DE102017113009B4 - Method and device for correcting a sensor signal in an exhaust duct of an internal combustion engine - Google Patents

Abstract

Method for correcting a sensor signal of a sensor (26, 28) in an exhaust system (12) of an internal combustion engine (10), wherein a dynamic correction of the sensor signal is carried out by correcting a measured sensor signal by a dynamic correction value, the dynamic correction value being derived from a sensor model of the sensor (26, 28), the sensor signal of which is to be corrected, the dynamic component from the sensor model being added to the measured signal of the sensor (26, 28), the sensor (26, 28) being a temperature sensor (26) is, wherein the sensor temperature of the temperature sensor (26) is determined and a dynamic temperature change of the temperature signal is dynamically corrected by estimating a temperature in the exhaust system (12), characterized in that the dynamic component of the difference between an estimated gas temperature and the signal of the sensor model corresponds, the dynamic correction value being weighted depending on the gradient of the temperature change measured at the temperature sensor (26), and a complementary filter for dynamic correction of the exhaust gas temperature sensor (26) being designed based on a model, the geometric dimensions of the temperature sensor (26) and the exhaust gas duct section (36), in which the temperature sensor (26) is arranged, flow into the sensor model.

Description

The invention relates to a method and a device for correcting a sensor signal in an exhaust duct of an internal combustion engine according to the preamble of the independent claims.

The current and increasingly strict exhaust gas legislation in the future places high demands on the engine's raw emissions and the exhaust gas aftertreatment of combustion engines. The demands for a further reduction in consumption and the further tightening of emissions standards represent a conflict of objectives for engine developers. In order to enable the most efficient and clean, low-emission combustion of the fuel or to improve the quality of the exhaust aftertreatment, there are sensors in the exhaust duct of the combustion engine , in particular temperature sensors, NOx sensors or lambda sensors, which are subject to highly dynamic stress. In particular, highly dynamic temperature measurement at ever higher peak temperatures in the exhaust gas represents a challenge for the sensors. Due to the finite heat transfer and the internal inertia of the thermal material, the measurement signal of the measured exhaust gas temperature only follows the actual temperature curve with a time delay. To control and/or monitor exhaust aftertreatment systems, one or more temperature sensors are arranged in an exhaust duct of the internal combustion engine. In dynamic engine operation, the measured temperature curve therefore has a time delay compared to the actual temperature curve. Particularly when monitoring and/or regulating components of the exhaust gas aftertreatment system, this dynamic inertia of the sensor results in deviations which can lead to a suboptimal result of the exhaust gas aftertreatment. It is particularly problematic if the exhaust gas temperature is set in relation to other parameters which are determined by calculation or dynamically better sensors.

One way to correct a sensor signal is to use multiple sensors. However, this leads to an increase in costs and the temporal derivation of the signals is also required, which is problematic in the case of noisy signals and due to the quantization of the signals.

Another possibility for correction is inverse filtering of the sensor signal. This solution also has the problem that the time derivative of the signals is required. Due to the coarse quantization of the signals, large temperature gradients are calculated when determining the derivative. Low-pass filtering the signal can only partially solve this problem. However, there are limits to the process when it comes to dynamic correction.

Another possibility for correcting a sensor signal is complementary filtering of the sensor signal. A fixed filter constant is used. However, this can be problematic because the resulting sensor signal does not have sufficient signal quality throughout the entire operating range. Another problem is that if the cutoff frequency is too high, disturbances in a slow signal are not adequately attenuated. A lower counter frequency attenuates the disturbances of the slow signal, but has a negative effect on the low-frequency components of the fast signal. From the DE 11 2014 000 060 T5 A method is known in which different complementary filters with different cutoff frequencies are used for different boundary conditions.

From the DE 101 08 181 A1 a method and a device for correcting a temperature signal from a temperature sensor in a motor vehicle are known, the correction being carried out by a first correction factor, which takes into account the response behavior of the sensor, and by a second correction factor, which takes into account a temporal behavior of the internal combustion engine and / or the internal combustion engine assigned components are taken into account. This makes it possible in particular to improve the dynamic behavior of the sensor signal when an operating parameter of the internal combustion engine changes.

From the DE 10 2005 003 832 A1 a method and a device for measuring the temperature of flowing fluids are known. Measuring the temperature of flowing fluids is particularly difficult when there are highly dynamic temperature changes. Due to the finite heat transfer and the internal inertia of the probe's thermal material, the measurement signal only followed the course of the fluid temperature with a delay. This is done in the DE 10 2005 003 832 A1 A method is proposed in which the signals from at least two thermal sensors are evaluated and the signal from the thermal sensors is corrected mathematically. For this purpose, the dynamics of the heat transfer are divided into parts that depend on the ambient conditions, the heat transfer coefficient and other factors and the measurement signal is corrected accordingly.

The DE 10 2014 208 751 A1 discloses a method for providing a modeled nominal signal of a lambda sensor. For this purpose a Sen sor signal of the lambda sensor is detected, an adapted target signal is determined and a modeled nominal signal is calculated from the sensor signal, the adapted target signal and a target value of the nominal signal.

The DE 103 00 939 A1 discloses a method and a device for monitoring the NOx signal of a NOx sensor in the exhaust system of an internal combustion engine. The NOx signal is examined for plausibility depending on the operating parameters of the combustion engine.

In addition, from the EP 2 458 167 A1 a method and a device for determining an exhaust gas temperature are known. A temperature sensor is arranged in an exhaust duct of an internal combustion engine, with a time delay in the change in the temperature signal being adjusted by a correction factor.

Furthermore, from the DE 10 2013 203 175 A1 a method for determining the temperature of hot gases with minimal errors is known. Temperature measurement values are continuously recorded using a temperature sensor, with an error-minimized, expected temperature value being determined from several successive temperature measurement values.

The invention is now based on the object of proposing a method for correcting a sensor signal in an exhaust duct of an internal combustion engine, in which the inertia caused by the measuring principle is dynamically corrected, the disadvantages known from the prior art, in particular the use of several temperature sensors, being overcome .

According to the invention, the object is achieved by a method for correcting a sensor signal of a sensor in the exhaust duct of an internal combustion engine, wherein a dynamic correction of the sensor signal is carried out by correcting a measured sensor signal by a dynamic correction value, the dynamic correction value being obtained from a sensor model of the sensor, whose sensor signal is to be corrected, whereby the dynamic component from the sensor model is added to the measured signal of the sensor. The aim of the invention is to correct a slow sensor using a dynamic model value in order to obtain a faster sensor signal at the sensor position. The measured temperature should be approximated to a temperature calculated using a sensor model. This is achieved using a sensor model that has the same time behavior as the real sensor. If this is the case, the dynamic component can be determined from the sensor model in conjunction with the fast model. In conjunction with the measured sensor signal of the real sensor, the desired dynamics of the corrected sensor signal results. The starting point for the dynamic correction is the fast model, which is used to estimate the measured variable on the sensor to be dynamically corrected. The real sensor cannot record this measurement precisely due to its transmission behavior. Using the estimated measurement variable, the dynamic behavior of the real sensor is depicted using a sensor model. This sensor model is modeled so that the modeled signal of the sensor model dynamically shows the same behavior as the real sensor. The difference between the estimated measured value and the signal of the sensor model corresponds to the dynamic correction value, which is superimposed on the signal of the real sensor, since the real sensor cannot adequately detect these dynamic changes due to its design-related, time-delayed response. To do this, the dynamic correction value is added to the measured value determined by the real sensor, so that the result is that the real gas state at the position of the real sensor can be approximated. This method can be used to calculate fast sensor signals that cannot be measured under real vehicle conditions.

According to the invention it is provided that the sensor is a temperature sensor, the sensor temperature of the temperature sensor being determined and a dynamic temperature change in the temperature signal being dynamically corrected by estimating a temperature in the exhaust duct. Particularly in the case of temperature sensors that are subject to high thermal loads, particularly in the exhaust duct of an internal combustion engine upstream of the turbine of the exhaust gas turbocharger, a failure is critical, since a failure of the temperature sensor causes the internal combustion engine to switch to emergency running mode. In order to reduce the temperature gradients of the temperature sensor and thus increase the durability and thermal shock resistance of the sensor, the thermal mass of the temperature sensor is increased. This results in the temperature signal losing significant momentum. This conflicts with the most accurate temperature measurement possible, even with dynamic temperature changes. Therefore, the application of the proposed method is particularly advantageous for a temperature sensor in order to improve the dynamic behavior through a dynamic correction factor.

According to the invention, it is provided that a complementary filter for dynamic correction of the exhaust gas temperature sensor is designed based on a model. This can be compared to A complementary filter with a fixed value can be used to correct the measured value more precisely.

Furthermore, it is provided according to the invention that the geometric dimensions of the temperature sensor and the exhaust duct section in which the temperature sensor is arranged are incorporated into the sensor model. In addition to the heat capacity and the size of the sensor, the immediate sensor environment also influences the dynamic behavior of the sensor. Therefore, the geometric dimensions of the exhaust duct section are also taken into account in the sensor model in order to generate an even more precise model.

According to the invention it is provided that the dynamic correction factor is weighted depending on the gradient of the temperature change measured at the temperature sensor. Since at a stationary operating point the exhaust gas temperature is essentially constant after a starting phase, a correction is provided, particularly at dynamic operating points.

The features specified in the dependent claims make advantageous improvements and developments of the method specified in the independent claim possible.

It is preferred that the dynamic correction factor is weighted more heavily if a strong temperature change is detected on the temperature sensor between two measurements that immediately follow each other in time. If the sensor detects that the exhaust gas temperature in the exhaust duct changes, the dynamic correction element can be adjusted accordingly depending on the change.

In an alternative embodiment of the method, which is not part of the invention, it is provided that the sensor is a NOx sensor, the NOx concentration being measured at the NOx sensor and a dynamic change in the NOx concentration by estimating a concentration in the exhaust duct dynamically is corrected. In addition to temperature sensors, other sensors in the exhaust system and in the fresh air path of the combustion engine also have a certain inertia. Therefore, the method according to the invention for dynamically correcting a measured value is not only limited to a temperature sensor, but can also be used on a sensor for measuring the nitrogen oxide concentration in the exhaust system. Analogous to the method described for the temperature, a correction of the measured value determined on the sensor can be carried out, particularly in dynamic operating situations, in order to adapt the exhaust gas aftertreatment, in particular the amount of metered reducing agent, to the operating situation.

In a further embodiment of the method that is not part of the invention, it is provided that the sensor is a lambda sensor, in particular a broadband lambda sensor, the oxygen concentration being measured at the lambda sensor and a dynamic change in the oxygen concentration being dynamically corrected by estimating an oxygen concentration in the exhaust gas duct . In principle, the method according to the invention can also be transferred to a lambda sensor in the exhaust system of the internal combustion engine in order to further improve the measurement accuracy of the lambda sensor in highly dynamic operating states and to enable improved control of the combustion air ratio in the combustion chambers of the internal combustion engine. This allows the raw emissions from the combustion engine to be further reduced.

According to the invention, a device for correcting a sensor signal with a sensor is proposed, with which a parameter, in particular a temperature, is measured in an exhaust duct of an internal combustion engine, the device having means with which a dynamic correction of the measured parameter is carried out, the means being the Take the response behavior of the sensor into account and adapt a time delay in the measured signal using a dynamic correction.

According to the invention, a control device is also proposed which is set up to carry out a method according to the invention for correcting a sensor signal when a machine-readable program code is executed by the control device. Using a corresponding program code on the control unit, a method according to the invention can be easily implemented in order to further improve the control of the internal combustion engine or the components for exhaust gas aftertreatment.

The various embodiments of the invention mentioned in this application can be advantageously combined with one another, unless stated otherwise in individual cases.

• 1 ein Ausführungsbeispiel eines Verbrennungsmotors, in dessen Abgasanlage eine Mehrzahl von Sensoren angeordnet ist, deren Genauigkeit in einem dynamischen Betrieb durch ein erfindungsgemäßes Verfahren verbessert wird;
• 2 ein Modell zur dynamischen Korrektur eines Messwertes am Beispiel eines Temperatursensors; und
• 3 ein Ablaufdiagramm zur Durchführung eines erfindungsgemäßen Verfahrens zur dynamischen Korrektur eines Messwertes eines Sensors in einem Abgaskanal eines Verbrennungsmotors.

The invention is explained below in exemplary embodiments using the associated drawings. The same components or components with the same function are marked with the same reference numbers. Show it:

• 1 an embodiment of an internal combustion engine, in the exhaust system of which a plurality of sensors are arranged, their accuracy in dynamic operation is improved by a method according to the invention;
• 2 a model for the dynamic correction of a measured value using the example of a temperature sensor; and
• 3 a flowchart for carrying out a method according to the invention for dynamically correcting a measured value from a sensor in an exhaust duct of an internal combustion engine.

In 1 an exemplary embodiment of an internal combustion engine 10 with an exhaust gas aftertreatment system is shown. The internal combustion engine 10 is preferably designed as a self-igniting diesel engine, but can also be designed as a spark-ignited internal combustion engine 10, in particular as an internal combustion engine 10 spark-ignited using spark plugs based on the Otto principle. An exhaust system 12 is connected to an outlet 14 of the internal combustion engine 10. In the exhaust system 12, in the direction of flow of an exhaust gas from the internal combustion engine 10 through the exhaust system 12, a turbine 16 of an exhaust gas turbocharger 18 is arranged downstream of the outlet 12 and an oxidation catalytic converter 32 is arranged downstream of the turbine 16. A metering module 20 is arranged downstream of the oxidation catalyst 32, with which a reducing agent, in particular aqueous urea solution, can be metered into the exhaust system 12. A particle filter 22 with a coating 24 for the selective, catalytic reduction of nitrogen oxides is arranged downstream of the metering module 20 in order to filter out particles from the exhaust gas and reduce nitrogen oxide emissions. As an alternative to a coated particle filter 22, 24, the particle filter 22 and a catalyst for the selective, catalytic reduction of nitrogen oxides can also be designed as separate components. In addition to or instead of the oxidation catalytic converter 32, a NOx storage catalytic converter 34 can be arranged in the exhaust system 12.

In order to optimize combustion in the combustion chambers of the internal combustion engine and to ensure the best possible function of the exhaust gas aftertreatment components 18, 20, 22, 24, several sensors 26, 28 are arranged in the exhaust system 12. For reasons of clarity, in the exemplary embodiment shown there is only one temperature sensor 26, which is arranged upstream of the exhaust gas turbocharger 18 and thus in the most thermally stressed area of the exhaust system 12, and one NOx sensor 28, which is arranged downstream of the particle filter 22. shown. In addition to the sensors 26, 28 shown, other sensors, in particular other temperature sensors, NOx sensors or pressure sensors, are also possible. The signals from the sensors 26, 28 are evaluated and processed in a control unit 30, in particular in an engine control unit of the internal combustion engine 10.

In 2 is a schematic representation of a dynamic correction of a measured value based on sensor models using the example of a temperature sensor 26. The aim of the method is to correct a slow sensor 26, 28 using a dynamic model value in order to obtain a faster gas temperature signal at the position of the sensor 26, 28. In addition to the dimensions and thermal capacities of the sensor 26, 28, the dimensions of the exhaust duct section 36, in which the sensor 26, 28 is arranged, are also included in the sensor model. A direct combination of the fast signal with the sensor 26, 28 is not possible. Therefore both signals must have comparable time behavior. This is realized in the proposed method using a sensor model which has the same time behavior as the real sensor 26, 28. If the time behavior of the model is essentially the same as the time behavior of the sensor 26, 28, the dynamic component of the change in temperature in the exhaust system 12 can be determined from the sensor model in conjunction with the fast model. In conjunction with the measured temperature signal from the real temperature sensor 26, the dynamic gas temperature in the exhaust system 12 at the position of the temperature sensor 26 can thus be determined at the desired speed. With this structure, a filter behavior can now be achieved that cannot be achieved with conventional complementary filter structures and is valid for all operating points across the map.

The starting point for the dynamic correction is a fast model with which an exhaust gas temperature I is estimated at the sensor 26, 28 to be dynamically corrected. A dynamic component III is estimated at the model level ME of the sensor model SM. The real sensor RS cannot detect this exhaust gas temperature I due to its transmission behavior. With the help of the estimated exhaust gas temperature I, the dynamic behavior of the sensor signal of the real sensor RS is mapped using a sensor model SM II. This sensor model is to be modeled in such a way that the modeled signal II dynamically shows the same behavior as the real sensor RS. The difference between the estimated gas temperature I and the signal from the sensor model SM corresponds to the dynamic component III, which the real sensor RS cannot detect due to its design-related time-delayed response. This dynamic component III is added to the measured temperature of the real sensor RS in order to obtain the real gas state IV to get to the position of the real sensor RS. This is done at the sensor level SE of the model.

In 3 a flowchart for correcting a sensor signal is shown. The starting point is a model of the sensor 26, 28, the dynamic behavior of which is estimated in a method step <100>. A measured value of the desired measurement variable is modeled at the corresponding position of the sensor in the exhaust system 12 in a method step <110>. In a method step <120>, the dynamic behavior of the real sensor 26, 28 is mapped using the sensor model. In a method step <130>, the dynamic behavior is added as a dynamic correction factor to the value measured by the sensor 26, 28 and thus a value is obtained which corresponds to the actual exhaust gas temperature at the position of the temperature sensor 26 in the event of a dynamic temperature change of the exhaust gas flow of the internal combustion engine 10 .

Reference symbol list

10

Internal combustion engine

12

Exhaust system

14

outlet

16

turbine

18

Exhaust gas turbocharger

20

Dosing module

22

Particulate filter

24

Coating for the selective, catalytic reduction of nitrogen oxides

26

Temperature sensor

28

NOx sensor

30

Control unit

32

Oxidation catalyst

34

NOx storage catalytic converter

36

Exhaust duct section

Claims (4)

Method for correcting a sensor signal of a sensor (26, 28) in an exhaust system (12) of an internal combustion engine (10), wherein a dynamic correction of the sensor signal is carried out by correcting a measured sensor signal by a dynamic correction value, the dynamic correction value being derived from a sensor model of the sensor (26, 28), the sensor signal of which is to be corrected, the dynamic component from the sensor model being added to the measured signal of the sensor (26, 28), the sensor (26, 28) being a temperature sensor (26) is, wherein the sensor temperature of the temperature sensor (26) is determined and a dynamic temperature change of the temperature signal is dynamically corrected by estimating a temperature in the exhaust system (12), characterized in that the dynamic component of the difference between an estimated gas temperature and the signal of the sensor model corresponds, the dynamic correction value being weighted depending on the gradient of the temperature change measured at the temperature sensor (26), and a complementary filter for dynamic correction of the exhaust gas temperature sensor (26) being designed based on a model, the geometric dimensions of the temperature sensor (26) and the exhaust gas duct section (36), in which the temperature sensor (26) is arranged, flow into the sensor model. Procedure according to Claim 1 , characterized in that the dynamic correction value is weighted more heavily if a strong temperature change is detected between two immediately successive measurements on the temperature sensor (26). Device for correcting a sensor signal, the device comprising a sensor (26, 28) with which a temperature in an exhaust system (12) of an internal combustion engine (10) is measured, the sensor signal being dynamically corrected by a measured sensor signal being passed through a dynamic correction value is corrected, the dynamic correction value being obtained from a sensor model of the sensor (26, 28), the sensor signal of which is to be corrected, the dynamic component from the sensor model being added to the measured signal of the sensor (26, 28), whereby the sensor (26, 28) is a temperature sensor (26), wherein the sensor temperature of the temperature sensor (26) is determined and a dynamic temperature change of the temperature signal is dynamically corrected by estimating a temperature in the exhaust system (12), characterized in that the dynamic Proportion of the difference between an estimated gas temperature and the signal of the sensor model corresponds, the dynamic correction value being weighted depending on the gradient of the temperature change measured at the temperature sensor (26), and a complementary filter for dynamic correction of the exhaust gas temperature sensor (26) being designed based on the model, wherein the geometric dimensions of the temperature sensor (26) and the exhaust duct section (36) in which the temperature sensor (26) is arranged are incorporated into the sensor model. Control device (30) for an internal combustion engine (10), the control device (30) following a method one of the Claims 1 or 2 executes when a machine-readable program code is executed by the control unit (30).

Images