DE102005062116A1 - Catalytic converter diagnosing method for internal combustion engine, involves changing lambda controller from rapid adaptation to long-term adaptation when catalytic converter diagnosis is completed - Google Patents

Abstract

The method involves making possible a rapid adaptation of a lambda controller (121) if a diagnosis-request signal is present. The lambda controller is switched-off during catalytic converter diagnosis, and a diagnosis-mixture with a controlled diagnosis mixture-lambda-desired-signal path is introduced into an internal combustion engine (110). The lambda controller is changed from the rapid adaptation to a long-term adaptation when the catalytic converter diagnosis is completed. Independent claims are also included for the following: (1) a device for diagnosing a catalytic converter in an exhaust gas area of an internal combustion engine (2) a control device program for executing all steps of a method for diagnosing a catalytic converter in an exhaust gas area of an internal combustion engine (3) a control device-program product with a program code for executing a method for diagnosing a catalytic converter in an exhaust gas area of an internal combustion engine.

Description

State of technology

The The invention relates to a method for monitoring an exhaust gas purification system an internal combustion engine with a in the emission control system arranged behind a catalyst, with a control device connected exhaust probe with jump characteristic, wherein for determination the oxygen storage capacity the catalyst targeted by means of a modeled lambda precontrol between a fat (λ <1) and a lean one (λ> 1) fuel-air mixture is changed.

The storage capacity an emission control system for Oxygen is used to absorb oxygen in lean phases and again in fat phases. This ensures that converts the pollutant gas components of the exhaust gas to be oxidized can be. As the emission control system ages, its storage capacity for oxygen decreases OSC (Oxygen Storage Capacity). This can be in the fat phases not enough anymore Oxygen available be set to purify the exhaust gas from the noxious gas components, and the lambda probe behind the emission control system detects this to oxidizing components. Furthermore, this lambda probe detects in longer Mealy phases the oxygen that is no longer from the emission control system can be stored. In many countries, a review of the Emission control system during driving required by the engine control by law (On-board diagnostics). An active catalyst diagnosis has the task of an inadmissible sinking the conversion that is too illegal increase the emission values leads, to recognize and over to display a warning light MIL (Malfunction Indicator Lamp).

One Known diagnostic method is the oxygen storage ability to determine the emission control system, as experience shows with the storage capacity also the conversion ability decreases.

The oxygen-measuring catalyst diagnosis is carried out with a broadband probe according to the following procedure:
The catalyst is first freed of oxygen by a rich mixture (λ <1). After this phase, referred to as conditioning, a lean exhaust gas (λ> 1) is subsequently introduced and the amount of oxygen introduced is integrated. If, during this measurement phase, the probe behind the catalyst indicates a lean, ie oxygen-containing, mixture, the integrated amount corresponds to the current oxygen storage capacity of the catalyst, which represents a measure of the quality. This procedure is applied several times in succession.

The Requirement for the accuracy of the catalyst diagnosis is such that a so-called GRENZKAT is actually diagnosed as a GRENZKAT becomes. As a CATEGORY, a catalyst is called the exhaust gas value just reached the legal limit. This one is in the Application as reference for the exhaust gas optimization and for the Diagnosis used.

to correct determination of the aging of the emission control system the amplitude of lambda oscillation behind the emission control system have to the average lambda value and the amplitude of the lambda oscillation the exhaust gas purification system, the exhaust gas mass flow and to what extent a stationary operating state the internal combustion engine is present with consideration. Special operating conditions as a coasting operation without fuel injection, in which the exhaust gas purification system 1.5 to 2 times the storage capacity as in lean operation, must be considered.

In the DE 41 12 478 C2 describes a method for assessing the aging state of a catalyst, in which the lambda values are measured in front of and behind the catalyst. It is examined whether the lambda value behind the catalyst shows a corresponding transition in the case of a control oscillation in front of the catalyst from rich to lean or, conversely, if this is the case, the gas mass flow flowing through the catalyst is determined as the time integral of the product from gas mass flow and lambda value before the catalyst is calculated, the temporal integral of the product of gas mass flow and lambda value is calculated behind the catalyst and as a measure of the aging state of the catalyst either the difference between the two integrals or the quotient of the two integrals or the quotient of the difference and one of the two integrals is used. A disadvantage of the method described is that the lambda value must be measured before the emission control system with a complex broadband lambda probe in order to determine the introduced or extracted amount of oxygen via the integration of the product of current lambda values and gas mass flow.

Another method for determining the oxygen storage capacity of an emission control system describes the EP 0546 318 B1 , The system is supplied with a Lambdaverlauf whose oxygen deficiency entry is initially higher than the oxygen storage capacity of the catalyst. The oxygen input is chosen so that the catalyst is filled in the lean phases each to its capacity limit. To determine the acid material storage capacity of the catalyst, the average lambda value before the catalyst during the lambda oscillations of the system is selectively shifted towards lean and thus reduces the oxygen removal from phase to phase. By determining the number of lean-to-rich transitions that the lambda probe located behind the catalytic converter indicates, the oxygen storage capability can be determined, with an increased number of phases resulting in reduced storage capability. A disadvantage of this method is that in the lean phases unpurified exhaust gas is discharged.

• – bei dem der Sauerstoffgehalt des Abgases hinter dem Katalysator bestimmt und
• – bei dem der mittlere Sauerstoffgehalt des Abgases vor dem Katalysator in eine Richtung verändert wird, die von dem zuvor bestimmten Sauerstoffgehalt hinter dem Katalysator wegführt und
• – bei dem die aus der Änderung des mittleren Sauerstoffgehaltes resultierende Änderung des Sauerstofffüllstandes des Katalysators bestimmt wird und mit einem vorbestimmten Grenzwert verglichen wird und
• – bei der eine Fehlermeldung unterbleibt, wenn der vorbestimmte Grenzwert überschritten wird bevor sich der Sauerstoffgehalt des Abgases hinter dem Katalysator ändert.

The DE 198 03 828 A1 describes a method and a device for checking an exhaust gas catalytic converter in internal combustion engines,

• - Determines the oxygen content of the exhaust gas behind the catalyst and
• - In which the average oxygen content of the exhaust gas before the catalyst is changed in a direction which leads away from the previously determined oxygen content behind the catalyst and
• - In which the resulting from the change in the average oxygen content change in the oxygen level of the catalyst is determined and compared with a predetermined limit and
• - In which an error message is omitted when the predetermined limit is exceeded before the oxygen content of the exhaust gas changes behind the catalyst.

at a two-point probe in front of the catalyst can the oxygen content of the input exhaust gas can not be determined quantitatively. This applies to both the fat mixture as well the lean mixture. An approach for the realization is to start from a correct precontrol and to model the oxygen content of the input exhaust gas. The punctual tolerances of the system have no effect activities inadmissible out, so the evaluation quality can not be enough to make a catalyst better or worse to diagnose as a BORDER CAT. In addition, in the further Course an additional Set deviation of feedforward, which additionally falsifies the result. at road measurements were observed in controlled operation mixture fluctuations, the in the same order of magnitude lay like the lambda lean. this leads to very big Scattering in the OSC determination.

It It is therefore an object of the invention to provide a method the accuracy of the feedforward control and thus the evaluation quality of the active catalyst diagnosis significantly improved especially with a two-point or jumping probe.

Advantages of invention

• – Überprüfung der Einschaltbedingungen
• – Bestimmung des mittleren Vorsteuerfehlers
• – Gesteuerte Fettphase
• – Gesteuerte Magerphase
• – Plausibilisierung eines modellierten Sondensprungs vor dem Katalysator sowie
• – Ermittlung und Ausgabe eines Mittelwertes

durchgeführt werden. Mit diesen Verfahrensschritten kann erreicht werden, dass eine deutlich verbesserte Genauigkeit der Lambda-Vorsteuerung erzielt werden kann, was insbesondere die Genauigkeit der Messung der Sauerstoffspeicherfähigkeit des Katalysators erhöht. Die aktive Katalysator-Diagnose kann dadurch im Hinblick auf vorhandene System-Toleranzen überprüft und bei Abweichungen adaptiert werden.The object is achieved in that the modeled lambda precontrol is specified such that the chronological steps

• - Checking the switch-on conditions
• - Determination of mean pilot control error
• - Controlled fat phase
• - Controlled lean phase
• - Plausibilisierung a modeled probe jump before the catalyst as well
• - Determination and output of a mean value

be performed. With these method steps can be achieved that a significantly improved accuracy of the lambda pilot control can be achieved, which in particular increases the accuracy of the measurement of the oxygen storage capacity of the catalyst. The active catalyst diagnosis can be checked with regard to existing system tolerances and adapted in case of deviations.

there is provided in a variant of the method that during the review of Switch-on conditions as an abort criterion an abort threshold for one Air mass change set as a relative deviation proportional to an air mass becomes. It can thus be ensured that the diagnosis is at a quasi-stationary one Operation performed can be.

A Variant provides that while the review of Switch-on conditions as a termination criterion the demolition threshold for the air mass change as a relative deviation proportional to a characteristic over the Air mass is set. This increases the accuracy for the determination the demolition threshold.

In A preferred variant of the method is used to determine the mean Pre-control error performed a quick adaptation and frozen a learning value. With that you can selective tolerances in a short time over several periods of the two-step control learned and, once the adaptation has settled, these stored become.

there it is envisaged that the rapid adaptation takes place when the general release conditions at the beginning of an active catalyst diagnosis Fulfills are. This ensures that the fast adaptation to a defined Time occurs and can be ruled out that due malfunctioning at the beginning of the active catalyst diagnosis faulty Adaptations performed and correspondingly incorrect learning values are stored.

If a fast adaptation is used, the learning value is frozen and a controller is deactivated Thereafter, the fat phases for the lambda may be selectively pre-controlled until the oxygen-free reference state of the catalyst is established, followed by the controlled lean phase for the lambda to perform the OSC measurement until the probe jump past the catalyst.

A preferred variant of the method provides that the plausibility of the modeled probe jump before the catalyst by comparison the dates for the modeled probe jump and a real probe jump the catalyst takes place. This allows system tolerances that to mixture errors in the lambda pilot control and thus to changes of the operating point can, if they are still within the tolerable range Limits are.

at one too big temporal deviation between the modeled probe jump and the real probe jump in front of the catalyst can diagnose the Oxygen storage capacity aborted, since then in general from an error the mixture formation must be assumed.

In preferred method variant is at a small temporal Deviation between the modeled probe jump and the real one Probe jump before the catalyst within the Plausi bilisierungsgrenzen a time model for the modeled lambda precontrol according to the deviation advanced or delayed and / or updated or overwritten. This will ensure that the real lambda lean skip with coincides with the modeled lean skip. At smaller, but still tolerable system tolerances can thus independently adapts the catalyst diagnostic procedure become, within certain limits, a self-learning fault tolerance equivalent.

at a simple on-board diagnosis, where the exact OSC value can not be determined can with the modeled probe jump behind the catalyst instead of with the modeled probe jump behind the catalyst can be started, which is a simplified OSC determination allows.

Become after several individual measurements, measurement outlier is eliminated and then averaging carried out, so can thus statistical fluctuations, as they occur in real systems can, be largely compensated. Short-term disturbances have only a very little influence on the measurement result for the OSC value.

drawing

The Invention will be described below with reference to one shown in the figure Embodiment explained in more detail. It shows:

1 a schematic representation of an internal combustion engine.

description the embodiments

1 shows by way of example a technical environment in which the method according to the invention runs. In the figure is an internal combustion engine 1 consisting of an engine block 40 and a supply air duct 10 that the engine block 40 supplied with combustion air, shown, with the amount of air in the supply air duct 10 with a Zuluftmesseinrichtung 20 is determinable. The exhaust gas of the internal combustion engine 1 is guided through an exhaust gas purification system, which as main components an exhaust duct 50 in which, in the flow direction of the exhaust gas, a first exhaust gas probe 60 in front of a catalyst 70 and a second exhaust gas probe 80 behind the catalyst 70 is arranged. Like in the 1 indicated, the emission control system can still a second catalyst 90 behind the second exhaust gas probe 80 exhibit.

The exhaust probes 60 . 80 are with a control device 100 connected from the data of the exhaust probes 60 . 80 and the data of the supply air measuring device 20 calculates the mixture and a fuel metering 30 for the metered addition of fuel with appropriate injectors in the supply air duct 10 controls. The control device can also, as in 1 shown with a display / storage unit 110 be connected.

With the in the exhaust duct 50 behind the engine block 40 arranged exhaust gas probe 60 can with the help of the control device 100 a lambda value can be adjusted, which is suitable for the exhaust gas purification system for achieving an optimal cleaning effect. The exhaust gas probe 60 can be designed as a simple jumping probe or as a complicated broadband probe, through which the air ratio lambda can be determined in a wide range. Is the exhaust gas probe 60 formed as a jump probe, this is less expensive, but allows only a control to a setpoint of λ = 1. Therefore, no lambda control can be active during the determination of the oxygen storage capacity. In this operating phase, only a pilot control of the fuel mixture with one of the control device takes place 100 predetermined lambda value. The in the exhaust duct 50 behind the first catalyst 70 arranged second exhaust gas probe 80 can also be in the controller 100 be evaluated and serves, in a method according to the prior art, the sow To determine erstoffspeicherfähigkeit the emission control system.

In general, during the on-board diagnosis, the internal combustion engine is determined to determine the oxygen storage ability 1 initially operated for a sufficiently long time with excess fuel ("rich mixture") to all the oxygen in the catalyst 70 to reduce. In the subsequent lean operation is in the catalyst 70 Oxygen stored, with the exhaust probe 80 is detected when oxygen-rich exhaust gas occurs, since the oxygen storage capacity OSC is then exceeded.

To the prior art, a sudden change between λ <1 and λ> 1 is specified. at this pilot control can Errors occur, for example, due to a scattering of the injectors or errors in the fill detection. These distort the lambda value and lead thus also errors in the determination of oxygen storage capacity.

• 1. Überprüfung der Einschaltbedingungen
• 2. Bestimmung des mittleren Vorsteuerfehlers
• 3. Gesteuerte Fettphase
• 4. Gesteuerte Magerphase
• 5. Plausibilisierung eines modellierten Sondensprungs vor dem Katalysator 70 sowie
• 6. Ermittlung und Ausgabe eines Mittelwertes

The timing of the diagnostic procedure is defined according to the invention in the following sub-steps:

• 1. Checking the switch-on conditions
• 2. Determination of mean pilot control error
• 3. Controlled fat phase
• 4. Controlled lean phase
• 5. Plausibility check of a modeled probe jump in front of the catalyst 70 such as
• 6. Determination and output of a mean value

in the In the first phase, as usual, the prerequisites for the procedure will be fulfilled active catalyst diagnostics. So be for example as switch-on conditions the parameters catalyst temperature, Gradient, speed, load, mass flow, and / or gradient checked. at known disorders a shutdown occurs already in this phase. The diagnosis will limited in time and at quasi-stationary Operation, i. in a restricted Area of an operating map or carried out at constant driving. While checking the Switch-on conditions become a termination criterion as a termination criterion for one Air mass change as a relative deviation proportional to an air mass (e.g., 20 kg / h). It can thus be ensured that the Diagnosis in a quasi-stationary Operation performed can be. Another variant provides that while checking the Switch-on conditions as a termination criterion the demolition threshold for the air mass change as a relative deviation proportional to a characteristic over the Air mass is set. This increases the accuracy for determining the Abort threshold. The demolition threshold for air mass changes can also remain absolutely constant in a map.

During the In the second phase, a quick adaptation is performed and a learning value is frozen. The fast adaptation can then take place if the general Release conditions at the beginning of an active catalyst diagnosis Fulfills are.

If the learning value is frozen and a controller is switched off in the case of a fast response, then the fat phases for the lambda can be selectively precontrolled in the third phase until the oxygen-free reference state of the catalyst is reached 70 followed by the controlled lean phase for the lambda (fourth phase) to the OSC measurement until probe exit of the exhaust gas probe 80 behind the catalyst. The controlled fat phase in the third phase is carried out, as hitherto customary, in that after a certain waiting time (typically 1 s) for a certain time, the lambda fat (for example, λ = 0.96) is piloted. This happens until it is ensured that even with a new catalyst 70 all the oxygen has been removed. This is achieved by having the exhaust gas probe 80 behind the catalyst 70 indicates the rich lambda value (ie, the probe voltage is> 600 mV), is still waiting for a certain time, which is determined by an integral of lambda deviation and air mass.

In the fourth phase, as hitherto customary, the controlled lean skip occurs, for example to a value of λ = 1.04. This turns oxygen into the catalyst 70 entered. The beginning of the OSC measurement takes place when the exhaust gas cloud according to a gas flow time model (without storage effect of the catalyst 70 ) on the exhaust gas probe 80 behind the catalyst 70 has arrived. The time from the start of the OSC measurement to the delayed response of the exhaust gas probe 80 behind the catalyst 70 (Probe voltage <450 mV) provides the information for the oxygen storage capacity.

Due to system tolerances, mixture deviations in the lambda precontrol can occur, causing slight changes in the operating point. Therefore, in the phase 5 a plausibility check of the runtime model with the real probe jump of the exhaust gas probe 60 in front of the catalyst 70 required, which determines the starting point. Mixture errors can be detected by means of the exhaust gas probe 60 in front of the catalyst 70 detected, for example, if the probe jump is too early compared to the modeled time. In this case, the mixture is too lean and the measured OSC capacity is assessed as being too small or too poor as a result. If the probe jump is too late, the mixture is too rich and the measured OSC capacity is judged to be too large or too good.

If the time deviates too much from the modeled start time, the diagnosis is aborted. With a small time difference between the modeled probe jump and the real probe jump in front of the catalyst 70 Within the plausibility limits, the time model for the modeled lambda precontrol according to the deviation is advanced or delayed and / or updated or overwritten. The temporal deviation of the real measurement to the transit time model can be stored symmetrically, as described above, or also be stored asymmetrically as well as by means of two characteristics over the modeled running time in the model.

Depending on the requirement, the runtime model can be more or less complex. For simple systems (eg for the European On-Board-Diagnostic EOBD), where the exact OSC value need not be specified, the OSC measurement can easily be done with the modeled probe jump in front of the catalyst 70 to be started.

Finally, in phase 6 After several individual measurements, the measuring outliers were eliminated and then averaging was carried out. Thus, statistical fluctuations, as they can occur in real systems, are largely compensated. Short-term disturbances thus have only a very small influence on the measurement result for the OSC value.

With the method and the corresponding device can be achieved that a higher accuracy the determination of the oxygen storage capacity of the emission control system with a simple structure allows becomes. This is particularly advantageous in terms of cost reduction for one On-board diagnostics.

Claims (11)

Method for monitoring an exhaust gas purification system of an internal combustion engine ( 1 ) with a in the exhaust gas purification system behind a catalyst ( 70 ), with a control device ( 100 ) connected exhaust gas probe ( 80 ) with jump characteristics, wherein for determining the oxygen storage capacity of the catalyst ( 70 ) is selectively changed by means of a modeled lambda pilot control between a rich (λ <1) and a lean (λ> 1) fuel-air mixture, characterized in that the modeled lambda precontrol is specified such that the chronological steps - Checking the switch-on conditions - Determining the mean pilot control error - Controlled rich phase - Controlled lean phase - Plausibility check of a modeled probe jump in front of the catalytic converter ( 70 ) as well as - determination and output of an average value. Method according to claim 1, characterized in that that while the review of Switch-on conditions as an abort criterion an abort threshold for one Air mass change set as a relative deviation proportional to an air mass becomes. Method according to claim 1, characterized in that that while the review of Switch-on conditions as a termination criterion the demolition threshold for the air mass change as a relative deviation proportional to a characteristic over the Air mass is set. Method according to one of claims 1 to 3, characterized that a fast adaptation is carried out to determine the average pilot control error and a learning value is frozen. Method according to one of claims 1 to 4, characterized that the quick adaptation then occurs when the general release conditions are fulfilled at the beginning of an active catalyst diagnosis. Method according to one of claims 1 to 5, characterized that the learning value is frozen in the case of a settled fast adaptation and a regulator is turned off. Method according to one of claims 1 to 6, characterized in that the plausibility of the modeled probe jump before the catalyst ( 70 ) by comparing the times for the modeled probe jump and a real probe jump before the catalyst ( 70 ) he follows. Method according to claim 7, characterized in that if there is too great a time difference between the modeled probe jump and the real probe jump in front of the catalytic converter ( 70 ) the oxygen storage ability diagnosis is aborted. Method according to claim 7, characterized in that with a small temporal deviation between the modeled probe jump and the real probe jump in front of the catalytic converter ( 70 ) within the plausibility limits a time model for the modeled lambda precontrol according to the deviation is preferred or delayed and / or updated or overwritten. Method according to one of claims 1 to 9, characterized in that in a simple on-board diagnosis with the modeled probe jump in front of the catalyst ( 70 ) instead of the modeled probe jump behind the catalyst ( 70 ) is started. Method according to one of claims 1 to 10, characterized that after several individual measurements, measurement outliers are eliminated and then one Averaging performed becomes.