Method for determining the loading of a soot filter

Abstract

The invention relates to a method for determining a loading of a soot filter with soot particles from an exhaust gas mass flow of an internal combustion engine in a motor vehicle, a control device for an internal combustion engine having a soot filter, and a computer program product for carrying out the method. In the first step 100 of the method a characteristic curve for the relationship between the exhaust gas mass flow, exhaust gas temperature, ambient pressure, and pressure drop across the soot filter without loading is determined; in the second step 200 a second exhaust gas mass flow and a second pressure drop that occurs during loading of the soot filter are determined; in the third step 300, from the characteristic curve the first pressure drop is determined for which the first and second exhaust gas mass flows have the same value; in the fourth step 400 an estimated value for the loading of the soot filter is computed via a real-time parameter estimation, preferably by use of the gradient method, based on the previously determined parameters. The method allows a reliable determination of the instantaneous loading of a particulate filter, regardless of the type of measuring signals used in each case for characterizing the loading behavior of the soot filter.

Claims

1. A method for determining a loading of a soot filter with soot particles from an exhaust gas mass flow of an internal combustion engine, in particular a gasoline engine, in a motor vehicle, comprising the steps: determining a characteristic curve for the relationship between the exhaust gas mass flow, exhaust gas temperature, ambient pressure, and pressure drop across the soot filter without loading, wherein a plurality of different first exhaust gas mass flows of an internal combustion engine are led through a soot filter without loading, and for each first exhaust gas mass flow a first pressure drop (y.sub.empty) that occurs is measured by means of at least one sensor while simultaneously detecting a first exhaust gas temperature and a first ambient pressure, and wherein the first pressure drops are provided, normalized by computer with regard to the first exhaust gas temperature and the first ambient pressure, in a form of a characteristic curve of the soot filter without loading; determining a second exhaust gas mass flow of the internal combustion engine that is led through the soot filter, with loading, during operation of the internal combustion engine, and measuring a second pressure drop (y.sub.meas) that occurs, by means of the at least one sensor while simultaneously detecting a second exhaust gas temperature and a second ambient pressure; selecting a first pressure drop (y.sub.empty) from the characteristic curve of the soot filter, without loading, for which the first and second exhaust gas mass flows have the same value, the first and second exhaust gas temperatures have the same value, and the first and second ambient pressures have the same value; computing an estimated value (y.sub.estimated) for the loaded soot filter via a real-time parameter estimation according to a reference model
y.sub.estimated(θ)=y.sub.empty.Math.θ.sub.1+θ.sub.2=y.sub.empty.Math.k+d where θ: estimated parameter, θ.sub.1=k: estimated amplification due to loading of the soot filter, θ.sub.2=d: estimated deviation of the at least one sensor in determining the pressure drop across the soot filter, wherein in a first step an error (e) between the second pressure drop (y.sub.meas), measured for the soot filter with loading, and the estimated value (y.sub.estimated) to be determined is computed:
e=y.sub.estimated(θ)−y.sub.meas=(k.Math.y.sub.empty+d)−y.sub.meas where θ.sub.2==1 for the soot filter without loading, and wherein in a second step the first estimated parameter (θ.sub.1=k) and the second estimated parameter (θ.sub.2=d) are determined by real-time parameter estimation.

2. The method according to claim 1, wherein the first estimated parameter (θ.sub.1=k) and the second estimated parameter (θ.sub.2=d) are determined by real-time parameter estimation according to the gradient method, wherein the estimated error (e) is used to determine the first and second estimated parameters (θ.sub.1=k, θ.sub.2=d) by integration according to
{dot over (k)}=γ.sub.k.Math.e.Math.y.sub.meas and {dot over (d)}=γ.sub.d.Math.e where {dot over (k)}: derivative of the first estimated parameter with respect to time, {dot over (d)}: derivative of the second estimated parameter with respect to time, γ.sub.k: estimation speed of the first estimated parameter, and γ.sub.d: estimation speed of the second estimated parameter.

3. The method according to claim 2, wherein the estimation speeds of the first and second estimated parameters (γ.sub.k, γ.sub.d) in the real-time parameter estimation are varied corresponding to the measuring accuracy of the first and second exhaust gas mass flows and the first and second pressure drops (y.sub.empty, y.sub.meas), where
γ.sub.k=γ.sub.d=0 for low measuring accuracy, and
γ.sub.k,γ.sub.d>>0 for high measuring accuracy, are used in determining the loading of the soot filter.

4. The method according to claim 1, wherein the first estimated parameter (θ.sub.1−k) and the second estimated parameter (θ.sub.1−d) are determined by real-time parameter estimation according to the least squares method, wherein the sum of the estimated error (e) squared is minimized.

5. The method according to claim 1, wherein the first and second pressure drops (y.sub.empty, y.sub.meas) are measured with a differential pressure sensor.

6. The method according to claim 5, wherein the measuring signals for the first and second pressure drops (y.sub.empty, y.sub.meas) obtained with the differential pressure sensor undergo digital filtering for noise suppression prior to further processing, in particular using a Butterworth filter.

7. The method according to claim 1, wherein the first and second exhaust gas mass flows are determined by ultrasonic measurement.

8. The method according to claim 7, wherein the measuring signals for the first and second exhaust gas mass flows, obtained via the ultrasonic measurement, undergo digital filtering for noise suppression prior to further processing.

9. The method according to claim 1, wherein the determination of the loading of the soot filter with soot particles during operation of the internal combustion engine takes place continuously, at predefined time intervals, or as a function of situation recognition, in particular of a fuel consumption and/or a distance traveled, with provision of the measured second pressure drop (y.sub.meas) and determined first pressure drop (y.sub.empty) for further use.

10. A control device for an internal combustion engine, in particular a gasoline engine, having a soot filter, characterized in that the control device is designed for carrying out the method according to claim 1.

11. A computer program product, comprising a program with instructions which, when executed by a computer, prompt the computer to carry out the method according to claim 1.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) The invention is explained in greater detail below in one exemplary embodiment, with reference to the associated drawings, which show the following:

(2) FIG. 1 shows a schematic flow chart of a method according to the invention; and

(3) FIG. 2 shows a graphical illustration of an estimated loading of the soot filter with soot particles over time, with the associated measured and estimated amplification of the pressure drop across the soot filter.

DETAILED DESCRIPTION OF THE INVENTION

(4) FIG. 1 shows a schematic flow chart of the method according to the invention. The illustrated method steps for determining loading of a soot filter with soot particles from an exhaust gas mass flow of an internal combustion engine are explained in greater detail below, with reference to the estimated result for the loading according to the method, shown in FIG. 2. In the example according to FIG. 2, a determination of the loading of a soot filter in a motor vehicle takes place with a gasoline engine during driving operation after a cold start of the engine, over a period of 120 seconds.

(5) To this end, in a first method step 100 a brand-new soot filter having the same design is connected to a gasoline engine on a test stand, so that the exhaust gas mass flow of the engine can flow through the soot filter corresponding to its use in the motor vehicle. In addition, a differential pressure sensor is installed between the inlet and the outlet of the soot filter in order to measure a pressure difference y.sub.empty between the filter inlet and the filter outlet. The soot filter as well as the differential pressure sensor are of the same type as used in the motor vehicle for the driving operation.

(6) The gasoline engine is then operated in each case for several minutes with progressively increasing load, and thus, increasing exhaust gas mass flow, until the entire flow range within which the exhaust gas mass flow through the soot filter can fluctuate during actual driving operation is covered.

(7) For each operating state of the engine that is thus run through, the pressure drop across the empty soot filter, the exhaust gas temperature, and the ambient pressure are measured within the holding period. With these measured values, the associated exhaust gas mass flows are then computed in the manner stated above and provided in the form of a characteristic curve for the empty soot filter for use in the subsequent method steps.

(8) For the soot filter and the differential pressure sensor, each with the same respective designs, in the motor vehicle having a gasoline engine, during driving operation corresponding to FIG. 2 the differential pressure y.sub.meas between the filter inlet and the filter outlet is then continuously measured and detected in each case in a second step 200, once again with the exhaust gas temperature and the ambient pressure in each case being detected at the same time. All measurements are carried out using an anti-aliasing filter (possibly analogous to an RC element, for example), and possibly a digital filter for noise suppression.

(9) In a third step 300, based on the characteristic curve that is determined for the empty soot filter in the first method step, the first pressure drop y.sub.empty is selected that was measured for the same values for the exhaust gas temperature and the ambient pressure; in the event of a deviation from the exhaust gas temperature and/or the ambient pressure during the measurement of the first pressure drop y.sub.empty and of the second pressure drop y.sub.meas, the first pressure drop y.sub.empty is corrected by computer with the values for the exhaust gas temperature and/or the ambient pressure that were measured during driving operation. This ensures that a deviation of the exhaust gas mass flows through the soot filter, with and without loading, is based essentially on a loading with soot particles.

(10) Lastly, in a fourth step 400 the two filtered values for the first and second pressure drop y.sub.empty, y.sub.meas determined in the third method step are used to compute the instantaneous loading of the soot filter at constant estimation speeds γ.sub.k, γ.sub.d. Based on the assumption that a loading of the soot filter with soot particles results in a measured value that is higher by an amplification factor k than with a corresponding empty filter, and the amplification resulting from the sensor offset d, the parameters k and d to be estimated are computed according to the gradient method as stated above, using the real-time parameter estimation according to the invention.

(11) The measured values thus obtained for the amplification 31 and the sensor offset 32 as well as the amplification 33 obtained by real-time parameter estimation are illustrated in the top portion of FIG. 2, in each case over the entire measuring period. The bottom portion of FIG. 2 shows the estimated loading of the soot filter with soot particles during driving operation over the measuring period, according to the method according to the invention, using the estimated amplification 33.

LIST OF REFERENCE NUMERALS

(12) 1 loading measurement protocol 2 loading (Pfil densSootLd-Pfil_VW) 21 estimated loading 3 amplification 31 measured amplification (IAV amplification) 32 sensor offset (IAV offset) 33 estimated amplification (Pfil_rSootLdAdp_VW) 4 time axis 100 first method step 200 second method step 300 third method step 400 fourth method step