DE102021209373A1 - Method for determining the current loading of a particle filter of an internal combustion engine - Google Patents

Abstract

The invention relates to a method for determining a loading of a particle filter according to the wall flow principle, comprising determining (304, 404) a first modeled flow resistance (k _flow,ident ) of the particle filter on the basis of an exhaust gas mass flow (ṁ _exh ), a differential pressure (Δp) over the particle filter and a temperature (T _us ) and a pressure (p _us ) in front of the particle filter; determining (375, 475) a second modeled flow resistance (k _flow,mod ) of the particle filter based on an exhaust gas mass flow, a modeled current loading of the particle filter, and a temperature (T _us ) and a pressure (p _us ) in front of the particle filter; Forming a difference between the first and the second modeled flow resistance, and calculating at least one correction value (ṁ _soot,in,corr , ṁ _{soot,layer,corr} , ṁ _{soot,wall,corr} ) for a particle mass flow into the particle filter from the difference formed, and determining (355, 365, 455, 465) a current adjusted loading of the particle filter using the at least one correction value for the particle mass flow.

Description

The present invention relates to a method for determining the current loading of a particle filter for an internal combustion engine, as well as a computing unit and a computer program for carrying it out.

Background of the Invention

The strict limit values of today's emissions legislation can no longer be achieved with measures internal to the internal combustion engine alone. In order to reduce exhaust emissions to the level required by law, particle filters are increasingly being used in the exhaust system of both diesel and petrol engines. These are often wall-flow filters, which become loaded with soot and ash particles over the course of the vehicle's life, which in turn significantly increase the filtration efficiency of the filter. Particles are deposited both within the porous wall and as a layer of particles on the wall surface. While the ash remains permanently in the filter and thus makes a constant contribution to filtration, the contribution to filtration by the soot is particularly dependent on the soot particle load. Optimum filtration efficiency can therefore be achieved through a targeted build-up or breakdown of a layer of soot particles under certain conditions, such as high filter temperatures or excess oxygen.

Empty particle filters only have a limited filtration efficiency. It is therefore important that these filters are preloaded with ash or soot as soon as possible to bring filtration efficiency to the required level.

Targeted pre-loading of the filter before installation or heavy sooting of the filter shortly after starting the engine are suitable measures for this. The filter efficiency increases as the journey progresses, as the individual pores in the filter wall are increasingly filled.

On the other hand, if the temperatures are sufficiently high and there is a large supply of oxygen, the layer of soot particles that forms on the filter wall can burn off very quickly, which leads to a severe loss in filtration efficiency. It is therefore of crucial importance to stop this burning process, i.e. the oxidation of the soot particle layer, in good time so that the particle filter is only regenerated up to a certain limit and the filtration efficiency is kept at the desired level.

Over the course of the vehicle's life, the particulate filter becomes increasingly clogged with ash, which, unlike the soot particles, cannot be burned off again and thus ensures permanently high filtration efficiency. If the load is too high, however, the exhaust back pressure increases, which in turn leads to disadvantages in fuel consumption. As a result, the filter should be replaced when the ash load reaches a critical level.

A method based on a measured differential pressure across the particle filter can be used to determine the filter loading. For this purpose, the difference in the pressures at the inlet and outlet of the filter is recorded and a change in the filter flow resistance is derived from the change in this differential pressure, which in turn allows an approximate conclusion to be drawn about the particle load of the entire filter. This rough loading value can be used to control measures to improve filtration efficiency. In certain driving situations, for example, it is possible to increase the build-up of soot, reduce soot or stop an exothermic soot burn-off reaction.

Disclosure of Invention

According to the invention, a method for determining, in particular dynamic modeling, of a loading of a particle filter according to the wall-flow principle, as well as a computing unit and a computer program for its implementation with the features of the independent patent claims are proposed. Advantageous configurations are the subject of the dependent claims and the following description.

This solution creates the possibility of being able to make a better statement about the proportion of the particle filter that is currently loaded with ash or soot and where the current efficiency of the filter actually lies. This makes it easier to keep the efficiency of the particle filter in an optimal range with suitable measures. In particular, with the proposed method, other influences on the filter loading, such as fuel quality, effects in the Regenera tion process can be included indirectly by using the described measured and modeled values and the previous mathematical models for loading while driving can be continuously corrected and updated.

A method is proposed in detail, in which a first modeled flow resistance of the particle filter is determined on the basis of an exhaust gas mass flow, a differential pressure across the particle filter and a temperature and a pressure in front of the particle filter, and a second modeled flow resistance of the particle filter is determined on the basis of an exhaust gas mass flow, a modeled current loading of the particle filter and a temperature and a pressure in front of the particle filter is determined. The difference between these first and second modeled flow resistances is then formed, and at least one correction value for a particle mass flow into the particle filter is formed from the difference formed. The values for the exhaust gas mass flow can be measured directly via a sensor or determined from other values such as temperature and pressure. The differential pressure can also be measured directly or, for example, determined from two pressure measurements before and after the filter. Likewise, the values for temperature and pressure can also be at least partially calculated.

In all cases, it is also possible to interpolate values or estimate them from approximation functions based on individual measurements for further points in time. An adapted current loading of the particle filter can then be determined using the at least one correction value for the particle mass flow. By using the correction value, a more precise determination of the actual loading of a particulate filter is achieved. This in turn makes it possible to initiate measures at an early stage (e.g. adjustments to the fuel system or overrun cut-off ban) in order to increase filtration efficiency in good time or to counteract excessive accumulation of soot particles in the filter for reasons of component protection (e.g. high temperature stress during combustion). A more targeted use of these measures can also prevent measures with unnecessarily high fuel consumption or increased carbon dioxide emissions being triggered due to inaccurate data on the filter behavior.

The determination of an adapted modeled filter loading can further include the calculation of an adapted particle mass flow into the particle filter based on a modeled filter efficiency and a particle mass flow into the filter calculated on the basis of a raw emission model, with a correction value for the particle mass flow being added to the particle mass flow calculated on the basis of the raw emission model becomes.

In exemplary embodiments, the determination of the current adjusted loading of the particle filter can include in particular the determination of a current adjusted loading of the filter wall and the determination of a current adjusted loading of the particle layer of the particle filter.

Such a determination of the adjusted loading of the wall and particle layer can be made by determining a filter efficiency of the filter wall and the particle layer depending on a current loading of the particle filter, and then determining the current adjusted loading of the filter wall based on the filter efficiency of the filter wall and a on the basis of a raw emission model calculated particle mass flow into the filter is calculated, and analogously the current adjusted loading of the particle layer is calculated on the basis of the filter efficiency of the particle layer and a particle mass flow calculated on the basis of a raw emission model into the filter. To calculate the filter efficiency, an adapted loading determined in a previous run can also be used in order to dynamically adapt the method.

Alternatively or in addition to using a correction value for the particle mass flow entering the filter from an untreated emission model, a first correction value can first be determined from the difference between the modeled flow resistances, and then this first correction value can be converted into a second correction value for a particle mass flow into a filter wall of the particle filter and a third correction value for a particle mass flow in a particle layer of the particle filter. In this way, the correction can be applied more accurately for wall and layer, especially in the case of very different mass flows, e.g. in the course of a full regeneration.

The currently adjusted loading of the filter wall can be calculated, for example, by integrating a first input value over time, the first input value being formed from the particle mass flow into the filter wall, the second correction value and a particle mass flow from the filter wall resulting from filter regeneration; and the current adjusted loading of the batch The particle layer can be calculated by integrating a second input value over time, the second input value being formed from the particle mass flow into the filter layer, the third correction value and a particle mass flow from the particle layer resulting from filter regeneration.

In all of the embodiments, a current, adapted loading of the particle filter, which was calculated with the aid of the method steps mentioned, can be used, for example, to control a regeneration of the particle filter. In addition, the currently adjusted loading can be used in all calculations and models that use the loading of the particle filter, so that a dynamic adjustment occurs during active driving.

To implement the modeling and calculations described, the method can include measuring at least one of the following parameter values continuously or at predetermined intervals: a pressure in front of the particle filter, a temperature in front of the particle filter, an exhaust gas mass flow in front of the particle filter, an exhaust gas mass flow in front of the particle filter, a Differential pressure across the particle filter. These values can be recorded by sensors, saved and optionally adjusted with further intermediate steps, such as approximation functions or parameter estimates, for use in the models.

A major advantage of the present invention is that, with precise information on the loading of the particle filter of a combustion engine (Otto or Diesel), one is able to take measures to improve filtration efficiency (e.g. adjustments in the fuel system so that larger soot particles are formed) and to protect components (e.g. overrun cut-off ban, i.e. injection despite the overrun phase to reduce excess oxygen). A high level of filtration efficiency is considered a decisive criterion in order to be able to comply with the demanding limit values in exhaust gas legislation. Furthermore, fuel consumption and carbon dioxide emissions can be reduced if no unnecessary fuel cutoff bans are issued. A further advantage of the present invention is that the physical models used are tracked in a targeted manner with the measured differential pressure. Since a deviation between the modeled and measured differential pressure often results in a deviation of the modeled soot mass from the actual soot mass, the accuracy of the entire model can be improved by adapting the measured differential pressure accordingly.

A computing unit according to the invention, e.g. a control unit of a motor vehicle, is set up, in particular in terms of programming, to carry out a method according to the invention.

The implementation of a method according to the invention in the form of a computer program or computer program product with program code for carrying out all method steps is advantageous because this causes particularly low costs, especially if an executing control unit is also used for other tasks and is therefore available anyway. Suitable data carriers for providing the computer program are, in particular, magnetic, optical and electrical memories, such as hard drives, flash memories, EEPROMs, DVDs, etc. It is also possible to download a program via computer networks (Internet, intranet, etc.).

Further advantages and refinements of the invention result from the description and the attached drawing.

The invention is shown schematically in the drawing using exemplary embodiments and is described below with reference to the drawing.

character list

• 1 shows an example of a particle filter in a vehicle in which a method according to the invention can be applied;
• 2a and 2 B represent an overview of the physical model for a particulate filter in which a method according to the invention can be used;
• 3a shows schematically the calculation of a correction value for the particle mass flow in a possible embodiment;
• 3b shows an exemplary modeling of parameters of the particle filter, in which a correction value for the particle mass flow can be incorporated;
• 4a shows schematically the calculation of divided correction values for the particle mass flows in the filter wall and soot layer according to a possible embodiment; and
• 4b shows an exemplary modeling of parameters of the particle filter, in which divided correction values for filter wall and layer are incorporated.

Embodiments of the invention

1 shows an example of a particle filter element 1 schematically in longitudinal section, in which embodiments of the invention can be used. In this case, exhaust gas 4 originating from an internal combustion engine and containing particles of different sizes and masses is introduced into the filter. The flow of exhaust gas through the filter is indicated by arrows. Inside the filter, there are wall elements 10 made of a porous or partially permeable material, which can, for example, form a honeycomb structure with channels 12 that are mutually closed off at their ends, so that the inflowing exhaust gas 4 has to pass through the channel walls 10 and the particles contained therein and be deposited in the walls 10. The layer of particles or soot that builds up in the filter on the walls as a particle layer due to the exhaust gases flowing through is referred to below as the particle layer or soot layer. The walls of the particle filter through which the exhaust gases flow and in which particles are also deposited due to their porous structure are referred to as the filter wall. The exhaust gas filtered in this way is then passed on in the exhaust system 6 .

Depending on the design, a particle filter element 1 can also be combined with a catalytic converter, or one or more catalytic converter elements can be provided at other points in the exhaust system.

Various sensors can be provided before and after the filter element 1 in order to determine measured variables of the exhaust gas flow or of the filter during operation. For example, a pressure sensor is used before and/or after the filter element to measure the pressure before (p _us ) or after (p _ds ) the filter element. A differential pressure can then be determined from the two values. It is also possible that, alternatively or additionally, there is a single pressure sensor for the direct measurement of the differential pressure Δp across the filter element.

A mass flow sensor, for example an air mass meter or flow sensor, can be used to measure the exhaust gas mass flow ṁ _exh . The oxygen concentration w _O2 in the exhaust gas flow can be determined by an oxygen sensor such as a lambda probe. The oxygen content is required, for example, to model the exhaust gas components using the raw emissions model. In addition, the temperature T of the exhaust gas can be recorded at various points in the exhaust system, for example before (T _us ) and/or after (T _ds ) the particle filter. A temperature sensor can also be installed in the particulate filter itself.

The measured sensor values recorded can then be sent to a suitable control device, such as the engine control unit or a separate control unit, where they can be further processed for various purposes, e.g. to control engine operation, to analyze relevant variables in the exhaust system, to control measures for exhaust gas treatment or for modeling other unknown quantities. Likewise, these measured values can also be stored or forwarded to other control devices. Conversely, it is also possible for measured values that are subsequently used for the calculation not to be received directly as an output signal from the respective sensors, but rather by a control unit or a memory element. In turn, the control unit has a computing unit such as a processor or a microcontroller, which is able to use the recorded measured values to carry out specified calculations and modeling of variables that are relevant for the control and also to store these modeled values and forward them to other units or continue to use it in further modelling.

In particular, the sensor readings can also be used to determine dynamic correction values for the filter loading of the particle filter, as described below.

2a and 2 B show an example of an overview of a physical model for determining various filter parameters, in particular the loading of a particle filter. The values for the filter wall are each denoted by the subscript "wall", while the values for the soot layer are denoted by the subscript "layer".

• Konstandopoulos A. G., „Flow Resistance Descriptors for Diesel Particulate Filters: Definitions, Measurements and Testing“, SAE Tech. Paper No. 2003-01-0846 (SP-1755), 2003;
• Konstandopoulos A. G. und Kostoglou M., „Periodically Reversed Flow Regeneration of Diesel Particulate Traps“, SAE Tech. Paper No. 1999 01-0469, 1999; Mohammed H, Triana AP, Yang SL, Johnson JH. An advanced 1D 2-layer catalyzed diesel particulate filter model to simulate: filtration by the wall and particulate cake, oxidation in the wall and particulate cake by NO2 and O2, and regeneration by heat addition. SAE technical paper 2006-01-0467; 2006; Konstandopoulos A. G., Kostoglou M., Skaperdas E., Papaioannou E., Zarvalis D., and Kladopoulou E., „Fundamental Studies of Diesel Particulate Filters: Transient Loading, Regeneration and Aging“, SAE Tech. Paper No. 2000-01-1016 (SP-1497), 2000.
• C.T. Huynh, J.H. Johnson, S.L. Yang, S.T. Bagley, J.R. Warner, A One-Dimensional Computation Model for Studying the Filtration and Regeneration Characteristics of a Catalyzed Wall-Flow Diesel Particulate Filter, SAE 2003-01-0841, 2003.

The basic physical models used here, such as a differential pressure model, an untreated emissions model for modeling the exhaust gas components that occur during engine operation, various models for regenerating particle filters, various models for parameters such as the flow resistance in the filter and other relationships are known in the art. For the details of the modeling and the mathematical relationships that are not essential to the idea of the invention, reference is expressly made to the following documents, for example:

• Konstandopoulos AG, "Flow Resistance Descriptors for Diesel Particulate Filters: Definitions, Measurements and Testing", SAE Tech. paper no. 2003-01-0846 (SP-1755), 2003;
• Konstandopoulos AG and Kostoglou M., "Periodically Reversed Flow Regeneration of Diesel Particulate Traps", SAE Tech. paper no. 1999 01-0469, 1999; Mohammed H, Triana AP, Yang SL, Johnson JH. An advanced 1D 2-layer catalyzed diesel particulate filter model to simulate: filtration by the wall and particulate cake, oxidation in the wall and particulate cake by NO2 and O2, and regeneration by heat addition. SAE technical paper 2006-01-0467; 2006; Konstandopoulos AG, Kostoglou M., Skaperdas E., Papaioannou E., Zarvalis D., and Kladopoulou E., "Fundamental Studies of Diesel Particulate Filters: Transient Loading, Regeneration and Aging", SAE Tech. paper no. 2000-01-1016 (SP-1497), 2000.
• Huynh CT, Johnson JH, Yang SL, Bagley ST, Warner JR, A One-Dimensional Computation Model for Studying the Filtration and Regeneration Characteristics of a Catalyzed Wall-Flow Diesel Particulate Filter, SAE 2003-01-0841, 2003.

To model the filtration efficiency of the particulate filter in blocks 210 and 220 of 2a both directly measurable and calculated variables are used, such as the exhaust gas mass flow ṁ _exh , pressure p _us and temperature T _us in front of the particle filter, the temperature in the particle filter, the oxygen concentration w _O2 in the exhaust gas mass flow, the loading of the filter with ash m _ash and the mass flow of raw emissions. It is also important to know the mass distribution of the soot between the filter wall and the built-up layer of soot particles. For this purpose, both the incoming soot mass flow ṁ _soot,in , which loads the filter, and the change in the soot mass or the soot mass flow due to regeneration for the filter wall, ṁ _{soot,wall,rgn} and the soot particle layer, ṁ _soot, are calculated in the model. _{layer, rgn} determined. In the present example, the soot particle mass is used; alternatively, however, the number of particles could also be used in all calculations and modelling.

To calculate the particle mass flow ṁ _soot,in with which the filter is loaded, as well as the negative particle mass flow due to a regeneration process or burn-off in the filter wall and in the soot particle layer, non-linear models are usually used, which are based on physical interrelationships and in the sources mentioned above are detailed.

In addition to the raw mass flow ṁ _soot,in of particles that flows into the filter, the efficiency of the filter wall and the soot particle layer also play an important role in calculating the filter loading.

für die Veränderung der Beladung der Filterwand, die in Block 240 der 2a modelliert wird, und $${\dot{\text{m}}}_{\text{soot},\text{layer},\text{stored}}={\mathrm{\eta }}_{\text{layer}}\cdot {\dot{\text{m}}}_{\text{soot},\text{in}}$$

für die Veränderung der Partikelschicht, die in Block 230 der 2a modelliert wird,
wobei ṁ_{soot,wall,stored} der Rußmassenstrom bzw. die Änderung der Partikelmasse in der Filterwand beim Beladen des Filters durch den eingehenden Abgasstrom ist, ṁ_{soot,layer,stored} der Rußmassenstrom bzw. die Änderung der Rußmasse in der Rußschicht beim Beladen des Filters ist,
η_layer der Wirkungsgrad der Rußschicht ist,
η_wall der Filterwirkungsgrad der Filterwand, und
ṁ_soot,in der Rußmassenstrom ist, der in den Filter eintritt, also der Massenstrom der Rohemissionen im Abgas.applies $${\dot{\text{m}}}_{\text{soot},\text{wall},\text{stored}}=\left(1-{\mathrm{n}}_{\text{layer}}\right)\cdot {\mathrm{n}}_{\text{wall}}\cdot {\dot{\text{m}}}_{\text{soot},\text{in}}$$

for changing the loading of the filter wall, which is carried out in block 240 of the 2a is modeled, and $${\dot{\text{m}}}_{\text{soot},\text{layers},\text{stored}}={\mathrm{n}}_{\text{layers}}\cdot {\dot{\text{m}}}_{\text{soot},\text{in}}$$

for the change in the particle layer, which is determined in block 230 of the 2a is modeled,
where ṁ _{soot,wall,stored is} the soot mass flow or the change in the particle mass in the filter wall when the filter is loaded by the incoming exhaust gas flow, ṁ _{soot,layer,stored is} the soot mass flow or the change in the soot mass in the soot layer when the filter is loaded ,
η _layer is the efficiency of the soot layer,
η _wall is the filter efficiency of the filter wall, and
ṁ _soot,in is the soot mass flow that enters the filter, i.e. the mass flow of raw emissions in the exhaust gas.

The efficiencies for the filter wall and soot particle layer depend on the exhaust gas mass flow, the pressure p _us in front of the particle filter and the temperature T _us in front of the particle filter, as well as the currently existing loading in the wall and soot particle layer.

und in Block 220 die Filtereffizienz der Filterwand, $${\mathrm{\eta }}_{\text{wall}}=\text{f}2\left({\text{m}}_{\text{soot},\text{wall}},{\dot{\text{m}}}_{\text{exh}},{\text{T}}_{\text{us}},{\text{p}}_{\text{us}}\right)$$

gebildet wird.This is done in the modeling 2a each calculated the filter efficiency as a function of these variables, where
in block 210 the filter efficiency of the particle layer in the filter, $${\mathrm{n}}_{\text{layers}}=\text{f}1\left({\text{m}}_{\text{soot},\text{wall}},{\text{m}}_{\text{soot},\text{layers}},{\text{m}}_{\text{ash}},{\dot{\text{m}}}_{\text{exh}},{\text{T}}_{\text{us}},{\text{p}}_{\text{us}}\right)$$

and in block 220 the filter efficiency of the filter wall, $${\mathrm{n}}_{\text{wall}}=\text{f}2\left({\text{m}}_{\text{soot},\text{wall}},{\dot{\text{m}}}_{\text{exh}},{\text{T}}_{\text{us}},{\text{p}}_{\text{us}}\right)$$

is formed.

wobei ṁ_{soot,layer,rgn} der durch die Regeneration entstehende Partikelmassenstrom der Rußpartikelschicht, m_soot,layer die aktuelle Beladung der Rußpartikelschicht, m_ash die Aschbeladung des Filters, T_us die Temperatur vor dem Partikelfilter, p_us der Druck vor dem Partikelfilter und w_O2 der Sauerstoffgehalt im Abgasstrom ist. 2 B shows schematically the modeling of the filter regeneration. When calculating the particle mass flow during the regeneration of the soot particle layer, i.e. the mass flow of soot particles that are burned off by the regeneration in the filter wall, the exhaust gas mass flow ṁ _exh , the pressure p _us and the temperature T _us play a role in addition to the current loading in the soot particle layer before the particle filter and the oxygen ratio w _O2 in the exhaust gas flow play an important role, so that the particle mass flow through regeneration is essentially a function of these variables, represented by block 201: $${\dot{\text{m}}}_{\text{soot},\text{layers},\text{rgn}}={\text{f}}_4\left({\text{m}}_{\text{soot},\text{layer}},{\text{m}}_{\text{ash}},{\dot{\text{m}}}_{\text{exh}},{\text{p}}_{\text{us}},{\text{T}}_{\text{us}},{\text{w}}_{\text{O}2}\right)$$

where ṁ _{soot,layer,rgn} is the particle mass flow of the soot particle layer resulting from the regeneration, m _soot,layer is the current loading of the soot particle layer, m _{ash is} the ash loading of the filter, T _us is the temperature in front of the particle filter, p _us is the pressure in front of the particle filter and w _O2 is the oxygen content in the exhaust stream.

Similarly, the particle mass flow during regeneration in the filter wall is a function of the exhaust gas mass flow, the pressure, the temperature in front of the particle filter and the ratio of oxygen in the exhaust gas mass flow, shown in block 202: $${\dot{\text{m}}}_{\text{soot},\text{wall},\text{rgn}}={\text{f}}_5\left({\text{m}}_{\text{soot},\text{wall}},{\dot{\text{m}}}_{\text{exh}},{\text{p}}_{\text{us}},{\text{T}}_{\text{us}},{\text{w}}_{\text{O}2}\right)$$

Equations (1) and (2) can thus be used to calculate the particle mass flows with which the filter wall or the soot layer is loaded by the inflowing exhaust gas, while the regeneration models for the filter wall and soot layer as in equations (5) and ( 6) the changes in the particle mass during a filter regeneration due to the burning off of the soot particles are described.

From the particle mass flow during loading, ṁ _{soot,wall,stored} or ṁ _{soot,layer,stored} , which is calculated in blocks 230 and 240 in 2a was modeled, and the particle mass flow during regeneration, ṁ _{soot,layer,rgn} and ṁ _{soot,wall,rgn} from blocks 201 and 202 in 2 B A difference can then be formed for the particle layer (function block 250) and filter wall (function block 260) and this difference can then be integrated over time in function blocks 255 and 265 in order to obtain the total mass of particles that accumulate in the particle filter in the filter wall or remain in the soot particle layer, ie the current loading of the filter wall, m _soot,layer , and the current loading of the particle layer, m _soot,wall .

If these loading values, ie the current mass or number of particles in the wall and in the soot particle layer of the filter, are known, a modeled value for the differential pressure can be calculated using the differential pressure model, which is also known per se. $$\mathrm{\Delta }\text{p}=\text{f}3\left({\text{m}}_{\text{soot},\text{layers}},{\text{m}}_{\text{soot},\text{wall}},{\dot{\text{m}}}_{\text{exh}},{\text{p}}_{\text{us}},{\text{T}}_{\text{us}}\right)$$

The exact functions f1 to f5 are not explained in more detail here and are known to the person skilled in the art from the models given in each case.

Usually, most of the unchangeable filter parameters for these models are recorded by the manufacturer and are known from the data sheets of a particle filter. These parameters can then be used directly for parameterization in the model. Some parameters, such as the permeability of the filter or activation energies, are also determined experimentally based on measurements for the respective filter, e.g. with a single motor, so that parameter values for this specific filter-motor combination are available and can also be used for the modelling . These values and presettings can also be stored in the control unit.

According to one possible embodiment, these modeled loading values for the soot particle layer and the filter wall can now be dynamically adjusted by further modelling. The following examples are based on the 3a and 3b explained where 3a shows an exemplary modeling of a correction value while 3b essentially the parameter modeling for a particle filter 2a corresponds, but now the in 3a formed correction value is included.

For this purpose, the relationship given above for the calculation of the differential pressure in equation (7) is simplified as follows or broken down into several components: $$\mathrm{\Delta }\text{p}={\text{k}}_{\text{flow}}\left({\text{m}}_{\text{soot},\text{layer}},{\text{m}}_{\text{soot},\text{wall}},{\text{m}}_{\text{ash}}\right)\cdot \text{G}1\left({\dot{\text{m}}}_{\text{exh}},{\text{p}}_{\text{us}},{\text{T}}_{\text{us}}\right)+\text{G}2\left({\dot{\text{m}}}_{\text{exh}},{\text{p}}_{\text{us}},{\text{T}}_{\text{us}}\right)$$

Here, k _flow is the flow resistance of the particle filter, which is directly dependent on the loading of the filter; g1 is a function of the exhaust gas mass flow and of the pressure and temperature in front of the particle filter, which, together with the flow resistance, describes the ideal pressure drop in the filter. The non-linear component g2 essentially describes the additional contraction and expansion losses at the inlet and outlet of the particle filter as a function of the exhaust gas mass flow, pressure and temperature in front of the filter and is mainly dependent on the filter design. Again, the corresponding models for the functions g1 and g2 can be found in the sources given above.

gewählt werden.This means that the display can also be used as $$\mathrm{\Delta }\text{p}-\text{G}2\left({\dot{\text{m}}}_{\text{exh}},{\text{p}}_{\text{us}},{\text{T}}_{\text{us}}\right)={\text{k}}_{\text{flow}}\left({\text{m}}_{\text{soot},\text{layers}},{\text{m}}_{\text{soot},\text{wall}},{\text{m}}_{\text{ash}}\right)\cdot \text{G}1\left({\dot{\text{m}}}_{\text{exh}},{\text{p}}_{\text{us}},{\text{T}}_{\text{us}}\right)$$

to get voted.

This means that the unknown parameter k _flow, which describes the flow resistance of the particle filter, in function block 304 3a can be determined as soon as the measured differential pressure Δp _sens , the measured exhaust gas mass flow and the measured pressure and temperature in front of the particle filter are available for a point in time. In this case, these values can be measured, for example, at predetermined intervals. In some cases, the values can also be calculated or determined from other measurements, eg the differential pressure from pressure measurements before and after the particle filter, or the exhaust gas mass flow can be determined from pressure and temperature. For a real-time determination of the parameter k _flow while a vehicle is being driven, all or part of these values can be determined by a parameter estimator based on the actually measured values, for example using the least squares method. The calculated component g2 and the measured differential pressure Δp _sens flow into function block 302 . In the subsequent function block 304, the likewise known linear component g1 is added, with a parameter estimator then being used to determine the flow resistance. The flow resistance determined in this way on the basis of equation (9) is denoted by k _flow,ident in the following and in the figures.

However, the flow resistance can also be modeled directly from the differential pressure model in function block 375, analogously to equation (7) for the differential pressure, i.e. as a function of the measured values for the exhaust gas mass flow ṁ _exh , the pressure p _us and the temperature T _us in front of the filter, the current loading of the filter wall, and the current loading of the particle layer, i.e $${\text{k}}_{\text{flow},\text{model}}=\text{G}3\left({\text{m}}_{\text{soot},\text{layer}},{\text{m}}_{\text{soot},\text{wall}},{\dot{\text{m}}}_{\text{exh}},{\text{p}}_{\text{us}},{\text{T}}_{\text{us}}\right)$$

The determination of the current filter loading has already been done in connection with 2a explained. The function block 375 essentially corresponds to the block 270 from FIG 2a .

Thus, two modeled values for the flow resistance were now determined; a first modeled flow resistance k _flow,ident , which results from the measured values as described in connection with equation (9) and which is calculated in block 304, and a second modeled flow resistance k _flow,mod from the differential pressure model in block 375, which together was explained with Equation (10).

In a further step 306, a difference can now be calculated from these two modeled values for the flow resistance, so that initially a correction value based on the flow resistance is obtained, which can be further used in various ways. In one possible embodiment, the resulting difference can, for example, _{be corrified} with a correspondingly selected factor k in step 306 of 3a converted into a correction value ṁ _soot,in,corr for the incoming soot mass flow.

In certain situations, this correction value ṁ _soot,in,corr can then be superimposed on the particle mass flow ṁ _soot,in , which has been calculated by the raw emissions model, as in 3b is shown. The basic modeling corresponds to the corresponding steps that have already been mentioned in connection with 2a have been shown and will therefore not be described again in detail. The filter efficiency is therefore modeled in steps 310 and 320, from which the change in the filter loading is then determined in steps 330 and 340 by the incoming particle mass flow. At this point, however, in step 380 the correction value ṁ _soot,in,corr is added to the modeled particle mass flow into the filter, so that a corrected particle mass flow is obtained for each layer and wall. Then, as already described and analogously to the function blocks 250, 255, 260 and 265, by subtracting the particle mass flow of the regeneration process (blocks 350, 360) and by integration over time (blocks 355, 365), a corrected or adapted current loading for the filter wall, m _soot,wall , and for the soot layer in the filter, m _soot,layer .

The current adjusted loads can then be reused in any controls, calculations or models. For example, the corrected loading value can be used again in the calculation of the filter efficiency in blocks 201 and 202 in 2a or included in equations (5) and (6). Also the calculation of the second flow resistance value k _flow,mod , as described in connection with 3a was described, uses the current filter loadings as input values in step 375, so that the adjusted values can also be used here. In this way, a dynamic, constantly self-adapting modeling of the filter loading and all associated parameters is achieved, since the corrected results flow directly into the next calculation cycle for the filter loading.

4 Figure 12 shows an alternative embodiment of the invention in which a split correction value for particulate mass flow is determined.

Both in 4a as well as in 4b most of the procedural steps analogous to the steps in 2a , or the 3a and 3b educated. Two modeled values for the flow resistance of the particulate filter are thus formed again, with the differential pressure Δp _sens determined by measurements forming the differential pressure Δp sens _,lin with the function component g2 in step 402, which then together with the exhaust gas mass flow, pressure and temperature-dependent pressure Function g1 is included in the first modeled value, so that in block 404 a first modeled flow resistance k _flow,ident is obtained by means of a parameter estimator. Also analogous to the embodiment 3 a second modeled value for the flow resistance is formed in block 475 on the basis of the differential pressure model, again as a function of the measured values for the exhaust gas mass flow ṁ _exh , the pressure p _us and the temperature T _us in front of the filter, the current loading of the filter wall, and the current loading of the particle layer. The difference between the flow resistance values is then formed in step 406 and converted to a correction value for the particle mass flow using a suitable factor in step 408 .

In contrast to the previous embodiment, however, this correction value ṁ _soot,in,corr is not directly included in the further modeling, but is divided into two separate correction values for the filter wall and particle layer. The distribution factors k _wall,corr and k _layer,corr for wall and layer can again depend on the measured exhaust gas mass flow, temperature and pressure in front of the particle filter as well as the existing soot mass or load in the filter wall and the layer of soot particles built up, so that they can, for example, correspond to the current correspond to the distribution ratio of the loads. However, any other distribution models are conceivable. As a result, a corrected value ṁ _{soot,wall,corr} for the particle mass flow into the filter wall is obtained in block 490, and in parallel in block 495 a corrected value ṁ _{soot,layer,corr} for the particle mass flow into the soot layer.

4b shows analogous to the 2a and 3b the modeling of various filter parameters to determine the filter loading. The filter efficiency is modeled in steps 410 (like step 210, 310) and 420 (like step 220, 320), from which the change in the filter loading is then determined in steps 430 and 440 by the incoming particle mass flow. While in the previous embodiment of the 3 here the correction value for the particle mass flow has been incorporated, in this alternative embodiment for steps 430 and 440 first the uncorrected particle mass flow ṁ _soot,in is used in the filter from the raw emission model together with the modeled filter efficiencies η _layer , η _wall . In the following step, in which the input values for the integrators 455 and 465 are formed separately for the filter layer and filter wall, the correction values are now calculated in steps 450 and 460 in addition to the negative particle mass flows from the regeneration 4a added. The input value for the integrator 455 is the sum of three components, namely the particle mass flow into the soot layer from block 430, ṁ _{soot,layer,stored} , the particle mass flow from the layer through the regeneration, ṁ _{soot,layer,rgn} , and the Correction value for the filter layer from block 495 in 4a , ṁ _{soot,layer,corr} . The input value for the integrator 465 results in the same way 4b from the particle mass flow ṁ _{soot,wall,stored} into the filter wall from block 440, the particle mass flow ṁ _{soot,wall,rgn} from the filter wall through regeneration, and the correction value ṁ _{soot,layer,corr} for the filter wall from block 490 in 4a .

As in the previous examples, the current adjusted loading values obtained in this way can be used to obtain, for example, a dynamically adjusted differential pressure 470, a dynamically adjusted flow resistance 475 or corrected efficiencies for the filter wall and layer, so that the loading values can be continuously corrected.

Another possible embodiment (not shown) can use both corrections from the previous examples of a suitable combination, i.e. both provide the soot mass flow from the raw emission model with the correction value and also correct the input of the integrators for the filter wall and/or soot layer in a suitable way. In principle it is also possible to correct only the values for the filter wall or only the values for the layer. In other cases, it is conceivable that the choice of the correction value to be used for the various particle mass flows is made dependent on conditions, for example a certain loading ratio or a special situation in the filter, e.g. after complete regeneration or when the end of the filter service life has been recognized.

In all possible embodiments, the parameter values corrected or adapted in this way, for example a corrected filter loading, corrected particle mass flows, corrected differential pressure or flow resistance, can then be used in the control and monitoring of the exhaust system. For example, if the filter load is too high, a regeneration process can be initiated by adjusting the exhaust gas temperature, or if the filter load is too low and the filter performance is therefore too low, parameters of fuel use and temperatures can be used to quickly load the filter. Other controls not mentioned here and for which the filter loading is relevant can also accept and apply the dynamically adjusted and therefore more precise values.

It is also understood that not all of the previous steps can be carried out by the same software or hardware module, but that, for example, the processing of sensor values or certain modeling can also take place separately. Several separate control devices or different modules and applications can also be used within a control device, between which the measurement data, the modeled data and the correction values described here are then exchanged.

Claims (11)

Method for determining a loading of a particle filter according to the wall flow principle, comprising: determining (304, 404) a first modeled flow resistance (k _flow,ident ) of the particle filter on the basis of a determined exhaust gas mass flow (ṁ _exh ), a determined differential pressure (Δp) across the particle filter and a temperature (T _us ) and a pressure (p _us ) in front of the particle filter; Determining (375, 475) a second modeled flow resistance (k _flow,mod ) of the particle filter on the basis of a determined exhaust gas mass flow, a modeled current loading of the particle filter and a temperature (T _us ) and a pressure (p _us ) in front of the particle filter; Forming a difference between the first and the second modeled flow resistance and calculating at least one correction value (ṁ _soot,in,corr , ṁ _{soot,layer,corr} , ṁ _{soot,wall,corr} ) for a particle mass senstrom in the particle filter from the formed difference, and determining (355, 365, 455, 465) a current adjusted loading of the particle filter using the at least one correction value for the particle mass flow. procedure after claim 1 , wherein the determination of a current adjusted loading comprises: calculating an adjusted particle mass flow into the particle filter on the basis of a modeled filter efficiency (η _layer , η _wall ) and a particle mass flow (ṁ _soot,in ) calculated on the basis of a raw emission model into the filter, with a correction value for the particulate mass flow is added to the particulate mass flow calculated on the basis of the raw emissions model. procedure after claim 1 or 2 , wherein determining the current adjusted loading of the particle filter comprises: determining (365, 465) a current adjusted loading of the filter wall (m _soot,wall ), and determining (365, 465) a current adjusted loading of the particle layer (m _soot,layer ) of the particle filter. procedure after claim 3 , further comprising: determining (320, 420) a filter efficiency of the filter wall as a function of a current loading of the particle filter; determining (310, 410) a filter efficiency of the particle layer as a function of a current loading of the particle filter; calculating the current adjusted loading of the filter wall based on the filter efficiency of the filter wall and a particle mass flow into the filter calculated on the basis of an untreated emissions model; and calculating the current adjusted loading of the particle layer based on the filter efficiency of the particle layer and a particle mass flow into the filter calculated on the basis of a raw emissions model. procedure after claim 3 or 4 , wherein the calculation of at least one correction value for a particle mass flow into the particle filter comprises: calculating a first correction value from the difference between the modeled flow resistances, and dividing the first correction value into a second correction value for a particle mass flow in a filter wall of the particle filter and a third correction value for a particle mass flow into a particle layer of the particle filter. procedure after claim 5 , wherein the currently adjusted loading of the filter wall is calculated by integrating a first input value over time, the first input value being formed from the particle mass flow into the filter wall, the second correction value and a particle mass flow from the filter wall resulting from filter regeneration; and wherein the currently adjusted loading of the particle layer is calculated by integrating a second input value over time, wherein the second input value is formed from the particle mass flow into the filter layer, the third correction value and a particle mass flow from the particle layer resulting from filter regeneration. Method according to one of the preceding claims, wherein the method comprises: using a current adapted loading of the particle filter to control a regeneration of the particle filter. Method according to one of the preceding claims, wherein the method comprises measuring at least one of the following parameter values continuously or at predetermined intervals: a pressure in front of the particle filter, a temperature in front of the particle filter, an exhaust gas mass flow in front of the particle filter, an exhaust gas mass flow in front of the particle filter, a Differential pressure across the particle filter. Arithmetic unit which is set up to carry out all method steps of a method according to one of the preceding claims. Computer program that causes a computing unit to carry out all the method steps of a method according to one of Claims 1 until 8th to be performed when it is executed on the computing unit. Machine-readable storage medium with a computer program stored on it claim 10 .

Images