US20200063633A1

US20200063633A1 - Method and system for compensating nox sensor measurement error

Info

Publication number: US20200063633A1
Application number: US16/109,287
Authority: US
Inventors: Luca Lauritano; Luis Daniel GUERRERO CRUZ; Giuseppe MAZZARA BOLOGNA; Andrea Dutto; Sarah Funk
Original assignee: GM Global Technology Operations LLC
Current assignee: GM Global Technology Operations LLC
Priority date: 2018-08-22
Filing date: 2018-08-22
Publication date: 2020-02-27
Also published as: DE102019113396A1

Abstract

Systems and associated methods of treating diesel exhaust in a vehicle are disclosed. An example method includes providing an exhaust aftertreatment device in an exhaust tailpipe, and a first nitrogen oxide (NOx) sensor upstream of the exhaust aftertreatment device, with the exhaust aftertreatment device configured to reduce NOx present in an exhaust flow through the exhaust aftertreatment device with a treatment fluid applied to the exhaust flow. This example method may also include measuring a concentration of NOx in the exhaust flow with the first sensor, determining an error in the measurement of the NOx concentration, and applying a learning corrective adjustment in a subsequent measurement of the NOx concentration in the exhaust flow in response to that determination. A first magnitude of the learning corrective adjustment may be based at least in part upon a second magnitude of the error in the measurement.

Description

INTRODUCTION

Internal combustion engines generally produce nitrogen-oxide (NOx) emissions which results from the combustion of hydrocarbon fuel. NOx production is a particular problem for diesel engine applications, and manufacturers have employed various devices and systems in an effort to reduce NOx production and/or reduce NOx concentration present in vehicle exhaust by way of after-treatment systems. Known after-treatment systems typically rely upon at least one NOx sensor to directly measure an amount or concentration of NOx in an exhaust flow.
Known sensors are subject to at least some error in measurement, or otherwise may degrade in accuracy over the life of the sensor. Inaccurate NOx measurement can result in excess NOx being released into the atmosphere from the vehicle, or may cause overuse of an associated after-treatment, resulting in either inefficiency or harmful environmental effects. Accordingly, there is a need for improved controls for an exhaust system and associated methods for reducing NOx emissions in a vehicle.

SUMMARY

In at least some examples, a method of treating diesel exhaust in a vehicle includes providing an exhaust aftertreatment device in an exhaust tailpipe, and a first nitrogen oxide (NO_x) sensor upstream of the exhaust aftertreatment device, with the exhaust aftertreatment device configured to reduce NO_xpresent in an exhaust flow through the exhaust aftertreatment device with a treatment fluid applied to the exhaust flow. This example method may also include measuring a concentration of NO_xin the exhaust flow with the first sensor, determining an error in the measurement of the NO_xconcentration, and applying a learning corrective adjustment in a subsequent measurement of the NO_xconcentration in the exhaust flow in response to that determination. A first magnitude of the learning corrective adjustment may be based at least in part upon a second magnitude of the error in the measurement.
In some examples, a method may also include determining that the vehicle is operating in a normal exhaust mode, and the error in the measurement of the concentration of NO_xis determined in response to the determination that the vehicle is operating in a normal exhaust mode,
In other examples, the learning corrective adjustment is applied in response to a determination that the error exceeds a predetermined magnitude.
In some examples, the exhaust aftertreatment device is either a Selective Catalytic
Reduction device (SCR) or a Selective Catalytic Reduction on Filter (SCRF) device. In such examples, the aftertreatment device may be configured to reduce at least a portion of the NOx present in the exhaust flow to nitrogen (N2) and water (H2O). Additionally, in some of these examples, a method may include passing the exhaust flow through at least one of a Diesel Oxidation Catalyst (DOC) and a Diesel Particulate Filter (DPF).
In some example methods, the treatment fluid includes urea.
The first sensor may in some examples be positioned downstream of an engine exhaust output.
Some example methods also include providing a second NO_xsensor positioned downstream of the exhaust aftertreatment device.
The treatment fluid may, in some examples, be supplied by a diesel exhaust fluid dosing system.
In at least some examples, a method may also include determining the error includes comparing a first integral of the measurement taken with the first sensor with a second integral of a model of an expected NOx concentration. In some of these examples, comparing the first and second integrals may include determining a ratio of the first and second integrals. In a subset of these examples, a method may further include applying a low-pass filter to the ratio of the first and second integrals.
In some example approaches, a method also includes determining a dead-band from a variation in at least one of (a) measurement by the first sensor and (b) a model of expected NOx production, with the learning corrective adjustment being applied in response to a determination that the error in the measurement exceeds the dead-band.
In some examples, a method also includes modeling an expected NOx concentration from one or more engine operating parameters.
In another example of a method of treating diesel exhaust in a vehicle, the method includes providing an exhaust aftertreatment device in an exhaust tailpipe, and a first nitrogen oxide (NO_x) sensor upstream of the exhaust aftertreatment device, with the exhaust aftertreatment device configured to reduce NO_xpresent in an exhaust flow through the exhaust aftertreatment device with a treatment fluid applied to the exhaust flow. The method may also include measuring a concentration of NO_xin the exhaust flow with the first sensor, and determining an error in the measurement of the NO_xconcentration. In this example, determining the error includes comparing a first integral of the measurement taken with the first sensor with a second integral of a model of an expected NOx concentration by determining a ratio of the first and second integrals, and applying a low-pass filter to the ratio of the first and second integrals. The method may further include applying a learning corrective adjustment in a subsequent measurement of the NO_xconcentration in the exhaust flow in response to the determination of an error in the measurement of NO_xconcentration, a first magnitude of the learning corrective adjustment based at least in part upon a second magnitude of the error in the measurement.
In some examples, the learning corrective factor is applied in response to a determination that the error exceeds a predetermined magnitude.
In at least some examples, a vehicle may include an exhaust tailpipe having an exhaust aftertreatment device configured to reduce nitrogen oxide (NO_x) present in an exhaust flow through the exhaust aftertreatment device with a treatment fluid applied to the exhaust flow. The tailpipe may further include a first NO_xsensor upstream of the exhaust aftertreatment device, the first NOx sensor configured to measure a concentration of NO_xin the exhaust flow, The vehicle may also include a controller communicatively linked to the sensor, the controller configured to determine an error in the measurement of the amount of the NO_xconcentration, and apply a corrective factor in a subsequent measurement of the NO_xconcentration in the exhaust flow, with a first magnitude of the corrective factor being based at least in part upon a second magnitude of the error in the measurement.
In some examples, the controller is configured to apply the learning corrective factor in response to a determination that the error exceeds a predetermined magnitude.
In at least some example approaches, the treatment fluid includes urea.

BRIEF DESCRIPTION OF THE DRAWINGS

One or more embodiments of the invention will hereinafter be described in conjunction with the appended drawings, wherein like designations denote like elements, and wherein:

FIG. 1 is a schematic illustration of a vehicle having an exhaust system, according to one example approach;

FIG. 2 is a schematic illustration of a controller for the vehicle of FIG. 1, according to an example; and

FIG. 3 is a process flow diagram for a method of reducing emissions in a vehicle, according to one example.

DETAILED DESCRIPTION

Example illustrations of a method of reducing NOx concentration in an exhaust flow, and associated exhaust system(s) and vehicle(s) are provided herein. Generally, a learning corrective adjustment may be made in response to a detected error in measurement of NOx concentration by a sensor in an exhaust system. An error may be detected by comparing direct measurements of an NOx concentration, e.g., with a vehicle sensor, with a model of expected NOx concentration that is based upon one or more engine or vehicle operating parameters. In this manner, errors in sensor measurement beyond expected variations may be compensated based upon a magnitude of the variation from an expected NOx concentration. Moreover, as will be discussed further below, in some examples adjustments to the measurement may be applied only where the error exceeds a threshold amount. In this manner, over-correction of minimal error may be reduced or prevented entirely.
Turning now to FIG. 1, an example vehicle 100 is illustrated. Vehicle 100 may have an internal combustion engine 102 for providing motive power to the vehicle 100. The engine 102 may be a diesel or compression-ignition (CI) engine, although the concepts disclosed herein are applicable to other combustion engine types where NOx is produced as a combustion byproduct, e.g., gasoline engines. The vehicle 100 may rely solely upon the engine 102 for providing power to the vehicle 100, or may alternatively include other power sources, e.g., an electric motor-generator. Thus, the vehicle 100 may be powered exclusively by the engine 102, or may be a hybrid vehicle employing other power sources in addition to the engine 102.
The vehicle 100 may include an exhaust system 104 receiving an exhaust flow from the engine 102. The exhaust system 104 may include one or more pipes or other means of directing exhaust flow from the engine 102 and into the ambient air around the vehicle or otherwise to the atmosphere. Moreover, the exhaust system 104 may include various components for reducing emissions in the exhaust flow before expelling the treated exhaust flow to the atmosphere. As illustrated in FIG. 1, the exhaust system 104 may expel the exhaust flow to a tailpipe 120, which may include one or more mufflers for reducing noise associated with the exhaust flow. The tailpipe 120 may, in turn, expel the exhaust flow into the ambient atmosphere about the vehicle.
The exhaust system 104 may include one or more aftertreatment devices or other components configured to reduce emissions, such as nitrogen oxides (NOx). Reduction of NOx may be accomplished by any devices or systems that are convenient and are not limited to the specific types or examples discussed herein or illustrated in FIG. 1. As illustrated in FIG. 1, the exhaust system 104 includes a NOx sensor 106, which as illustrated is positioned to detect concentration or amount of NOx in the exhaust flow from the engine 102. The NOx sensor 106 is positioned upstream with respect to the exhaust flow from one or more exhaust aftertreatment devices, so as to thereby measure an NOx concentration entering the aftertreatment device(s).
In the example illustrated, a first exhaust aftertreatment device 110 downstream of the sensor 106 is a diesel exhaust aftertreatment device, and more particularly a an aftertreatment device that reduces or eliminates NOx present in the exhaust flow by way of a catalyst present in the aftertreatment device 110 and a treatment fluid that may be injected into the exhaust flow as needed, as will be discussed further below. In the example illustrated, the aftertreatment device 110 is a Selective Catalytic Reduction on Filter (SCRF) device 110 that uses an SCR process occurring on a filter contained within the SCRF 110. A second aftertreatment device 112 is also provided. In the example of FIG. 1, the aftertreatment device 112 is a Selective Catalytic Reduction (SCR) device 112, may also be provided to further reduce NOx concentration in the exhaust flow using a selective catalytic reduction process occurring on a flow-through substrate within the SCR 112. While the example illustrated in FIG. 1 illustrates a single SCRF 110 and a single SCR 112, there is generally no limitation on the number or type(s) of selective catalytic reduction devices that may be used in the exhaust system 104. Example exhaust systems may be provided with any number of SCRF or SCR devices, and may in some examples have only one type of selective catalytic reduction device (e.g., only SCRF-type device(s), or only SCR-type device(s)), or may have multiple types of selective catalytic reduction devices (e.g., at least one SCRF and at least one SCR). Thus, while two aftertreatment devices 110, 112 are provided in the example illustrated in FIG. 1, any number or type of aftertreatment device(s) may be employed, though the following examples are discussed in conjunction with the system 104 having a single SCRF 110 and a single SCR 112 illustrated in FIG. 1. The examples hereing may thus be applied in the context of any exhaust system having at least one aftertreatment device using a selective catalytic reduction process.
A second NOx sensor 108 may be provided in the exhaust system 104 positioned downstream from the SCRF 110 and SCR 112. The second NOx sensor 108 may thereby detect NOx levels in the exhaust flow after the exhaust flow is passed through the SCRF 110 and SCR 112. The measurement of NOx concentration or levels in the exhaust flow provided by the SCRF 110 may be used to determine effectiveness of the SCRF 110 and/or SCR 112 at reducing concentration of NOx in the exhaust flow. While two NOx sensors 106 and 108 are illustrated in FIG. 1, in some examples additional sensors may be provided, e.g., where measurement of NOx concentration at other areas of the exhaust system 104 is desired.
The SCRF 110 and the SCR 112 may each include a catalyst contained within the devices 110 and 112 which reduces NOx concentration in the exhaust flow, as noted above, when the treatment fluid is applied to the exhaust flow. More specifically, an SCRF, e.g., SCRF 110, may inject treatment fluid into the SCRF 110. Reduction of NOx concentration may also occur by way of application of treatment fluid to the exhaust flow upstream of an exhaust aftertreatment device, e.g., in the case of an SCR device. Varying amounts of the DEF may be injected in direct proportion to a detected concentration of NOx in the exhaust flow. Moreover, as will be described further below, a learning error adjustment based upon expected NOx concentration in the exhaust flow may be used in combination with a NOx concentration measured by a sensor to reduce the effects of any error in the sensor.
In the example exhaust system 104 illustrated in FIG. 1, a diesel exhaust fluid (DEF) dosing system 114 is provided which is configured to detect conditions in the exhaust system 104 such as the concentration of NOx via the sensor 106, and inject an appropriate amount of treatment fluid into the exhaust flow, downstream of the first NOx sensor 106. In one example, the DEF is an aqueous urea solution (AUS). Merely by way of example, an aqueous urea solution (AUS) of AUS 32 may be employed, one example of which is commercially available as ADBLUE™. The DEF may be used as a consumable in a selective catalytic reduction reaction in the SCRF 110 and/or SCR 112, thereby lowering NOx concentration in the exhaust flow.
The DEF dosing system 114 may include a tank 124 or other supply of the DEF, as well as a pump 126 configured to draw the DEF from the tank 124, thereby supplying the DEF to one or more injectors. A differential pressure sensor 128 a may be provided that is configured to measure a differential pressure across the SCRF 110. In the example exhaust system 104 illustrated, a single injector 128 b is provided. The injector 128 b may inject the DEF into the exhaust flow at any location in the exhaust system 104 that is convenient for applying, injecting, mixing, or otherwise dispersing treatment fluid into the exhaust flow. In the case of the example illustrated in FIG. 1, the injector 128 b is positioned downstream of the sensor 106 and upstream of the SCRF 110. The DEF dosing system 114 may also include a DEF controller 122. The DEF controller 122 may generally manage delivery of the DEF to the exhaust system 104, as well as the conditioning of the DEF within the tank 124, e.g., based on a dosing command provided from an engine control module (ECM) 130. The ECM may generally control various aspects of operation of the engine 102, e.g., fuel ignition, ignition timing, etc., and other aspects of the vehicle 100 and/or the exhaust system 104. The DEF controller 122 and ECM 130 may each include a processor and any computer-readable memory, e.g., a non-transitory computer-readable memory, which includes instructions that, when executed by the processor, are configured to control injection of the DEF into the exhaust flow and operational aspects of the engine 102 and exhaust system 104. The ECM 130 may be communicatively linked with any components of the vehicle 100, e.g., the engine 102, exhaust system 104, sensors 106 and/or 108, that may be convenient for determining an appropriate amount of DEF to be injected into the exhaust flow. Moreover, as will be described further below, the controller 130 may employ a control strategy or methodology for accounting for error or other inaccuracies in NOx concentration measured within the exhaust system 104, e.g., by the sensor 106.
The exhaust system 104 may include any other additional components for reducing emissions or particulates emitted from the engine 102 that are convenient. For example, as illustrated in FIG. 1 the exhaust system may include a diesel oxidation catalyst (DOC) 116 that is positioned upstream of the SCRF 110, and downstream of the first NOx sensor 106. The DOC 116 may be configured to oxidize carbon monoxide, gas phase hydrocarbons, and diesel particulate matter to carbon-dioxide (CO₂) and water (H₂O). The exhaust system 104 may include particulate filters, e.g., a Diesel Particulate Filter (DPF), particularly where the aftertreatment device(s) 110 and 112 themselves lack a filter. Generally, a single filtering device within the exhaust system 104 will be sufficient for trapping or filtering out particulates from the exhaust flow, and therefore an additional particulate filter may not be needed where the system 104 includes at least one SCRF, e.g., SCRF 110. Nevertheless, this should not be construed as any limitation on the number of filtering devices that may be employed in vehicle 100.
The exhaust flow from the diesel engine 102 may generally contain sufficient amounts of oxygen necessary for the reactions in the DOC 116 (and/or the SCRF 110/112). Concentration of oxygen in the exhaust gases from engine 102 may vary depending on the engine load and amount of exhaust flow through the exhaust system 104. The catalyst activity of the SCRF 110/112 and the DOC 116 may increase with temperature. In one example, a minimum exhaust temperature is necessary for the catalyst in the DOC 116 and/or the SCRF 110/112 to “light off” At elevated temperatures, conversions depend on the catalyst size and design, and generally increase with temperature. In an example, a minimum exhaust temperature of about 200 degrees Celsius (C) is needed in order for the catalyst of the DOC 116 and/or SCRFs 110 and 112 to be effective.
Exhaust system 104 may also include a rear oxidation catalyst (ROC), which receives the exhaust flow exiting the SCR 112. The ROC 118 may be an ammonia (NH₃) slip catalyst or clean up catalyst. The DEF injected into the exhaust flow in exhaust system 104 may generally convert to NH3 within the exhaust system 104, and in some cases ammonia (NH3) may be released from the SCR components, i.e., SCRF 110 or SCR 112, under certain operating conditions. The ROC 118 may include a filter or other mechanism configured to reduce the levels of ammonia released from the exhaust system 104 and the vehicle 100 to the environment.
Turning now to FIG. 2, an example control methodology 200, e.g., for use by the ECM 130, is illustrated. The control 200 may generally compensate for error in an NOx sensor, e.g., sensor 106, based upon predicted or expected NOx concentration that is determined from operating parameters of the engine 102, the exhaust system 104, or other components of the vehicle 100 that are measured in real-time. In one example, a vehicle or powertrain controller employing the control 200, e.g., ECM 130, may communicate with other components of the vehicle 100 to implement the control methodology 200.
The control 200 may receive the measured NOx concentration at block 202, i.e., from the NOx sensor 106, as one input. The control 200 may also receive, as another input, a NOx model determination at block 204. The NOx model determination may calculate an expected NOx concentration in the exhaust flow based upon real-time measurements of operating parameters of the engine 102, the exhaust system 104, or other components of the vehicle 100 that bear upon production of NOx in the exhaust flow. Accordingly, various inputs such as fuel flow in the engine 102, exhaust temperature, load being placed upon the engine 102, ambient air temperature, or any other factors affecting expected NOx concentration in the exhaust flow may be used to determine an expected NOx concentration. Any model or set of equations may be used to determine expected concentration or levels of NOx in the exhaust flow based upon operating parameters of the engine 102, exhaust flow 104, or other components of the vehicle 100 that are convenient. Merely by way of example, the Zel'dovich mechanism generally describes oxidation of nitrogen and formation of NOx by the following chemical equations:
N+O₂=NO+O
N+OH=NO+H
Inputs to the NOx model with respect to real-time measured operating parameters may include, but are not limited to, fuel injection timing with respect to a main injection and a pilot injection, fuel rail pressure, engine intake manifold temperature, ambient humidity around the vehicle 100, exhaust equivalence ratio, oxygen concentration, environmental pressure, engine speed, and an intake/exhaust pressure ratio. Accordingly, the ECM 130 may receive these inputs, e.g., from sensors of the vehicle 100, as well as any other inputs needed to model NOx formation according to measured conditions of the engine 102 and/or exhaust system 104.
In one example, the control 200 is a proportional-integral control. Accordingly, the two input blocks 202 and 204 may each be integrated at blocks 202′ and 204′, respectively. The integrated inputs may then be compared at block 206. In one example comparison, a ratio of the integrated inputs 202′ to the integrated input 204′ is determined. The ratio output from block 206 may then be input to a low-pass filter at block 208. The low-pass filter may generally facilitate a smoother measurement or comparison of the measured NOx concentration and modeled NOx concentration by attenuating frequencies above a predetermined frequency threshold. The low-pass filter may employ any frequency threshold that is convenient.
The output from the low-pass filter may be received at block 210, which may compare the ratio with a calculated “dead-band” application configured to reduce or delay adjustments to the sensor 106 unless an expected variation in NOx measurement by the sensor 106 (i.e., block 202) and/or the model (i.e., block 204) is exceeded. Example dead-band calculations, which will be discussed further below, may generally improve robustness of the error detection by preventing adjustments to a learning error value when the detected error is within a variation guideline.
Proceeding to block 212, the measurement of NOx concentration at sensor 106 may be adjusted with a learning error factor or adjustment. Generally, the learning factor may be used to adjust measured NOx concentration at the NOx sensor 106, thereby compensating to at least some extent an error in the measured NOx concentration, e.g., by the sensor 106. An example learning factor will be described in further detail below.
Turning now to FIG. 3, an example process 300 for treating an exhaust flow is illustrated. Process 300 may begin at block 305, where process 300 queries whether an associated vehicle is operating under normal conditions, or it is otherwise appropriate to employ an adjustment for sensor error in an exhaust treatment system. Generally, it may be desirable to apply a learning error adjustment in situations where an engine is operating normally and under non-extreme conditions, e.g., extreme cold/hot temperatures, extreme ambient air pressure, etc.
Examples of conditions that may be checked at block 305 to determine whether the relating to the vehicle 100 and/or engine 102 are operating normally will now be described in further detail. In an example, process 300 may determine whether the engine 102 is on and running or whether a key of the vehicle 100 is on. In another example, process 300 may check for errors or faults in any components of the vehicle 100 relevant to NOx detection and/or reduction. In one such approach, the vehicle 100 may check for on-board diagnostic (OBD) codes relevant to NOx detection and/or reduction, e.g., OBD codes associated with the sensor 106.
Other example conditions associated with the vehicle 100 and/or engine 102 may be employed to determine that the engine 102 and/or exhaust system 104 is operating normally. For example, process 300 may check whether the engine 102 speed is normal. In one specific example, process 300 checks whether the engine speed is above idle (i.e., to ensure the engine is running and under a load). At the same time, relatively high engine speeds may not be ideal conditions for modelling the NOx concentration, and thus process 300 may alternatively or additionally check whether the speed of the engine 102 is above a threshold speed, e.g., 6000 rotations per minute.
In another example, process 300 may determine whether the engine coolant temperature is indicative of the engine 102 being warmed up and not operating in extreme conditions. In an example, process 300 may determine whether the engine temperature is between approximately 180 and 230 degrees Fahrenheit. This temperature range is merely an example, and any temperature range may be employed that is convenient to indicate the vehicle 100 has been “warmed up.”
Other example conditions employed in block 305 may be used to determine that the ambient conditions surrounding vehicle 100 are not extreme, e.g., extreme cold or hot temperatures, extreme altitude, extreme high/low atmospheric air pressure, or the like. In one example, temperature and pressure ranges to identify extreme operating conditions may align with on-board diagnostic regulatory requirements, e.g., a threshold of 10,000 feet altitude to identify high altitude (and corresponding low atmospheric pressure), and −20 degrees Fahrenheit as a threshold for cold temperature, merely as examples.
In one example, all of the above conditions are used to determine that the engine 102/vehicle 100 is operating normally in nominal ambient conditions. More specifically, in this example process 300 answers the query in block 305 in the affirmative only if it is determined that (1) a key of the vehicle 100 is on, (2) no faults or errors are detected in any components of the exhaust system 104, (3) engine speed is above idle and also below a high-speed threshold such as that discussed above, (4) engine coolant temperature indicates the engine 102 and/or vehicle 100 is sufficiently warmed up, and (5) that ambient temperature and pressure measurements about the vehicle 100 are in nominal ranges.
Once the query at block 305 is answered affirmatively, i.e., the relevant condition(s) indicating normal operation of the engine 102 are satisfied, process 300 may proceed to block 310.
At block 310, process 300 queries whether a predetermined amount of time has passed, at which time the process 300 may proceed to determine a learning error adjustment or corrective factor. Generally, it may be desirable to allow a predetermined amount of time to pass while the engine 102 is running normally, e.g., to allow a sufficient concentration of NOx in the exhaust flow to be generated by the engine. The time period employed at block 310 may be, merely as one example, 800 seconds. Any other time period may be used that is convenient for determining that a sufficient level of NOx has been generated by the engine 102. When it is determined at block 310 that this time period has not yet expired, process 300 may proceed to block 315.
At blocks 315 and 320, process 300 may determine an integral of the engine-out NOx concentration according to (1) a model calculation and (2) direct measurement by the NOx sensor 106, respectively. As noted above, the model calculation may be based upon one or more operating parameters of the engine 102, exhaust system 104, and/or other components of the vehicle 100 measured generally in real-time. Proceeding to block 325, process 300 may increment a timer associated with the determination in block 310. Process 300 may then proceed back to block 305 and 310 to determine whether normal operating conditions (still) apply and whether the time window threshold at block 310 has been reached. Thus, process 300 may continue in a “loop” through blocks 305 and 310 until the time window threshold has been reached.
Upon determination at block 310 that the time window threshold has been reached, process 300 may proceed to block 330. At block 330, process 300 determines a learning error adjustment. In an example, process 300 may employ the control methodology as described above in FIG. 2, e.g., using the ECM 130.
Proceeding to block 335, process 300 may query whether the learning error adjustment (K_error) is greater than a predetermined magnitude. This determination may be used to prevent application of a learning error adjustment if the magnitude of the error in the NOx sensor measurement is relatively small. In one example, process 300 compares the learning error adjustment with a representation of variation in both the model and sensor measurement. This variation may be referred to as a “dead band,” within which no adjustment is made to the learning error adjustment. In one example, the dead band δ_Kmay be determined from the following equation:
δ_K K√{square root over ((δ_∫NOx _EOModel)²+(δ_∫NOx1 _Sensor)²)}
Where: K=ratio of (a) the integrated NOx model to (2) the output from the NOx sensor measurement, e.g., from sensor 106;
δ∫NOx_EOModel=variation of the integral of the output from the NOx model; and
δ_∫NOx1 _Sensor=variation of the integral of the output from the NOx sensor measurement, e.g., from sensor 106.
The variation or dispersion included above may be a statistical sigma of a measurement of the sensor 106 and the model measurements. In these examples, when the learning error adjustment exceeds the magnitude of the dead band, process 300 may proceed to block 340. If the learning error adjustment does not exceed the magnitude of the dead band, process 300 may instead proceed to block 350.
At block 340, a new adjustment to the learning error factor or adjustment may be determined. Proceeding from block 340 to block 345, process 300 may apply the determined learning error adjustment to a (previous) measurement taken from the first NOx sensor 106. An example learning error adjustment Lxr may be calculated using the following equation:
L _xr =x*[1+((r−1)*Coeff)]
Where: x=previous value of the learning coefficient;

- r=ratio of the integrated NOx model to sensor reading (same as the variable “K” in the preceding equation for determining the variation); and
- Coeff=adjustable learning coefficient.

The adjustable learning coefficient may be employed to adjust the effect of the learning error adjustment over time, e.g., by reducing or increasing an effect of the adjustment, e.g., in response to a determination that the learning error adjustment is adjusting sensor measurements too quickly or too slowly, respectively. In one example, the learning coefficient may be adjusted from 0.5 to 0.9, with higher coefficient values resulting in a quicker response, and a lower coefficient value resulting in greater robustness. Process 300 may then proceed to block 350.
At block 350, the control methodology may be reset, e.g., by resetting the timer employed in block 310 and integrals of the sensor measurement and sensor model. Process 300 may then proceed to block 355. In this manner, upon determination and application of a learning error adjustment, the calculation may be reset to allow subsequent measurement for any error, if still present, and further adjustment if needed.
At block 355, process 300 may determine whether the method should terminate, e.g., if the engine is no longer running. In one example, if engine 102 is not running, process 300 terminates. If, on the other hand, the engine 102 is still running or it is otherwise determined that process 300 should continue, the method may proceed back to block 305.
It is to be understood that the foregoing is a description of one or more embodiments of the invention. The invention is not limited to the particular embodiment(s) disclosed herein, but rather is defined solely by the claims below. Furthermore, the statements contained in the foregoing description relate to particular embodiments and are not to be construed as limitations on the scope of the invention or on the definition of terms used in the claims, except where a term or phrase is expressly defined above. Various other embodiments and various changes and modifications to the disclosed embodiment(s) will become apparent to those skilled in the art. All such other embodiments, changes, and modifications are intended to come within the scope of the appended claims.
As used in this specification and claims, the terms “e.g.,” “for example,” “for instance,” “such as,” and “like,” and the verbs “comprising,” “having,” “including,” and their other verb forms, when used in conjunction with a listing of one or more components or other items, are each to be construed as open-ended, meaning that the listing is not to be considered as excluding other, additional components or items. Other terms are to be construed using their broadest reasonable meaning unless they are used in a context that requires a different interpretation.

Claims

What is claimed is:

1. A method of treating diesel exhaust in a vehicle, comprising:

providing an exhaust aftertreatment device in an exhaust tailpipe, and a first nitrogen oxide (NO_x) sensor upstream of the exhaust aftertreatment device , the exhaust aftertreatment device configured to reduce NO_xpresent in an exhaust flow through the exhaust aftertreatment device with a treatment fluid applied to the exhaust flow;

measuring a concentration of NO_xin the exhaust flow with the first sensor;

determining an error in the measurement of the NO_xconcentration; and

applying a learning corrective adjustment in a subsequent measurement of the NO_xconcentration in the exhaust flow in response to the determination in step (c), a first magnitude of the learning corrective adjustment based at least in part upon a second magnitude of the error in the measurement.

2. The method of claim 1, further comprising

determining that the vehicle is operating in a normal ambient condition;

wherein the error in the measurement of the concentration of NOx is determined n response to the determination that the vehicle is operating in a normal exhaust mode.

3. The method of claim 1, wherein the learning corrective adjustment is applied in step (d) in response to a determination that the second magnitude of the error exceeds a predetermined magnitude.

4. The method of claim 1, wherein the exhaust aftertreatment device is one of a Selective Catalytic Reduction (SCR) device and a Selective Catalytic Reduction on Filter (SCRF) device.

5. The method of claim 4, wherein the exhaust aftertreatment device is configured to reduce at least a portion of the NOx present in the exhaust flow to nitrogen (N₂) and water (H₂O).

6. The method of claim 4, further comprising at least one of a Diesel Oxidation Catalyst (DOC) and a Diesel Particulate Filter (DPF).

7. The method of claim 4, wherein the treatment fluid includes urea.

8. The method of claim 1, wherein the first sensor is positioned downstream of an engine exhaust output.

9. The method of claim 1, further comprising providing a second NO_xsensor positioned downstream of the exhaust aftertreatment device.

10. The method of claim 1, wherein the treatment fluid is supplied by a diesel exhaust fluid dosing system.

11. The method of claim 1, wherein determining the error includes comparing a first integral of the measurement taken with the first sensor with a second integral of a model of expected NOx concentration.

12. The method of claim 11, wherein comparing the first and second integrals including determining a ratio of the first and second integrals.

13. The method of claim 12, further comprising applying a low-pass filter to the ratio of the first and second integrals.

14. The method of claim 1, further comprising determining a dead-band from a variation in at least one of (a) measurement by the first sensor and (b) a. model of expected NOx production based upon one or more engine operating parameters; wherein the learning corrective adjustment is applied in response to a determination that the error in the measurement exceeds the dead-band.

15. The method of claim 1, further comprising modeling an expected NOx concentration from one or more operating parameters of the engine.

16. A method of treating diesel exhaust in a vehicle, comprising:

providing an exhaust aftertreatment device in an exhaust tailpipe, and a first nitrogen oxide (NO_x) sensor upstream of the exhaust aftertreatment device, the exhaust aftertreatment device configured to reduce NO_xpresent in an exhaust flow through the exhaust aftertreatment device with a treatment fluid applied to the exhaust flow;

measuring a concentration of NO_xin the exhaust flow with the first sensor;

determining an error in the measurement of the NO_xconcentration, including:

comparing a first integral of the measurement taken with the first sensor with a second integral of a model NOx reduction by determining a ratio of the first and second integrals; and

applying a low-pass filter to the ratio of the first and second integrals; and

17. The method of claim 16, wherein the learning corrective factor is applied in step (d) in response to a determination that the error exceeds a predetermined magnitude.

18. A vehicle, comprising:

an exhaust tailpipe, including:

an exhaust aftertreatment device configured to reduce nitrogen oxide (NO_x) present in an exhaust flow through the exhaust aftertreatment device with a treatment fluid applied to the exhaust flow; and

a first NO_xsensor upstream of the exhaust aftertreatment device, the first NOx sensor configured to measure a concentration of NO_xin the exhaust flow; and

a controller communicatively linked to the sensor, the controller configured to determine an error in the measurement of the amount of the NO_xconcentration, and apply a corrective factor in a subsequent measurement of the NO_xconcentration in the exhaust flow, wherein a first magnitude of the corrective factor is based at least in part upon a second magnitude of the error in the measurement.

19. The vehicle of claim 18, wherein the controller is configured to apply the learning corrective factor in response to a determination that the error exceeds a predetermined magnitude.

20. The vehicle of claim 18, wherein the treatment fluid includes urea.