second number of further test cycles. Method according to at least one of the preceding claims, characterized in that in the first step (a) a lambda value in the range of 0.95 to 0.98 or less is set. Method according to at least one of the preceding claims, characterized in that in the second step (b) a lambda value in the range of 1.02 to 1.05, in particular 1.03, is set. Method according to at least one of the preceding claims, characterized in that the determination of at least one parameter of the lambda probes comprises a test of a lean stress and / or a fatigue voltage of the first and / or second lambda probe.

Images