DE102005032623A1 - Cylinder selective soot and nitrogen oxide emission method for diesel internal combustion engine involves selecting perimeters and determining emission during combustion process by means of multiple input single output model - Google Patents

Abstract

The method involves selecting parameters, described by given number of combustion processes, so that they show heat rate (dQH) and fluid mixture composition of combustion process. Soot and nitrogen oxide (NOx) emission of combustion process of a cylinder is then determined from these measured parameter values during combustion process by means of multiple input single output (MISO) model. An independent claim is also included for the use of the method.

Description

The The invention relates to a method for determining the cylinder-selective Soot and NOx emissions of a Diesel engine according to the preamble of claim 1 and the use of the method in the engine control and / or in the Control of exhaust gas purification.

Of the Proportion of passenger cars with low-consumption DI diesel engine on new registrations is around 40% in Germany and is still rising. To this Trend also among future ones to maintain or even accelerate demanding emission standards to be able to will be intense after action to reduce the most significant particulate and NOx emissions. In particular internal engine measures move in the focus of aggregate research, as is an active exhaust aftertreatment mostly negative impact on manufacturing costs and consumption. The looming availability cost-effective cylinder pressure sensors as well as powerful processor technology enable the replacement of the map-out Control of combustion by a control based on combustion chamber functions, such as the cylinder pressure. Advantages of such a cylinder pressure-based Motoneregelung, in which each cylinder with a cylinder pressure sensor equipped, the compensation is production and aging Disturbing influences the combustion, an equalization of combustion in the individual Cylinders and a significantly reduced application and diagnostic effort.

A another interesting measure to reduce soot and NOx emission provides engine control based on emissions is a so-called closed loop, also referred to as closed-loop control. For one such regulation requires a mathematical model which Based on information from the combustion chamber, the soot and NOx emissions in real time calculated.

The Soot and Hang NOx emissions of local parameters, e.g. the local fuel / air ratio or from the maximum flame temperature. These sizes are only with special Measuring technology to be determined and derived models are not satisfactory. Furthermore, it is known that individual global No sizes Have an influence on the emissions. It is not possible, for example, solely because of the global gas temperature to make a statement about the level of NOx emissions.

Out A. Opalinski, M. Willmann, "Relationships of with global thermodynamic sizes the soot and NOx emissions in the TDI engine ", The working process of the internal combustion engine (9th session), Univ. Prof. Dr. H. Eichlseder, VKM-THD Messages Issue 83, Graz 2003, pp. 135-148, Two 0-dimensional models are known in which the soot and NOx emissions with the help of of six characteristic parameters, the burning rate and describe the mixture composition, can be predicted. The disadvantage is that the known models only in a small Operating area are applicable and validatable. In addition, a real-time calculation The burning rate according to the current state of the art not possible. Further based the firing rate on assumptions for wall heat transfer, which is difficult and can be verified inaccurately. misperceptions the wall heat losses, which can be up to 15% to 20% of the energy used, act result in significantly over-calculated NOx emissions.

Of the The invention is therefore based on the object, a method for determining the cylinder-selective soot and NOx emissions of a diesel engine with a significant larger operating range and the use of the method in the engine control and / or to create in the control of the exhaust gas purification. In particular should different injection strategies (without / with pre-injection), various injection systems (pump-nozzle, common rail, amplified piston Common Rail), a variation of the charge motion (fully variable Ventiltrieb) as well as different combustion processes (homogeneous, partially homogeneous and heterogeneous combustion) from the process and it should continue to different engine configurations (different displacements, Number of cylinders, etc.).

These Task is solved by a method having the features of claim 1, and by the use of the method in the engine control and / or at the control of the exhaust gas purification with the features of the claim 11. Preferred embodiments of the method and use of the method are the subject of the dependent claims.

In the inventive method for determining the cylinder-selective soot and NOx emis For example, in a diesel engine, the soot and NOx emissions of the diesel engine are calculated using a predetermined number of parameters describing the combustion process, the parameters being selected to reflect the heating rate and mixture composition of the combustion process, and those determined during a combustion process Parameter values the soot emission and NOx emission of the combustion process of a cylinder by means of a MISO model (Multiple Input - Single Output) are calculated.

Preferably the number of parameters used is greater than or equal to four and less or eight. In particular, six parameters are used for description the heating rate or its course and the mixture composition used.

• a) der Zündverzug τ_zv,
• b) der Anstieg der Heizrate α₁
• c) der Abfall der Heizrate α₂,
• d) der maximale Verbrennungsdruckanstieg dp_max oder die Schwerpunktlage der Verbrennung Ai_50%
• e) der Sauerstoffgehalt im Abgas O2 oder das Luft/Kraftstoffverhältnis λ, und
• f) das Verhältnis der Luftmasse zur gesamten Gasmasse m_l/m_g,

wobei die Parameter a) bis d) die Heizrate und die Parameter e) und f) die Gemischzusammensetzung betreffen.Preferably, the following parameters describing the heating rate and the mixture composition are used to calculate the soot and NOx emissions of a cylinder:

• a) the ignition delay τ _zv ,
• b) the increase of the heating rate α ₁
• c) the drop in the heating rate α ₂ ,
• d) the maximum combustion pressure increase dp _max or the center of gravity of combustion Ai _50%
• e) the oxygen content in the exhaust gas O2 or the air / fuel ratio λ, and
• f) the ratio of the air mass to the total gas mass m _l / m _g ,

wherein the parameters a) to d) relate to the heating rate and the parameters e) and f) the mixture composition.

at the choice of parameters must be taken to ensure that for all parameters both with the soot as well as with the NOx emissions a physical connection exists, so for the soot as well as the NOx model the same parameters are used can. As the number of parameters increases, soot and NOx emissions become on the one hand always better represented, on the other hand takes the complexity and the Calculation of the models and it comes to an overfitting. The Using six parameters gives the optimal compromise between accuracy and computing time.

A exact explanation or definition of said parameters or parameters takes place in connection with the figures. This is where the parameters stand out by assuming that they are within the framework of future engine concepts as well can be determined during vehicle operation.

Especially For example, a polynomial or a neural network can be used as the MISO model become.

Preferably can the MISO model for the soot emission also by a potency product approach of the above parameters be formed, each parameter with an exponent e1 to e6 is provided and the potency product approach still a constant Term K1.

The same can for the NOx emission carried out Here, too, preference is given to the MISO model for NOx emission by a potency product approach of the above, the heating rate formed parameters describing the combustion process, wherein each parameter is provided with an exponent f1 to f6 and the power product approach still has a constant term K2.

wobei hier die Schwerpunktlage der Verbrennung und das Verbrennungsluftverhältnis in den Formeln 1 und 2 angegeben ist. Es kann stattdessen auch das Maximum des Zylinderdruckanstiegs und/oder der Sauerstoffgehalt verwendet werden, was zu den folgenden Formeln führt:

Therefore, the following formulas are available for the calculation of soot and NOx emissions:

Here, the center of gravity of the combustion and the combustion air ratio is given in the formulas 1 and 2. Instead, the maximum of the cylinder pressure rise and / or the oxygen content may also be used, resulting in the following formulas:

Preferably become the exponents e1 to e6 and f1 to f6 and the constants K1 or K2 according to the least squares method from a predetermined number of measurements determined. To determine the Exponents and the constants can therefore test bench measurements be performed in different map points. In particular, can the selection of the map points with the methods of statistical Experimental design.

Preferably are the exponents and the constant for the soot emission and the exponents and the constant for the NOx emission for a multi-cylinder diesel engine as averages of formed individual cylinder, the values for various diesel internal combustion engines are different.

The method according to the invention described above is preferably used in an engine control unit Diesel engine for real-time calculation of soot and / or NOx emissions.

Further can the inventive method to carry out an on-board diagnosis of the diesel engine due to determined soot and NOx emissions are used.

Further can with the method according to the invention calculated soot and NOx emissions as a guide to Determination of recirculated exhaust gas in exhaust gas recirculation systems be used.

Especially can with the method according to the invention calculated NOx emissions as reference variables for the dosing systems used by SCR catalysts.

One Another preferred application of the method according to the invention is the use of the calculated soot emissions as a guide to Determination of the regeneration time for diesel particulate filters.

By the use as a "virtual Sensors "can the explained above Models real soot and NOx sensors replace and so significantly reduce the cost of production contribute to a vehicle.

A preferred embodiment The invention is explained below with reference to the drawings. there shows

1 the heating rate of a combustion process as a function of the crank angle, and

2 the corresponding heating process.

1 shows a typical course of the heating rate dQ _H in J / ° KW of a combustion process in a cylinder of a diesel engine as a function of the crank angle KW in degrees. The injection phase or injection period during which the fuel is introduced into the cylinder is in 1 symbolically represented by the bar labeled E. The injection phase E begins in this example at a crank angle of about -9 ° and is completed at an angle of about -5 °. It comes to the actual onset of combustion to an ignition delay ZV and at an angle of about 5 ° starts in this example, the combustion. This point is referred to as start of burning BG. The heating rate generated by the combustion is in 1 represented by the graph 1, wherein the heating rate increases sharply after the start of combustion after the combustion of the pilot injection quantity and after the maximum heating rate dQHmax drops again to zero at the burning BE. The three points start of combustion BG, maximum heating rate dQHmax and burning end BE define a triangle enclosing the graph I of the heating rate with the angles α1 and α2, which are also used as parameters for describing the heating rate.

2 shows the heating curve Q, shown as graph II, wherein the heating curve Q is formed by the integral of the heating rate dQH on the crank angle KW. The graph II of the heating curve has the start of burning BG, which in 2 is defined by the point QHmin where the integral heating history has its global minimum. Further, the curve II shows the combustion 50% boost point U associated with the Ai50% combustion center of gravity, which is the crank angle at which 50% of the span between the global minimum and the global integral heating maximum is exceeded. Finally, curve II shows the Burning BE, which is assigned the point Ai95%, which is the crank angle at which 95% of the span between global minimum and global maximum is exceeded.

• α1: Die Größe α1 beschreibt den Winkel zwischen dem Brennbeginn BG und der maximalen Heizrate dQHmax. Für den Brennbeginn können verschiedene Definitionen verwendet werden, beispielsweise: a) Kurbelwinkel, bei dem der integrale Heizverlauf sein globales Minimum aufweist (QHmin), b) Kurbelwinkel, bei dem 5 % der Spanne zwischen globalem Minimum und globalem Maximum des integralen Heizverlaufsüberschritten werden (Ai 5% Wert), c) Kurbelwinkel, bei dem 85 % des Minimums im negativen Abschnitt des integralen Heizverlaufs überschritten werden, oder d) Kurbelwinkel, bei dem die Zylinderdruckdifferenz zwischen dem Zylinderdruck bei gefeuertem und geschlepptem Motor größer als 2 bar ist

The following explanations serve to explain and define the in 1 and 2 used parameters and parameters for the description of the heating rate:

• α1: The variable α1 describes the angle between the start of combustion BG and the maximum heating rate dQHmax. Different definitions can be used for the start of combustion, for example: a) crank angle at which the integral heating curve has its global minimum (QHmin), b) crank angle at which 5% of the span between global minimum and global integral heating maximum is exceeded (Ai 5% value), c) crank angle exceeding 85% of the minimum in the negative portion of the integral heating history, or d) crank angle where the cylinder pressure difference between the cylinder pressure with the engine fired and towed is greater than 2 bar

mit:

dQHmax [J/°KW]:

maximale Heizrate

AdQHmax [°KW]:

Kurbelwinkel bei der maximalen Heizrate

Ai5% [°KW]:

Kurbelwinkel bei Brennbeginn,

wobei in der Formel (3) für den Brennbeginn die Definition b) verwendet wurde. Bei einer Verwendung der anderen Definitionen ist die Formel (3) entsprechend zu ändern.

• α2: Der Parameter α2 beschreibt den Winkel zwischen der maximalen Heizrate dQHmax und dem Brennende: Für das Brennende können ebenfalls verschiedene Definitionen verwendet werden, beispielsweise: e) Kurbelwinkel, bei dem der integrale Heizverlauf sein globales Maximum aufweist, oder f) Kurbelwinkel, bei dem 95 % der Spanne zwischen globalem Minimum und globalem Maximum des integralen Heizverlaufsüberschritten werden (Ai 95% Wert).

The angle α1 is calculated as follows:

With:

dQHmax [Y / ° CA]:

maximum heating rate

AdQHmax [° CA]:

Crank angle at the maximum heating rate

Ai5% [° CA]:

Crank angle at the start of burning,

wherein in the formula (3) for the start of combustion, the definition b) was used. When using the other definitions, the formula (3) should be changed accordingly.

• α2: The parameter α2 describes the angle between the maximum heating rate dQHmax and the end of combustion: Different definitions can also be used for the end of combustion, for example: e) crank angle, where the integral heating curve has its global maximum, or f) crank angle, at which 95% of the span between the global minimum and the global integral heating maximum is exceeded (Ai 95% value).

mit:

dQHmax [J/°KW]:

maximale Heizrate

AdQHmax [°KW]:

Kurbelwinkel bei der maximalen Heizrate

Ai95% [°KW]:

Kurbelwinkel bei Brennende

The angle α2 is calculated 1 as follows:

With:

dQHmax [Y / ° CA]:

maximum heating rate

AdQHmax [° CA]:

Crank angle at the maximum heating rate

Ai95% [° CA]:

Crank angle at the end of combustion

there in formula (4) the burning end according to definition f) has been used. When using other definitions, formula (4) is appropriate to change.

Ai50%: The in 2 characteristic Ai50% shown is called the center of gravity of the combustion. It is the crank angle at which 50% of the span between the global minimum and the global maximum of the integral heating curve Q is exceeded. Instead of the center of gravity, the formulas for calculating the soot and NOx emissions may also use the maximum increase in combustion pressure, in which case, of course, the corresponding exponents must be adjusted.

Ai 5% [KW]:

Kurbelwinkel bei Brennbeginn

SB [°KW]:

Kurbelwinkel bei Spritzbeginn oder elektrischer Förderbeginn

τ _zv : The ignition delay τ _zv describes the time span between the start of injection SB and the start of combustion BG. Different definitions for the start of firing have been mentioned above. The start of injection is the time at which the first droplets of fuel from the injector enter the combustion chamber. Since this point in time is difficult to determine with sufficient accuracy without measuring technology, the electrical start of delivery is taken as a substitute variable. τ zv = Ai5% - SB With:

Ai 5% [KW]:

Crank angle at the start of combustion

SB [° CA]:

Crank angle at start of injection or electrical start of delivery

λ: The combustion air ratio λ can with Help a mounted in the exhaust system lambda probe can be determined.

ml / mg: The relationship the air mass to the total gas mass ml / mg also describes the Gas composition. It is a substitute for the EGR rate since there is no close-to-production solution for their determination in vehicle operation. Models for calculation the EGR rate, which partly also takes into account the internal EGR, are subject to a much more complex calculation rule.

The Air mass ml can be measured with the aid of a hot film air flow meter (HFM) be determined.

mit:

psaug [bar]:

Saugrohrdruck

VHub [m3]:

Hubvolumen

RL [J/(kgK)]:

Spezifische Gaskonstante

TSaug [K]:

Saugrohrtemperatur

For example, the total mass mg can be calculated approximately with the ideal gas equation:

With:

psaug [bar]:

Intake manifold pressure

VHub [m3]:

displacement

RL [J / (kgK)]:

Specific gas constant

TSaug [K]:

Intake manifold

mit:

m [kg]:

Masse pro Arbeitsspiel und Zylinder

ṁ [kg/h]:

Gesamtmassenstrom

n [U/min]:

Drehzahl

a [–]:

Kurbelwellenumdrehungen je Arbeitsspiel

z [–]:

Zylinderzahl

Mass flows can be converted via the following relationship in masses per working cycle and cylinder:

With:

m [kg]:

Mass per working cycle and cylinder

ṁ [kg / h]:

Total mass flow

n [rpm]:

rotation speed

a [-]:

Crankshaft revolutions per working cycle

z [-]:

number of cylinders

In The formula for calculating soot and NOx needs both λ and ml / mg, since the same lambda either by a change the amount of fuel or the amount of recirculated exhaust gas can be achieved can, which has different effects on emissions.

As already mentioned above, instead of the center of gravity of the combustion Ai50%, the maximum of the cylinder pressure increase dp _max and instead of lambda λ, the oxygen content 02 in the exhaust gas can also be used.

In order to calculate the emissions from the above six parameters, several MISO models (multiple input - single output), such as polynomials or artificial neural networks, are conceivable. In addition to its great robustness, the good interpretability of classical polynomials is advantageous. However, the limited ability to map strong nonlinearities is disadvantageous. Neural networks, on the other hand, are good at mapping such nonlinearities, but are more difficult to interpret and more likely to overfill. Because of the low complexity, the short computation time and the good interpretability, a power product approach was chosen here, whereby the six parameters were given exponents e1 to e6 and f1 to f6, which represent an indicator for the direction and magnitude of the influence of the parameter , The exponents are determined by the method of least squares. Each formula also contains a constant K, which serves to align the different dimensions.

This then leads to the formulas 1 and 2 or 1 'and 2' already given above.

to Determination of the exponents and the constants were made on a 1-cylinder diesel engine 200 operating points with heterogeneous combustion and a long start of injection and measure AGR variation. The results are engine-specific and the exponents and constants must be determined once for each motor. The selection of suitable operating points is advantageously carried out with the Methods of statistical experimental design.

dQh

heating rate

dQHmax

maximum heating rate

Q

heat curve

ZV

ignition delay

KW

crank angle

BG

combustion start

e

Injection duration

BE

burning

U

50% sales point

I

graph the heating rate

II

graph the heating process

Claims (16)

Method for determining the cylinder-selective soot and NOx emissions of a diesel internal combustion engine, wherein the soot and NOx emissions of the diesel internal combustion engine are calculated by means of a predetermined number of parameters describing the combustion process, characterized in that the parameters are selected such that they correspond to the Represent the heating rate and the mixture composition of the combustion process, and from the determined during a combustion process parameter values, the soot emission and NOx emission of the combustion process of a cylinder can be calculated by means of a MISO model. Method according to claim 1, characterized in that that the predetermined number of parameters is four to eight. A method according to claim 2, characterized in that the following parameters describing the heating rate and the mixture composition are used to calculate the soot and NOx emissions of a cylinder: the ignition delay τ _zv , the increase of the heating rate α1, the decrease of the heating rate α2, maximum combustion pressure increase dpmax or the center of gravity of the combustion Ai50% of the oxygen content in the exhaust gas O2 or the combustion air ratio λ, and the ratio of the air mass to the total gas mass ml / mg Method according to one of the preceding claims, characterized characterized in that as a MISO model a polynomial or a neural Network is used. Method according to one of claims 1 to 3, characterized that the MISO model for the soot emission is formed by a power product approach of the parameters, where each parameter is provided with an exponent e1 to e6 and the power product approach still has a constant term K1. Method according to one of claims 1 to 3, characterized that the MISO model for the NOx emission is formed by a power product approach of the parameters, where each parameter is provided with an exponent f1 to f6 and the power product approach still has a constant term K2. Method according to claim 5 or 6, characterized that the exponents e1 to e6 and f1 to f6 and the constants K1 or K2 according to the least squares method from a predetermined number of measurements are determined. Method according to claim 7, characterized in that that to determine the exponents and the constants test bench measurements be performed in different map points. Method according to claim 8, characterized in that that the selection of the map points using the methods of statistical Experimental design is done. Method according to one of claims 5 to 9, characterized that the exponents and the constant for the soot emission and the exponents and the constant for the NOx emission for a multi-cylinder diesel engine as averages of individual cylinders are formed. Use of the method according to one of the preceding claims in an engine control unit a Diesel engine for real-time calculation of soot and / or NOx emission. Use according to claim 11, characterized that the calculated soot and NOx emissions are used to control the engine. Use according to claim 1 or 12, characterized that an on-board diagnosis the diesel engine is performed based on the detected soot and NOx emissions. Use according to one of claims 11 to 13, characterized that the calculated soot and NOx emissions as a guide to Determination of recirculated exhaust gas in exhaust gas recirculation systems be used. Use according to one of Claims 11 to 14, characterized that the calculated NOx emissions are used as a guide for the metering systems of SCR catalysts become. Use according to one of claims 11 to 15, characterized that the calculated soot emissions as a guide to Determination of the regeneration time used in diesel particulate filters become.