DE102009029011A1 - Method and device for determining the composition of a fuel mixture for operating an internal combustion engine - Google Patents

Abstract

The invention relates to a method for determining the composition of a fuel mixture from a first fuel and at least one second fuel for operating an internal combustion engine, the fuels differing in their knock resistance, in their lean running limit and / or in their exhaust gas recirculation compatibility. It is provided that a first composition value of the fuel mixture is determined from an output signal of a knock sensor or from a knock control of the internal combustion engine, that the lean running limit of the internal combustion engine is determined for the fuel mixture, that a second composition value is determined from the lean running limit and that the composition of the Fuel mixture is determined from a combination of the first composition value and the second composition value. The invention also relates to a corresponding device. The method and the device enable an inexpensive and reliable determination of the composition of fuel mixtures for internal combustion engines.

Description

The invention relates to a method for determining the composition of a fuel mixture of a first fuel and at least a second fuel for operating an internal combustion engine, wherein the fuels differ in their anti-knock properties, in their lean running limit and / or in their Abgasrückführverträglichkeit.

The invention further relates to a device for determining the composition of a fuel mixture of a first fuel and at least a second fuel for operating an internal combustion engine, wherein the fuels in their anti-knock, in their lean-burn limit and / or in their Abgasrückführverträglichkeit differ and wherein the internal combustion engine is a control unit assigned.

State of the art

Internal combustion engines based on gasoline engines are generally operated with a fuel from hydrocarbons from fossil fuels based on refined crude oil. To this fuel is increasingly from renewable raw materials (plants) produced alcohol, such as ethanol or methanol, mixed in different mixing ratios. In the US and in Europe, a mixture of 75-85% ethanol and 15-25% gasoline is often used under the brand name E85. The internal combustion engines are designed so that they can be operated with both pure gasoline and with mixtures up to E85; this is called "flex-fuel operation". For economical operation with low pollutant emissions coupled with high engine performance, the operating parameters in flex-fuel operation must be adapted to the particular fuel mixture present. By way of example, there is a stoichiometric air / fuel ratio at 14.7 parts by weight of air per part of gasoline, but when using ethanol, an air content of 9 parts by weight must be set.

In flex-fuel operation, an adapted enrichment factor must be specified as a function of the mixing ratio due to the different temperature-dependent evaporation properties of ethanol and gasoline at the start of the internal combustion engine. The ignition timing must also be adjusted depending on the fuel mixture. The knowledge of the present fuel mixture ratio is therefore of fundamental importance for the operation of the internal combustion engine.

To determine the composition of the fuel mixture, different fuel type sensors, also referred to as "fuel composition sensors" can be used. Fuel type sensors use the different properties of the fuels used, for example, alcohol and gasoline, to determine the fuel composition. For example, ethanol is a protic solvent which contains hydrogen ions and has a large dielectric constant, which is dependent on the water content. Gasoline is an aprotic solvent with a small dielectric constant. Based on this, there are fuel type sensors which determine the fuel composition based on the dielectric properties of the fuel mixture. Other fuel type sensors utilize the different electrical conductivity or different optical properties of the fuels, such as, for example, the different refractive indices. The disadvantage here is that the system costs are increased by the use of fuel type sensors. Another disadvantage is that according to existing regulations, the correct functionality of the fuel type sensors must be monitored, which requires additional effort.

Therefore, software-based systems for determining fuel composition have been developed which do not use special fuel type sensors but which evaluate the signals of the sensors present on the internal combustion engine. These systems can be realized more cost effectively than systems with fuel type sensors.

From the DE 3036107 C3 a control device is known for a Kraftstoffzumesssystem in an internal combustion engine consisting of a fuel supply device (fuel injection valve), a lambda probe, means (timer) for forming a Grundzumesssignals the operating characteristic variable dependent ultimately determines the drive signal (ti) of the fuel supply device, a lambda controller, the starting from a signal measured by the lambda probe (λ) determines a correction factor which multiplicatively affects the Grundzumesssignal (tp) with the correction factor. It is provided that the lambda correction is dependent not only on the correction factor (KR λ) but also on an additive (KA λ) and / or a multiplicative (KL λ) correction variable which is determined as a function of the correction factor and operating parameters.

The control device makes it possible to compensate for systematic deviations of the fuel metering, which is specified by the basic metering signal, ie the so-called pilot control, from the value determined by the lambda control by an adaptation intervention with a corresponding long-term correction. Systematic deviations can be caused, for example, by aging influences or by manufacturing influences. On average corresponds to the amount of fuel defined by the corrected pilot control of the actually required amount. Short-term deviations can be compensated with the lambda controller, to which the entire control range is now available again. The underlying process is also known as mixture adaptation.

The determination of the fuel mixture ratio can be carried out without additional fuel-type sensor by means of a fuel adaptation. The fuel adaptation is activated after a refueling detected by the tank level sensor. A fuel composition changed by refueling results in a changed stoichiometric air / fuel ratio. By a corresponding intervention of the lambda control on the operating parameters of the internal combustion engine, in particular on the set air / fuel ratio and the ignition timing, this change in the fuel properties is taken into account in the context of fuel adaptation. From the intervention of the lambda control or the fuel adaptation can be concluded that the stoichiometric ratio and hence the composition of the fuel mixture. The composition of the fuel mixture can be determined inexpensively by a pure software solution accordingly.

The disadvantage here is that after a refueling process during the fuel adaptation, the mixture adaptation must be temporarily deactivated so that changed combustion properties of the fuel mixture are not erroneously compensated by the long-term correction of the mixture adaptation. The deactivation of the mixture adaptation is undesirable and contradicts the requirements of CARB in the USA in particular for continuous monitoring and diagnosis of the fuel supply system.

Another disadvantage of the method is that in a refueling operation with small amounts of fuel, the tank level sensors do not always reliably recognize this. If such minute amounts of refueling are repeatedly carried out, it may happen, in particular with a low filling level of the fuel tank, that the fuel composition changes significantly, without this being recognized and compensated by the fuel adaptation. The change in combustion characteristics is then erroneously corrected by the mixture adaptation.

Another method for determining the fuel mixture ratio without using a specific fuel type sensor is described in applicant's unpublished document R.318835. The document describes a method for determining the composition of a fuel mixture of a first fuel and a second fuel for operating an internal combustion engine, wherein the first and the second fuel have different octane numbers and wherein the internal combustion engine has at least one knock sensor and a knock control. The method is characterized in that the composition of the fuel mixture is determined by means of an output signal of the knock sensor.

Another method for determining the composition of a fuel mixture of a first and at least a second fuel for operating an internal combustion engine, wherein fuel mixtures of different composition have a different knock resistance, characterized in that the composition of the fuel mixture is determined from the ignition angle, in which just no Engine knock occurs. The method is described in the not yet published document R.320246 of the applicant.

In the not yet published document R.325571 the applicant is a method for determining a fuel composition of a fuel mixture of a first and at least a second fuel for operating an internal combustion engine with known air-fuel ratio or a lambda value at known fuel, in which by Interventions the smoothness of the internal combustion engine is at least temporarily negatively affected described. It is provided that the influence is carried out by means of a variation of an ignition angle and / or by means of a variation of the air-fuel ratio, wherein from a resulting uneven running or from a resulting gradient for the uneven running a parameter for the composition of currently present fuel mixture or a lambda value for the cylinders of the internal combustion engine are derived.

The mentioned methods for determining the composition of a fuel mixture based on various, the internal combustion engine or its control unit already present characteristics and signals have the disadvantage that the achievable uncertainty in the determination of the fuel composition is sometimes significantly greater than the accuracy required to control the internal combustion engine.

In the not yet published document R.322094 the applicant a method for controlling an internal combustion engine is described, which is operated with a fuel mixture of a first and at least a second fuel, wherein the internal combustion engine Fuel metering device, a tank level gauge for determining the tank contents and a change in the tank contents, a sensor for detecting the cylinder filling for determining an air mass supplied to the internal combustion engine and at least one exhaust gas probe for determining and regulating the oxygen content in the exhaust gas. The method is characterized in that, in a first method step, a first value for the composition of the fuel mixture from the supplied air mass and the measured change in the tank contents, taking into account the oxygen content in the exhaust gas, is determined that in a second method step a second value for the composition of the Fuel mixture is determined from the supplied during idling of the internal combustion engine air mass and the amount of fuel supplied by the fuel metering device that the first and the second value are compared to match within a predetermined limit and that at divergent values on an error in the fuel metering device , in which determination of the supplied air mass or tank level gauge is closed. The method therefore serves to monitor the fuel metering device, to monitor the supplied air mass or to monitor the tank level gauge; the accuracy in determining the two values for the composition of the fuel mixture is limited to the accuracy of the individual underlying methods for determining the composition and thus subjected to great statement blurring.

It is an object of the invention to provide a method which allows a reliable and cost-effective detection of the composition of a fuel mixture of at least two fuels with a sufficiently low statement uncertainty.

Disclosure of the invention

The object of the invention relating to the method is achieved by determining a first composition value of the fuel mixture from an output signal of a knock sensor or from knock control of the internal combustion engine, determining the lean burn limit of the internal combustion engine for the fuel mixture, determining a second composition value from the lean burn limit and that the composition of the fuel mixture is determined from a combination of the first composition value and the second composition value.

The composition values describe the composition of the fuel mixture, as determined from the respective individual parameters.

Modern internal combustion engines often have knock sensors whose output signal is used for knock control. If fuels differ in their octane number or in their anti-knock properties, the octane number and the anti-knocking properties of a fuel mixture derived therefrom are determined by the mixing ratio of the fuels involved. From the output signal of the knock sensor or a control operation of the knock control, for example the Zündwinkelausgabe, can be closed to the composition of the fuel mixture and the first composition value formed.

An emaciation of the air-fuel mixture, d. H. an increase in the lambda value leads to an increasing uneven running in internal combustion engines. Different fuels and fuel mixtures composed of these have a typical so-called lean running limit. If the internal combustion engine supplied air-fuel mixture emaciated over the lean-burn limit, so both the ignition delay is increased and the ignition slowed down greatly or the ignition does not begin. This leads to extremely delayed burns or combustion misfires, which then result in a significantly increased uneven running. In the case of a fuel mixture of, for example, gasoline and ethanol, the lean running limit shifts with increasing ethanol content towards larger lambda values. The knowledge of the lean running limit is therefore an indication of the mixing ratio of the fuel mixture used and can be used for the determination of the second composition value. It can easily be determined by a corresponding variation of the air-fuel ratio supplied to the internal combustion engine, wherein the combustion stochastics thereby changed can be evaluated by evaluating the crankshaft speed, for example by means of appropriate rest rest algorithms.

To control the internal combustion engine, the composition of the fuel mixture must be known with sufficient accuracy. This accuracy can not be reliably achieved either by determining the composition of the knock sensor output signal or knock control, or by determining the composition of the lean burn limit. Only through the combination according to the invention of the composition values derived from the output signal of the knock sensor or the knock control and from the lean running limit can the composition of the fuel mixture be determined with an accuracy sufficient for the control of the internal combustion engine. The advantage here is that the composition values are determined from independent characteristics and stochastically distributed around the actual composition of the fuel mixture.

If a knock sensor or a knock control is already present for controlling the internal combustion engine, no additional components, such as additional fuel-type sensors, need to be provided for carrying out the method, which enables a cost-effective implementation of the method as a pure software solution. In this case, the lean running limit and, derived therefrom, the second composition value can also be determined by a software adaptation.

In order to carry out the method according to the invention, diagnostic functions required by the legislator, such as, for example, the mixture adaptation, do not have to be deactivated, as a result of which the acceptance of the method significantly increases compared to existing software solutions.

Another advantage over existing solutions is that the determination of the fuel composition is not tied to the recognition of a refueling operation. If existing software-based methods for determining fuel compositions, which presuppose the detection of a refueling operation, frequently refuel small quantities that are no longer detectable by the tank level sensing system, this can lead to a noticeable change in the composition of the fuel mixture, especially if the tank is low that this is recognized by the method. This disadvantage is avoided by the method according to the invention.

If it is provided that a limit for the exhaust gas recirculation compatibility of the internal combustion engine is determined, a third composition value is determined from the limit for the exhaust gas recirculation compatibility and the composition of the fuel mixture from a combination of the first composition value and the third composition value or a combination of the second composition value and the third composition value or a combination of the first composition value, the second composition value and the third composition value, the accuracy in determining the composition of the fuel mixture can be further increased.

As an emaciation of the internal combustion engine supplied air-fuel mixture and a change in the internal combustion engine via the exhaust gas recirculation supplied amount of exhaust gas leads to a dependent on the composition of the fuel mixture influencing the smoothness of the internal combustion engine. With a fuel mixture of gasoline and ethanol, the exhaust gas recirculation compatibility increases with increasing content of ethanol. By varying the exhaust gas recirculation and an evaluation of the combustion stochastics, for example, by evaluating the crankshaft speed by appropriate rest restraint algorithms, the limit for Abgasrückführverträglichkeit can be determined and deduced therefrom on the composition of the fuel mixture. Again, it is advantageous that the limit for Abgasrückführverträglichkeit can be determined without additional components by a software adaptation.

A determination of the composition of the fuel mixture from the individual composition values, which can be carried out by a simple arithmetic operation, can be achieved by taking the combination of the composition values by averaging over the composition values.

A more accurate determination of the composition of the fuel mixture may be obtained by specifying weighting factors for the compositional values in dependence on the accuracy of the respective compositional value, and determining the composition of the fuel mixture from a weighted combination of the compositional values. The weighting factors indicate how great the accuracy is in determining the respective compositional value. The more accurate composition value is weighted more heavily than the more inaccurate composition value in the combination of the composition values for determining the composition of the fuel mixture according to the weighting factor.

The weighting factors can be determined from the reciprocal of the percentage accuracy of the composition value, and the Gaussian error propagation method can be used to calculate the most likely composition of the fuel mixture and the mean deviation from the true value. If the statement accuracy for a composition value is 20% by way of example, the weighting factor 1 / 0.2 = 5 must be used. In the case of two composition values Z1 and Z2 and two weighting factors W1 and W2, the most probable composition is wZ wZ = (W1 ² · Z1 + W2 ² · Z2) / (W1 ² + W2 ² ).

Thus, the composition of the fuel mixture may be determined from the sum of the composition values multiplied by the squares of the associated weighting factors divided by the sum of the squares of the weighting factors. After this calculation after the Method of Gaussian error propagation is the most probable value for composition wZ.

The accuracy m for the most probable composition wZ for a maximum ethanol content of 85% is calculated as

The accuracy m thus achieved in determining the composition of the fuel mixture is significantly better than can be achieved by the individual composition values.

The accuracy of the determination of the composition values from the output signal of a knock sensor, from a knock control of the internal combustion engine, from the lean running limit and from the limit for the exhaust gas recirculation compatibility is in each case dependent on the operating range in which the internal combustion engine is currently being operated. Therefore, it can be provided that the weighting factors of the individual composition values are predefined as a function of the present operating range of the internal combustion engine so as to obtain high accuracy in the determination of the composition of the fuel mixture over all operating ranges.

The object of the invention relating to the device is achieved by providing in the control unit a program sequence for determining a first composition value from an output signal of a knock sensor or from a knock control of the internal combustion engine, that a program sequence for determining the lean running limit is provided, that a program sequence for Determination of a second composition value from the lean-burn limit is provided and that a program sequence for determining the composition of the fuel mixture from a combination of the first composition value and the second composition value is provided.

With an existing knock sensor and existing head control of the internal combustion engine, the composition of a fuel mixture can thus be determined by a pure software solution without additional components without having to temporarily disable predetermined by the legislature diagnostic functions. In this case, the output signal of the knock sensor and thus the knock control as well as the determined by appropriate evaluation of the running of the engine with changed, the internal combustion engine supplied air / fuel ratio certain lean boundary directly with the composition of the fuel mixture, so that determines the individual composition values by appropriate program sequences can be. By combining the composition values, the composition of the fuel mixture can be determined much more accurately than is possible by the individual composition values.

According to a preferred embodiment variant of the invention it can be provided that in the control unit a program sequence for determining a limit for the Abgasrückführverträglichkeit the internal combustion engine for the fuel mixture is provided that a program sequence for determining a third composition value from the limit for Abgasrückführverträglichkeit is provided and that a routine for determining the composition of the fuel mixture from a combination of the first composition value and the third composition value or a combination of the second composition value and the third composition value or a combination of the first composition value, the second composition value and the third composition value is provided. By additionally obtaining from the limit for the exhaust gas recirculation third composition value, the accuracy can be further improved in the determination of the composition of the fuel mixture.

According to a preferred embodiment variant of the invention, it can be provided that the program sequence for determining the composition of the fuel mixture comprises a program routine for determining the average of the first composition value and the second composition value or the first composition value and the third composition value or the second composition value and the third composition value of the first composition value and the second composition value and the third composition value. The mean value of the composition values can be determined easily and without much memory and computation and, as a result, leads to an accuracy in the determination of the composition of the fuel mixture, which is significantly better than the accuracy in the determination of the individual underlying composition values.

According to a particularly preferred embodiment variant of the invention, it may be provided that weighting factors are predetermined as a function of the statement accuracy of the composition values in the respective operating ranges of the internal combustion engine that a memory area for Storing the weighting factors of the composition values as a function of the operating ranges of the internal combustion engine, and in that the program sequence for determining the composition of the fuel mixture contains a program routine for determining the composition of the fuel mixture from the composition values weighted with the respective weighting factors. The weighting factors make it possible, when determining the composition of the fuel mixture, to take into account the different accuracy of the compositional values used. Since the degree of inaccuracy of the composition values depends on the particular operating range in which the engine is operated, the weighting factors can be stored depending on the operating range and taken into account in the combination of the composition values.

The method and apparatus may preferably be used to determine the composition of a fuel mixture of gasoline and alcohol, preferably a fuel mixture of gasoline and ethanol.

Short description of the drawing

The invention will be explained below with reference to an embodiment shown in FIGS. It shows:

1 a flowchart for determining a composition of a fuel mixture for operating an internal combustion engine

2 a diagram for determining a composition of a fuel mixture for operating an internal combustion engine.

1 shows a flowchart for determining an in 2 shown composition 30 a fuel mixture for operating an internal combustion engine using the example of a gasoline-ethanol fuel mixture.

In a first branch of the flowchart are a block ZW1 10 and a block WF 12 and in a second branch a block ZW2 11 and a block WF2 13 intended.

The block ZW1 10 is a signal as input 20 a knock control, the block ZW2 11 a lean running limit determined in a program routine (not shown) 21 supplied to the internal combustion engine. The blocks WF1 12 and WF2 13 is the operating area 22 , in which the internal combustion engine is currently being operated, fed.

The blocks WF1 and WF2 is a block Z 14 downstream.

Owing to the higher octane number of ethanol compared to gasoline, an indication of the ethanol content in the fuel mixture can be obtained by evaluating the behavior of the knock control on different ignition angle outputs. In block ZW1 10 can therefore out of the signal 20 the knock control a in 2 shown first composition value 32 be determined for the fuel mixture.

Internal combustion engines show a lean running limit depending on the combustion characteristics of the fuel used 21 , If the internal combustion engine supplied air-fuel ratio over the lean limit 21 emaciated, this results in an increased running noise of the internal combustion engine. In the case of a gasoline-ethanol fuel mixture, the lean limit shifts 21 with increasing ethanol content towards larger lambda. By evaluating the combustion stochastics, for example the standard deviation of the mean indicated pressure in at least one of the cylinders of the internal combustion engine, the lean running limit 21 be determined. For this purpose, for example, the crankshaft speed can be evaluated via running rest algorithms. Due to the correlation of the thus determined lean running limit 21 with the ethanol content of the fuel mixture may be in the block ZW2 11 from the lean-burn limit 21 a in 2 shown second composition value 33 be determined.

The one out of the signal 20 the knock control derived first composition value 32 and the one from the lean-burn limit 21 derived second composition value 33 are afflicted with a statement uncertainty that is greater than the maximum total deviation required for the control of the internal combustion engine 34 in determining the composition 30 of the fuel mixture. In this case, the respectively achievable accuracy of the operating range 22 , in which the internal combustion engine is operated, dependent. In block WF1 12 therefore becomes the first composition value 32 a in 2 shown first weighting factor 36a and in block WF2 13 becomes the second composition value 33 a likewise in 2 shown second weighting factor 36b depending on the operating range 22 assigned to the internal combustion engine.

In the block Z 14 then the determination of the composition takes place 30 of the fuel mixture by one with the respective weighting factors 36a . 36b weighted combination of the first composition value 32 and the second composition value 33 , The less the statement inaccuracy of a composition value 32 . 33 is, the larger the associated weighting factor 36a . 36b The corresponding composition value is correspondingly stronger and stronger 32 . 33 in the combination for determining the composition 30 considered the fuel mixture.

2 shows a diagram for determining a composition 30 a fuel mixture for operating an internal combustion engine using the example of a gasoline-ethanol fuel mixture.

Along a first axis 31 the percentage of ethanol in a gasoline-ethanol fuel mixture is plotted. To a desired total deviation 34 in determining the composition 30 to reach, be out of the signal 20 the knock control a first composition value 32 and the one from the lean-burn limit 21 a second composition value 33 certainly. The composition values 32 . 33 are shown as expected with standard deviation and along a second axis 17 arranged side by side. The standard deviation of the composition values 32 . 33 are greater than the desired maximum total deviation 34 for the determination of the composition 30 of the fuel mixture.

According to the invention, the composition 30 of the fuel mixture by a combination of the compositional values 32 . 33 certainly. The combination is done by forming a weighted average over the compositional values 32 . 33 , The composition values 32 . 33 are weighting factors 36 which is a measure of the accuracy of the respective composition value 32 . 33 are. By way of example, the first weighting factor expresses 36a with the value 5 an accuracy of 1/5 = 20%; the weighting factor 36b with the value 3 means an accuracy of 1/3 = 33.3%.

Because the composition values 32 . 33 were determined independently of each other and thus by a stochastic distribution of the composition values 32 . 33 about the composition 30 can be assumed, by suitable combination of the composition values 32 . 33 the required and by the total deviation 34 illustrated accuracy in determining the composition 30 of the fuel mixture, even if the statement blurs the underlying compositional values 32 . 33 are significantly larger.

The improvement of the statement accuracy by the use of the weighted averaging is explained in the following with an example calculation:
The first composition value 32 40%, the second composition value 33 amount to 20% of ethanol in a gasoline-ethanol fuel mixture. The weighting factors 36 , which reflect the error in the determination of the respective characteristic amount 5 for the first weighting factor 36a respectively 3 for the second weighting factor 36b , According to the known method of Gaussian error analysis, the composition determined on the basis of the above two value pairs 30 : Composition 30 = (40% × 5 ² + 20% × 3 ² ) / (5 ² + 3 ² ) = 34.7%

Likewise, the method of Gaussian error analysis allows the accuracy in the determination of the composition 10 at the two weighting factors 18 with the values 5 and 3 and a maximum ethanol content of 85%:

The combination of the two composition values 12 . 13 with weighting factors 18 with the values 5 and 3 , which correspond to accuracies of 85% / 5 = 17% and 85% / 3 = 28.3%, thus gives a statement with an improved accuracy of 14.6%.

According to the invention, the weighting factors 36 from the operating area 22 the internal combustion engine are determined dependent. This improves the reliability of the method and its accuracy, since in some operating states characteristics that make up the composition values 32 . 33 can be determined, can be determined very precisely and therefore the composition values obtained therefrom 32 . 33 can be weighted higher than in other operating areas.

QUOTES INCLUDE IN THE DESCRIPTION

This list of the documents listed by the applicant has been generated automatically and is included solely for the better information of the reader. The list is not part of the German patent or utility model application. The DPMA assumes no liability for any errors or omissions.

Cited patent literature

• DE 3036107 C3 [0007]

Claims (10)

Method for determining the composition ( 30 ) of a fuel mixture of a first fuel and at least a second fuel for operating an internal combustion engine, wherein the fuels in their anti-knock, in their lean running limit ( 21 ) and / or differ in their Abgasrückführverträglichkeit, characterized in that from an output signal of a knock sensor or from a knock control of the internal combustion engine, a first composition value ( 32 ) of the fuel mixture, it is determined that the lean-burn limit ( 21 ) of the internal combustion engine for the fuel mixture is determined that from the lean-burn limit ( 21 ) a second composition value ( 33 ) and that the composition ( 30 ) of the fuel mixture from a combination of the first composition value ( 32 ) and the second composition value ( 33 ) is determined. A method according to claim 1, characterized in that a limit for the Abgasrückführverträglichkeit the internal combustion engine for the fuel mixture is determined, that is determined from the limit for exhaust gas recirculation compatibility, a third composition value and that the composition ( 30 ) of the fuel mixture from a combination of the first composition value ( 32 ) and the third composition value or a combination of the second composition value ( 33 ) and the third composition value or a combination of the first composition value ( 32 ), the second composition value ( 33 ) and the third composition value. Method according to claim 1 or 2, characterized in that the combination of the composition values ( 32 . 33 ) by averaging over the composition values ( 32 . 33 ) he follows. Method according to claim 1 or 2, characterized in that weighting factors ( 36 ) for the composition values ( 32 . 33 ) depending on the accuracy of the respective composition value ( 32 . 33 ) and that the composition ( 30 ) of the fuel mixture from one with the associated weighting factors ( 36 ) weighted combination of composition values ( 32 . 33 ) is determined. Method according to claim 4, characterized in that the weighting factors ( 36 ) of the individual composition values ( 32 . 33 ) depending on the operating range ( 22 ) of the internal combustion engine can be specified. Device for determining composition ( 30 ) of a fuel mixture of a first fuel and at least a second fuel for operating an internal combustion engine, wherein the fuels in their anti-knock, in their lean running limit ( 21 ) and / or differ in their Abgasrückführverträglichkeit and wherein the internal combustion engine is assigned a control unit, characterized in that in the control unit a program sequence for determining a first composition value ( 32 ) is provided from an output signal of a knock sensor or from a knock control of the internal combustion engine that a program sequence for determining the lean running limit ( 21 ), it is provided that a program sequence for determining a second composition value ( 33 ) from the lean-burn limit ( 21 ) and that a program procedure for determining the composition ( 30 ) of the fuel mixture from a combination of the first composition value ( 32 ) and the second composition value ( 33 ) is provided. Apparatus according to claim 6, characterized in that in the control unit a program sequence for determining a limit for the Abgasrückführverträglichkeit the internal combustion engine for the fuel mixture is provided that a program sequence for determining a third composition value from the limit for Abgasrückführverträglichkeit is provided and that a program flow for Determination of composition ( 30 ) of the fuel mixture from a combination of the first composition value ( 32 ) and the third composition value or a combination of the second composition value ( 33 ) and the third composition value or a combination of the first composition value ( 32 ), the second composition value ( 33 ) and the third composition value. Apparatus according to claim 6 or 7, characterized in that the program sequence for determining the composition ( 30 ) of the fuel mixture, a program routine for determining the mean value of the first composition value ( 32 ) and the second composition value ( 33 ) or the first composition value ( 32 ) and the third composition value or the second composition value ( 33 ) and the third composition value or the first composition value ( 32 ) and the second composition value ( 33 ) and the third composition value. Device according to claim 6 or 7, characterized in that weighting factors ( 36 ) depending on the accuracy of the composition values ( 32 . 33 ) in the respective operating areas ( 22 ) of the internal combustion engine are specified in that in the control unit Memory area for storing the weighting factors ( 36 ) of the composition values ( 32 . 33 ) depending on the operating areas ( 22 ) of the internal combustion engine is provided and that the program sequence for determining the composition ( 30 ) of the fuel mixture, a program routine for determining the composition ( 30 ) of the fuel mixture from the with the respective weighting factors ( 36 ) weighted composition values ( 32 . 33 ) contains. Use of the method and the device according to one of the preceding claims for determining the composition ( 30 ) of a fuel mixture of gasoline and alcohol, preferably a fuel mixture of gasoline and ethanol.

Images