DE102014201589B3 - Method for monitoring an internal combustion engine having a particle filter, control unit, internal combustion engine, computer program product and machine-readable medium - Google Patents

Abstract

A method is proposed for monitoring an internal combustion engine (1) having a particle filter (9) with the following steps: time-resolved detection of a differential pressure across the particle filter (9) in a predetermined time interval (Δt); Determining a first time-resolved soot loading course (21) of the particulate filter (9) in the predetermined time interval (Δt) from the detected differential pressure; Fitting a first function (25) to the first soot loading course (21) in the predetermined time interval (Δt); Determining a second time-resolved Rußladungsverlaufs (23) in the predetermined time interval (.DELTA.t) based on a loading model in response to at least one parameter; Adjusting a second function (27) to the second soot loading course (23) in the predetermined time interval; Determining a first value (29) of a variable relevant for a future soot loading of the particulate filter (9) from the first function (25) and a second value (31) of the variable relevant for the future soot loading of the particulate filter (9) from the second function ( 27), and comparing the first value (29) with the second value (31).

Description

The invention relates to a method for monitoring a particle filter having an internal combustion engine according to claim 1, a control device for an internal combustion engine according to the preamble of claim 8, an internal combustion engine with a particulate filter according to the preamble of claim 9, a computer program product according to claim 10, and a machine-readable medium according to claim 11th

Methods for monitoring an internal combustion engine having a particle filter of the type discussed here are known. In particular, it is known to determine a soot charge of the particulate filter of such an internal combustion engine based on a measured differential pressure which drops across the particulate filter during operation of the internal combustion engine. The falling over the particulate filter differential pressure increases with increasing soot loading. A gradient of the time-resolved detected differential pressure is evaluated to determine if the engine is operating within a desired operating mode or if it has a fault condition. For example, if the gradient is too steep, it indicates that the engine is running. A disadvantage of such a method, however, is that a very expensive calibration is necessary.

From the European patent application EP 1 602 806 A1 shows an exhaust gas monitoring device for an internal combustion engine, which in particular has a first calculating means which calculates a first loading amount of a particulate filter from a dropping across the particulate filter differential pressure, wherein a second calculating means is provided which from an operating condition of the internal combustion engine one of the internal combustion engine per unit time emitted Calculating the amount of exhaust particles, wherein the second calculating means integrates the amount of exhaust particles, so as to calculate a second load amount, further comprising an evaluation means, which detects the presence or absence of an abnormality in an exhaust flow through the particulate filter in response to a Calculated relationship between the first load amount and the second load amount.

From the international patent application WO 2013/074023 A1 is a diagnostic method for an internal combustion engine, wherein an exhaust aftertreatment system is provided with at least one particulate filter to treat an exhaust gas flow resulting from combustion in the internal combustion engine, wherein in the process, a particle sensor disposed upstream of the particulate filter is used, to determine, at a location upstream of the particulate filter, a first particulate fraction of the exhaust gas flow from the internal combustion engine, and wherein the first particulate fraction is used as a basis to determine if the internal combustion engine is malfunctioning. It can be provided that a malfunction of the internal combustion engine is detected when the first particle fraction exceeds a second particle fraction by more than a first value in a first number of determinations, wherein it can be provided that the second particle fraction on the basis of prevailing operating conditions of the internal combustion engine is determined.

From the German publication DE 11 2007 002 658 T5 For example, a problem detection device in an exhaust emission control system has a filter that is disposed in an exhaust path in an internal combustion engine and traps particulates in the exhaust gas. A differential pressure sensor is provided which detects a pressure difference generated between the upstream side and the downstream side of the filter, the problem detecting device including a differential pressure calculating means that calculates a value of a pressure difference between the upstream and downstream sides of the filter based on an operating state of the internal combustion engine. Further, the problem detecting device includes judging means that calculates a difference between an actual value of the pressure difference detected by the differential pressure sensor between the upstream side and the downstream side of the filter and the calculation differential pressure value from the differential pressure value calculating means, compares the difference with a threshold, and detects that Filter has a problem.

The invention has for its object to provide a method, a control unit, an internal combustion engine, a computer program product and a machine-readable medium, in which the disadvantages mentioned do not occur.

The object is achieved by providing a method with the steps of claim 1. In this case, a differential pressure which drops during operation of the internal combustion engine above the particle filter is detected in a time-resolved manner in a predetermined time interval. From the detected differential pressure, a first time-resolved course of soot loading of the particulate filter in the predetermined time interval is determined. In particular, the first time-resolved soot load curve is calculated from the detected differential pressure. A first function is adapted to the first soot loading course in the predetermined time interval. A second time-resolved soot loading course in the predetermined time interval is determined on the basis of a loading model as a function of at least one parameter. In particular, the second time-resolved soot load course is calculated on the basis of the load model. A second function in the predetermined time interval is adapted to the second soot loading course. From the first function, a first value of a variable relevant for future soot loading of the particulate filter is determined. The first value is preferably calculated from the first function. From the second function, a second value of the variable relevant for the future soot loading of the particulate filter is determined. The second value is preferably calculated from the second function. The first value is compared to the second value. The fact that the method does not rely solely on the detected differential pressure, but in addition to the preferably independent of the differential pressure load model, wherein an evaluation and thus a monitoring of the internal combustion engine is based on a comparison of differential pressure based calculation on the one hand and the calculation based on the load model on the other hand, it does not require any elaborate calibration. Rather, the method is based on two independently derived forecasts of future soot loading, comparing the predictions of both forecasts to monitor the engine.

Under a differential pressure that drops across the particulate filter, a pressure difference in the exhaust gas flow upstream and downstream of the particulate filter is to be understood. This results in the differential pressure in particular as the difference of a pressure upstream, in particular immediately upstream of the particulate filter, and a downstream pressure, in particular immediately downstream of the particulate filter.

A determination of the soot load of the particulate filter from the differential pressure or a soot loading curve from a time-resolved detected differential pressure is known as such, so it will not be discussed further here.

The differential pressure is preferably determined by a differential pressure sensor which directly detects a differential pressure between a first measuring point upstream of the particle filter and a second measuring point downstream thereof. Alternatively, it is possible that a first pressure value upstream of the particulate filter is measured by means of a first pressure sensor arranged upstream of the particulate filter, wherein a second pressure value downstream of the same is measured by means of a second pressure sensor which is arranged downstream of the particulate filter. The differential pressure then results as the difference between the first and the second pressure value.

The time-resolved detection of the differential pressure is preferably carried out with a resolution of 5 ms to 10 s, more preferably with a resolution of 1 s. Preferably, the first and the second time-resolved soot loading course in the predetermined time interval with a resolution of 5 ms to 10 s, particularly preferably determined with a resolution of 1 s, in particular calculated.

The predetermined time interval is preferably chosen so that short-term fluctuations in the soot loading processes have little or no effect on the result and are averaged out in particular. The predetermined time interval is preferably at least 1 s to at most 1 h, preferably at least 1 min to at most 30 min, preferably at least 5 min to at most 15 min, particularly preferably 10 min.

As part of the method, a normalized differential pressure is preferably detected. It is possible, for example, that the differential pressure is related to a differential pressure value when the particle filter is unloaded or to a differential pressure value when the particle filter is loaded to the maximum. In this way, properties of the concrete particulate filter can be eliminated in favor of a simpler, in particular calibration-independent, approach.

The loading model is preferably a mathematical model which indicates the soot loading course of the particulate filter as a function of at least one parameter of the internal combustion engine and / or an exhaust gas flow thereof, the loading model preferably not or at least not explicitly depending on the differential pressure across the particulate filter. In this way, the soot load history calculated by differential pressure and the soot load history calculated based on the load model are independent of each other. The loading model furthermore preferably does not depend or at least does not depend directly on the condition of the particulate filter. However, it preferably allows a statement about the soot content in the exhaust gas of the internal combustion engine, based on the at least one parameter. Such loading models are known per se, so that will not be discussed in detail.

The loading model is preferably obtained on the basis of a simulation of the internal combustion engine virtually or computer-aided. In this case, a virtual or computer-aided calibration of the loading model is preferably carried out. This can very simple, time-consuming and cost-effective.

An embodiment of the method is preferred, which is characterized in that a desired mode of operation or a fault condition of the internal combustion engine is determined based on the comparison. The monitoring of the internal combustion engine is thus based on the comparison of the two independently determined values which are characteristic of a future soot loading of the particulate filter and thus represent a prognosis. If the independent forecasts differ significantly, the two values differ significantly, indicating that the engine is in an error state. On the other hand, if the two values coincide-preferably within a confidence interval-they therefore have a comparatively small deviation from one another, it is concluded that a desired mode of operation of the internal combustion engine is involved.

The fact that the method uses two different and preferably independent methods for determining the soot loading of the particulate filter, more different error conditions can be determined, as is the case with recourse to a single method. For example, in a method based solely on the sensed differential pressure, it is not possible to identify a fault condition in which the particulate filter is defective and allows too much soot to pass through. In this case, this results in a systematically too small gradient of the differential pressure, which does not appear as a fault condition in its sole evaluation. Even a fault condition of the particulate filter, in which this too little soot burns off, can not be determined by evaluating the differential pressure gradient alone, since this permanently sets a high differential pressure level. In addition, by comparing the differential pressure-based soot load history with the load-based model of soot load, errors can be detected by sensors whose measurements are included in the load model. These, in turn, would not be detectable as such in a method which relies exclusively on a loading model because this presupposes the error-free functioning of the sensors. An error state of an oxidation catalyst upstream of the particulate filter and thus too low a reactivity for combustion of the soot in the particulate filter can also be determined only by comparing the differently determined soot loadings. In that regard, the same applies as for a fault condition of the particulate filter, in which this burns too little soot.

Preferably, measures are taken in the context of the method when a fault condition of the internal combustion engine is detected. It is possible that a warning, for example in the form of lighting a warning light or generally a warning signal, is delivered to an operator of the internal combustion engine. Alternatively or additionally, it is possible that due to the detected error condition, a regeneration of the particulate filter is started in shorter time intervals. Alternatively or additionally, it is also possible that - in particular depending on the severity of the detected fault condition - the internal combustion engine is switched off. In one embodiment of the method, it is also possible that a normal regeneration of the particulate filter is controlled in the context of the method based on the comparison of the first value and the second value.

The first value used in a preferred embodiment of the method is a gradient of the first function at an end time of the predetermined time interval. Alternatively or additionally, it is possible that a gradient of the first function averaged over the predetermined time interval is used as the first value. Accordingly, in a preferred embodiment of the method, a gradient of the second function at the end time of the predetermined time interval is used as the second value. Alternatively or additionally, it is possible that a gradient of the second function averaged over the predetermined time interval is used as the second value. Preferably, matching values are used for the first value and for the second value, which however relate to the first function for the first value and to the second function for the second value.

Incidentally, a gradient of a derivative of a function under consideration is referred to here and in the following as a gradient.

An embodiment of the method is preferred, which is characterized in that the first value is determined by extrapolating the first function to a predetermined time after the predetermined time interval. The second value is preferably determined by extrapolating the second function to the predetermined time. It is obvious that in this case the first value and the second value are characteristic for a future soot loading of the particulate filter, in particular for soot loading at the predetermined time, which is later than the predetermined time interval. A distance of the predetermined time from the end of the predetermined time interval is preferably at least the length of the predetermined time interval, preferably 1.5 times the length of the predetermined time interval, preferably twice the length of the predetermined time interval Time interval. However, larger distances of the predetermined time to the end of the predetermined time interval are possible.

In a preferred embodiment of the method, the time interval between the end of the predetermined time interval and the predetermined time is 20 minutes. In particular, an embodiment of the method is preferred in which additionally the length of the predetermined time interval is 10 minutes, the distance of the predetermined time point to the end of the predetermined time interval being just twice its length.

An embodiment of the method is also preferred, which is characterized in that the first function is continuous, differentiable, or continuous and differentiable. Alternatively or additionally, the second function is preferably continuous, differentiable, or continuous and differentiable. Such functions can be adapted particularly easily to the determined soot loadings.

Particularly preferred is an embodiment of the method in which the first function is a straight line. Alternatively or additionally, the second function is preferably a straight line. The adaptation of a straight line to at least one of the two soot load profiles is particularly simple. Preferably, the method of least squares fitting is used for this purpose. The adaptation of a straight line to a soot load curve also has the advantage that an averaging of its course is performed in the predetermined time interval, so that short-term outliers make particularly little noticeable. Rather, by adjusting the straight line, a trend is obtained, which is then preferably extrapolated to a future, predetermined time point. The extrapolation is also particularly easy to carry out if a straight line is adapted to a soot loading course as a function.

An embodiment of the method is also preferred, which is characterized in that a deviation between the first and the second value is determined. This is compared with a predetermined limit deviation. It is then possible, for example, to determine a desired operation of the internal combustion engine when the deviation is smaller than the predetermined limit deviation. In contrast, a fault condition of the internal combustion engine can be determined if the deviation is greater than the predetermined limit deviation. Depending on the configuration of the method, it is possible to determine a desired mode of operation or a fault state of the internal combustion engine, if the deviation just corresponds to the predetermined limit deviation. Finally, by determining the deviation and comparing it with the predetermined limit deviation, it is determined whether the two soot loading history calculation methods correlate with each other within a specified confidence interval, assuming, in the positive case, that the system is functioning properly. On the other hand, if the deviation exceeds the predetermined limit deviation, a missing correlation of the two methods is found, from which an error can be deduced.

In this context, an embodiment of the method is preferred, which is characterized in that a difference between the first value and the second value is calculated as a deviation. In this case, the amount of the difference is preferably compared with the predetermined limit deviation. In this way, it can basically be determined whether the two methods correlate with each other, or whether a relevant deviation is found.

In a particularly preferred embodiment of the method, the sign of the difference is used to identify a fault condition of the internal combustion engine. If, for example, the particulate filter allows too much soot, the result is a systematically too small gradient of the differential pressure and the soot load calculated therefrom. In this case, the first value is likely to be significantly smaller than the second value, so that the corresponding sign results depending on the order of the difference formation. If, for example, the second value is subtracted from the first value, the result is a negative sign. On the other hand, if the particulate filter burns too little soot, a permanently high level of the pressure gradient and the soot charge calculated from it results, whereby the load model - which is based on a defect-free system - does not predict a correspondingly high load. Therefore, in this case, if the second value is subtracted from the first value, a positive sign results. If a sensor whose measured value is included in the load model has an error, the load model results in a systematically wrong soot load, which deviates from the soot load determined by the differential pressure. However, the sign of the difference between the first value and the second value in this case depends on the specific sensor error. A defect in an oxidation catalyst upstream of the particle filter leads to a too low burnup in the particle filter, resulting in a positive sign analogous to the case in which the defective particle filter burns off too little soot when the second value is subtracted from the first value.

It is also preferred an embodiment of the method, which is characterized in that the at least one parameter that enters the loading model is a parameter that is characteristic of combustion conditions in a combustion chamber of the internal combustion engine. In this way, the parameter says something about the development of soot during combustion, which allows conclusions about the soot content in the exhaust gas of the internal combustion engine and thus also allows conclusions about a course of soot loading in the particulate filter. It will be appreciated that such a parameter is independent of the actual differential pressure across the particulate filter, as well as the instantaneous state of the particulate filter.

Alternatively or additionally, the at least one parameter for the loading model is selected from a group consisting of a nitrogen oxide concentration in the exhaust gas, a lambda value and an exhaust gas temperature. Each of these values is characteristic of combustion conditions present in a combustion chamber of the internal combustion engine and thus also says something about the formation of soot during combustion. Alternatively or additionally, it is also possible to use a temperature of the internal combustion engine, in particular a coolant temperature thereof, in the context of the loading model. Preferably, the loading model is based on more than one parameter, in particular a plurality of parameters, in particular a plurality of the parameters mentioned here.

The method is preferably carried out continuously, with a new predetermined time interval in particular following each completed predetermined time interval with subsequent extrapolation and evaluation of the result. In this way, it is possible to continuously monitor the internal combustion engine during operation.

The object is also achieved by providing a control device for an internal combustion engine having the features of claim 8. The control unit is set up to carry out a method according to one of the previously described embodiments. In this case, it is possible for the control unit to be set up by suitable electronic circuits, thus by suitably designed hardware, for carrying out the method, this being implemented in the hardware of the control device. Alternatively, a control device is preferably provided in whose memory a computer program product is loaded, which has instructions on the basis of which the method is carried out when the computer program product is executed on the control device. In this case, the control unit is set up on the software side to carry out the method. In connection with the control unit, the advantages that have already been explained in connection with the method are realized.

The object is also achieved by providing an internal combustion engine with a particle filter having the features of claim 9. This is characterized by a control unit according to one of the embodiments described above. In this case, realized in connection with the internal combustion engine, the advantages that have already been described in connection with the control unit or with the method.

An embodiment of the internal combustion engine is preferred in which it has a sensor arrangement for detecting a differential pressure over the particle filter. The sensor arrangement preferably has a differential pressure sensor designed and arranged for this purpose. Alternatively, it is possible for the sensor arrangement to have a first sensor upstream of the particle filter and a second sensor downstream of the particle filter. The sensor arrangement is operatively connected to the control unit in such a way that a differential pressure can be detected directly by the differential pressure sensor and transmitted to the control unit, or a first pressure value can be detected by the first sensor and transmitted to the control unit, wherein a second pressure value can be detected by the second sensor and can be transferred to the control unit. In this last case, the control unit is preferably designed to calculate the differential pressure from a difference between the first pressure value and the second pressure value.

An internal combustion engine is also preferred which has at least one sensor for detecting the at least one parameter for the loading model. The at least one sensor is preferably selected from a group consisting of a nitrogen oxide sensor, a lambda probe, and a temperature sensor, in particular an exhaust gas temperature sensor. The at least one sensor is preferably operatively connected to the control unit, wherein the control unit is set up to calculate the loading model on the basis of the at least one parameter detected by the at least one sensor.

The internal combustion engine is preferably designed as a reciprocating engine. In a preferred embodiment, the internal combustion engine is used to drive in particular heavy land or water vehicles, such as mine vehicles, trains, the internal combustion engine is used in a locomotive or a railcar, or ships. It is also possible to use the internal combustion engine to drive a defense vehicle, for example a tank. An embodiment of the Internal combustion engine is preferably also stationary, for example, used for stationary power supply in emergency power, continuous load operation or peak load operation, the internal combustion engine in this case preferably drives a generator. A stationary application of the internal combustion engine for driving auxiliary equipment, such as fire pumps on oil rigs, is possible. Furthermore, an application of the internal combustion engine in the field of promoting fossil raw materials and in particular fuels, for example oil and / or gas, possible. It is also possible to use the internal combustion engine in the industrial sector or in the field of construction, for example in a construction or construction machine, for example in a crane or an excavator. The internal combustion engine is preferably designed as a diesel engine, as a gasoline engine, as a gas engine for operation with natural gas, biogas, special gas or another suitable gas. In particular, when the internal combustion engine is designed as a gas engine, it is suitable for use in a cogeneration plant for stationary power generation.

The particulate filter is formed in a preferred embodiment as a diesel particulate filter (DPF).

The object is also achieved by providing a computer program product having the features of claim 10. This has instructions on the basis of which a method is carried out according to one of the previously described embodiments, in particular if the computer program product - partially or completely - in the memory of a computer, in particular in the memory of an engine control unit, loaded, the instructions the computer, in particular on the control unit for the internal combustion engine, are executed. The advantages associated with the computer program product have already been explained in connection with the method.

Finally, the object is also achieved by providing a machine-readable medium having the features of claim 11. This has a computer program product according to the embodiment described above. In this case, the advantages that have already been explained above in connection with the method, the control unit, the internal combustion engine and the computer program product are realized.

The invention will be explained in more detail below with reference to the drawing. Showing:

1 a schematic representation of an embodiment of an internal combustion engine;

2 a schematic representation of an embodiment of the method for monitoring the internal combustion engine in a desired mode of operation, and

3 a schematic representation of the embodiment of the method when an error condition exists.

1 shows a schematic representation of an embodiment of an internal combustion engine 1 , This has an engine block here 3 on, in which at least one combustion chamber, here eight combustion chambers are arranged, of which for better clarity, only one with the reference numeral 5 is marked. Exhaust gas is from the combustion chambers 5 via an exhaust pipe 7 dissipated, the internal combustion engine 1 along the exhaust pipe 7 a particle filter 9 having.

It is a sensor arrangement 11 provided for the time-resolved detection of a differential pressure across the particulate filter 9 , For this purpose, the sensor arrangement 11 in the illustrated embodiment, a first, immediately upstream of the particulate filter 9 in the exhaust pipe 7 arranged pressure sensor 13 and a second, immediately downstream of the particulate filter 9 in the exhaust pipe 7 arranged pressure sensor 15 on. Alternatively, it is possible that a differential pressure sensor is provided, which with a first measuring point immediately upstream of the particulate filter 9 and connected to a second measuring point downstream thereof for detecting the differential pressure.

A control unit 17 is with the engine block 3 for controlling the internal combustion engine 1 operatively connected. In addition, the controller 17 with the first pressure sensor 13 and also with the second pressure sensor 15 operatively connected, via the first pressure sensor 13 a first pressure value and via the second pressure sensor 15 a second pressure value can be detected. About the active compounds are the pressure values to the control unit 17 forwarded, where from these pressures the differential pressure is calculated, which is above the particulate filter 9 drops. The control unit 17 is also set up to calculate a time-resolved soot loading curve of the particulate filter 9 from the detected differential pressure.

By means of a sensor 19 is a parameter detectable, which is characteristic of combustion conditions in the combustion chambers 5 , In the illustrated embodiment, the in the exhaust pipe 7 arranged sensor 19 preferably a lambda probe or an exhaust gas temperature sensor. The control unit 17 is with the sensor 19 operatively connected to capture the parameter. It is also set up to calculate a Loading model for soot loading of the particulate filter 9 depending on a measured value of the sensor 19 ,

As part of the process is by the sensor assembly 11 and the controller 17 a differential pressure across the particulate filter 9 detected at a predetermined time interval. The detected differential pressure becomes a first time-resolved soot load course of the particulate filter 9 in the predetermined time interval by the controller 17 calculated. In addition, by the control unit 17 a second time-resolved soot loading course in the predetermined time interval for the particulate filter 9 on the basis of a loading model as a function of at least one parameter, in particular as a function of the measured value of the sensor 19 , calculated.

2 shows a schematic, diagrammatic representation of an embodiment of the method for a case in which a desired operation of the internal combustion engine 1 is present. In this case, a soot loading B is plotted against the time t. A first, time-resolved soot loading course 21 , which is calculated from the detected differential pressure is shown as a dash-dotted curve. A second time-resolved soot loading course 23 calculated on the basis of the load model is shown as a dashed curve. Both soot loadings 21 . 23 are shown in a predetermined time interval between a time t = 0 and a time t = t ₁ , thus in a predetermined time interval At.

At the first soot loading course 21 becomes a first function in the predetermined time interval Δt 25 , here a straight line, adjusted. Accordingly, the second Rußbeladungsverlauf is 23 a second function 27 , here also a straight line, adapted.

The straights 25 . 27 are extrapolated to a predetermined time t ₂ , whose time interval from the end of the predetermined time interval .DELTA.t, ie here from the time t ₁ , just twice the length of the time interval .DELTA.t, ie 2Δt. By extrapolating the first function 25 becomes a first value 29 as the function value of the first function 25 obtained at the time t ₂ . Accordingly, a second value 31 by extrapolating the second function 27 as their function value at the time t ₂ .

The first value 29 comes with the second value 31 compared. In particular, a difference of the first value 29 from the second value 31 formed by, for example, the second value 31 from the first value 29 is deducted. The amount of the difference is preferably compared to a predetermined limit deviation. If the amount is smaller than the predetermined limit deviation, is of a desired operation of the internal combustion engine 1 went out. If, on the other hand, the amount exceeds the predetermined limit deviation, an error condition of the internal combustion engine becomes 1 detected. At the in 2 In the case shown, the amount of the difference is smaller than the predetermined limit deviation, so that a desired operation of the internal combustion engine 1 is detected.

It becomes clear that in the context of the method preferably two linear models for the soot loading of the particulate filter 9 which are extrapolated into the future.

The deviation of the two independent models in the future will determine if they correlate with each other or if they differ significantly. This can contribute not only to the detection of a fault condition but also to its characterization and preferably also to its identification.

3 shows a schematic, diagrammatic representation of the embodiment of the method according to 2 for a case where a fault condition of the internal combustion engine 1 is present. Identical and functionally identical elements are provided with the same reference numerals, so that reference is made to the preceding description. In this case it turns out that the first value 31 and the second value 29 vary greatly, the amount of their difference is greater than the predetermined limit deviation. So there is a fault condition. At the same time it shows that here is the second value 31 well above the first value 29 so that the sign is at subtraction of the second value 31 from the first value 29 is negative. In the case illustrated concretely, therefore, the loading model for the future predicts a greater soot loading of the particulate filter than the calculation based on the differential pressure. Such a situation may, for example, indicate a defective particulate filter that lets too much soot through.

For example, is the first value instead 29 well above the second value 31 , wherein the amount of the difference is greater than the predetermined limit deviation, this may be due to an erroneously strong sooting internal combustion engine 1 The difference-based calculation predicts a significantly higher soot loading for the future than the error-free system-based calculation using the loading model.

Based on 2 and 3 also shows that by adjusting the functions 25 . 27 which are straight lines here, a trend of both models is taken into account. Short-term outliers in the soot loading processes 21 . 23 do not lead to an error in the process.

Overall, it turns out that with the aid of the method, the control unit, the internal combustion engine, the computer program product and the machine-readable medium, it is possible to monitor an internal combustion engine with particle filter in a simple and reliable manner, whereby no time- and cost-intensive and expensive calibration is required.

Claims (11)

Method for monitoring a particulate filter ( 9 ) having internal combustion engine ( 1 ), comprising the following steps: time-resolved detection of a differential pressure above the particle filter ( 9 ) in a predetermined time interval (Δt); Determining a first time-resolved soot loading course ( 21 ) of the particulate filter ( 9 ) in the predetermined time interval (Δt) from the detected differential pressure; - Customize a first function ( 25 ) to the first soot loading course ( 21 ) in the predetermined time interval (Δt); Determining a second time-resolved soot charge history ( 23 ) in the predetermined time interval (Δt) on the basis of a loading model in dependence on at least one parameter; - customize a second feature ( 27 ) to the second soot loading course ( 23 ) in the predetermined time interval; - determining a first value ( 29 ) one for a future Rußbeladung the particulate filter ( 9 ) relevant size from the first function ( 25 ) and a second value ( 31 ) for the future soot loading of the particulate filter ( 9 ) relevant size from the second function ( 27 ), and comparing the first value ( 29 ) with the second value ( 31 ). A method according to claim 1, characterized in that based on the comparison of a desired operation or a fault condition of the internal combustion engine ( 1 ) is detected. Method according to one of the preceding claims, characterized in that the first value ( 29 ) by extrapolating the first function ( 25 ) is determined at a predetermined time (t ₂ ) after the predetermined time interval (Δt), the second value ( 31 ) by extrapolating the second function ( 27 ) is determined to the predetermined time (t ₂ ). Method according to one of the preceding claims, characterized in that the first and / or the second function ( 25 . 27 ) is continuous, differentiable, or continuous and differentiable, the first and / or the second function ( 25 . 27 ) is preferably a straight line / straight line / are. Method according to one of the preceding claims, characterized in that a deviation between the first value ( 29 ) and the second value ( 31 ) and compared with a predetermined limit deviation. A method according to claim 5, characterized in that as a deviation, a difference between the first value ( 29 ) and the second value ( 31 ), wherein preferably the amount of the difference is compared with the predetermined limit deviation, wherein the sign of the difference is particularly preferably used to determine an error state of the internal combustion engine ( 1 ) to identify. Method according to one of the preceding claims, characterized in that the at least one parameter for the loading model is a parameter which is characteristic of combustion conditions in a combustion chamber ( 5 ) of the internal combustion engine ( 1 ), and / or that the at least one parameter is selected from a group consisting of a nitrogen oxide concentration in the exhaust gas, a lambda value, and an exhaust gas temperature. Control unit ( 17 ) for an internal combustion engine ( 1 ), characterized in that the control unit ( 17 ) is arranged for carrying out a method according to one of claims 1 to 7. Internal combustion engine ( 1 ) with a particle filter ( 9 ), characterized by a control device ( 17 ) according to claim 8. A computer program product comprising instructions on the basis of which a method according to one of claims 1 to 7 is carried out when the instructions are executed on a computer, in particular on a control unit ( 17 ) for an internal combustion engine ( 1 ). A machine readable medium comprising a computer program product according to claim 10.

Images