Method for operating an internal combustion engine, computing device and computer program product

Abstract

The invention relates to a method for operating an internal combustion engine with an exhaust-gas aftertreatment system, which has an exhaust-gas catalytic converter and at least two exhaust-gas sensors, wherein at least one first exhaust-gas sensor is arranged upstream of the exhaust-gas catalytic converter and at least one second exhaust-gas sensor is arranged downstream of the exhaust-gas catalytic converter, which method comprises: determining the filling level of at least one exhaust-gas component which can be stored in the exhaust-gas catalytic converter in the exhaust-gas catalytic converter using a theoretical catalytic converter model, at least one signal of the first exhaust-gas sensor as a first signal being entered into the theoretical catalytic converter model as an input variable; detecting a signal of the second exhaust-gas sensor downstream of the exhaust-gas catalytic converter as a second signal; calculating a deviation of the second signal from a target signal; reinitializing the catalytic converter model when the deviation of the second signal from the target signal exceeds a predetermined threshold value.

Description

Technical Field

[0001] The present invention relates to a method for operating an internal combustion engine, and a computing device and computer program product for performing the method. Background Technology

[0002] In the internal combustion engines of motor vehicles, such as diesel engines, gasoline engines, or rotary piston engines, when the air-fuel mixture is not completely burned, a large number of combustion products are emitted in addition to nitrogen (N2), carbon dioxide (CO2), and water (H2O). These combustion products include at least hydrocarbons (HC), carbon monoxide (CO), and nitrogen oxides (NOx). X This is subject to legal restrictions. With current technology, effective exhaust emission limits for motor vehicles can be met solely through catalytic converters. By using, for example, a three-way catalytic converter, these harmful substances can be converted into relatively harmless exhaust components such as carbon dioxide, nitrogen, and water.

[0003] In three-way catalytic converters, targeting HC, CO, and NO is achieved only within a narrow λ range around the stoichiometric operating point (λ=1), the so-called "catalytic window." X Simultaneously, high conversion rates are achieved. For the catalytic converter to operate within the catalytic converter window, λ adjustment is typically used, based on signals from oxygen sensors located before and after the catalytic converter. To adjust the λ value before the catalytic converter, an oxygen sensor measures the oxygen content of the exhaust gas before the catalytic converter. This measurement is used to adjust the amount of fuel delivered to the internal combustion engine. For more accurate adjustment, an additional oxygen sensor is used to analyze the exhaust gas after the catalytic converter. This signal is used for steering adjustment, which overlaps with the λ adjustment before the catalytic converter. A step-type oxygen sensor is typically used as the oxygen sensor after the catalytic converter, as it has an extremely steep characteristic line at λ=1 and therefore can indicate λ=1 with very high accuracy.

[0004] In addition to the guide adjustment, which is usually designed to be slow and only adjusts to a very small deviation from λ=1, current motor control systems typically have a function that is responsible for quickly reaching the catalyst window again in the form of λ pre-control according to a large deviation from λ=1.

[0005] A drawback of many current regulation schemes is that they only detect departures from the catalyst window with lag, relying on the voltage of a step oxygen sensor located behind the catalyst.

[0006] An alternative approach for regulating a three-way catalytic converter based on signals from an oxygen sensor downstream of the converter is to adjust the average oxygen fill level of the converter. Since this average fill level is unmeasurable, it is modeled using a road segment model. This type of regulation can identify potential breakdowns in a timely manner and react to them before they actually occur. Model-based regulation of the three-way catalytic converter fill level based on the kinetics of the most important reactions occurring in the converter and the oxygen storage capacity is described in DE 10 2016 222 418 A1. Several sets of stored model parameters can also be incorporated into this model-based catalytic converter regulation. Adapting the catalytic converter's storage capacity to the current operating point is also possible. This method is known, for example, by DE 10 2018 216 980 A1 and DE 10 2018 251 720 A1. Summary of the Invention

[0007] According to the present invention, a method for operating an internal combustion engine, a computing device for performing the method, and a computer program product are proposed. Advantageous designs are the subject of this disclosure.

[0008] A method for controlling an internal combustion engine with an exhaust aftertreatment system according to the present invention is proposed. The exhaust aftertreatment system has an exhaust catalyst and at least two exhaust sensors, wherein at least one first exhaust sensor is arranged upstream of the exhaust catalyst and at least one second exhaust sensor is arranged downstream of the exhaust catalyst. The method includes: determining the fill degree of at least one exhaust component that can be stored in the exhaust catalyst using a theoretical catalyst model, wherein at least one signal from the first exhaust sensor as a first signal is input into the theoretical catalyst model as an input parameter; detecting the signal from the second exhaust sensor downstream of the exhaust catalyst as a second signal; and determining the second... The deviation of the second signal from the rated signal, where the rated signal corresponds to the expected signal at a specific fill level; when the deviation of the second signal from the rated signal exceeds a predetermined threshold, the catalyst model is reinitialized, so that the specific fill level, after reinitialization, should result in the rated signal corresponding to the detected second signal; the fill level is adjusted based on the rated fill level by adjusting the air-fuel mixture supplied to the internal combustion engine; the deviation between the first signal and the second signal is calculated after the catalyst model is reinitialized and after the fill level is adjusted to the rated fill level; and the first signal is corrected by means of a correction value calculated based on the deviation between the first signal and the second signal, thus reducing the deviation.

[0009] The basic advantage of the adjustment scheme according to the present invention is that, with the help of modeled fill degree, it is possible to identify the exit from the catalyst window in a timely manner, and thus generally contribute to less harmful emissions or more effective exhaust purification.

[0010] Within the scope of this invention, re-initialization refers to a method step in which the measured value (in particular the second signal) is used to modify the parameters of the calculation rules (of the theoretical catalyst model) so that the catalyst fill degree is obtained when processing the input parameters in the calculation rules, i.e., the measured signal is matched with the modeled fill degree.

[0011] Here, the first exhaust sensor is advantageously a broadband oxygen sensor and / or the second exhaust sensor is a step oxygen sensor and / or the first signal and / or the second signal includes the λ value of the exhaust from the internal combustion engine. Broadband oxygen sensors are particularly suitable for use upstream of the exhaust catalyst because a wide value range can be traversed here during internal combustion engine operation, while downstream of the catalyst, the high accuracy of the step oxygen sensor within a narrow value range helps to reliably monitor exhaust purification. This reduces uncertainty in the catalyst model by adapting a signal with greater uncertainty (based on the particularly reliable signal adaptation of the step oxygen sensor (with a particularly steep characteristic line in the λ=1 range) to the broadband oxygen sensor (with a significantly flatter characteristic line and consequently greater measurement accuracy). However, other exhaust sensors can also be used in conjunction with this invention, particularly those that measure or determine the concentration of exhaust components, such as nitrogen oxide sensors.

[0012] At least one component of the exhaust gas includes oxygen. Oxygen is particularly crucial to the performance of the exhaust catalytic converter, especially its conversion efficiency.

[0013] The correction value is preferably calculated as the product of the deviation between the first and second signals and the attenuation factor, which is specifically selected from the range of values ​​between 0 and 1. This avoids overcompensation and thus avoids a “sudden increase” in the correction.

[0014] Here, the smaller the attenuation factor is preferably chosen, the larger the gap between the second signal and the rated signal. Therefore, the fact that the oxygen sensor behind the catalyst provides particularly reliable values ​​when the λ value is around 1 (i.e., near the rated signal in normal operating mode) is taken into account, while the measured value becomes increasingly unreliable as the λ value moves further away from 1 (i.e., further away from the normal rated signal).

[0015] The first signal is preferably corrected by adding the correction value to the first signal. This correction is particularly computationally efficient when both signals already include a λ value or when there is a linear correlation between the signal and the associated λ value.

[0016] The invention will now be explained using a three-way catalytic converter as an example. The implementation methods can also be applied to other catalytic converter types, and the invention is not limited to the use of three-way catalytic converters. In principle, any conceivable type of exhaust catalytic converter capable of storing at least one exhaust component can be used.

[0017] The core of this invention is to first bring the modeled fill level of the catalyst to a defined state based on the signal from the oxygen sensor located behind the catalyst, where the actual catalyst fill level and the modeled fill level are at least approximately identical. This reinitialization is followed by adjusting the fill level to a point where minimum emissions and λ=1 are expected behind the catalyst. After adjusting this fill level, the deviation of λ from 1, measured by the actual oxygen sensor located behind the catalyst, is detected. This deviation corresponds to the existing offset between λ in front of and behind the catalyst. By using low-pass filtering and an attenuation factor to account for fluctuations in λ behind the catalyst during dynamic driving and the tolerances of the oxygen sensor behind the catalyst, the attenuation factor ensures that the calculated offset is received or adapted only proportionally. This improves the robustness of the method and avoids overcompensation for the offset. The method can be executed repeatedly, directly sequentially, until the deviation of the actual λ value measured behind the catalyst from 1 is sufficiently small. Therefore, the existing offset can be gradually and completely adapted in a short period of time.

[0018] The advantage of model-based tuning of the catalytic converter is that it can identify the impending departure from the catalytic converter window earlier than during guided tuning, based on signals from exhaust sensors located behind the catalytic converter. This allows for timely and targeted adjustments to the air-fuel mixture to counteract the departure of the catalytic converter window (before it actually occurs). The robustness of model-based tuning can be further improved by expanding compensation for measurement and model uncertainties around the rapid λ offset-fit according to the invention. In particular, it can simultaneously and more robustly adapt to numerically large λ offsets. This further reduces emissions in real-world driving operations. More stringent legal requirements can be met with a smaller catalytic converter cost.

[0019] The invention is illustrated here using an exhaust device as an example, which includes a broadband oxygen sensor, a three-way catalytic converter, and a step-type oxygen sensor sequentially along the flow direction. However, additional or other catalytic converters, sensors, and auxiliary components, such as particulate filters, may also be provided, which at least will not negatively affect the application of the method.

[0020] This invention is based on a self-adaptive catalyst model. For example, a catalyst model can be configured that implements multi-level adaptation to compensate for uncertainties in measurement parameters or model parameters entering a road segment model (Streckenmodell) based on the model, as well as uncertainties in the road segment model itself.

[0021] This multi-level adaptation combines, for example, extremely accurate adaptation that works continuously for small deviations with rapid correction for discontinuous deviations.

[0022] Continuous adaptation and discontinuous correction can be based on signal values ​​from different sensor ranges, particularly those from the oxygen sensor located downstream of the catalytic converter and positioned at the output end. However, these signal values ​​derive two fundamentally different pieces of information. This model allows for consideration of the varying persuasiveness of signal values ​​from different ranges when relating to exhaust composition and the fill level of at least one exhaust component in the catalytic converter.

[0023] In addition, multiple signal value ranges can be set, in which only continuous adaptation is active, only discontinuous correction is active, or both are active.

[0024] In discontinuous corrections, for example, when the voltage of the oxygen sensor at the output end indicates a breakthrough in rich or lean exhaust gas downstream of the catalyst and thus indicates an excessively low or high actual (oxygen) fill level, the modeled fill level is corrected according to the actual fill level. This correction is performed discontinuously to allow for evaluation of the response of the oxygen sensor voltage downstream of the catalyst. Because this response is based on stretchentotzeit and the catalytic converter's storage performance delay, it is particularly possible to configure an adapted catalyst model that performs a correction only when the λ value of the signal from the oxygen sensor located downstream of the catalyst allows for inference of the actual (oxygen) fill level in the catalyst.

[0025] This invention is built upon this discontinuous correction and supplements it with λ-offset adaptation, because it has been identified that, exactly after this correction to the modeled fill factor, there exists a particularly well-defined state that allows the offset of the oxygen sensor upstream of the catalyst to be determined as accurately as possible using the signal from the oxygen sensor downstream of the catalyst.

[0026] In the aforementioned continuous adaptation, for example, the λ signal from the step oxygen sensor downstream of the catalyst is compared with a modeled λ signal downstream of the catalyst. This comparison allows the derivation of the λ offset between the λ values ​​upstream and downstream of the catalyst. This λ offset is used, for example, to correct the λ rating formed by pre-control. However, this continuous adaptation works significantly slower than the discontinuous correction described above and is therefore unsuitable for eliminating large offsets, as it would severely delay the arrival of the catalyst window. The present invention effectively reduces the gap in a particularly advantageous manner when the offset of the oxygen sensor upstream of the catalyst is large.

[0027] The computing device according to the invention, such as a controller for a motor vehicle, is particularly configured in terms of programming techniques to execute the method according to the invention.

[0028] It is also advantageous to implement the method according to the invention in the form of a computer program or a computer program product with program code for performing all the method steps, because this results in particularly low costs, especially when the operating controller is also used for other tasks and therefore already exists. Suitable data carriers for providing the computer program are, in particular, magnetic, optical, and electrical memories, such as hard disks, flash memory, EEPROM, DVDs, etc. The program can also be downloaded via computer networks (Internet, Ethernet, etc.).

[0029] Further advantages and design schemes of the present invention can be derived from the specification and accompanying drawings. Attached Figure Description

[0030] The present invention is schematically illustrated in the accompanying drawings with reference to embodiments, and will now be described with reference to the accompanying drawings:

[0031] Figure 1 A schematic diagram of a vehicle is shown, in which the method according to the invention can be used;

[0032] Figure 2 An advantageous design of the method according to the invention is shown in the form of a flowchart. Detailed Implementation

[0033] exist Figure 1 The vehicle 100 is schematically shown as a block diagram, in which the method according to the invention can be used. The vehicle 100 is preferably configured to perform the... Figure 2 The method 200 includes an internal combustion engine 120 such as a gasoline motor, a catalytic converter 130, and a computing device 140. Furthermore, the vehicle 100 may also include a fuel treatment device 110, for example, in the form of an injection pump, a turbocharger, or a combination thereof.

[0034] In addition, this vehicle also has (exhaust) sensors 145, 147, and in particular oxygen sensors, which are arranged upstream and downstream of the catalytic converter 130 in the exhaust system of the vehicle 100.

[0035] The computing device also controls the operation of the internal combustion engine 120, for example, by controlling the ignition timing, valve opening time, and the composition, quantity, and / or pressure of the air-fuel mixture provided by the fuel treatment device 110.

[0036] The computing device 140 is advantageously configured for, and also performs, Figure 2 The method 200 described herein, according to a preferred embodiment of the invention.

[0037] The exhaust gas produced during the operation of the internal combustion engine 120 is delivered to the catalytic converter 130. In the first step 210 of the method 200, the number of air particles λ in the exhaust gas is measured upstream of the catalytic converter 130 by means of a first oxygen sensor 145 and the first λ value is transmitted to the computing device 140.

[0038] In step 220, the fill degree of at least one exhaust component in the catalyst 130 is calculated based on the λ value upstream of the catalyst 130 obtained in step 210. This involves, for example, the oxygen fill degree, but the corresponding fill degree in the catalyst 130 can also be determined for other exhaust components, such as nitrogen oxides.

[0039] The catalyst accelerates or enables the reaction between exhaust components, thereby converting harmful components such as carbon monoxide, nitrogen oxides, and incompletely burned hydrocarbons into relatively harmless products such as water vapor, nitrogen, and carbon dioxide. In step 230, a second λ value is obtained downstream of the catalyst 130 by a second oxygen sensor 147 and this second λ value is transmitted to the computing device 140.

[0040] The first and second λ values ​​can differ from each other temporarily or permanently because the composition of the exhaust gas upstream and downstream of the catalyst 130 differs from each other due to the reactions within the catalyst 130. Furthermore, the exhaust gas requires a certain amount of time to flow through the catalyst 130 (the so-called settling time). This settling time depends in particular on the current volumetric flow rate of the exhaust gas, i.e., on the current operating state of the internal combustion engine 120. When the internal combustion engine 120 is operating at full load, it produces a higher exhaust gas volume per unit time, for example, than when idling. The corresponding settling time therefore varies according to the operating state of the internal combustion engine 120, since the volume of the catalyst 130 remains constant.

[0041] In these interpretations, the starting point is, for example, simplified, the minimum emission value when λ is 1. However, the invention can still be used when the rated fill level used to achieve the minimum emission corresponds to a different λ rating than λ=1.

[0042] As described, the present invention is based on the signal from the exhaust gas sensor 147 downstream of the catalytic converter 130. The invention is particularly useful here for associating a voltage signal with the current (oxygen) fill level in the catalytic converter 130 when the oxygen sensor 147 downstream of the catalytic converter 130 clearly indicates a high or low voltage. This is especially true when the sensor voltage does not correspond to λ within the range of 1. In this case, the catalytic converter 130 is oxygen-free or oxygen-filled, allowing rich or lean exhaust gases to pass through. According to the invention, when the oxygen sensor 147 downstream of the catalytic converter 130 shows a clearly high or low voltage, this is used to reinitialize one or more modeled fill levels, such as modeled oxygen fill levels in multiple axial regions of the catalytic converter 130.

[0043] Correspondingly, in step 240, the deviation between the sensor signal of the oxygen sensor 147 downstream of the catalyst 130, obtained in step 230, and the rated signal corresponding to the expected signal based on the catalyst fill level obtained in step 220, is calculated. The rated signal, under the static operating conditions of the internal combustion engine 120, i.e., under operating conditions with constant load requirements, may specifically correspond to a λ value of 1 or a λ value near 1.

[0044] In step 250, the deviation between the actual λ value downstream of catalyst 130, obtained in step 240, and the expected value is compared with a threshold. If the deviation is less than the threshold, then method 200 returns to step 210. Conversely, if the deviation reaches or exceeds the threshold, then method 200 continues with step 260, in which the catalyst model is reinitialized based on the sensor signals obtained downstream of catalyst 130. This reinitialization 260 brings the modeled fill factor of catalyst 130 to a defined state in which the modeled fill factor at least approximately matches the corresponding fill factor of the actual catalyst 130.

[0045] This discontinuous correction or reinitialization 260 of the modeled fill level results in a deviation between the modeled average fill level and the predetermined nominal value. This deviation is then adjusted in step 270. This deviation causes the air-fuel mixture to adjust in the direction of the fill level adjustment nominal value and rapidly moves the catalyst 130 toward the catalyst window. This deviation therefore directly leads to improved emissions and simultaneously brings the catalyst 130 into a defined state where λ=1 (or a λ nominal value near 1) should appear as expected after the catalyst 130. However, λ=1 only truly appears after the catalyst 130 if the signal from the oxygen sensor 145 before the catalyst 130 (on which the modeled oxygen fill level is based) has no offset. Otherwise, instead of λ=1, a λ value deviates from λ=1 by this offset.

[0046] Once the oxygen fill level is adjusted after reinitialization (in step 260), in step 280, the deviation of the actual λ value downstream of the catalyst 130 from λ=1 (or the nominal λ value) is detected by means of an oxygen sensor 147 disposed downstream of the catalyst 130. A correction value 285 is calculated based on this deviation, and this correction value is taken into account when subsequently calculating the λ value upstream of the catalyst 130. This correction value 285 is specifically a number added to the λ value upstream of the catalyst 130 measured in step 210. It goes without saying that the detection of the deviation is meaningful only if the signal from the oxygen sensor 147 downstream of the catalyst 130 is reliable and this sensor 147 is, in particular, ready to operate. Optionally, a waiting time or minimum displacement can also be set, which must be estimated or implemented before detecting the deviation. This allows for the consideration, in particular, of the already mentioned settling time.

[0047] After step 280, the method 200 returns to step 210, wherein the correction value 285 obtained when determining the λ value upstream of the catalyst 130 is taken into account in the renewed step 210.

[0048] Because this can be taken as a starting point during dynamic driving operations—that is, the λ value behind the catalytic converter 130 is not constant but may fluctuate around the average value—the signal from the oxygen sensor 147 behind the catalytic converter 130 is preferably low-pass filtered. This means that the deviation between the low-pass filtered λ signal from the sensor 147 behind the catalytic converter 130 and λ=1 is detected and interpreted as the offset between the λ value in front of the catalytic converter 130 and the λ value behind the catalytic converter 130.

[0049] Furthermore, because the accuracy of the signal from the step oxygen sensor 147 downstream of the catalyst 130, away from λ=1, may be limited by temperature effects, lateral sensitivity, and the flattening characteristics of the voltage-λ characteristic line, it is stipulated that the calculated offset is received only as a correction value 285 by means of an attenuation factor, for example, only 50%. Attenuation is preferably enhanced further away from λ=1, and the closer the measured oxygen sensor signal is to λ=1, the greater the attenuation reduction, because λ accuracy is highest there. The robustness of the offset-correction is improved due to the proportional reception, as this reliably avoids overcompensation of the offset that could lead to a sudden increase in offset adaptation and emissions.

[0050] The correction value 285 obtained in this way is preferably used to adapt and correct the signal of the oxygen sensor 145 in front of the catalyst 130. Assuming an attenuation factor of 50%, the offset between the λ value in front of the catalyst 130 and the λ value behind the catalyst 130 after the first adaptation step is only half of the original.

[0051] When the voltage of the oxygen sensor 147 downstream of the catalyst 130 re-indicates a clearly high or low voltage, method 200 is repeated, and repeated multiple times if necessary. Method 200 may also be repeated if the signal from the oxygen sensor 147 downstream of the catalyst 130 moves in one direction (which does not correspond to the direction predicted based on reinitialization), for example, when the sensor voltage is reinitialized at a very low sensor voltage and then moves initially in the direction of a higher sensor voltage, but then moves again in the direction of a low oxygen sensor voltage. Because the more drastically the attenuation factor decreases, the closer the signal measured by the oxygen sensor 147 downstream of the catalyst 130 is to λ=1, each step adapts a larger proportion of the offset. In this way, larger λ offsets can also be quickly and robustly fully adapted through multiple successive reinitialization and adaptation steps.

Claims

1. A method (200) for operating an internal combustion engine (120) with an exhaust aftertreatment system, the exhaust aftertreatment system having an exhaust catalyst (130) and at least two exhaust sensors, wherein, At least one first exhaust sensor (145) is disposed upstream of the exhaust catalyst (130), and at least one second exhaust sensor (147) is disposed downstream of the exhaust catalyst (130), the method comprising: Using a theoretical catalytic converter model, the filling degree of at least one exhaust component that can be stored in the exhaust catalyst (130) is determined, and at least one signal from the first exhaust sensor (145) as a first signal is input into the theoretical catalytic converter model as an input parameter. The signal of the second exhaust sensor (147) downstream of the exhaust catalyst (130) is detected as a second signal. The deviation between the second signal and the rated signal is calculated, where the rated signal corresponds to the expected signal at a specific fill level in the exhaust catalyst (130). When the deviation between the second signal and the rated signal exceeds a predetermined threshold, the catalyst model is reinitialized. Therefore, the specific fill level, after reinitialization, should result in a rated signal corresponding to the detected second signal. By adjusting the air-fuel mixture supplied to the internal combustion engine (120) according to the rated fill degree, the fill degree in the exhaust catalyst (130) is adjusted based on a specific fill degree. After the catalyst model is reinitialized and the fill level is adjusted according to the rated fill level, the deviation between the first and second signals is calculated, and The first signal is corrected by using a correction value (285) obtained based on the deviation between the first signal and the second signal, thereby reducing the deviation between the first signal and the second signal.

2. The method (200) according to claim 1, wherein, The first exhaust sensor (145) is a broadband oxygen sensor and / or the second exhaust sensor (147) is a step oxygen sensor and / or the first signal and / or the second signal includes a λ value.

3. The method (200) according to claim 1 or 2, wherein, The at least one component of the exhaust gas includes oxygen.

4. The method (200) according to claim 1 or 2, wherein, The correction value (285) is calculated as the product of the deviation between the first signal and the second signal and the attenuation factor.

5. The method (200) according to claim 4, wherein, The attenuation factor is selected from the range of values ​​between 0 and 1.

6. The method (200) according to claim 4, wherein, The smaller the attenuation factor is chosen, the greater the distance between the second signal and the rated signal.

7. The method (200) according to claim 1 or 2, wherein, The first signal is corrected by adding the correction value (285) to the first signal.

8. A computing device configured to perform all the method steps of the method (200) according to any one of the preceding claims.

9. A computer program product, when implemented on a computing device, causes the computing device to perform all the method steps of the method (200) according to any one of claims 1 to 7.

10. A machine-readable storage medium having a computer program stored thereon, which, when executed, implements all the method steps of the method according to any one of claims 1 to 7.

Images