Method for correcting air volume differences in the air intake path of an internal combustion engine

Abstract

Method (100) for correcting air quantity differences in the air intake path of an internal combustion engine (501), comprising: obtaining (120) measurement data (131, 132) indicating a deviation of an air mass drawn in by the internal combustion engine (501) from a modeled air mass over the rotational speed of the internal combustion engine (501); determining (140) a correction relationship, comprising multiplying a predefined relationship (111) between the deviation over the rotational speed by a scaling factor (S), wherein the correction relationship approximates the measurement data (131, 132); correcting (150) an actual air quantity in the air intake path based on the correction relationship.

Description

[0001] The invention relates to a method for correcting air volume differences due to manufacturing tolerances and aging effects of internal combustion engines.

[0002] The technical field lies in the area of ​​engine technology, in particular the optimization and correction of air volume variations in the air intake system of internal combustion engines. Manufacturing tolerances and aging effects are addressed, which can influence the air supply and thus the performance and emissions of engines, causing them to deviate from the target values.

[0003] In an internal combustion engine, such as a gasoline or diesel engine, the gas mixture burned in the cylinders is crucial for the engine's efficiency and emissions. This mixture typically consists of air and a fuel such as gasoline or diesel. The ratio between air and fuel, also known as the air-fuel ratio or combustion air ratio, significantly influences combustion. An ideal mixture, referred to as the stoichiometric ratio, ensures complete combustion of the fuel without excess oxygen or unburned hydrocarbons. To measure the composition of the gas mixture, sensors such as lambda sensors are used, which, for example, determine the oxygen content in the exhaust gases.These measurements allow the air-fuel ratio to be monitored and optimized to ensure the most efficient combustion possible, with low levels of harmful emissions.

[0004] Manufacturing tolerances in the air intake system of an internal combustion engine refer to deviations in the production of the components responsible for supplying air to the engine. These tolerances are crucial because they affect the airflow into the cylinders and can therefore vary the mass of air drawn in under otherwise identical conditions. This, in turn, affects the engine's efficiency and performance.

[0005] Excessive deviations can lead to turbulence, pressure losses or uneven air distribution, which reduces performance and increases fuel consumption and emissions.

[0006] Aging effects can have a similar impact, for example if deposits alter the geometry in the intake tract, thereby also changing the airflow.

[0007] DE 10 2014 211 162 A1 relates to a method for measuring the filling level in a cylinder of an internal combustion engine and a device for measuring the filling level in a cylinder of an internal combustion engine.

[0008] DE 10 2015 210 761 A1 relates to a method for determining an air quantity within a cylinder of an internal combustion engine, an engine control unit which is configured to carry out the method, and an internal combustion engine which has the engine control unit.

[0009] DE 10 2008 012 607 A1 relates to a method and a device for determining an adaptation value for an air-fuel ratio of an injection system of an internal combustion engine.

[0010] WO 97 / 35 106 A2 describes a method for determining, using a model, the mass of fresh air flowing into the cylinders of an internal combustion engine with external exhaust gas recirculation.

[0011] The object of the present invention is to provide an improved method, an improved control unit and an improved vehicle.

[0012] This problem is solved by the method according to claim 1. This problem is further solved by a control unit according to claim 19 and by a vehicle according to claim 20.

[0013] Further advantageous embodiments of the invention will become apparent from the dependent claims and the following description of preferred embodiments of the present invention.

[0014] The invention encompasses the development and application of mathematical models and adaptation methods for correcting flow variations caused by manufacturing tolerances and coking effects. In such cases, control based on the air-fuel ratio using a lambda controller can regulate emissions, but not the generated torque. With few exceptions, the target air-fuel ratio in modern gasoline engines is always set to λ = 1. The actual air volume from the filling model is used to determine the fuel quantity. A deviation of the intake air volume from the modeled air volume manifests as a mixture error, which is compensated for by the mixture controller / lambda controller via the fuel quantity. The air-fuel ratio is thus continuously regulated.

[0015] Therefore, the models are designed to detect deviations in the intake air mass from the air mass modeled for a given driving situation and driver request, due to manufacturing tolerances and coking effects, and to differentiate these deviations from other variations. Measurement data from an air mass meter, for example, located in the intake manifold, can be used for this purpose. Furthermore, values ​​from a lambda controller can also be used as measurement data, since this indicates the mixture error between air and fuel quantities when properly regulated. Other deviations can arise, for example, from injection system errors or seal degradation, such as in cases of increased blow-by. The goal is to adapt the engine control unit to account for these manufacturing deviations in order to meet specifications regarding generated torque and emissions.Furthermore, the aim is to account for aging effects through adaptation in the engine control unit in order to ensure long-term compliance with specifications. Additionally, the ability to diagnose aging effects can be a further objective, for which the model's ability to distinguish these effects can be utilized.

[0016] An inventive method for correcting air quantity differences in the air intake path of an internal combustion engine comprises: obtaining measurement data indicating a deviation of an air mass drawn in by the internal combustion engine from a modeled air mass over the rotational speed of the internal combustion engine; determining a correction relationship comprising multiplying a predetermined relationship between the deviation over the rotational speed by a scaling factor, wherein the correction relationship approximates the measurement data; correcting an actual air quantity in the air intake path based on the correction relationship.

[0017] An internal combustion engine is a motor that generates mechanical energy through the combustion of fuel. In this process, chemical energy is converted into mechanical energy. This energy is then used, for example, to power a vehicle. The term "internal combustion engine" can refer to an engine that mixes and burns air and fuel in a cylinder to move the pistons and thus perform mechanical work. Typical examples of internal combustion engines are gasoline engines (Otto engines) and diesel engines.

[0018] The air intake system is the part of the internal combustion engine through which air flows to the cylinders. This system begins at the air intake, passes through various components such as the air filter, the turbocharger (if present), the throttle body, and ends at the intake valves of the cylinders. The air intake system is designed to provide the correct amount of air for combustion in the engine. Precise control of the air supply may be necessary to ensure an optimal air-fuel ratio, which in turn affects the engine's efficiency and emissions. Furthermore, the amount of incoming air, along with the amount of fuel, can be adjusted to produce the desired torque for the driver.

[0019] The air intake system and the internal combustion engine can be optimized using this method so that, despite manufacturing tolerances and aging effects, the correct amount of air reaches the cylinders. This is achieved by using measurement data to detect deviations between the intake air mass and the modeled air mass and to determine a correction relationship. This correction relationship is then used to adjust the actual air volume in the intake system so that the engine's performance and emissions remain within the desired limits.

[0020] Obtaining measurement data can mean that the relevant data from the operation of the internal combustion engine are measured and made available to the process, for example, without a process unit executing the process having to collect the data. The data can also be read from memory. This data can include parameters such as the value of a lambda controller or its deviation from a reference, which is controlled based on measurements taken by sensors such as lambda probes. Furthermore, its dependence on the engine speed, detected by speed sensors, or on an applied load, quantified by the applied torque or intake manifold pressure, is considered. The deviation of the lambda controller value from the reference can be recorded as a percentage or as a ratio. This value is subsequently referred to as the lambda control value.The measurement data can, for example, be transmitted (in real time) to the engine control unit (ECU), where it is stored and analyzed to detect deviations in the lambda control value at different speeds (and loads).

[0021] Determining a correction relationship refers to the process of developing a mathematical relationship (function or mapping) to model and correct the deviations of the lambda control value as a function of the internal combustion engine's rotational speed, for example, by adapting an engine model. Mathematically, a relationship is a function or mapping that describes a dependency between two or more variables. In this case, a predefined relationship, or dependency, is used to approximate the deviations of the lambda control value across the rotational speed. This approximation is achieved by adjusting a scaling factor, which is multiplied by the predefined relationship. The scaling factor is determined through regression analysis or a similar method that minimizes the deviations in the measurement data to achieve the best possible fit.

[0022] The correction relationship can be directly determined by multiplying the given relationship by the scaling factor. Furthermore, an adaptation factor can be determined based on the scaling factor and then multiplied by the given relationship to obtain the correction relationship. The adaptation factor can also be determined based on other factors, such as the mileage already driven or a percentage that determines how much of the difference between the current adaptation value and a specific scaling factor may actually be adapted. This percentage can also depend on the engine speed at which the measurement data was taken or the mileage. Repeated execution of the procedure can lead to a robust adaptation.

[0023] This correction relationship allows for precise adjustment of the actual air volume in the intake manifold, thus optimizing engine performance and emissions (both in the short and long term). The predefined relationship ensures that the adjustment can only successfully modify effects that correspond to the predefined relationship at different scales. Furthermore, adjustments due to other effects or errors can be excluded from this adaptation.

[0024] Correcting the actual air volume refers to the process of adjusting the calculated amount of air intended to enter the cylinders of an internal combustion engine based on a previously determined correction relationship. This correction is necessary to ensure that, despite manufacturing tolerances and aging effects, the optimal amount of air is provided for combustion. The engine control unit (ECU) uses the correction relationship to dynamically adjust the actual air volume across the engine speed by modifying control parameters such as intake manifold pressure, throttle position, valve timing, valve lift, and boost pressure. These adjustments optimize the air-fuel ratio, resulting in more efficient combustion and ensuring compliance with emissions regulations.Correcting the actual air volume thus helps to stabilize the engine's performance, reduce emissions, and ensure the longevity of the engine components.

[0025] There are versions where the measurement data includes at least three measured values.

[0026] Measured values ​​can be specific, quantifiable data points acquired through measurements. Regarding the method for correcting air mass variations in the intake air system of an internal combustion engine, the measured values ​​refer to the acquired parameters used to determine and correct the air mass. These measured values ​​can include various physical quantities relevant for analyzing and optimizing the deviation of the intake air mass from the modeled air mass or the lambda control value.

[0027] In detail, the measured values ​​could include the following: the lambda control value, the engine speed, and the engine load. The corresponding engine speed and load can be assigned to the lambda control value. This can be achieved, for example, with a delay that models the path of the exhaust gas from the cylinder to the sensor that measures the air-fuel ratio or information indicative of the air-fuel ratio (e.g., an oxygen concentration in the exhaust gas), and an additional delay resulting from the information flow to the lambda controller and a time delay before the lambda controller sets the lambda control value.

[0028] The engine speed indicates how fast the engine's crankshaft rotates, measured in revolutions per minute (rpm). The amount of air flowing into the cylinder can depend on the engine speed or the piston speed. The engine load can be quantified by the applied torque, intake manifold pressure, or the work output (mean pressure). The load affects the amount of air and fuel required for efficient combustion.

[0029] Other measurements could also be relevant, such as intake manifold pressure and intake air temperature. Intake manifold pressure is the pressure in the intake tract that forces air into the cylinders. The intake air temperature affects the air's density and therefore the amount of air drawn in.

[0030] By recording and analyzing these measured values, the method can detect deviations between the intake air mass and the modeled air mass or the lambda control value under different engine operating conditions and make appropriate corrections to optimize the engine's performance and emissions.

[0031] There are versions where the measurement data is obtained from a lambda controller.

[0032] A lambda controller receives measurement data from a lambda sensor, with this data being indicative of the air-fuel ratio during combustion. If the intake air mass deviates from the modeled air mass, where the injected fuel quantity is calculated based on the modeled air mass, a deviation in the air-fuel ratio occurs. This deviation is measured by the lambda sensor and compensated for by the lambda controller. Thus, the control values ​​of the lambda controller can be used as measurement data to quantify a deviation between the intake air quantity and the modeled air quantity. The calculation of the injected fuel mass can also be incorrect (e.g., calculating the injection mass in relation to the opening duration). This would also lead to a deviation in the lambda control value. However, due to the error pattern (the engine speed trend), this source of error can be ruled out, indicating a pure air modeling error.

[0033] A lambda sensor, also known as an oxygen sensor, is a sensor that can be installed in the exhaust system of an internal combustion engine. Its main function is to measure the oxygen content in the exhaust gas. This measurement makes it possible to determine the air-fuel ratio (also known as the lambda value) present in the engine's cylinders.

[0034] The air-fuel ratio is the ratio of the mass of air to the mass of fuel introduced into the combustion process. An ideal air-fuel ratio for complete combustion without excess oxygen or unburned hydrocarbons is called the stoichiometric ratio. For gasoline engines, this ratio might be around 14.7:1, meaning that there are 14.7 parts air to 1 part fuel.

[0035] The lambda sensor continuously provides data on the oxygen content in the exhaust gas to the engine control unit (ECU). This information is used to monitor and adjust the air-fuel ratio. If the ratio is too rich (too much fuel) or too lean (too much air), the ECU can adjust the amount of fuel injected or the air supply to ensure optimal combustion. This helps to maximize engine efficiency, minimize emissions, and extend engine life.

[0036] There are two main types of lambda sensors. The narrowband sensor (also known as a switching sensor) can provide a signal that fluctuates widely around the stoichiometric point. It is simple and inexpensive, but is only suitable for controlling the air-fuel ratio within a narrow range around the stoichiometric point. The wideband lambda sensor can accurately measure the air-fuel ratio over a wide range. It provides a more precise signal and allows for more accurate control of the air-fuel ratio, which is particularly important in modern engines with variable valve timing and direct injection.

[0037] The lambda sensor can be important for emissions control and optimizing engine performance. It enables modern engine management systems that aim to reduce environmental impact while maximizing engine efficiency and performance.

[0038] There are versions where the measurement data is obtained from an air mass meter.

[0039] This allows measurement data to be available that quantifies the amount of air drawn in and thus enables a calculation of the deviation from a modeled amount of air.

[0040] There are versions where the given relationship is a parabolic function.

[0041] A parabolic function is a special type of quadratic function, mathematically described by an equation of the form: y(x)=ax2+bx+c The parabola is described as follows, where a, b, and c are constants and x is the independent variable. The shape of the parabola is determined by the coefficient a; if a > 0, the parabola opens upwards, and if a < 0, the parabola opens downwards.

[0042] The parabolic function can be used to model the deviations of the lambda control value as a function of rotational speed. For this purpose, a parabolic function with predefined constants a and b (and c) can be used, multiplied by a scaling factor S: y(x)=(ax2+bx)∗S.

[0043] Here, x is, for example, the rotational speed and y(x) the lambda control value to which the parabola is approximated using the scaling factor. An adaptation factor can be determined from the scaling factor S, which is then incorporated into an engine model of the internal combustion engine. In addition to the scaling factor, the adaptation factor can be based on other influences, such as the robustness of the adaptation.

[0044] Thus, the predefined relationship adapted to the measurement data yields a parabolic function, which, multiplied by the adaptation factor, results in a correction relationship. This correction relationship can be used to perform a speed-dependent adjustment or adaptation in an engine model (for example, a filling model). In this way, the actual intake air volume is adjusted to the air volume modeled in the engine model by correcting parameters of the intake tract. This adjustment is speed-dependent.

[0045] By adapting the system to a predefined general relationship and a specific parabolic function, these effects do not need to be compensated for by a controller. Such a controller would have to adjust the manipulated or controlled variable every time the rotational speed changes rapidly. Therefore, the emission and torque specifications can be maintained even during acceleration or gear changes.

[0046] There are versions where determining a correction relationship involves adding an offset to approximate it to the measurement data.

[0047] Thus, a constant value (offset d) can be added to a given relationship, for example, before or after multiplication by the scaling factor. In the special case of a parabola as the given relationship, this leads to the following formula: y(x)=(ax2+bx)∗S+d.

[0048] This allows for an improved fit between a predefined relationship and the measurement data. For example, this can be particularly useful if the measurement data exhibits a systematic shift caused by another source of error. The offset can also be used solely for fitting to the measurement data. However, the correction can be based on a correction relationship that itself is based on the fitted predefined relationship with a scaling factor (adaptation factor), but without the offset. Thus, the fitting is performed with an offset to the predefined relationship, while the correction or adaptation is performed without this offset to reduce the influence of other errors on this adaptation.

[0049] Adding the offset makes the correction relationship more flexible and allows it to be more accurately adapted to the actual measurement data. This enables a more precise correction of the actual air volume in the intake manifold and helps to optimize the performance and emissions of the internal combustion engine.

[0050] There are versions where the measurement data is collected in a measuring range that can be defined by load and speed limits.

[0051] The measurement range can refer to the specific range of operating conditions under which the measurement data is collected. In an internal combustion engine, the measurement range for the lambda control value can be defined by various parameters such as load and engine speed. The measurement range is crucial to ensure that the collected data is representative of the actual operating conditions of the internal combustion engine. By defining a suitable measurement range, the measurement data can be made consistent and comparable, thereby improving the accuracy and reliability of the correction relationship. The measurement range can be selected to cover typical driving conditions under which the internal combustion engine operates, enabling realistic and practical adaptation. Furthermore, load and engine speed ranges where other errors are apparent can be excluded.

[0052] The load and speed limits can define the specific operating conditions within which the measurement data is collected. The load limits (upper and lower limits) can define a range of applied torque or intake manifold pressure within which the measurement data is used for the procedure. The speed limits (upper and lower limits) can define a range of engine speeds within which the measurement data is used for the procedure. These limits ensure that the internal combustion engine is in a stable operating condition without large speed or load fluctuations, and that the measurement data is comparable.

[0053] A load-speed diagram is a graphical representation that shows the measurement range in relation to the engine's load and speed. In this diagram, the load is plotted on the y-axis and the speed on the x-axis. The measurement range can be defined by a window or within this diagram, in which the lambda control value measurement data is collected. These windows represent specific combinations of load and speed that are relevant for adapting the correction relationship.

[0054] There are versions in which several measuring ranges are provided, each separated from the others by lower speed limits, upper speed limits, lower load limits and upper load limits.

[0055] Several measuring ranges can be provided, each defined by specific speed and load limits. These enable detailed and precise measurement of the lambda control value across a wide variety of operating conditions of the internal combustion engine. These operating conditions can include, for example, comparable load ranges as well as different speed ranges, allowing the lambda control value to be measured across the engine speed.

[0056] These measurement ranges can be defined as separate measurement windows in a load-speed diagram, with each measurement window covering a specific range of load and speed.

[0057] This enables a differentiated analysis and adaptation of the correction relationship, taking into account both low and high speeds as well as different load conditions. Separating the measurement ranges through specific lower and upper limits for speed and load ensures that the data is collected under stable operating conditions, thereby increasing the accuracy and reliability of the correction relationship.

[0058] Specifically, when a parabola is used as the given relationship, the deviation increases with higher rotational speeds. Therefore, the scaling factor can be determined more accurately using measurement ranges at higher speeds. If measurement ranges with higher rotational speeds are involved in the adaptation, a higher percentage (the difference between the applied value and the value determined by the method) can be allowed for correction between the scaling factor and the adaptation factor, resulting in faster adaptation.

[0059] There are designs in which the correction of the actual air volume to the calculated air volume is carried out successively based on the correction relationship, and an already used correction relationship is only changed by part of the difference between a current scaling factor and the scaling factor determined by the correction relationship.

[0060] Successive correction can be a step-by-step adjustment process in which the actual air volume in the intake manifold is adjusted not abruptly, but in small, controlled increments. This method allows for a gradual approach to the optimal actual air volume, thus avoiding sudden changes and potential instabilities in engine operation. Successive correction enables the engine control unit (ECU) to continuously monitor performance and emissions parameters and ensure that the adjustments are consistent with the engine's operating conditions. This step-by-step approach contributes to increased robustness and reliability of the correction by allowing for smooth and progressive adjustments.

[0061] The previously used or current scaling factor can be the value used to correct the actual airflow at the time the measurement data is taken in the engine control unit (ECU). This scaling factor may be based on previous measurements and adjustments and represents the current state of adaptation. The current scaling factor is continuously monitored and adjusted as needed to ensure that the actual airflow corresponds to the engine's actual operating conditions. By taking the current scaling factor into account, the procedure can ensure that corrections are consistent and coordinated, which can contribute to stable and efficient engine performance.

[0062] The difference can be the difference between the current scaling factor and the new scaling factor, determined by the correction relationship. This difference indicates how far the current scaling factor deviates from the optimal value required for the precise correction of the actual air volume. By adjusting the factor incrementally, only a portion of this difference is corrected in each step, ensuring a smooth and controlled adjustment. Considering the difference allows adjustments to be made in small, well-controlled increments, thus avoiding sudden changes and potential instabilities. This method helps to increase the accuracy and reliability of the correction and ensures that the actual air volume is continuously optimized.The procedure can therefore be repeated at (predetermined) time intervals, with each execution leading to adaptation, and the final scaling factor or adaptation value then becoming the current scaling factor upon a subsequent execution.

[0063] There are versions that provide a fast correction range, which is applicable within a specified service life and in which a larger part of the difference is used to correct the actual air volume than in a slow correction range outside the specified service life.

[0064] The rapid correction phase can be an adaptation period that occurs immediately after the internal combustion engine is commissioned, typically within the first few days of operation or a limited number of kilometers driven. During this phase, a larger portion of the difference between the current scaling factor and the new scaling factor can be adapted. The scaling factor determined by the correction relationship is used to adjust the actual air volume. This allows for a quick and efficient correction of air volume variations caused by manufacturing tolerances. The rapid correction phase can be designed to rapidly minimize initial deviations, enabling the internal combustion engine to quickly reach optimal operating conditions. In this phase, deviations due to manufacturing variations can be corrected.

[0065] The slow correction range can come into effect after the fast correction range and refers to a phase of adaptation that continues throughout the entire service life of the internal combustion engine. In this range, only a smaller portion of the difference between the current scaling factor and the new scaling factor is used to adjust the actual air volume. This gradual adjustment allows for a smooth and continuous correction of air volume variations that can arise from aging effects such as carbon buildup. The slow correction range helps ensure the long-term stability and efficiency of the internal combustion engine by providing a progressive adjustment that avoids sudden changes and potential instabilities in engine operation.

[0066] The specified service life can be a specific period or a certain number of operating hours or kilometers driven, during which the adaptation of the actual air volume in the intake manifold of the internal combustion engine is carried out particularly intensively and rapidly (in larger increments). This phase can be designed to quickly correct initial deviations caused by manufacturing tolerances in order to rapidly bring the internal combustion engine to an optimal operating state. The specified service life could, for example, cover the first 4,000 kilometers or the first 50 operating hours. Within this period, a larger portion of the difference between the current scaling factor and the new scaling factor determined by the correction relationship is used to adjust the actual air volume.After this predetermined service life has elapsed, the process can switch to the slow correction phase, in which adjustments can be made gradually and in smaller steps to ensure long-term stability and efficiency. If the internal combustion engine has been serviced in a workshop and aging effects such as carbon deposits have been removed, the predetermined service life can be reactivated to allow for rapid adaptation to the new condition of the air intake system.

[0067] There are designs where the correction of the air quantity in the air intake path is achieved by correcting the filling model of the internal combustion engine based on the correction relationship.

[0068] The filling model of an internal combustion engine can be a mathematical model that describes the amount of air flowing into the engine's cylinders. It can take into account various parameters, such as engine speed, intake pressure, intake air temperature, and the geometry of the intake tract. The filling model can thus influence the air-fuel ratio, as it forms the basis for determining the required amount of air to ensure optimal combustion. Continuously adapting the filling model to the engine's actual operating conditions ensures consistent and reliable engine performance.

[0069] Correcting the air volume in the intake tract of an internal combustion engine is the process of adjusting the calculated or measured amount of air flowing into the cylinders to compensate for variations due to manufacturing tolerances or aging effects. Air volume correction can be achieved by adjusting the control of various engine components, such as the throttle valve, valve timing, or boost pressure.

[0070] The engine's filling model can be corrected by using a correction relationship based on measurement data and a predefined mathematical function such as a parabola. This correction relationship allows the actual air volume to be adjusted as a function of engine speed to compensate for deviations due to manufacturing tolerances or aging effects, thereby optimizing engine performance and emissions.

[0071] There are versions where the correction of the combustion engine's filling model adjusts a target pressure in the air intake path.

[0072] The target pressure in the air intake system of an internal combustion engine can be the desired pressure (intake manifold pressure) that should prevail in the intake tract to ensure optimal air supply. This pressure can be influenced by various factors, such as the throttle position, the turbocharger boost pressure (if present), and the valve timing. Precise control of the target pressure may be necessary to optimize the air-fuel ratio and thus control the engine's efficiency and emissions.

[0073] By correcting the filling model based on the correction relationship, the target pressure in the air intake tract can be adjusted to compensate for deviations due to manufacturing tolerances or aging effects. This means the engine control unit (ECU) can dynamically adjust the target pressure by modifying control parameters such as throttle position, boost pressure, and valve timing. These adjustments ensure the correct amount of air enters the cylinders, resulting in more efficient combustion and reduced emissions. Continuous adjustment of the target pressure stabilizes engine performance and ensures the longevity of engine components.

[0074] There are versions where the procedure further includes: diagnosing an exceedance of a predetermined tolerance limit, correcting the actual air volume, and indicating that the predetermined tolerance limit has been exceeded.

[0075] Diagnosing an exceedance of a predetermined tolerance limit for the actual air volume correction can be the process by which the system continuously monitors whether the actual air volume adjustments remain within acceptable limits. This monitoring can be performed by the engine control unit (ECU), which analyzes the correction relationship and the current operating conditions of the engine. If the actual air volume corrections exceed a predetermined tolerance limit, this may indicate serious deviations, which could be caused by aging effects (or manufacturing tolerances). The diagnostic system can detect these deviations and generate an appropriate diagnostic message to draw attention to the problem.

[0076] The tolerance limit can be a predetermined threshold that defines the maximum permissible correction of the actual air mass. This limit can be set based on technical specifications, legal requirements, or empirical data, and ensures that corrections remain within an acceptable range, thus protecting components, for example. The tolerance limit can be defined for both short-term and long-term deviations to account for both sudden and gradual changes.

[0077] Furthermore, a pattern can be recognized in the correction over time, for example, that the correction always adapts in the same direction and does not settle around a constant value.

[0078] Indicating that a predetermined tolerance limit has been exceeded can serve as a notification to the driver or maintenance personnel about a problem. This indication can be provided through various methods, such as a warning light on the instrument panel, a message on the vehicle's display, or a notification in the vehicle's diagnostic software. The indication can provide specific information about the nature and extent of the deviation, allowing appropriate action to be taken, such as a visit to a workshop and walnut blasting to remove carbon deposits. Timely indication of a tolerance limit exceedance ensures that potential problems are identified and addressed early, thus maintaining engine performance and reliability.

[0079] There are versions where the procedure further includes: verifying whether the measurement data corresponds to the multiplication of the given relationship with the scaling factor, whereby the correction of the air volume is carried out if the measurement data corresponds to the multiplication of the given relationship with the scaling factor.

[0080] Plausibility checks ensure that the measured data actually correspond to the expected physical properties and patterns described by the predefined relationship. This predefined relationship, such as a parabolic function, represents a theoretical model of the lambda control value deviations as a function of engine speed. During plausibility checks, the measured data can be compared with the correction relationship determined by multiplying the predefined relationship by the scaling factor. If the measured data matches this correction relationship, the air mass is corrected. This means that the measured deviations are indeed caused by the effects modeled in the predefined relationship and thus provide a valid basis for adaptation.Should the plausibility check reveal that the measurement data does not correspond to the correction relationship, the air volume correction is not performed to avoid mismatches. This prevents other error patterns that also affect the lambda control value from being corrected by an incorrect adaptation. This step ensures that only relevant and physically plausible deviations are corrected, thereby increasing the accuracy and reliability of the procedure.

[0081] There are versions where plausibility checking includes verifying the regression quality between the multiplication of the given relationship with the scaling factor and measurement data, and an adjustment to the measurement data is plausible if the regression quality exceeds a given value.

[0082] The goodness of fit of regression, often expressed as the coefficient of determination (R²), is determined by the coefficient of determination (R²). 2Regression goodness of fit (expressed as ) is a statistical measure that quantifies the quality of a model's fit to the observed data. It can indicate how well the given relationship (e.g., a parabolic function) matches the measured data after being multiplied by a scaling factor.

[0083] The correlation between measurement data and the multiplication of the given relationship by the scaling factor (model) can be determined using the regression goodness of fit. This can be achieved by a high R². 2 -value displayed. An R 2 A value of 1 indicates a perfect fit, while a value of 0 means that the model has no explanatory power for the variability of the data. The regression goodness of fit R 2 is defined as: R2=1−∑i=1n(yi−y^i)2∑i=1n(yi−y¯)2.

[0084] Here, y i the observed values, ŷ iThe values ​​predicted by the model, y the mean of the observed values, n the number of observations. The R 2 The R-value indicates what proportion of the variance of the dependent variable is explained by the model. Alternative methods for determining whether the model can explain the measured values ​​include the Mean Absolute Error (MAE), the Mean Squared Error (MSE), the Root Mean Squared Error (RMSE), the Mean Absolute Percentage Error (MAPE), the adjusted R-squared, which takes into account the number of predictors in the model, the Akaike Information Criterion (AIC), and the Bayesian Information Criterion (BIC).

[0085] To correct for differences in air volume, it can be verified whether the measured data are comparable to the predefined relationship (e.g., parabolic function) multiplied by a scaling factor. If the regression goodness R 2If a predefined threshold is exceeded, the measurement data is assumed to conform to the predefined relationship. This means that the deviations in the lambda control value are indeed caused by the effects modeled in the predefined relationship and thus provide a valid basis for adaptation.

[0086] There are implementations where the given relationship is a function, a family of functions is generated by multiplying several different scaling factors with the given relationship, and plausibility checking further includes determining whether the measurement data corresponds to a function from the family of functions.

[0087] The given relationship can be used to model the deviations of the lambda control value as a function of rotational speed. This given relationship is a function that, when multiplied by different scaling factors, generates a family of functions. A family of functions is a set of functions that can take on different forms by varying one or more parameters (in this case, the scaling factors). Plausibility checks involve determining whether the measured data correspond to a function from this family of functions. This means that the measured data must match one of the functions modified by the scaling factors (e.g., using the regression goodness of fit R). 2), to be used as a valid basis for the adaptation. This method ensures that only the deviations actually caused by the effects modeled in the given relationship are corrected, thereby increasing the accuracy and reliability of the procedure. The fitting process can thus be simplified, and the scaling factor whose function exhibits the highest regression goodness of fit can be easily determined, for example, if the regression goodness of fit exceeds a threshold.

[0088] There are versions where the procedure further includes: checking whether preconditions for collecting the measurement data are met; whereby the correction of an air quantity is carried out if the preconditions for collecting the measurement data are met.

[0089] Prerequisites for acquiring measurement data may include specific operating conditions of the internal combustion engine, ensuring that the collected data is representative and consistent. These preconditions may include the engine running in a stable operating condition, the absence of sudden load or speed changes, and the absence of additional influences such as fuel tank venting or other temporary disturbances. Verifying these preconditions ensures that the measurement data is free of interfering factors and thus provides an accurate and reliable basis for determining the correction relationship.

[0090] There are versions where the preconditions include at least one of the following: fluctuations in the performance of the internal combustion engine, tank venting within a specified range, crankcase venting within a specified range, or correction of an actual air quantity above a predetermined limit.

[0091] The prerequisites for acquiring measurement data can include specific operating conditions of the internal combustion engine to ensure that the collected data is representative and consistent. One such prerequisite might be that the internal combustion engine operates in a stable condition, without sudden fluctuations in power or speed. This means that the engine control unit (ECU) verifies that the engine's power output remains constant within a specific range and that no abrupt changes occur that could distort the measurement data. By ensuring a stable operating condition, the measurement data can provide an accurate and reliable basis for determining the correction relationship.

[0092] Another prerequisite may be that fuel tank venting or crankcase ventilation occurs within a specified range while the measurement data is being collected. The specified range can define a normal operating range in each case. Increased fuel tank venting can cause temporary fluctuations in the air-fuel ratio by introducing additional vapors into the intake system. Similarly, crankcase ventilation can introduce oil droplets and other particles into the intake system, which could distort the measurement data. By checking and ensuring that fuel tank venting or crankcase ventilation occurs within its respective specified range during data collection, the accuracy and reliability of the measurement data are increased.Furthermore, a prerequisite can be that the correction factor used to adjust the actual air volume lies within a predetermined limit. This ensures that the adaptation is not based on extreme deviations that might be due to other sources of error. By adhering to these prerequisites, the procedure can ensure that the air volume correction is performed precisely and effectively.

[0093] A control unit according to the invention is designed to execute a method according to one of the above descriptions.

[0094] The control unit can be an engine control unit (ECU) specifically designed to execute the procedure for correcting air volume variations in the intake air tract of an internal combustion engine, as described above. This control unit can receive data from a variety of sensors that continuously acquire data about the engine's operating state, including the lambda control value, engine speed, load conditions, and other relevant parameters. The control unit can be equipped with a microprocessor capable of processing complex mathematical models and algorithms in real time (deterministically). By implementing the described correction relationship, based on a predefined mathematical function, the control unit can make precise adjustments to the actual air volume (in the charge model) to compensate for deviations due to manufacturing tolerances and aging effects.In addition, the control unit may have diagnostic functions that make it possible to detect exceedances of tolerance limits and to send appropriate warnings to the driver or maintenance personnel.

[0095] A vehicle according to the invention comprises the control unit according to the above design and the internal combustion engine.

[0096] The vehicle may include the control unit designed to execute the procedure for correcting air volume variations in the air intake system of an internal combustion engine, as described above. This control unit is integrated into the vehicle and operates in real time to continuously monitor and adjust the engine's operating parameters.

[0097] Real-time in this context can mean that the control unit deterministically calculates models in different computational intervals (1 ms, 2 ms, 5 ms, 10 ms, etc., or speed-synchronized). Thus, for example, a new value is available for each work cycle.

[0098] The vehicle may include an internal combustion engine. This engine can be either a gasoline or diesel engine and is controlled by the control unit, for example, using a filling model or a torque model. The internal combustion engine can be equipped with various sensors that detect important operating parameters such as engine speed, load, and air-fuel ratio, and the resulting lambda control value, and transmit this information to the control unit.

[0099] The vehicle can be, for example, a passenger car or a commercial vehicle. The vehicle can include: trucks, semi-trailer trucks, vans, tippers, garbage trucks, fire engines, ambulances, tow trucks, construction vehicles, buses, motorboats, yachts, fishing boats, ferries, cargo ships, tankers, tugboats, patrol boats, propeller aircraft, helicopters, agricultural aircraft, military aircraft, tractors, and combine harvesters.

[0100] Exemplary embodiments of the invention will now be described by way of example and with reference to the accompanying drawing. Fig. Figure 1 shows a functional setup of a method for correcting air volume differences in the air intake path of an internal combustion engine according to an exemplary embodiment; Fig. Figure 2 shows a characteristic map of an internal combustion engine with measuring ranges according to an exemplary embodiment; Fig. Figure 3 shows a family of standard parabolas according to an exemplary embodiment; Fig. Figure 4a shows a regression of a given relationship with a scaling factor on measurement data according to an exemplary implementation; Fig. Figure 4b shows a regression of a given relationship with a scaling factor on measurement data according to an exemplary implementation; Fig. Figure 5 shows a predefined relationship with an offset according to an exemplary embodiment; Fig. Figure 6 shows a successive correction with a fast and a slow correction range according to an exemplary embodiment; Fig. Figure 7 shows a successive correction with a fast and a slow correction range according to an exemplary embodiment; Fig. Figure 8a shows a speed-dependent actual torque according to an exemplary embodiment; Fig. Figure 8b shows a speed-dependent calculated actual filling level according to an exemplary embodiment; Fig. Figure 8c shows a speed-dependent target intake manifold pressure according to an exemplary embodiment; Fig. Figure 8d shows a speed-dependent lambda control value according to an exemplary embodiment; Fig. Figure 9 shows speed-dependent power outputs of internal combustion engines with differences in cylinder filling; Fig. Figure 10a shows a valve stroke-dependent flow coefficient for an inlet valve; Fig. Figure 10b shows a valve stroke-dependent flow coefficient for an outlet valve; Fig. Figure 11a shows a valve of an internal combustion engine; Fig. Figure 11b shows a pressure-ratio-dependent flow function through the valve; Fig. Figure 11c shows valve lifts and mass flows as a function of the crankshaft angle of an internal combustion engine; Fig. 12 shows an internal combustion engine; Fig. 13a shows a lambda control value over load and speed; Fig. Figure 13b shows the amount of injected fuel over the injection time; Fig. 14a shows a lambda control value over load and speed; Fig. Figure 14b shows the amount of injected fuel over the injection time; Fig. 15a shows a lambda control value over load and speed; Fig. 15b shows a lambda control value over the speed; Fig. 16a shows a lambda control value over load and speed; Fig. Figure 16b shows an actual air mass flow versus a calculated air mass flow; Fig. 17a shows a lambda control value over load and speed; Fig. Figure 17b shows an air mass across a phase position of an intake camshaft; Fig. 18a shows a lambda control value over load and speed; Fig. 18b shows an air mass versus an exhaust pressure; and Fig. Figure 19 shows a vehicle according to an exemplary embodiment;

[0101] Fig. Figure 1 shows a functional setup of a method for correcting air quantity differences in the air intake path of an internal combustion engine according to an exemplary embodiment.

[0102] Correcting air volume variations in the air intake tract of an internal combustion engine aims to correct deviations in air volume caused by manufacturing tolerances and aging effects. Method 100 is based on mathematical modeling and adaptation that ensures the optimal air volume is provided for combustion to optimize engine performance and emissions.

[0103] The procedure 100 includes a set of normal parabolas 110, a residual determination 120, a plausibility check 130, a determination of a degree of air volume difference 140 and a correction of a filling model 150.

[0104] The family of standard parabolas 110 forms the mathematical basis function for correcting air volume differences, modeling the deviations of the lambda control value from its reference and as a function of the engine speed. This basis function is generally a parabolic function, which is described by an equation of the form already discussed above: y(x)=ax2+bx+c The parabola is described as follows, where a, b, and c are constants and x is the independent variable. The shape of the parabola is determined by the coefficient a; if a > 0, the parabola opens upwards, and if a < 0, the parabola opens downwards.

[0105] To form the family of functions, the basic function is multiplied by various scaling factors S to make the model more plausible for the actual deviations of the measurement data and to fit the filling model based on the basic function (predefined relationship) and the scaling factor. The scaled basic function is given by the equation: y(x)=(ax2+bx)∗S. described.

[0106] The Residual Determination 120 measures or receives measurement data, for example, in specific measurement ranges defined by load and speed limits. These measurement ranges are referred to as adaptation windows and represent specific combinations of load and speed relevant for adapting the correction relationship. The load can always remain within the same limits or range, but the speed increases from one measurement range to the next. This is because the air mass difference to be adapted (corrected) increases with the speed. The residuals are the deviations between the measurement data and a reference, for example, a previously determined adaptation and the current adaptation at the time of execution of the procedure, derived from the basic function with a scaling or adaptation factor.

[0107] The plausibility check 130 verifies whether the measurement data correspond to the scaled basis function. This is determined by the regression goodness of fit R. 2 The regression accuracy is checked to determine how well the basic function matches the measured data. If the regression accuracy exceeds a predefined threshold, the air volume correction is performed. Additionally, an offset can be considered during the adjustment and determination of the regression accuracy, which is not taken into account for the subsequent adjustment of the filling model. Offset error sources—that is, error sources that affect the lambda control value but are not speed-dependent or whose speed dependency lies outside the measurement ranges—therefore have no influence on the final correction of the air volume differences.

[0108] Determining the degree of air volume difference (140) determines the scaling factor with which the basic function (predefined relationship) best approximates the measured data. The result of the regression goodness of fit R can be used for this purpose. 2 The value can be used or recalculated. Then, the amount of the air volume difference (140) to be corrected is determined, thus implementing a robust continuous correction function. The desired correction is therefore defined as a correction relationship between the basis function and an adaptation value. The adaptation value can correspond to the determined scaling factor; however, greater robustness of the procedure is achieved if the adaptation value corresponds to a percentage of the scaling factor.

[0109] The correction of a filling model 150 is based on a specific correction relationship. The filling model of the internal combustion engine is adjusted using this correction relationship. The filling model describes the amount of air flowing into the engine's cylinders and thus influences the air-fuel ratio. For example, correcting the filling model adjusts the actual air volume in the intake manifold to compensate for deviations due to manufacturing tolerances or aging effects.

[0110] Furthermore, by measuring the residual relative to a reference, a successive adaptation or correction is possible, with the last adaptation becoming the reference for a subsequent adaptation. Thus, the adaptation of the actual air volume occurs successively, that is, in small, controlled steps (since the adaptation factor then corresponds to a percentage of the scaling factor). A previously used correction relationship is only modified by a portion of the difference to the new correction relationship with the determined scaling factor.

[0111] Furthermore, there can be two correction ranges: a fast correction range, applicable within a predetermined service life (operating hours, kilometers driven) and allowing for larger adjustments, and a slow correction range, applicable outside this service life and allowing only smaller adjustments. The fast correction range enables rapid correction of manufacturing tolerances, while the slow correction range allows for continuous adjustment to aging effects such as coking.

[0112] Furthermore, a diagnostic function can be provided. If the corrections to the actual air volume exceed a predetermined tolerance limit, this is diagnosed and a corresponding warning is sent to the driver or maintenance personnel. This allows for the early detection and resolution of potential problems to ensure the engine's performance and reliability.

[0113] Fig. Figure 2 shows a characteristic map of an internal combustion engine with measuring ranges according to an exemplary embodiment.

[0114] The characteristic curve illustrates the residual determination 120. A lambda control value is plotted against the load on the y-axis and the rotational speed on the x-axis. The lambda control value below the load limit 121 is represented by lines 122a to f, each representing the same lambda control value. Ideally, the lambda control value should be 1 (when calculated as a quotient of its reference). For example, line 122a represents a lambda control value of 1, line 122b a value of 0.98, line 122c a value of 0.95, line 122d a value of 0.94, line 122e a value of 0.88, and line 122f a value of 0.8.

[0115] The measuring ranges F1 to F5 each have the same lower load limits 124 and upper load limits 123. Furthermore, the measuring ranges F1 to F5 have different upper speed limits 126a to 126e and lower speed limits 125a to 125e.

[0116] The lambda control value deviates further and further from its ideal value of 1 as the engine speed increases. This error pattern corresponds to the air volume differences that need to be corrected. Therefore, measuring ranges F1 to F5 can measure this difference, which increases with engine speed. Using the given relationship, e.g., the parabolic function, this air volume difference can be modeled and then corrected.

[0117] An internal combustion engine does not necessarily need to be operated within all speed ranges defined for a measurement range in order to perform the procedure. However, measurement values ​​from the first three measurement ranges, F1 to F3, should be available to ensure sufficient accuracy. Measurement ranges F1 to F3 can be below 4000 revolutions per minute, which may correspond to fuel-efficient driving.

[0118] Fig. Figure 3 shows a family of standard parabolas according to an exemplary embodiment.

[0119] The diagram shows a relative air volume difference compared to the reference on the y-axis and the rotational speed on the x-axis. Thus, the y-axis is based on the deviation of the intake air volume from the modeled air volume, and therefore also on the deviation of the lambda control value from its reference. This deviation can be converted into an air volume using known methods. This relationship applies generally to all figures.

[0120] The family of standard parabolas 110 includes, for example, parabolas 111a to 111g. The depicted parabolas 111a to 111g have, from bottom to top, scaling factors of 1 to -0.4 in steps of 0.2. Thus, the standard parabola (basic function or predefined relationship) opens downwards and exhibits a dependence on the rotational speed, similar to the air volume difference. The air volume difference is represented, for example, by the lines (122a to f in Fig. 2) of the lambda control value related to the air volume difference is illustrated. The constant a is therefore a negative value and the constant b can, for example, be a positive value. The constants a and b are predetermined values. These are determined, for example, by fitting a parabola to a large number of data or simulations that exhibit the error of the air volume difference.

[0121] By multiplying by the scaling factor, the standard parabola (basis function or predefined relationship) can be scaled depending on aging effects such as the degree of coking. As described above, the scaling factor can be used for adaptation; however, an adaptation factor can also be determined based on the scaling factor, thus enabling a successive correction or approximation to an ideal state. This successive approximation improves the robustness of the method. If an adaptation factor is determined, it is multiplied by the standard parabola (basis function or predefined relationship) to establish a correction relationship, which is then used to...

[0122] Fig. Figure 4a shows a regression of a given relationship with a scaling factor on measurement data according to an exemplary implementation.

[0123] The diagram shows a relative filling difference to the reference on the y-axis and the rotational speed on the x-axis. Thus, the y-axis is based on the deviation of the lambda control value from its reference.

[0124] Regression can be used both for plausibility checks 130 and for determining the correction relationship 140.

[0125] The parabola 111f is approximated to the measurement points 131 by varying the scaling factor until an error function, which quantifies the distance of all measurement points 131 to the parabola 111f, reaches a minimum.

[0126] The measured values ​​131 are distributed across the engine speed range, and it is noticeable that fewer measured values ​​are present at higher engine speeds, suggesting economical driving behavior.

[0127] Fig. Figure 4b shows a regression of a given relationship with a scaling factor on measurement data according to an exemplary implementation.

[0128] The diagram shows a relative filling difference to a reference on the y-axis and the rotational speed on the x-axis. Thus, the y-axis is based on the deviation of the lambda control value from a reference.

[0129] Unlike in Fig. 4a are the measured values ​​131 in Fig. 4b in the measuring ranges F1 to F5, as these measuring ranges are in Fig. Figure 2 is shown. Furthermore, an average value 132 is determined for the measured values ​​131 of each measurement range F1 to F5. Thus, the mean values ​​can also be used for the regression of the parabola 111f with a scaling factor.

[0130] Not all measurement ranges F1 to F5 need to be considered or contain measured values ​​131 in order to perform the procedure. Furthermore, the regression can also be performed directly on the measurement data 131 from the measurement ranges. Moreover, it is not always useful to calculate an average, for example, if only one value was measured in a measurement range (F5).

[0131] Fig. Figure 5 shows a predefined relationship with an offset according to an exemplary embodiment.

[0132] The diagram shows the difference in air volume relative to the reference value on the y-axis and the rotational speed on the x-axis. Thus, the y-axis is based on the deviation of the lambda control value from its reference.

[0133] Parabola 111f can be shifted using offsets, represented by arrows 114 and 115, to allow for improved adaptation to measurement data that exhibit further deviations from the lambda control value due to other error sources. This results in the shifted parabolas 111f1 and 111f2.

[0134] Fig. Figure 6 shows a successive correction with a fast and a slow correction range according to an exemplary embodiment.

[0135] The diagram shows a flow adaptation factor on the y-axis and the distance traveled on the x-axis. The flow adaptation factor is the adaptation factor based on the scaling factor.

[0136] The correction of a filling model 150 can therefore be a successive correction. The applied adaptation factor 153 is changed in larger increments up to the specified service life 154 than thereafter. Before the specified service life 154, there is thus a rapid correction range, and after the specified service life 154, there is a slower correction range.

[0137] Since the x-axis represents a distance traveled, for example since the vehicle or internal combustion engine was delivered, the specified service life is also a distance traveled, for example, a number of kilometers. The procedure is therefore executed in the fast correction range for the first 4000 km and then in the slow correction range.

[0138] Limits 151 and 152 define a range where the optimal adaptation factor lies compared to no adaptation at all. This optimal adaptation factor may be present due to manufacturing tolerances, but may not be known under certain circumstances.

[0139] The successive correction changes the applied adaptation factor 153 in the fast correction range in larger steps into the range between the limits 151 and 152.

[0140] If the specified service life 154 is exceeded, the applied adaptation factor 153 oscillates between the limits 151 and 152 in small increments (steady state). This is the slow correction range. The larger increments in the fast correction range allow for rapid initial adjustment to air volume variations due to manufacturing tolerances, and the slow correction range increases the robustness of the process.

[0141] Fig. Figure 7 shows a successive correction with a fast and a slow correction range according to an exemplary embodiment.

[0142] The diagram shows a flow adaptation factor on the y-axis and the distance traveled on the x-axis. The flow adaptation factor is the adaptation factor based on the scaling factor.

[0143] Here, a successive correction of the applied adaptation factor 153 is shown, which does not settle at a constant value. This can occur due to aging effects such as coking. Only the slow correction range is shown.

[0144] If such a pattern is detected or a threshold for the applied adaptation factor is exceeded, a diagnostic function can display this as an error, for example, coking of the intake manifold near the intake valve.

[0145] Fig. Figure 8a shows a speed-dependent actual torque according to an exemplary embodiment.

[0146] The diagram shows a moment plotted on the y-axis and the rotational speed on the x-axis.

[0147] If the actual amount of air entering the cylinder deviates from the calculated amount, a difference in air volume results, which increases with engine speed. This means that the actual torque generated (155b) will, for example, fall short of the calculated torque (155a), which is dependent on engine speed and may be determined by the driver's input. The deviation also increases with engine speed. If the filling model is corrected using this method, the internal combustion engine will then deliver the calculated torque (155a) as the actual torque.

[0148] The curves shown here and in the following Fig. 8b, Fig. 8c and Fig. 8d with the ending b can also be shown reflected on the curves marked with a and thus deviate in the other direction, since this error is also due to manufacturing tolerances.

[0149] Fig. Figure 8b shows a speed-dependent calculated actual filling according to an exemplary embodiment.

[0150] The diagram shows a calculated fill level on the y-axis and the rotational speed on the x-axis.

[0151] The calculated filling 156a initially increases with the rotational speed and then decreases more slowly. With such a calculated filling 156a, the internal combustion engine can generate the torque (155a in Fig. 8a) generate.

[0152] In the case of errors such as coking or slight production deviations, for example, due to a burr forming on a wall in the intake manifold, the cylinder is filled with less air than calculated. This effect increases with engine speed, as illustrated by the actual fill level in Figure 156b.

[0153] The lambda control value makes the difference in cylinder filling measurable according to the procedure and allows it to be attributed to a cylinder filling error (and not a fuel quantity error). The calculated cylinder filling value 156a can then be achieved again across all engine speed ranges by means of the correction.

[0154] Fig. Figure 8c shows a speed-dependent target intake manifold pressure according to an exemplary embodiment.

[0155] The diagram shows intake manifold pressure on the y-axis and rotational speed on the x-axis.

[0156] The intake manifold pressure 157a initially rises with engine speed and then falls again more slowly. With such an intake manifold pressure 157a, the internal combustion engine can achieve the calculated charge (156a in Fig. 8b) under normal conditions - no coking or production deviations - achieve and thus the moment (155a in Fig. 8a) generate.

[0157] In the case of errors such as coking or slight production deviations, for example because a burr has formed on a wall in the intake tract, the cylinder is filled with a smaller amount of air than calculated, since the errors limit the airflow, which creates an increasingly greater deviation at higher flow velocities (and thus also higher piston speeds and rotational speeds).

[0158] The lambda control value makes the difference in cylinder filling measurable according to the procedure and allows it to be attributed to a cylinder filling error (and not a fuel quantity error). By means of the correction, the target pressure in the intake manifold can be increased depending on the engine speed, as represented by the adapted target intake manifold pressure 157b. This allows the calculated cylinder filling 156a to be achieved again in all engine speed ranges. Thus, the limitation of airflow into the cylinder is compensated for by increasing the pressure in the intake manifold.

[0159] The pressure in the intake manifold is provided, for example, by a turbocharger. Such components typically have protection functions that ensure compliance with load limits (157c), such as thermal load limits due to exhaust gas temperature or load limits due to turbocharger speed.

[0160] Fig. Figure 8d shows a speed-dependent lambda control value according to an exemplary embodiment.

[0161] The diagram shows the lambda control value on the y-axis and the rotational speed on the x-axis.

[0162] The ideal value of the lambda control value 158a is 1 or at least fluctuates around 1. This suggests that no significant control is needed to achieve a stoichiometrically ideal combustion ratio.

[0163] The deviation from the lambda control value 158b, i.e., the difference in air volume, is represented as a parabolic deviation. Through the correction described above, the lambda control value can return to the value of lambda control value 158a in all engine speed ranges, as the deviation in the actual air volume is corrected.

[0164] The method in general, but especially the exemplary embodiment from Fig. 8a to 8d thus allow the target torque and therefore the target power output of the internal combustion engine to be regulated, while the speed-dependent deviation of the air-fuel mixture is adapted. This ensures compliance with both maximum deviation limits for power output and emission standards. Lambda control alone could only meet emission standards and might potentially worsen the torque deviation.

[0165] Fig. Figure 9 shows speed-dependent power outputs of internal combustion engines with differences in filling.

[0166] The diagram shows the power output of an internal combustion engine on the y-axis and the rotational speed on the x-axis.

[0167] The performance curves 161 also refer to a specific type of internal combustion engine and are caused by variations in cylinder filling. Thus, during the operation of internal combustion engines, there are differences in the flow of air into the cylinder, and therefore differences in air volume, which the procedure corrects.

[0168] The deviations of the performance curves 116 increase with the engine speed and are particularly pronounced in a plateau area 160 at high engine speed.

[0169] Fig. Figure 10a shows a valve stroke-dependent flow coefficient for an inlet valve.

[0170] The diagram shows a flow coefficient (alpha). k ) is plotted on the y-axis by an inlet valve and the valve lift on the x-axis.

[0171] The course of the flow coefficients (alpha) k , α k) 162 over the valve stroke shows a scatter of the flow coefficients (Alpha) k , α k ) 162 increases towards larger valve strokes.

[0172] Fig. Figure 10b shows a valve stroke-dependent flow coefficient for an outlet valve.

[0173] The diagram shows a flow coefficient (alpha). k , α k ) is plotted on the y-axis by an exhaust valve and the valve lift on the x-axis.

[0174] The observation of the flow coefficients (alpha) k , α k ) 163 at the exhaust valve corresponds to those at the intake valve as in Fig. 10a shown and with reference to Fig. 10a discussed.

[0175] Fig. Figure 11a shows a valve of an internal combustion engine.

[0176] The valve 200 comprises a valve seat in the cylinder head 201 and a valve disc with valve tappet 202.

[0177] The valve is shown in an open position. The incoming air 203 is represented by lines. The air flows through the cylinder head 201 with valve seat and the valve head with valve tappet 202, following a predefined geometry that may be subject to production variations. Furthermore, this geometry may also be subject to variations due to deposits, such as carbon buildup.

[0178] The geometry creates a geometric cross-section 205 (A geom Furthermore, the incoming air focuses further after passing through the valve, creating an effective flow cross-section 204 (A eff The relationship between the two is given by the flow coefficient (α). k ) described in the following formula: Aeff=αkAgeom

[0179] The flow coefficient (α) k ) is the one in Fig. Flow coefficient shown in 10a.

[0180] Fig. Figure 11b shows a pressure-ratio-dependent flow function through the valve.

[0181] The diagram shows a flow function through an inlet valve on the y-axis and a pressure ratio of the pressure in the inlet channel divided by the pressure in the cylinder on the x-axis.

[0182] The flow function has a constant section 211 up to a certain pressure ratio 212. Above the pressure ratio 212, the curve 210 of the flow function becomes parabolic.

[0183] The change in the effective flow cross-section ( Fig. 11a), for example due to production variations or aging effects, has a direct influence on the mass flow rate that flows through the intake valves into the cylinders. The smaller the effective flow cross-section, the greater the pressure difference between the intake port and the cylinder. As a result, the speed of sound is reached more quickly at the throttle point, which limits the mass flow rate. This limitation of the mass flow rate becomes increasingly pronounced at higher engine speeds, as the piston speed, which acts as the driving force for the intake of air, also increases. Thus, a difference in the effective flow area and its effects become particularly noticeable at higher engine speeds.

[0184] This explains the speed dependency of the error sources corrected in the process and why the given relationship can be corrected, for example, with a parabola.

[0185] Fig. Figure 11c shows valve lifts and mass flows as a function of the crankshaft angle of an internal combustion engine.

[0186] The diagram shows valve lifts and mass flows plotted on the y-axes and a crankshaft angle on the x-axis.

[0187] The valve lifts show that the exhaust valve lift 214 before the crankshaft angle of 360° follows an approximately normal distribution. This also applies to the intake valve lift 215 after the crankshaft angle of 360°. In the region around the crankshaft angle of 360°, there is an overlap area in which both valves are slightly open.

[0188] Accordingly, before the crankshaft angle of 360°, a mass flow 216 of exhaust gas from the cylinder occurs, and after the crankshaft angle of 360°, a mass flow 217 of air enters the cylinder. In the overlap region, there is a back-and-forth sloshing of the exhaust gas and air.

[0189] However, the valve strokes are small in the overlap area, so the sources of error (in Fig. (Discussed in sections 9 to 11b) production variations and aging effects in this area do not have a significant impact. However, they do have a significant impact in the area of ​​the maximum valve lift 215 of the intake valve.

[0190] Fig. Figure 12 shows an internal combustion engine.

[0191] The internal combustion engine 250 comprises lines from a tank vent 251 and crankcase vent 252, an intake manifold 253, a compressor 254a of a turbocharger, a pressure sensor 255, an intake manifold 265, an intake valve 257, an injector 258, an exhaust valve 255, an exhaust manifold 260, a turbine 254b of a turbocharger, a lambda sensor 261 upstream of a catalytic converter, the catalytic converter 262, a lambda sensor 263 downstream of the catalytic converter, a cylinder head 264, a cylinder 265, a piston 266 and a connecting rod 267.

[0192] Vapors from the fuel tank vent 251 and crankcase ventilation 252 can be routed to the intake manifold 253. Fresh air is drawn in through the intake manifold 235. The fresh air, and possibly the vapors, are compressed in the compressor 254a and fed into the intake manifold 256. The boost pressure is monitored by the pressure sensor 255.

[0193] From the intake manifold 256, the compressed air flows through the open intake valve 257 into the cylinder (space between cylinder head 264, cylinder 265, and piston 266). During the intake stroke, the piston moves downwards. The piston compresses the drawn-in air, causing it to move upwards with the intake valve 257 closed. The injector 258 injects fuel. The air-fuel mixture is ignited, moving the piston downwards during the power stroke. Then, during another upward movement with the exhaust valve 259 open, the exhaust gas is directed into the exhaust manifold 260. The exhaust valve 259 closes, and the intake valve 257 opens, thus resulting in the [missing information - likely a specific value or value]. Fig. 11c discussed overlap area.

[0194] The exhaust gas drives the turbocharger's turbine 254b, which in turn drives the compressor 254a via a shaft (not shown). After turbine 254b, the exhaust gas flows through lambda sensor 261 upstream of the catalytic converter, the catalytic converter 262, and lambda sensor 263 downstream of the catalytic converter. Here, the air-fuel ratio of the combustion process is determined, and the resulting lambda control value is calculated. If deviations in the lambda control value are detected by this process, a difference in air volume can be determined and corrected by adjusting the target intake manifold pressure as a function of engine speed. Any of the aforementioned adjustment options for the intake tract can be used for this purpose.

[0195] In Fig. Figure 13a shows a lambda control value over load and speed.

[0196] The diagram shows a lambda control value versus the load on the y-axis and the rotational speed on the x-axis.

[0197] The lambda control value is shown below the load limit 301 by the lines 302 to 308, which here each represent a lambda control value of 0.98.

[0198] Thus, a 2 percent deviation in the lambda control value is measured across the entire load-speed diagram. There is no speed dependency. Such a deviation is adjusted as an offset in the procedure, but not corrected, as it is caused, for example, by an incorrect slope in the amount of injected fuel over the injection time.

[0199] In Fig. Figure 13b shows the amount of injected fuel over the injection time.

[0200] The diagram shows an injection quantity on the y-axis and an injection time on the x-axis.

[0201] The error pattern that leads to the in Fig. The deviation discussed in section 13a is a different slope of the injected fuel quantity over the injection time 401 compared to a reference 400. Here, the fuel quantity can be adapted, but not the air quantity, at least if the moment desired by the driver is to be met.

[0202] Such a deviation is adjusted as an offset in the process, but not corrected, since it is caused, for example, by an offset in the amount of injected fuel over the injection time. Such an error cannot be detected by the measuring ranges in this process.

[0203] In Fig. Figure 14a shows a lambda control value over load and speed.

[0204] The diagram shows a lambda control value versus the load on the y-axis and the rotational speed on the x-axis.

[0205] The lambda control value is calculated as an error between a real amount of injected fuel and a calculated amount of fuel (for example, a quotient).

[0206] The lambda control value below the load limit 301 is represented by lines 302 to 308. Lines 302 to 305 represent a lambda control value of 1. Furthermore, line 306 represents a lambda control value of 0.98 and line 307 a lambda control value of 0.94.

[0207] Thus, a deviation in the lambda control value was measured at low loads. There is no speed dependency.

[0208] In Fig. Figure 14b shows the amount of injected fuel over the injection time.

[0209] The diagram shows an injection quantity on the y-axis and an injection time on the x-axis.

[0210] The error pattern that leads to the in Fig. The deviation discussed in section 14a is an offset of the injected fuel quantity over the injection time 401 compared to a reference 400. Here, the fuel quantity can be adapted, but not the air quantity, at least if the moment desired by the driver is to be met.

[0211] In Fig. 15a shows a lambda control value over load and speed.

[0212] The diagram shows a lambda control value versus the load on the y-axis and the rotational speed on the x-axis.

[0213] The lambda control value below the load limit 301 is represented by lines 302 to 308. For example, line 302 represents a lambda control value of 1, line 303 a value of 0.98, line 304 a value of 0.95, line 305 a value of 0.94, line 306 a value of 0.88, and line 307 a value of 0.8.

[0214] Thus, a deviation of the lambda control value from its reference value, which increases with engine speed, is measured. This deviation is corrected in the procedure.

[0215] In Fig. Figure 15b shows a lambda control value over the engine speed.

[0216] The diagram shows a lambda control factor on the y-axis and an injection time on the x-axis.

[0217] The error pattern that leads to the in Fig. The deviation discussed in section 15a is a difference in air volume, which leads to a deviation in the lambda control value 401 (analogous to the lambda control factor). The procedure corrects this deviation towards the reference 400, which corresponds to the x-axis here.

[0218] In Fig. Figure 16a shows a lambda control value over load and speed.

[0219] The diagram shows a lambda control value versus the load on the y-axis and the rotational speed on the x-axis.

[0220] The lambda control value below the load limit 301 is represented by lines 302 to 305. Here, for example, line 302 illustrates a lambda control value of 1, line 303 one of 0.98, line 304 one of 0.95, and line 305 one of 0.9.

[0221] This indicates a deviation in the lambda control value, which increases with low engine speed and low load. Such a deviation suggests leakage or blow-by problems.

[0222] In Fig. Figure 16b shows an actual air mass flow versus a calculated air mass flow.

[0223] The diagram shows an actual air mass flow on the y-axis and a calculated air mass flow on the x-axis.

[0224] The error pattern that leads to the in Fig. The deviation discussed in section 16a is one that can be caused by leakage or blow-by problems. Such a deviation is not corrected via the air-fuel mixture, but the procedure must recognize it as uncorrectable. The deviation of the actual air mass flow 401 from the reference 400 (diagonal) is particularly noticeable at low air mass flows and is therefore not relevant at high engine speeds. Consequently, it is also not relevant to the procedure and can only affect the determination of the offset.

[0225] In Fig. Figure 17a shows a lambda control value over load and speed.

[0226] The diagram shows a lambda control value as a function of load on the y-axis and rotational speed on the x-axis.

[0227] The lambda control value below the load limit 301 is represented by lines 302 to 305. Here, for example, line 302 illustrates a lambda control value of 1, line 303 one of 0.98, and line 304 one of 0.95.

[0228] This indicates a deviation in the lambda control value, which forms an island at low engine speed and low load. Such a deviation suggests valve train tolerances or adaptation errors.

[0229] In Fig. Figure 17b shows an air mass over a phase position of an intake camshaft.

[0230] The diagram shows an air mass on the y-axis and a phase position of an intake camshaft on the x-axis.

[0231] The error pattern that leads to the in Fig. The deviation discussed in section 17a is a valve train tolerance or adaptation error. Such a deviation is not corrected by the method according to the invention, but must be recognized by the method as not being correctable.

[0232] In Fig. Figure 18a shows a lambda control value over load and speed.

[0233] The diagram shows a lambda control value as a function of load on the y-axis and rotational speed on the x-axis.

[0234] The lambda control value below the load limit 301 is represented by lines 302 to 305. Here, for example, line 302 illustrates a lambda control value of 1, line 303 one of 0.98, and line 304 one of 0.95.

[0235] This indicates a deviation in the lambda control value, which forms an island at low engine speed and high load. Such a deviation suggests a modeling error in the filling model or a pressure sensor error.

[0236] In Fig. Figure 18b shows an air mass versus an exhaust pressure.

[0237] The diagram shows an air mass on the y-axis and exhaust pressure on the x-axis.

[0238] The error pattern that leads to the in Fig. The deviation discussed in section 18a is either a modeling error in the filling model or a pressure sensor error. Such a deviation is not corrected via the air-fuel mixture but must be recognized by the procedure as uncorrectable. The deviation of air mass 401 from the reference 400 shows a different slope with respect to exhaust pressure.

[0239] Fig. Figure 19 shows a vehicle according to an exemplary embodiment.

[0240] The vehicle 500 comprises a control unit 502 and an internal combustion engine 501. The control unit 502 is designed to execute the procedure. Reference symbol list 100 procedures 110 sets of standard parabolas 120 Residual Determination 130 Plausibility Check 140 Determination of the degree of air volume difference 150 Correction of a filling model 121 Load limit 122a-f lines of the lambda control value 123 Maximum load 124 Lower load limit 125a-e lower speed limits 126a-e Maximum speed limits F1-F5 measuring ranges 111a-g Parabolas 131 measuring points 132 Average value 114, 115 arrows (offset) 111f1, 111f2 Shifted Parabolas 151, 152 Borders 153 Applied adaptation factor 154 Specified service life 155a calculated moment 155b Actual generated moment 156a Calculated filling 156b Actual Filling 157a Intake manifold pressure 157b Adapted intake manifold pressure 157c Load limits 158a Ideal value of the lambda control value 158b Deviation of the lambda control value 160 plateau area 161 performance curves 162 flow coefficients 200 valve 201 Cylinder head 202 Valve plates with valve tappets 203 Incoming air 204 Effective flow cross-section 205 Geometric cross-section 210 Flow function curve 211 Constant section 212 pressure ratio 214 Valve lift of the exhaust valve 215 Valve lift of the inlet valve 216 Mass flow of exhaust gas 217 Mass flow of air 250 internal combustion engine 251 Tank vent 252 Crankcase ventilation 253 Intake manifold 254a Compressor of a turbocharger 254b Turbine of a turbocharger 255 Pressure sensor 256 Intake manifold 257 Inlet valve 258 injector 259 Exhaust valve 260 Outlet collector 261 Lambda sensor before the catalytic converter 262 Catalyst 263 Lambda sensor after the catalytic converter 264 Cylinder head 265 cylinders 266 pistons 267 connecting rods 301 Maximum load 302-308 lines of the lambda control value 400 Reference 401 Curve 500 vehicles 501 Internal combustion engine 502 Control unit

Claims

[1] Method (100) for correcting air quantity differences in the air intake path of an internal combustion engine (501), comprising: Received (120) from measurement data (131, 132) indicating a deviation of an air mass drawn in by the internal combustion engine (501) from a modeled air mass as a function of the rotational speed of the internal combustion engine (501); Determining (140) a correction relationship comprising a multiplication of a given relationship (111) between the deviation over the rotational speed by a scaling factor (S), wherein the correction relationship approximates the measurement data (131, 132); Correcting (150) an actual air volume in the air intake path based on the correction relationship. [2] Method (100) according to claim 1, wherein the measurement data (131, 132) comprise at least three measured values. [3] Method (100) according to one of the preceding claims, wherein the measurement data (131, 132) are obtained from a lambda controller. [4] Method (100) according to one of claims 1 or 2, wherein the measurement data (131, 132) are obtained from an air mass meter. [5] Method (100) according to any of the preceding claims, wherein the specified relationship (111) is a parabolic function. [6] Method (100) according to one of the preceding claims, wherein determining (140) a correction relationship comprises an addition with an offset (114, 115) to approximate it to the measurement data (131, 132). [7] Method (100) according to one of the preceding claims, wherein the measurement data (131, 132) are collected in a measurement range (F1-F5) which can be defined by load and speed limits (123, 124, 125, 126). [8] Method (100) according to claim 6, wherein several measuring ranges (F1-F5) are provided, each of which is separated from each other by lower speed limits (125a-e), upper speed limits (126a-e), lower load limits (124) and upper load limits (123). [9] Method (100) according to one of the preceding claims, wherein the correction (150) of the actual air quantity to the calculated air quantity is carried out successively based on the correction relationship and an already used correction relationship is changed only by a part of the difference from a current scaling factor to a scaling factor (S) determined by the correction relationship. [10] Method (100) according to claim 9, wherein a fast correction range is provided which is applicable within a predetermined service life (154) and in which a larger part of the difference is used to correct the actual air volume than in a slow correction range outside the predetermined service life (154). [11] Method (100) according to one of the preceding claims, wherein the correction (150) of the actual air quantity in the air intake path is carried out by a correction of the filling model of the internal combustion engine (501) based on the correction relationship. [12] Method (100) according to claim 11, wherein the correction (150) of the filling model of the internal combustion engine (501) adjusts a target pressure in the air intake path. [13] Method (100) according to any one of the preceding claims, further comprising: Diagnosing an exceedance of a predetermined tolerance limit of the correction (150) of the actual air volume and Indicates that the predetermined tolerance limit has been exceeded. [14] Method (100) according to any one of the preceding claims, further comprising: To verify (130) whether the measurement data (131, 132) correspond to the multiplication of the given relationship (111) by the scaling factor (S), wherein The correction (150) of the air quantity is carried out when the measurement data (131, 132) correspond to the multiplication of the given relationship (111) with the scaling factor (S). [15] Method (100) according to claim 14, wherein the plausibility check (130) is a check of the regression goodness (R) 2 ) between the multiplication of the given relationship (111) with the scaling factor (S) and measurement data (131, 132) and an adjustment to the measurement data (131, 132) is plausible if the regression goodness of fit (R) 2 ) exceeds a predetermined value. [16] Method (100) according to claim 14 or 15, wherein the predetermined relationship (111) is a function, by multiplying several different scaling factors (S) with the given relationship (111) a family of functions (111a-g) is generated and plausibility checking (130) further includes, Determine whether the measurement data (131, 132) correspond to a function from the family of functions (111ag). [17] Method (100) according to any one of the preceding claims, further comprising: Check whether the prerequisites for collecting the measurement data (131, 132) are met; where the correction (150) of the actual air volume is carried out when the preconditions for the collection of the measurement data (131, 132) are met. [18] Method (100) according to claim 17, wherein the preconditions comprise at least one of fluctuations in the power of the internal combustion engine (501), tank venting within a predetermined range, crankcase venting within a predetermined range or correcting (150) an actual air quantity above a predetermined limit. [19] Control unit (100) configured to perform a method (100) according to any of the preceding claims. [20] Vehicle (500) comprising the control unit (502) according to claim 19 and the internal combustion engine (501).

Images