DE112014006704T5 - Emission control system for an internal combustion engine - Google Patents

Abstract

The object of the present invention is to provide an exhaust gas purification system for an internal combustion engine, wherein the fine dust discharge amount and the HC discharge amount can be optimized. The exhaust purification system includes an LNT for oxidizing HC, a fuel injection valve that can perform a main injection and a post injection, a fuel injection control means that controls the fuel injection amount and the fuel injection timing from the fuel injection valve, a temperature detection means that detects the LNT carrier temperature, and an oxidation ability estimation means estimates the HC oxidation ability of the LNT based on the detected carrier temperature. The main injection post-injection is defined as a fuel injection in a range in which the fine dust combustion amount in the exhaust pipe increases with increasing distance from the main injection. The higher the oxidation characteristic characteristic parameter Pox indicating the HC oxidation ability, the more the fuel injection control means increases the distance between the main injection timing and the post injection timing.

Description

Exhaust gas purification systems for internal combustion engines serve to remove HC (hydrocarbon), CO (carbon monoxide) and NOx (nitrogen oxide) contained in the exhaust gas of the engine. Most widely used are exhaust gas purification systems which utilize a reaction on various types of catalysts provided in the exhaust passage to remove the aforementioned three constituent elements from the exhaust gas. Catalysts having various functions have been proposed as catalysts for exhaust gas purification, including an oxidation catalyst (hereinafter also referred to as "DOC" (short for "Diesel Oxidation Catalyst")), a three-way catalyst (hereinafter also referred to as "TWC" (abbreviated for "three-way catalyst"), a lean NOx trap (hereinafter also referred to as "LNT" (short for "lean NOx trap")) and a selective reduction catalyst (hereinafter also referred to as "SCR catalyst" ( short for "Selective Catalytic Reduction Catalyst").

The oxidation catalyst has an oxidation function for HC and CO purification by causing an oxidation reaction of HC and CO in the oxygen-rich exhaust gas (lean exhaust gas) in a lean air-fuel mixture. This oxidation catalyst also has a three-way purifying function because, in a stoichiometric air-fuel mixture in exhaust gas (stoichiometric air-fuel ratio exhaust gas), the oxidation reaction for HC and CO and a reduction reaction for NOx occur simultaneously with high efficiency. The three-way catalyst is a catalyst in which the described oxidation catalyst is added to an oxygen storage material (hereinafter also referred to as "OSC" ("Oxygen Storage Capacity")), and has a larger three-way compared to the oxidation catalyst Cleaning window on, so a wider range of air-fuel ratio at which the three-way cleaning function can be achieved. This effect results from the fact that the range of change in the air-fuel ratio in the catalyst is reduced in proportion to the margin of the air-fuel ratio before the catalyst by the oxygen storage effect of the OSC material.

In the selective reduction catalyst, NOx is reduced in the presence of a reducing agent such as NH ₃ or HC or the like supplied from the outside or present in the exhaust gas. The lean NOx trap stores the NOx of exhaust gas at a lean air-fuel ratio, and in stoichiometric or richer air-fuel ratio exhaust gas, reduces the stored NOx using a reducing agent. In an exhaust gas purification system of a lean-burn gasoline engine or a diesel engine or other lean-burn engines, to ensure the NOx purification performance of lean air-fuel ratio exhaust gas, a DeNOx catalyst such as the selective reduction catalyst, the lean NOx catalyst, is often used. Trap or the like also be used in combination with the above-described oxidation catalyst or three-way catalyst.

The Japanese Patent Laid-Open Publication No. 2010-116781 and 2009-293585 propose an exhaust gas purification system combining a three-way catalyst and a DeNOx catalyst. In the exhaust gas purification system of JP-O No. 2010-116781, stoichiometric stoichiometric air-fuel mixture driving, and the three-way, is performed during the engine warm-up process when the DeNOx catalyst provided on the underbody can not achieve sufficient NOx purification performance Purification reaction of the three-way catalyst exploited to counteract a reduction in the NOx purification rate during the engine warm-up process. In the exhaust gas purification system of JP-O No. 2009-293585, in a driving operation in which the NOx discharge amount exceeds a certain allowable value, a stoichiometric driving operation is performed and NOx is effectively removed by utilizing the three-way cleaning reaction. With respect to such stoichiometric driving, the JP-O No. 2010-127251 an exhaust gas purification system wherein, in SOx discharge control for separating SOx stored in a DeNOx catalyst, by performing post injection after the main injection, the air-fuel mixture is made stoichiometric.

Documents of the prior art

Patent documents

• Patent document 1: Japanese Patent Laid-Open Publication No. 2010-116781
• Patent document 2: Japanese Patent Laid-Open Publication No. 2009-293585
• Patent 3: Japanese Patent Laid-Open Publication No. 2010-127251

In this way, in an exhaust gas purification system with DeNOx catalyst, by performing a stoichiometric driving operation, the overall NOx emission amount of the system further continues can be reduced than in the case that the NOx removal takes place in the continued lean driving mainly by DeNOx catalyst. However, it is known that in stoichiometric driving, the amount of particulate matter emitted from an internal combustion engine actually increases, unlike NOx. Therefore, in the stoichiometric driving operation preferably a post-injection is carried out, whose technique is approximately in JP-O No. 2010-127251 is described. Through post-injection, the fine dust produced by the main injection can be burnt in the exhaust pipe, so that the amount of fine dust ejected from the internal combustion engine can be reduced.

32 shows the post-injection quantity in stoichiometric mode (mg / stroke) or the distance between main injection and post-injection (degree) and the ratio between particulate matter output rate (mg / stroke) and HC output quantity (mg / stroke). In 32 The solid line represents the particulate matter discharge amount, while the broken line represents the HC discharge amount. At the fine dust discharge amount and the HC discharge amount off 32 these are in each case the quantity ejected from the internal combustion engine per unit of time.

Like the solid line in 32 shows, the larger the post-injection amount or the larger the distance of the post-injection, the more decreases the particulate matter emission. The broken line in 32 shows, however, that the larger the post-injection amount or the larger the post-injection interval, the more the HC discharge amount increases. That is, in stoichiometric operation, there is a trade-off between particulate matter output and HC output. However, the prior art does not consider this relationship between PM emission amount and HC discharge amount with respect to the Nacheinspritzmenge or the time of the post injection rigidly according to a predetermined map according to the driving condition of the internal combustion engine, which is by no means of an optimization of particulate matter emission rate and HC output can be the question.

The present invention has been made in consideration of the above points, and its object is to provide an exhaust gas purification system for an internal combustion engine, wherein the particulate matter discharge amount and the HC discharge amount can be optimized.

Brief description of the invention

• (1) An exhaust gas purification system according to the invention (for example, exhaust gas purification system 2 ) of an internal combustion engine (for example engine 1 ) comprises an HC oxidation catalyst (for example LNT 41 ), in the exhaust duct (for example, exhaust duct 11 ) of the internal combustion engine and serves to oxidize HC in the exhaust gas, a fuel injection valve (for example, fuel injection valve 13 ) capable of performing a main injection and a post-injection after the main injection, which is a fuel injection in a region in which the fine dust combustion amount in the exhaust pipe increases with increasing distance from the main injection, a fuel injection control means (e.g. 3 ) controlling the fuel injection amount and the fuel injection timing of the fuel injection valve, a temperature detection means (for example, a temperature sensor in front of the catalyst 53 , Temperature sensor after the catalyst 54 and ECU 3 ), which detects the temperature of the HC oxidation catalyst, and an oxidizing ability estimating means (for example, ECU 3 and the later with reference to 10 performing processing from S52 to S54) which estimates the HC oxidation ability of the HC oxidation catalyst based on the temperature of the HC oxidation catalyst, the higher the HC oxidation ability, the more the fuel injection control means increases the distance between the main injection timing and the post injection timing.

31 FIG. 14 is an explanatory view showing the difference between the post injection, which is a fuel injection after the main injection, and the following injection. FIG. 31 FIG. 15 is a ratio chart showing the injection timing and the injection amount of the post-main injection fuel injection, and the particulate matter discharge amount (left) and the HC discharge amount (right) from the exhaust pipe. 31 shows the case of a constant engine running operation wherein the fuel injection timing after the main injection is changed in a range of about 45-75 (degree NOT) and the fuel injection amount after the main injection is between 2 mg (thin line), 6 mg (middle thickness line) and 8 mg (thick line) is changed. In the example off 31 the total amount of fuel injection is kept constant while the fuel injection amount is changed after the main injection, or the main injection timing is kept constant while the fuel injection timing after the main injection is changed.

As in 31 right, the more the fuel injection amount increases after the main injection, the more the HC discharge amount increases. In addition, the larger the distance between the main injection timing and the subsequent fuel injection timing, the more the HC discharge amount increases. As in 31 On the left side, the more the fuel injection amount increases after the main injection, the more the particulate matter discharge amount decreases. This decrease in the particulate matter discharge amount is due to the fact that reducing the main injection quantity, the amount of particulate matter produced in the exhaust pipe decreases and takes place by the post-main injection additional fuel fine dust combustion at low temperature in the exhaust pipe.

As in 31 On the left hand side, when retarding the fuel injection timing after the main injection to about 65 degrees, the particulate matter discharge amount decreases in the course of deceleration (that is, the particulate matter combustion amount in the exhaust pipe increases), while the particulate matter discharge amount does not decrease in deceleration beyond 65 degrees. The reason for this is that when the fuel injection timing is retarded after the main injection up to about 65 degrees, the temperature of the combustion gas decreases too much, so that no more low-temperature particulate matter combustion takes place. Thus, when the fuel injection is delayed by more than 65 degrees after the main injection, a corresponding reduction effect of the particulate matter discharge amount is not expected, and the HC discharge amount is even increased.

• (2) In diesem Fall wird bevorzugt, dass im Abgaskanal außerdem ein DeNOx-Katalysator vorgesehen ist, der eine DeNOx-Funktion aufweist, indem er unter Anwesenheit eines dem Abgaskanal zugesetzten Reduktionsmittels NOx aus dem Abgas entfernt, während der HC-Oxidationskatalysator eine Drei-Wege-Reinigungsfunktion aufweist, indem er bei einem stöchiometrisch Luft-Kraftstoff-Gemisch HC, CO und NOx aus dem Abgas entfernt, und wobei das Abgasreinigungssystem außerdem ein NOx-Überschussbeurteilungsmittel aufweist, das entsprechend einem oder beiden von dem Motorzustand und dem Zustand im Abgaskanal beurteilt, ob ein NOx-Überschusszustand herrscht, wobei sich der Anteil der in den NOx-Entfernungskatalysator einströmenden NOx-Menge im Verhältnis zu der durch die DeNOx-Funktion entfernbaren NOx-Menge vergrößert, wobei das Kraftstoffeinspritzsteuerungsmittel für den Fall, dass kein NOx-Überschusszustand herrscht, das Luft-Kraftstoff-Verhältnis auf mager statt stöchiometrisch regelt, und für den Fall, dass ein NOx-Überschusszustand herrscht, das Luft-Kraftstoff-Verhältnis stöchiometrisch regelt.
• (3) In diesem Fall wird bevorzugt, dass der HC-Oxidationskatalysator (beispielsweise LNT 41) außerdem eine Drei-Wege-Reinigungsfunktion, wobei bei stöchiometrischem Luft-Kraftstoff-Gemisch HC, CO und NOx aus dem Abgas entfernt werden, und eine DeNOx-Funktion aufweist, wobei unter Anwesenheit eines dem Abgaskanal zugesetzten Reduktionsmittels NOx aus dem Abgas entfernt wird, wobei das Abgasreinigungssystem außerdem ein NOx-Überschussbeurteilungsmittel (beispielsweise ECU 3 und das an späterer Stelle beschriebene Mittel zum Ausführen von S23 bis S25 aus 6) aufweist, das entsprechend einem oder beiden von dem Motorzustand und dem Zustand im Abgaskanal beurteilt, ob ein NOx-Überschusszustand herrscht, wobei ein Anteil der in den NOx-Entfernungskatalysator einströmenden NOx-Menge im Verhältnis zu der durch die DeNOx-Funktion entfernbaren NOx-Menge sich vergrößert, wobei das Kraftstoffeinspritzsteuerungsmittel für den Fall, dass kein NOx-Überschusszustand herrscht, das Luft-Kraftstoff-Verhältnis auf mager statt stöchiometrisch regelt, und für den Fall, dass ein NOx-Überschusszustand herrscht, das Luft-Kraftstoff-Verhältnis stöchiometrisch regelt.
• (4) Ein erfindungsgemäßes Abgasreinigungssystem (beispielsweise Abgasreinigungssystem 2) eines Verbrennungsmotors (beispielsweise, Motor 1) umfasst einen HC-Oxidationskatalysator (beispielsweise LNT 41), der eine HC-Oxidationsfunktion, wobei HC aus dem Abgas oxidiert wird, und eine Drei-Wege-Reinigungsfunktion aufweist, wobei bei stöchiometrischem Luft-Kraftstoff-Gemisch HC, CO und NOx aus dem Abgas entfernt werden, ein Kraftstoffeinspritzventil (beispielsweise Kraftstoffeinspritzventil 13), das eine Haupteinspritzung und eine nach der Haupteinspritzung erfolgende Nacheinspritzung durchführen kann, wobei es sich um eine Kraftstoffeinspritzung in einem Bereich handelt, in dem mit zunehmendem Abstand von der Haupteinspritzung die Feinstaubbrennungsmenge im Abgasrohr zunimmt, ein Kraftstoffeinspritzsteuerungsmittel (beispielsweise ECU 3), das die Kraftstoffeinspritzmenge und den Kraftstoffeinspritzzeitpunkt des Kraftstoffeinspritzventils steuert, ein Temperaturerfassungsmittel (beispielsweise Temperatursensor vor dem Katalysator 53, Temperatursensor nach dem Katalysator 54 und ECU 3), das die Temperatur des HC-Oxidationskatalysators erfasst, ein Oxidationsfähigkeitsschätzungsmittel (beispielsweise ECU 3 und ein Mittel, das die an späterer Stelle unter Bezugnahme auf 10 beschriebene Verarbeitung von S52 bis S54 durchführt), das anhand der Temperatur des HC-Oxidationskatalysators die HC-Oxidationsfähigkeit des HC-Oxidationskatalysators schätzt, und ein NOx-Überschussbeurteilungsmittel (beispielsweise ECU 3 und das an späterer Stelle beschriebene Mittel zum Ausführen von S23 bis S25 aus 6), das entsprechend einem oder beiden von dem Motorzustand und dem Zustand im Abgaskanal beurteilt, ob ein NOx-Überschusszustand herrscht, wobei sich ein Anteil der in den NOx-Entfernungskatalysator einströmenden NOx-Menge im Verhältnis zu der durch die DeNOx-Funktion entfernbaren NOx-Menge vergrößert, wobei das Kraftstoffeinspritzsteuerungsmittel für den Fall, dass geurteilt wird, dass ein NOx-Überschusszustand herrscht, das Luft-Kraftstoff-Verhältnis stöchiometrisch regelt, während die Nacheinspritzmenge so erhöht wird, dass sich die HC-Oxidationsfähigkeit steigert.
• (5) In diesem Fall wird die HC-Oxidationsfähigkeit vorzugsweise anhand von Zahlen dargestellt, nämlich einem Basisfaktor (Pox_bs) in Abhängigkeit von der Temperatur des HC-Oxidationskatalysators und dem Abgasvolumen, und Verschlechterungsfaktoren (Kmod, Kox, Krd, Dtw) in Abhängigkeit vom Verschlechterungsgrad des HC-Oxidationskatalysators.
• (6) In diesem Fall schätzt das Oxidationsfähigkeitsschätzungsmittel die HC-Oxidationsfähigkeit vorzugsweise anhand eines Wärmeerzeugungskoeffizienten (c) des HC-Oxidationskatalysators, der den Beitrag des durch die Nacheinspritzung zugeführten Kraftstoffs zum Temperaturanstieg des HC-Oxidationskatalysators anzeigt.
• (7) In diesem Fall schätzt das Oxidationsfähigkeitsschätzungsmittel die HC-Oxidationsfähigkeit vorzugsweise anhand der Reduktionsmittelmenge, die zum Abscheiden des im HC-Oxidationskatalysator gespeicherten Sauerstoffs benötigt wird, oder anhand der Reduktionsmittelmenge (Crd_hat), die zum Reduzieren des im HC-Oxidationskatalysator gespeicherten NOx benötigt wird.
• (8) In diesem Fall weist das Abgasreinigungssystem vorzugsweise außerdem einen O₂-Sensor (beispielsweise O₂-Sensor nach dem Katalysator 52) auf, der die Sauerstoffkonzentration im Abgas stromabwärts vom HC-Oxidationskatalysator erfasst, und das Oxidationsfähigkeitsschätzungsmittel schätzt die HC-Oxidationsfähigkeit anhand eines Luft-Kraftstoff-Verhältnis-Sollwerts (AFcmd), der derart festgelegt ist, dass der Ausgabewert (Vout) des O₂-Sensors auf einem bestimmten Sollwert (Vcmd) gehalten wird.

The fuel injection performed after the main injection is divided into the post injection and a post injection carried out after this post injection, but there is no definition clearly separating these two later injections. In this regard, in the present invention, a demarcation between post-injection and follow-up injection occurs because a significant change is observed in the effect of injection timing retardation. Accordingly, as a post-injection, the fuel injection after the main injection is defined, which takes place in the region in which the fine dust combustion amount in the exhaust pipe increases with increasing delay (in the example of FIG 31 45 to 65 (degree NOT)). In turn, as follow-up injection, the fuel injection after the main injection and the post-injection takes place in a region where the particulate matter combustion amount in the exhaust pipe does not increase despite deceleration, and only the HC discharge amount increases (in the example FIG 31 65 (degree NOT) or higher).

• (2) In this case, it is preferable that a DeNOx catalyst having a DeNOx function is also provided in the exhaust passage by removing NOx from the exhaust gas in the presence of a reducing agent added to the exhaust passage, while the HC oxidation catalyst carries out a three-way reaction. Having a path cleaning function by removing HC, CO and NOx from the exhaust gas in a stoichiometric air-fuel mixture, and wherein the exhaust gas purifying system further comprises NOx surplus judgment means judging according to one or both of the engine condition and the condition in the exhaust passage whether there is an NOx excess state, wherein the proportion of the NOx amount flowing into the NOx removal catalyst increases in proportion to the amount of NOx removable by the DeNOx function, the fuel injection control means being in the case of no NOx excess state , which regulates air-fuel ratio to lean instead of stoichiometric, and in the event that there is an excess NOx state, stoichiometrically controlling the air-fuel ratio.
• (3) In this case, it is preferable that the HC oxidation catalyst (for example, LNT 41 ) also has a three-way cleaning function, wherein stoichiometric air-fuel mixture HC, CO and NOx are removed from the exhaust gas, and has a DeNOx function, wherein in the presence of a reducing agent added to the exhaust gas NOx is removed from the exhaust gas, wherein the exhaust gas purifying system further includes NOx excess judging means (for example, ECU 3 and the later-described means for executing S23 to S25 6 ) which, according to one or both of the engine condition and the condition in the exhaust passage, judges whether or not there is an excess NOx condition, a proportion of the amount of NOx flowing into the NOx removal catalyst relative to the NOx removal capability of the DeNOx function. Amount increases, wherein the fuel injection control means in the event that there is no excess NOx state controls the air-fuel ratio to lean rather than stoichiometric, and in the event that there is an excess NOx state, stoichiometrically controls the air-fuel ratio ,
• (4) An exhaust gas purification system according to the invention (for example, exhaust gas purification system 2 ) of an internal combustion engine (for example, engine 1 ) comprises an HC oxidation catalyst (for example LNT 41 ) having an HC oxidation function, wherein HC is oxidized from the exhaust gas, and a three-way cleaning function, wherein in stoichiometric air-fuel mixture HC, CO and NOx are removed from the exhaust gas, a fuel injection valve (for example, fuel injection valve 13 ) capable of performing a main injection and a post-injection after the main injection, which is a fuel injection in a range in which as the distance increases from the main injection increases the fine dust combustion amount in the exhaust pipe, a fuel injection control means (for example ECU 3 ) controlling the fuel injection amount and the fuel injection timing of the fuel injection valve, a temperature detection means (for example, a temperature sensor in front of the catalyst 53 , Temperature sensor after the catalyst 54 and ECU 3 ), which detects the temperature of the HC oxidation catalyst, an oxidizing ability estimating means (for example, ECU 3 and a remedy, referring to the later reference to 10 performing processing from S52 to S54) that estimates the HC oxidation ability of the HC oxidation catalyst based on the temperature of the HC oxidation catalyst, and a NOx surplus judgment means (for example, ECU 3 and the later-described means for executing S23 to S25 6 ) which judges whether there is a NOx surplus condition in accordance with one or both of the engine condition and the condition in the exhaust passage, whereby a proportion of the NOx amount flowing into the NOx removal catalyst is increased in proportion to the NOx NOx removal capability of the DeNOx function. Increased amount, wherein the fuel injection control means in the event that it is judged that a NOx excess state prevails, stoichiometrically controls the air-fuel ratio, while the post-injection amount is increased so that the HC oxidation ability increases.
• (5) In this case, the HC oxidation ability is preferably represented by numbers, namely, a base factor (Pox_bs) depending on the temperature of the HC oxidation catalyst and the exhaust gas volume, and deterioration factors (Kmod, Kox, Krd, Dtw) depending on Degree of deterioration of the HC oxidation catalyst.
• (6) In this case, the oxidizing ability estimating means preferably estimates the HC oxidizing ability from a heat generation coefficient (c) of the HC oxidation catalyst indicating the contribution of the fuel supplied by the post injection to the temperature rise of the HC oxidation catalyst.
• (7) In this case, the oxidizing ability estimating means preferably estimates the HC oxidizing ability by the amount of reducing agent needed to separate the oxygen stored in the HC oxidizing catalyst or by the reducing agent amount (Crd_hat) needed to reduce the NOx stored in the HC oxidizing catalyst becomes.
• (8) In this case, the exhaust gas purifying system preferably further includes an O ₂ sensor (for example, O ₂ sensor after the catalyst 52 ), which detects the oxygen concentration in the exhaust gas downstream of the HC oxidation catalyst, and the oxidizing ability estimating means estimates the HC oxidizing ability based on an air-fuel ratio target value (AFcmd) set such that the output value (Vout) of the O ₂ Sensor is kept at a certain setpoint (Vcmd).

Effect of the invention

• (1) When the distance between the main injection timing and the post-injection timing is increased, although the particulate matter discharge amount decreases, the HC discharge amount increases at the same time. If, on the other hand, the distance between the main injection time and the post-injection time is reduced, the fine dust ejection quantity increases, but at the same time the HC ejection quantity decreases (see FIG 31 as discussed above). In the present invention, the HC oxidation ability of the HC oxidation catalyst is estimated from the temperature of the HC oxidation catalyst, and the larger the HC oxidation ability, the more the distance between the main injection timing and the post injection timing is increased. For example, as described above, when a stoichiometric driving operation is performed using a main injection and a post injection, and thus the distance between the main injection timing and the post injection timing is controlled according to the HC oxidation ability of the HC oxidation catalyst, maintaining the amount of HC not removed, downstream of the HC oxidation catalyst, the amount of PM discharged from the engine is kept to a minimum.
• (2) According to the present invention, it is judged according to one or both of an engine condition and a condition in the exhaust passage whether or not there is a NOx surplus condition, wherein the amount of NOx flowing into the NOx removal catalyst relative to the NOx Amount that can be removed by the lean NO x purifying function increases. Specifically, the NOx surplus state means, for example, an accelerating driving operation in which the amount of NOx discharged from the engine increases or a timing immediately after starting the engine when the temperature of the NOx removing catalyst is low and the lean NOx purifying function is weakened. In the present invention, when there is no NOx excess state, a lean running operation is performed wherein the air-fuel ratio is controlled not to be stoichiometric but lean to remove NOx by the lean NOx purifying function, and when a NOx Excessive state prevails, a stoichiometric driving operation is performed, wherein the air-fuel ratio is controlled to stoichiometrically, with the common application of main injection and post-injection by means of the three-way cleaning function to remove NOx. By thus switching between lean running and stoichiometric driving according to the state of the engine or the state inside the exhaust passage, a high NOx purification rate for the exhaust gas purification system as a whole can be maintained. By controlling the distance between the main injection timing and the post-injection timing according to the HC oxidizing ability in the stoichiometric driving operation, while maintaining the amount of unreacted HC exiting the HC oxidation catalyst downstream, the amount of particulate matter discharged from the engine can be kept to a minimum. In summary, according to the present invention, NOx and HC can be removed with high efficiency while keeping the particulate matter discharge amount to a minimum.
• (3) The present invention differs from the invention in (2) above in that the HC oxidation catalyst has a NOx purifying function. According to the present invention, therefore, the same effect as with the invention can be obtained in (2).
• (4) When the post-injection amount is increased, although the particulate matter discharge amount decreases, the HC discharge amount increases. If the post-injection amount is reduced, although the particulate matter ejection amount increases, at the same time, the HC ejection amount decreases (refer to FIG 31 as discussed above). In the present invention, when it is judged that there is a NOx surplus state, a stoichiometric driving operation is performed with the air-fuel ratio being stoichiometrically controlled to effectively remove NOx from the exhaust gas by the three-way purifying function. According to the present invention, moreover, in the stoichiometric driving operation, the larger the HC oxidizing ability, the more the post injection quantity is increased. Thus, while maintaining the amount of HC removed within the allowable range exiting downstream of the HC oxidation catalyst, the amount of PM discharged from the engine can be kept to a minimum.
• (5) In the present invention, based on a basic factor that varies depending on the engine condition such as temperature of the oxidation catalyst or exhaust gas volume, and the state in the exhaust passage, and a deterioration factor that changes depending on the deterioration of the HC oxidation catalyst, the HC oxidation ability of the HC oxidation catalyst is shown in numbers. In this way, since the HC oxidizing ability can always be accurately detected, the distance between the main injection timing and the post-injection timing and the post injection quantity can be appropriately controlled.
• (6) During the post-injection, particulate matter generated by the main injection is burned, and as the particulate matter discharge amount decreases, the HC discharge amount increases. As the HC discharge amount increases, the heat generated by the HC oxidation function in the HC oxidation catalyst also increases accordingly. That is, a heat generation coefficient of the HC oxidation catalyst indicating the degree of contribution of the fuel supplied by the post injection to the temperature rise of the HC oxidation catalyst may serve as an index of the HC oxidation function. By using the heat generation coefficient having this characteristic in the present invention, the HC oxidation ability can be maintained with high accuracy, whereby the distance between the main injection timing and the post injection timing and the post injection quantity can be appropriately controlled.
• (7) It happens that a catalyst having a storage function for oxygen or NOx from the exhaust gas is used as the HC oxidation catalyst. It is believed that this storage function decreases as the HC oxidation catalyst degrades to the same extent as the HC oxidation function. The amount of reducing agent required for eliminating the oxygen stored in the HC oxidation catalyst, which is necessary for reducing the amount of NOx stored in the HC oxidation catalyst or the like, can serve not only as an index for the storage performance of the HC oxidation catalyst but also for the HC oxidation performance. By using this amount of reducing agent in the present invention, the HC oxidation ability can be maintained with high accuracy, whereby the distance between the main injection timing and the post injection timing and the post injection quantity can be appropriately controlled.
• (8) In the present invention, the HC oxidation ability is estimated from an air-fuel ratio command value set such that an output value of an O ₂ sensor provided downstream of the HC oxidation catalyst is maintained at a target value. Incidentally, the state in which the output value of the O ₂ sensor is slipped to its target value corresponds to a state where a slight amount of the reducing agent reaches the downstream side of the HC oxidation catalyst. The amount of reducing agent needed to maintain this reducing agent slip state decreases as the reducing ability of the HC oxidation catalyst decreases. As the reduction capability of the HC oxidation catalyst decreases, the air-fuel ratio command value required to maintain the output value of the O ₂ sensor at its target value shifts too lean. It is believed that this reducing ability of the HC oxidation catalyst decreases as the HC oxidation catalyst deteriorates to the same extent as the HC oxidation function. thats why the air-fuel ratio target value not only an index for the reduction ability of the HC oxidation catalyst but also for the HC oxidation performance. By using this air-fuel ratio target value in the present invention, the HC oxidizing ability can be maintained with high accuracy, whereby the distance between the main injection timing and the post-injection timing and the post-injection amount can be appropriately controlled.

Brief description of the figures

Show it:

1 a schematic view illustrating the structure of an engine and an exhaust gas purification system thereof according to an embodiment of the present invention;

2 a view showing the post-injection amount in the stoichiometric operation and the ratio between post-injection timing and particulate matter discharge amount and HC discharge amount;

3 FIG. 12 is a view showing the relationship of the temperature on the intake side of the engine side catalyst, the temperature on the exhaust side, the carrier temperature of the engine bottom catalyst, the exhaust gas volume, and the like to the HC oxidation performance; FIG.

4 an example of a map for determining the correction coefficient for the post-injection amount;

5 a main flow chart illustrating the concrete procedure of the fuel injection control;

6 a flowchart illustrating the concrete flow of the processing for judging the stoichiometric operating conditions;

7 an example of a map for determining the value of the stoichiometric operating mode flag (for the three-way cleaning mode);

8th an example of a map for determining the value of the stoichiometric operating mode flag (for the common mode);

9 a flowchart illustrating the concrete flow of the feedback calculation of the air-fuel ratio before the catalyst;

10 a flowchart illustrating the concrete procedure for determining the post-injection amount and the Nacheinspritzzeitpunkt in stoichiometric driving;

11 an example of a map for determining the standard value for the oxidation characteristic parameter;

12 an example of a map for determining the default value for the post-injection amount;

13 an example of a map for determining the correction coefficient for the post-injection amount;

14 an example of a map for determining the default value for the post-injection time;

15 an example of a map for determining the correction value for the post injection timing;

16 a flowchart illustrating the concrete sequence of the feedback calculation of the air-fuel ratio after the catalyst;

17 an example of a map for determining the target air-fuel ratio in the low-fat operating mode;

18 an example of a map for determining the output target value of the O ₂ sensor after the catalyst;

19 5 is a timing chart showing a concrete example of the change of the output of the LAF sensor before the catalyst and the output of the O ₂ sensor after the catalyst in the feedback calculation of the air-fuel ratio after the catalyst;

20 Fig. 10 is a flowchart showing the concrete procedure of the engine underside catalyst reduction characteristic judgment processing;

21 FIG. 10 is a timing chart illustrating a concrete example of the change of the reduction treatment end flag and the like in the execution of the engine bottom catalyst reduction characteristic judgment processing; FIG.

22 Fig. 10 is a flowchart illustrating the concrete procedure of the adaptive calculation for the three-way catalyst characteristic;

23 Fig. 11 is a view showing an example of the form of the weighting function for the three-way catalyst characteristic;

24 a flowchart illustrating the concrete procedure of the adaptive calculation for the catalyst reduction characteristic;

25 an example of a map for determining the standard reducing agent supply amount;

26 Fig. 10 is a view showing an example of the shape of the reducing agent supply amount weighting function;

27 FIG. 3 is a flowchart illustrating the concrete procedure of the recursive identification calculation of a heat model of the engine bottom catalyst; FIG.

28 FIG. 4 is a flowchart illustrating the concrete procedure of the adaptive calculation of the catalyst oxidation characteristic; FIG.

29 FIG. 12 is a view illustrating an example of the shape of the oxidation characteristic weighting function; FIG.

30 a schematic view illustrating the structure of an engine and an exhaust gas purification system thereof according to an embodiment of the present invention;

31 an explanatory view illustrating the difference between post-injection and follow-up injection; and

32 a view showing the ratio of the post-injection amount in the stoichiometric operation and the distance between the main injection and post-injection and particulate matter discharge amount and HC discharge amount.

Embodiment of the invention

An embodiment of the present invention will be described below with reference to the figures. 1 is a view showing the structure of an internal combustion engine 1 (hereinafter referred to as "engine") and an exhaust gas purification system 2 the same according to an embodiment shows. At the engine 1 it is an engine with basically lean combustion, the air-fuel ratio is not regulated to stoichiometric, but lean, or more specifically to a diesel engine, a lean-burn gasoline engine and the like.

The emission control system 2 is designed to be a lean NOx trap (hereinafter "LNT") 41 and an exhaust gas purification filter (hereinafter "DPF") 43 in the exhaust duct 11 of the motor 1 are provided, an exhaust fuel injection device 45 that use fuel as a reducing agent in the exhaust duct 11 injects, and an electronic control unit (hereinafter "ECU") 3 includes the engine 1 and the exhaust fuel injection device 45 controls.

At the engine 1 is a fuel injection valve 13 provided inject the fuel into the individual cylinders. This fuel injector 13 is via an unillustrated drive device with the ECU 3 connected. The ECU 3 determined by one of the following with reference to 2 to 29 described fuel injection control fuel injection amount, fuel injection timing and the like of the fuel injection valve 13 , and the drive device actuates the fuel injection valve 13 such that the fuel injection state is realized.

The LNT 41 has at least three functions, namely an oxidation function, a DeNOx function and a three-way cleaning function. The oxidation function is a function whereby in the lean driving operation in which the air-fuel ratio is not regulated to stoichiometric but lean, HC and CO contained in the exhaust gas are oxidized. The DeNOx function is a function of absorbing NOx contained in the exhaust gas under lean driving and absorbing exhaust gas by exhaust fuel injection from the exhaust gas injection device 45 or by subsequent injection from the fuel injection valve 13 or the like fuel is supplied, this serves as a reducing agent to reduce NOx. The three-way cleaning function is a function wherein in stoichiometric operation with the air-fuel ratio stoichiometrically controlled, HC, CO and NOx contained in the exhaust gas are removed together.

In this way, the NOx in the exhaust gas in the lean driving using the DeNOx function of the LNT 41 removed and in stoichiometric operation using the three-way cleaning function of the LNT 41 away. A comparison of NOx purification using the DeNOx function and the three-way cleaning function shows that NOx purification is more effective with the three-way cleaning function. So if, for example, when driving at high load, when the engine 1 discharged NOx amount is large, or if the LNT 41 their activation is not achieved and a sufficient DeNOx function can not be achieved (that is, in an NOx excess state discussed later), it is changed from the lean driving to the stoichiometric driving to perform the cleaning using the three-way cleaning function. The concrete procedure of switching between lean driving and stoichiometric driving will be described later with reference to FIG 6 described.

When the exhaust gas passes through the fine openings in the filter wall of the DPF 43 occurs, particulate matter in the exhaust gas, the main component of which is carbon, is collected by accumulating on the surface of the filter wall and in the openings in the filter wall. As a material for forming the filter wall, a porous material such as aluminum titanate or cordierite is used.

The exhaust duct 11 is divided into an area inside the engine room (not shown) (engine under-side area) and an area under the floor of the vehicle (not shown) (floor bottom area). The engine base area is closer to the engine 1 as the bottom base area. Therefore, the average temperature of the engine underside region is higher than that of the bottom surface region, and also its temperature after starting the engine 1 rises faster. To achieve the oxidation function, three-way cleaning function and DeNOx function as effectively as possible, the LNT 41 therefore in the engine base area of the exhaust duct 11 intended. Because the LNT 41 is provided in the engine base area, the LNT 41 also referred to as engine bottom catalyst.

If the DPF 43 Particulate matter has collected until it reaches its absorption limit, the pressure loss is high. Therefore, compulsory regeneration is suitably performed to control the filtering function of the DPF 43 restore by burning the collected particulate matter. In this forced regeneration, for example, a follow-up injection or a fuel injection from the exhaust fuel injection device 45 performed, and by the increase in the temperature of the DPF 43 inflowing exhaust gas, the accumulated fine dust is burned within a short time.

Exhaust fuel injector 45 has a fuel tank 451 , one in the exhaust duct 11 upstream of the LNT 11 provided exhaust fuel injection nozzle 452 and a pressure pump 453 on that the fuel in the fuel tank 451 under pressure to the injector 452 promoted. The exhaust fuel injection nozzle 452 is electromagnetically connected to the ECU via an unillustrated drive device 3 connected. When the exhaust gas using the DeNOx function of the LNT 41 is cleaned, the ECU determines 3 by means of an exhaust fuel injection control (not shown), the exhaust fuel injection amount and the Exhaust fuel injection timing of the exhaust fuel injection nozzle 452 and the drive device actuates the exhaust fuel injection nozzle 452 to achieve the exhaust fuel injection state.

More recently, however, it has been found that at the LNT 41 hydrocarbon-derived intermediates can be produced by which high efficiency NOx purification can be achieved when injecting fuel from the exhaust fuel injector 452 and using the DeNOx function of the LNT 41 for reducing NOx, the exhaust fuel injection amount from the exhaust fuel injection nozzle 452 is increased or decreased at a frequency of 5 Hz or more within a certain range. However, this occurs at a carrier temperature of the LNT 41 of about 350 ° C or less an injection in the state described above, it happens that unnecessary components are generated, which do not contribute to the removal of NOx (for example, N ₂ O), and downstream of the LNT 41 be ejected. Therefore, in the exhaust fuel injection control, the fuel is injected intermittently in the above-described manner only when the carrier temperature of the LNT 41 is about 350 ° C or more with an upper limit of about 630 to 700 ° C.

With the ECU 3 are as sensors for detecting the condition in the exhaust passage 11 and the condition of the engine 1 a LAF sensor in front of the catalytic converter 51 , an O ₂ sensor after the catalyst after the catalyst 52 , a temperature sensor in front of the catalyst 53 , a temperature sensor after the catalyst 54 , a crankshaft position sensor 55 , Accelerator opening sensor 56 , an air flow sensor 57 and air temperature sensor 58 and the like connected.

The O ₂ sensor after the catalyst 52 is in the exhaust duct 11 between LNT 41 and DPF 43 intended. The O ₂ sensor after the catalyst 52 detects the oxygen concentration (the air-fuel ratio) in the exhaust gas downstream of the LNT 41 and sends a signal to the ECU according to the detected value 3 , The strength of the O ₂ sensor after the catalyst 52 transmitted signal has a substantially bivalent characteristic (see, for example 19 later) and is low in the case where the air-fuel ratio is not stoichiometric but rich, high (for example, 1), and low in the case where it is not stoichiometric but lean ("Low") (for example 0). Hereinafter, changing the output of the O ₂ sensor after the catalyst 52 between low and high called output reversal.

The LAF sensor in front of the catalytic converter 51 is in the exhaust duct 11 upstream of LNT 41 and exhaust fuel injection nozzle 452 intended. The LAF sensor 51 detects the air-fuel ratio of the exhaust gas upstream of the LNT 41 to which the exhaust fuel injection nozzle 452 has not yet injected fuel, and sends a signal substantially proportional to the detection value to the ECU 3 , That from the LAF sensor 51 output signal has, unlike the O ₂ sensor after the catalyst 52 a linear characteristic for a wider range of the air-fuel ratio from the rich to the lean region.

The temperature sensor in front of the catalytic converter 53 is in the exhaust duct 11 upstream of the LNT 41 provided, and the temperature sensor after the catalyst 54 is in the exhaust duct 11 downstream of the LNT 41 intended. These temperature sensors 53 . 54 each capture the temperature of the LNT 41 inflowing exhaust gas and that from the LNT 41 outgoing exhaust gas and send a signal substantially proportional to the detection value to the ECU 3 , The carrier temperature of the LNT 41 is from the ECU 5 as a weighted average of the outputs of the temperature sensors 53 . 54 estimated.

The crankshaft position sensor 55 detects the angle of rotation of the crankshaft of the engine 1 , generates pulses for the respective crankshaft angles and sends the pulse signal to the ECU 3 , The ECU 3 calculates the speed of the motor based on this pulse signal 1 , The accelerator opening sensor 56 detects a passage degree of an accelerator pedal of the vehicle (not shown) (hereinafter "accelerator opening degree") and sends a signal substantially proportional to the detection value to the ECU 3 , The ECU 3 calculates the driver demanded torque based on the accelerator opening degree and the engine speed and the like. The air flow measuring device 57 is in the intake channel 12 intended. The air flow measuring device 57 captures the amount of intake air passing through the intake passage 12 flows, and sends a signal substantially proportional to the detection value to the ECU 3 , The ECU 3 Calculates the volume of exhaust gas based on the amount of intake air.

Next, referring to 2 to 29 the contents of the fuel injection control will be described in detail. 2 is a view showing the post injection amount in stoichiometric operation (mg / stroke) and the relationship between post injection timing (degree NOT) and particulate matter ejection amount (mg / stroke) and HC ejection amount (mg / stroke). In 2 represents the solid line the Particulate emission rate, while the broken line represents the HC discharge amount. In 2 Further, the particulate matter discharge amount is the amount of particulate matter discharged from the engine per unit time, and the HC discharge amount is the amount of HC discharged without being removed on the downstream side of the engine side catalyst per unit time.

As with reference to 31 described increases as the Nacheinspritzmenge or delay of Nacheinspritzzeitpunkts the particulate matter emitted by the engine, while the amount of HC ejected from the engine increases. In the exhaust gas purification system off 1 For example, HC exhausted from the engine in stoichiometric driving is oxidized by the three-way purifying function of the engine bottom catalyst. Thus, the HC discharge amount changes downstream of the engine bottom catalyst as indicated by the three broken line in FIG 2 shown, depending on the HC oxidation performance of the engine bottom catalyst at this time. That is, the higher the HC oxidation performance of the engine bottom catalyst, the lower the HC discharge amount downstream of the engine bottom catalyst.

In stoichiometric operation, the HC discharge amount downstream of the engine bottom catalyst is an in 2 determined by the dot-dash line HC discharge amount upper limit. In order to reduce the load of the DPF as much as possible in stoichiometric operation, the amount of fine dust is preferably always low. Since the above-mentioned conditions are set for the HC discharge amount and the particulate matter discharge amount in the stoichiometric operation, an upper limit depending on the HC oxidation performance of the engine side catalyst at the respective time point is set for the post injection quantity and the post injection time. In other words, depending on the HC oxidation performance of the engine-side catalyst, the post-injection amount and the post-injection timing may be adjusted to bring the HC discharge amount below a certain upper limit value and minimize the particulate matter discharge amount. Next, a simple description will be made of the procedure for determining the post-injection amount and the post-injection timing taking this point into consideration (steps 1 to 3).

In step 1, a parameter indicative of the HC oxidation performance of the engine side catalyst is calculated (hereinafter, the oxidation characteristic parameter Pox). The factors for determining the HC oxidation performance of the engine bottom catalyst are roughly divided into the environment in which the engine bottom side catalyst is used and the deterioration degree and individual deviation of the engine bottom catalyst. In step 1, the oxidation characteristic parameter is numerically represented by two factors, namely, a base factor (hereinafter Pox_bs) determined by the environment of use of the engine bottom catalyst and a deterioration factor determined by the deterioration degree and individual variations of the engine bottom catalyst (hereinafter Kmod).

3 FIG. 12 is a view showing the relationship of the temperature on the intake side of the engine lower side catalyst, the temperature on the exhaust side, the carrier temperature of the engine lower side catalyst, the exhaust gas volume, and the like to the HC oxidation performance. As in 3 As shown, as the temperature at the outlet of the engine bottom catalyst increases, the HC oxidation performance of the engine bottom catalyst tends to increase with increasing temperature at the outlet thereof or with increasing carrier temperature. On the other hand, as the exhaust gas volume increases, the HC oxidation performance of the engine bottom catalyst tends to decrease. That is, the temperature at the inlet of the engine bottom catalyst, the temperature at the outlet thereof, the carrier temperature and the exhaust gas volume, and the like can be used as parameters for determining the basic factor. In step 1, these parameters are used as inputs and the base factor is determined by searching the map and certain calculation equations.

Deterioration degree and individual variations of the engine bottom catalyst, in turn, can not be directly represented in numbers. Therefore, attention is paid to the deterioration factor that the deterioration degree, etc., is correlated with various characteristics of the engine bottom catalyst, and determination is made by, for example, one of the methods 1 to 3 below.

For Type 1, the deterioration factor is determined based on the oxidation characteristics of the engine bottom catalyst. When a post-injection is executed, an amount of HC flows into the engine down-side catalyst, which is approximately proportional to the post-injection amount. In the engine bottom catalyst, the incoming HC is oxidized and heat is generated. Therefore, the heat generation coefficient, the represents the degree of contribution of the fuel supplied by the post-injection to the temperature rise of the engine bottom catalyst, can be used as a deterioration factor.

For Type 2, the deterioration factor is determined based on the storage characteristics of the engine bottom catalyst. The engine bottom catalyst has a storage function for storing oxygen or NOx in the exhaust gas. The amount of oxygen or NOx stored in the engine bottom catalyst, or the amount of reductant required to eliminate the oxygen or NOx, can therefore be used as a deterioration factor.

In Type 3, the deterioration factor is determined by setting the air-fuel ratio target value set during the later-described air-fuel ratio post-catalyst feedback control set such that the output value of the O ₂ provided downstream of the engine underside catalyst Sensor is kept at a setpoint, or the actual air-fuel ratio is used.

In step 2, based on the oxidation characteristic parameter calculated in step 1, by searching a map as shown in FIG 4 2, a correction coefficient of the post injection quantity and a correction value of the post injection timing are calculated. In 4 the horizontal axis shows the oxidation characteristic parameter calculated in step 1. The correction coefficient of the post-injection amount on the vertical axis is defined as a positive real number multiplied by the basic injection amount. The correction value of the post-injection timing on the vertical axis is defined as a real number subtracted from a given base time. As in 4 As shown in FIG. 2, the amount of post-injection increases as the HC oxidation performance of the engine bottom catalyst increases. In turn, the post-injection timing is corrected late as the HC oxidation performance of the engine bottom catalyst increases. That is, the distance between main injection and post injection increases with increasing HC oxidation performance of the engine bottom catalyst.

In step 3, the post injection amount and the post injection timing are determined based on the correction coefficient and correction value calculated in step 2. More specifically, the final post-injection amount is calculated by multiplying a predetermined basic injection amount by the correction coefficient. The final post-injection time, in turn, is calculated by subtracting the correction value from a particular base time. In this way, according to the HC oxidation performance of the engine side catalyst, post-injection amount and post-injection timing in stoichiometric operation can be adjusted so that the HC discharge amount is kept below a certain upper limit value and the fine dust discharge amount is kept as small as possible.

Next, referring to 5 to 29 a concrete procedure of the fuel injection control will be described. 5 FIG. 14 is a main flowchart illustrating the concrete procedure of the fuel injection control that determines the fuel injection state of the fuel injection valves of the individual cylinders. In the 5 The processing shown is executed by the ECU in synchronism with an OT timing of the individual cylinders per combustion cycle. In the following, values which are updated or scanned in the ECU in OT synchronism are marked with a "k" in brackets.

In S1, by searching a predetermined map (not shown) in accordance with the operating state of the engine, a basic fuel injection amount Gfuel_bs (k) is determined, whereupon a transition to S2 occurs. This basic fuel injection amount corresponds to the lean fuel injection amount (see step S13 below). As will be described in detail below, in stoichiometric driving, the fuel base injection amount is multiplied by an air-fuel ratio correction coefficient KAF (k) determined by feedback control based on the outputs of the pre-catalyst LAF sensor and the O ₂ sensor Catalyst is determined (see S8, described later). As input parameters representing the operating state of the engine and used for determining the basic fuel injection amount, for example, the torque requested by the driver, the engine speed, and the like can be cited.

In S2, it is judged whether or not the devices are normal in the context of the fuel injection control. The devices associated with the judgment of S2 are, for example, intake throttle and EGR valve (not shown), the pre-catalyst LAF sensor required for stoichiometric driving, the post-catalyst O ₂ sensor, the temperature sensor, and the like. If the judgment of S2 is "Yes" (the devices are normal), a transition is made to S3, and if if it is "No" (the devices are abnormal), a transition is made to S13 and lean driving is performed.

In S3, it is judged whether or not the engine underside catalyst is in the activation state. More specifically, in S3, an estimated value of the carrier temperature of the engine bottom catalyst is calculated, and when this estimated value is equal to or greater than a certain activation temperature (for example, 200 ° C), it is judged that the activation state exists while otherwise judging that the activation state is not present , The concrete procedure for calculating the estimated value of the carrier temperature of the engine bottom catalyst becomes S51 10 described. If the judgment of S3 is "Yes", a transition to S5 occurs, and if it is "No", a transition to S13 and a lean running operation is made. If the engine underside catalyst is not in the activated state, it may happen that even in the stoichiometric driving operation can not be achieved sufficient exhaust gas cleaning effect.

In S5, processing for judging the stoichiometric operating conditions is performed to judge whether or not the stoichiometric driving is performed, and a transition is made to S6. As later referring to 6 is described in this processing for judging the stoichiometric operating conditions on the basis of the operating condition of the engine, the condition in the exhaust passage, the condition of the engine underside catalyst or the like, whether the stoichiometric driving operation is performed or not. As a result, it is judged that one for carrying out If the state of stoichiometric driving is appropriate, a stoichiometric operation mode flag F_Stoic_mode (k) indicating this is set to "1", otherwise the flag F_Stoic_mode (k) is set to "0".

In S6, it is judged whether or not the stoichiometric operation mode flag F_Stoic_mode (k) is set to "1". When the judgment of S6 is "Yes", a transition to S7 is made, and the stoichiometric driving operation is performed, and if it is "No", a transition to S13 is made, and a lean running operation is performed. In the flow chart 5 corresponds to the processing of S7 to S12 of the fuel injection control in the stoichiometric driving mode, and the processing of S13 to S14 corresponds to the fuel injection control in the lean driving mode.

In S7, the air-fuel ratio feedback calculation described later is performed before the catalyst, and a transition is made to S8. As later with reference to 9 in detail, in the pre-catalyst air-fuel ratio feedback calculation, an air-fuel ratio correction coefficient KAF (k) is calculated based on the output of the LAF sensor upstream of the catalyst to determine the air-fuel ratio in or near the stoichiometric range (stoichiometric or low fat).

By multiplying the fuel base injection amount Gfuel_bs (k) obtained in S1 by the air-fuel ratio correction coefficient KAF (k) in S8, the total fuel injection amount K Gfuel (k) for the stoichiometric driving operation is determined (see Equation 1 below) a transition is made to S9. Here, the "total fuel injection amount" refers to the total amount of fuel supplied to the in-cylinder combustion during one combustion cycle and corresponds to the sum of the fuel injected in pilot injection, main injection, and post injection.

[Equation 1]

• Gfuel (k) = KAF (k) Gfuel_bs (k) (1)

In S9, the post-injection amount Gfuel_aft (k) and the post-injection time Θ_aft (k) are calculated, and a transition to S10 is made. The concrete procedure of calculating the post-injection amount and the post-injection timing will be described later with reference to FIG 10 described.

In S10, the fuel amount Gfuel_pi (k) (hereinafter, "pilot injection amount") supplied in the pilot injection, the timing for executing the pilot injection Θ_pi (k) (hereinafter, "pilot injection timing"), and the timing for executing the main injection Θ_main (k) ( hereinafter "main injection timing"), and a transition to S11 occurs. The pilot injection amount Gfuel_pi (k), the pilot injection timing Θ_pi (k), and the main injection timing Θ_main (k) include the engine speed and load parameters (eg, the effective mean pressure, required torque, fuel injection amount, engine torque estimated value and exhaust gas volume, and others) Parameters that increase in proportion to engine load) and are calculated using known methods such as map search.

In S12, by subtracting pilot injection quantity Gfuel_pi (k) and post-injection quantity Gfuel_aft (k) from the fuel injection quantity Gfuel (k), the fuel amount Gfuel_main (hereinafter, "main injection amount") supplied by the main injection is calculated, thus ending this processing.

In S13, the fuel base injection amount Gfuel_bs (k) obtained in S1 is set as the total fuel injection amount Gfuel (k) under the lean running, and a transition is made to S14. In S14, based on an algorithm for lean running, the fuel injection state is determined, thus ending this processing.

6 FIG. 12 is a flowchart showing the concrete flow of stoichiometric operating condition judgment processing for updating the stoichiometric operation mode flag F_Stoic_mode. In other words 6 a flowchart of the determination of whether stoichiometric driving or lean driving is to be performed. In the 6 Processing shown becomes a subroutine of the main processing 5 with the same synchronization (OT synchronization).

In S21, it is judged whether certain engine underside catalyst protection conditions set for protecting the engine bottom catalyst from heat are met. In stoichiometric driving, the exhaust gas temperature rises, and also the carrier temperature of the catalyst in the exhaust passage increases. Since the engine underside catalyst is close to the engine, its temperature increase in stoichiometric driving is large. Engine bottom catalyst protection conditions are conditions that are set to prevent deterioration of the engine bottom catalyst. More specifically, in S21, an estimated value of the carrier temperature of the engine bottom catalyst is calculated, and if this estimated value is below a certain catalyst protection temperature set to, for example, about 630 to 700 ° C, it is judged that the protection conditions are met, otherwise judged that the Protection conditions are not satisfied If the judgment in S21 is "No", a transition is made to S22, and a flag F_Stoic_mode showing that the stoichiometric driving operation is prohibited is set to "0", followed by a return to S6 in FIG 5 he follows. When the judgment in step S21 is "Yes", a flow goes to step S23.

In S23, it is judged whether the LNT is in a state where the DeNOx function can be sufficiently achieved. A "state in which the DeNOx function can be sufficiently achieved" is a state in which, in the context of the exhaust gas purification system 1 using an LNT as the engine bottom catalyst, in the presence of reducing agent injected by the exhaust fuel injection nozzle, no components not required by the LNT (for example, the above-mentioned N ₂ O, etc.) are discharged, and the NOx removal can be performed with appropriate efficiency. More specifically, in S23, it is judged that the LNT can sufficiently achieve its DeNOx function when the estimated value of the LNT carrier temperature is equal to or higher than a certain cleaning temperature set at 350 to 400 ° C, for example. For example, the judgment in S23 is "No" when the engine is in the warm-up process immediately after the start, or when the LNT temperature decreases in the extra-urban driving operation.

When the judgment in S23 is "NO", that is, when the amount of NOx that can be removed using the DeNOx function of the engine bottom catalyst is small, a transition is made to S24, based on a map for the three-way purge operation mode in which the focus of the exhaust gas purification is not on the DeNOx function but on the three-way cleaning function, the stoichiometric operating mode flag F_Stoic_mode (k) is updated. More specifically, the engine speed and engine load parameters (eg, the effective mean pressure) are detected, and from these input parameters, as in FIG 7 3, the value of the flag F_Stoic_mode (k) is determined by searching the map for the three-way cleaning operation mode. As in 7 Shown by the broken line, the operating state of the engine can be roughly divided into four areas, wherein in a low-speed and high-load area, a high-speed and low-load area and a high-speed and high-load area in which the The amount of NOx discharged from the engine and flowing into the engine side catalyst is high, the stoichiometric driving operation is selected (F_Stoic_mode ← 1), while in a low-speed, low-load region in which the amount of NOx flowing into the engine bottom catalyst is small lean driving mode is selected (F_Stoic_mode ← 0).

When the judgment in S23 is "Yes", that is, when the amount of NOx that can be removed using the DeNOx function of the engine bottom catalyst is large, a transition is made to S25 and a map for the sharing operation mode of DeNOx function and three-way cleaning function, the stoichiometric operating mode flag F_Stoic_mode (k) is updated. More specifically, the engine speed and engine load parameters (eg, the effective mean pressure) are detected, and from these input parameters, as in FIG 8th The value of the flag F_Stoic_mode (k) is determined by searching the shared operation mode map. A comparison of the map for the three-way cleaning mode of operation 7 and the sharing mode map 8th shows that the area in which the stoichiometric driving operation is selected (F_Stoic_mode ← 1) in the shared operation mode map 8th is smaller. The reason for this is that if the judgment in S23 is "Yes", the NOx amount that can be removed by means of the DeNOx function of the engine bottom catalyst is larger than "No".

A state in which, with respect to the amount of NOx that can be removed per unit time by the DeNOx function of the engine bottom catalyst, the amount of NOx flowing into the engine bottom catalyst per unit time is higher than a certain value is defined as an NOx surplus state , In the flow chart 6 is determined by the processings of S23 to S25 and the schemes 7 and 8th judges whether or not there is a NOx surplus state, and when there is an excess NOx state, the stoichiometric driving operation is selected in which NOx is removed by the three-way purifying function (F_Stoic_mode ← 1), while the lean running mode is selected and NOx is used the DeNOx function is removed (F_Stoic_mode ← 0) when there is no NOx excess state.

9 FIG. 10 is a flowchart showing the concrete flow of the pre-catalyst air-fuel ratio feedback calculation, wherein the stoichiometric air-fuel ratio correction coefficient KAF is determined. FIG. In the 9 Processing shown becomes a subroutine of the main processing 5 with the same synchronization (OT synchronization).

In S31, it is judged whether the LAF sensor before the catalyst has reached the activation or not. When the judgment in S31 is "No", the correction coefficient KAF (k) is set to 1 without the feedback calculation described below (S32), and a return to S8 is made 5 ,

When the judgment in S31 is "Yes", the correction coefficient KAF (k) is determined using a known feedback algorithm such that a deviation E_af (k) between the output value of the LAF sensor before the catalyst AFact (k) and a target value AFcmd (k) (see Equation (2-1) below) becomes 0 (S33), then return to S8 5 he follows. The set point AFcmd (k) for the output of the LAF sensor upstream of the catalytic converter is shown in the post-catalyst air-fuel ratio feedback calculation (see below) 16 below) by re-sampling a value which is updated in a control cycle of about 10 to 50 ms. As an example of the calculation in S33, a calculation formula is shown, wherein in the equations (2-1) to (2-3) below, the correction coefficient KAF (k) is determined using a sliding mode algorithm. In equation (2-2), "Pole_af" is the switching function parameter and is set to a value greater than -1 and less than 0 (for example, -0.65). In Equation (2-3), the two feedback gains "Krch_af" and "Kadp_af" are set to a negative value. Preferably, the pre-catalyst feedback offset error compensation rate is faster than the post-catalyst feedback feedback from S77 described later 16 set. [Equation 2]

10 FIG. 12 is a flowchart showing the concrete procedure for determining the post-injection amount and the post-injection timing in stoichiometric driving. FIG. In the 10 Processing shown becomes a subroutine of the main processing 5 executed with OT synchronization.

In S51, an estimated value of the carrier temperature of the engine side catalyst Tcc_hat (k) and an estimated value of the exhaust gas volume Gex_hat (k) are acquired, and a transition to S52 occurs. The estimated value Tcc_hat (k) of the engine sub-catalyst temperature is calculated as shown in Equation (3), for example, by the output Tup (k) of the sensor of the temperature upstream of the engine bottom catalyst and the output Tds (k) of the sensor of the temperature downstream of the engine bottom catalyst of a certain weighting coefficient Wt (0 ≦ Wt ≦ 1, for example 0.3) are weighted.

[Equation 3]

• Tcc_hat (k) = (1 - Wt) Tup (k) + Wt · Tds (k) (3)

In S52, a standard value Pox_bs (k) of the oxidation characteristic parameter Pox (k) of the engine bottom catalyst is calculated, which represents the HC oxidation ability of the engine bottom catalyst in numbers, and a transition is made to S53. The oxidation characteristic parameter Pox (k) is a product of a later-described standard value Pox_bs (k) which, as shown in Equation (4) below, corresponds to a base factor which depends on the carrier temperature of the engine bottom catalyst and the exhaust gas volume, and Oxidation characteristic correction coefficient Kmod (k) defined, which corresponds to a deterioration factor, which depends on the deterioration and individual deviations of the engine underside catalyst. In S52, from the estimated values Tcc_hat (k), Gex_hat (k) obtained in S51, by searching a predetermined map, a standard value Pox_bs (k) (0 ≦ Pox_bs (k) ≦ 1) is calculated.

11 FIG. 12 shows an example of a map for determining the default value for the oxidation characteristic parameter Pox_bs (k). As in 11 As shown, the HC oxidation performance of the engine bottom catalyst occurs when its carrier temperature exceeds the activation temperature (about 150 to 200 ° C) and increases with increasing temperature. The higher the exhaust gas volume, that is the more exhaust gas per unit of time flows through the engine bottom side catalytic converter, the more the HC oxidation power decreases. The default value Pox_bs (k) is 0 for the case where the engine bottom catalyst does not have HC oxidation ability, and is normalized to take a certain value (for example, 1) when the HC oxidation performance of the engine bottom catalyst is the highest.

Returning to 10 At S53, the oxidation characteristic correction coefficient Kmod (k) is calculated, and a transition is made to S54. As in 10 3, there are three methods of types 1 to 3 for calculating the oxidation characteristic correction coefficient Kmod (k). However, in each method, the correction coefficient Kmod (k) is obtained by resampling a parameter that is updated in a processing executed in a cycle other than the OT synchronously executed present processing in the OT synchronous manner.

In the method of the type 1, an adaptive catalyst oxidation characteristic coefficient Kox which is updated in a cycle tn in a processing which will be described later with reference to 28 is re-sampled, and this value is used as the correction coefficient Kmod and used in the subsequent processing. In the method of type 2, an adaptive catalyst reduction characteristic coefficient Krd updated in a cycle tm is processed as described later with reference to FIG 24 is re-sampled, and this value is used as the correction coefficient Kmod and used in the subsequent processing. In the method of the type 3, an adaptive catalyst three-way characteristic correction value Dtw which is updated in a cycle tm in a processing which will be described later with reference to 22 is re-sampled, and the value thus obtained is multiplied by -1, whereupon 1 is added, and this value is taken as the correction coefficient Kmod (= 1-Dtw) and used in the subsequent processing.

Although the coefficients Kox, Krd, and Dtw are values calculated based on different characteristics of the engine side catalyst in different methods, but all of them change along with the deterioration and individual deviation of the engine bottom catalyst, they can be used as a deterioration factor of the oxidation characteristic parameter Pox become.

By multiplying in S54 the base value Pox_bs (k) and the oxidation characteristic correction coefficient Kmod (k), the oxidation characteristic parameter Pox (k) is calculated (see Equation (4) below), and there is a transition to S55. The post-injection amount and the post-injection timing are finely adjusted based on the thus obtained oxidation characteristic parameter Pox (k).

[Equation 4]

• Pox (k) = Pox_bs (k) * Kmod (k) (4)

In S55, a default value Gfuel_aft_bs (k) of the post injection quantity Gfuel_aft (k) is calculated, and a transition is made to S56. More specifically, the standard value Gfuel_aft_bs (k) is calculated in S55 based on the engine speed and a load parameter by searching a predetermined map.

12 FIG. 15 shows an example of a map for calculating the default value Gfuel_aft_bs (k) of the post injection quantity. As in 12 shown, the Nacheinspritzmenge is increased with increasing engine speed. In addition, the post-injection amount is increased with increasing engine load. The reason for this is that as the load increases, the main injection amount increases, which also increases the amount of particulate matter to be combusted by the post-injection.

In S56, based on the oxidation characteristic parameter Pox (k), a correction coefficient Kg_aft (k) of the post injection quantity Gfuel_aft (k) is calculated, and a transition is made to S57. More specifically, the correction coefficient Kg_aft (k) is calculated in S56 from the oxidation characteristic parameter Pox (k) by searching a predetermined map.

13 FIG. 14 shows an example of a map for determining the correction coefficient Kg_aft (k) for the post-injection amount. The correction coefficient Kg_aft calculated in S56 is used as the multiplicative correction coefficient for the standard value Gfuel_aft_bs of the post injection quantity (see Equation (5) below). As in 13 As shown in FIG. 2, the more the post injection quantity, the oxidation characteristic parameter Pox, that is, the HC oxidation performance of the engine bottom catalyst, increases, the post injection quantity is increased to be corrected. In this way, since the particulate matter discharge amount corresponding to the HC oxidation performance of the engine side catalyst can be reduced as much as possible at each time point, the regeneration interval of the DPF can be prolonged and thus the fuel consumption can be improved. As with reference to 31 described increases with increasing Nacheinspritzmenge the flowing through the engine underside catalyst HC amount. However, by increasing the post-injection amount according to the oxidation characteristic parameter Pox, the amount of HC discharged downstream of the engine bottom catalyst can be kept below a certain upper limit.

In S57, the standard value Θ_aft_bs (k) of the post injection timing Θ_aft (k) is calculated based on the oxidation characteristic parameter Pox (k), and a transition to S58 occurs. More concretely, the standard value Θ_aft_bs (k) is calculated in S57 based on the engine speed and a load parameter by searching a predetermined map.

14 FIG. 14 shows an example of a map for calculating the standard value Θ_aft_bs (k) (degree AOT) of the post injection timing. As in 14 shown, the default value of Nacheinspritzzeitpunkts delayed with increasing engine speed. In addition, the post injection timing is delayed with increasing engine load. The reason for this is that as the load increases, the main injection amount increases, which also increases the amount of particulate matter to be combusted by the post-injection.

In S58, the correction value D_aft (k) of the post injection timing Θ_aft (k) is calculated based on the oxidation characteristic parameter Pox, and a transition to S59 occurs. More specifically, the correction value D_aft (k) is calculated in S58 from the oxidation characteristic parameter Pox (k) by searching a predetermined map.

15 FIG. 14 shows an example of a map for determining the correction value D_aft (k) of the post-injection timing. The last post-injection timing Θ_aft is calculated by subtracting the correction value D_aft (k) calculated in S58 from the standard value Θ_aft_bs (see Equation (6) below). As in 15 For the post-injection time point, the last post-injection time Θ_aft is delayed-corrected, the greater the oxidation-parameter parameter Pox becomes, that is, the more the HC-oxidation power increases. The main injection timing, unlike the post-injection timing, is not corrected by the oxidation characteristic parameter Pox. Delaying correction of the post injection quantity, the larger the oxidation characteristic parameter Pox becomes, by the correction value D_aft, as in FIG 15 is thus equivalent to increasing the distance between main injection time and post-injection time, the larger the Oxidizing characteristic parameter Pox becomes. In this way, since the particulate matter discharge amount corresponding to the HC oxidation performance of the engine side catalyst can be reduced as much as possible, the regeneration interval of the DPF can be lengthened to improve the fuel consumption. As with reference to 31 described, increases with a greater distance of the post-injection, the amount of HC flowing through the engine underside catalyst. However, by increasing the distance of the post injection according to the oxidation characteristic parameter Pox, the amount of HC discharged downstream of the engine bottom catalyst can be kept below a certain upper limit.

Returning to 10 at S59, by multiplying the standard value Gfuel_aft (k) and the correction coefficient Kg_aft (k), the post-injection amount Gfuel_aft (k) is calculated (see Equation (5) below), and a transition is made to S60.

[Equation 5]

• Gfuel_aft (k) = Kg_aft (k) · Gfuel_aft (k) (5)

In S60, the correction value D_aft (k) is subtracted from the standard value Θ_aft_bs (k), whereby the post injection timing Θ_aft (k) is calculated (see Equation (6) below), whereupon a return to S10 is made 5 he follows.

[Equation 6]

• θ_aft (k) = θ_aft_bs (k) - D_aft (k) (6)

16 FIG. 12 is a flowchart illustrating the concrete procedure of the post-catalyst air-fuel ratio feedback calculation to determine the target AFcmd of the LAF sensor output AFact before the catalyst. In the 16 The processing shown is executed by the ECU in a certain control cycle tm (10 to 50 ms). In the following, values which are updated or sampled in the cycle tm are provided with an "m" in brackets.

In S71, it is judged whether the O ₂ sensor after the catalyst has reached the activation or not. If the judgment in S71 is "No", the feedback calculation below is not performed, and the target value AFcmd (m) is taken as a certain standard value AFcmd_bs (a fixed value of, for example, 14.5) (S72), thus ending this processing. When the judgment in step S71 is "Yes", a flow goes to step S73.

In S73, it is judged whether or not the stoichiometric operation mode flag F_Stoic_mode (m) is set to "1". If the judgment in S73 is "Yes", it goes to S74, and if it is "No", it goes to S72, and as described above, Fcmd (m) = AFcmd_bs.

In S74, it is judged whether or not a later-described reduction-treatment-end flag F_CRD_Done (m) is "1". As described above, as a result of the stoichiometric operating mode flag F_Stoic_mode (m) changing from "0" to "1" during lean running, the stoichiometric driving operation is initiated. However, since the meager driving has taken place until then, the engine bottom catalyst has stored a lot of oxygen, and therefore, the three-way cleaning function of the engine bottom catalyst can not be put to good use despite start of the stoichiometric driving operation. Immediately after changing the flag F_Stoic_mode (m) from "0" to "1", therefore, instead of stoichiometrically, the air-fuel ratio is slightly regulated in the direction of rich ("weakly rich"), and a reduction treatment is performed to obtain the Engine bottom catalyst to release stored oxygen within a short time. A reduction treatment end flag F_CRD_Done (m) indicates that the reduction treatment is completed immediately after the start of the stoichiometric driving operation, and will be explained later by referring to FIG 20 updated processing for the engine underside catalyst reduction characteristic judgment. Hereinafter, the driving mode immediately after the start of the stoichiometric driving operation that promotes the reduction of the engine bottom catalyst will be referred to as a "weak rich operation mode". In addition, the driving operation mode in which the target air-fuel ratio AFcmd (m) is determined in the stoichiometric operation based on the output of the O ₂ sensor after the catalyst is called the "post-catalyst air-fuel feedback mode".

When the judgment in S74 is "No", a transition is made to S75, and the target air-fuel ratio AFcmd (m) in the low-rich operating mode is determined. More specifically, in S75, the estimated value Tcc_hat (m) of the carrier temperature of the engine side catalyst and the estimated value Gex_hat (m) of the exhaust gas volume are obtained, and on the basis of these two estimated values Tcc_hat (m) and Gex_hat (m), by searching a predetermined map, the air mass is obtained. Target fuel ratio AFcmd (m) determines what this processing ends.

17 FIG. 14 shows an example of a map for determining the target air-fuel ratio in the low-rich operating mode. As in 17 2, the target air-fuel ratio AFcmd is in the low-rich operating mode within the low-rich region (about 14.5 to 13.5) and is set to a value corresponding to the estimated value Tcc_hat of the engine bottom catalyst temperature and the exhaust gas volume estimated value Gex_hat. More specifically, the more the carrier temperature of the engine lower side catalyst increases and the more the exhaust gas volume decreases, the more the target air-fuel ratio AFcmd is further regulated.

If returning to 16 when the judgment in S74 is "Yes", a transition is made to S76, and the target air-fuel ratio AF_cmd (m) is determined in the air-fuel feedback mode after the catalyst. In S76, the estimated value Tcc_hat (m) of the carrier temperature of the engine side catalyst and the estimated value Gex_hat (m) of the exhaust gas volume are obtained, and based on these two estimated values Tcc_hat (m) and Gex_hat (m), a target value Vcmd (m ) is determined for the output Vout (m) of the O ₂ sensor after the catalyst, and a transition is made to S77.

18 FIG. 14 shows an example of a map for determining the output target value Vcmd (m) of the O ₂ sensor after the catalyst. As in 18 ₂ , the set value Vcmd of the O ₂ sensor is in a range farther toward rich than a reverse judgment threshold VIn which judges the reversal of the output of the O ₂ sensor (for example, 0.1), and is further corrected for rich the more the carrier temperature of the engine bottom catalyst increases. The more the exhaust gas volume increases (or, in other words, the higher the load), the more the amount of NOx discharged from the engine increases, and the more the exhaust gas flow rate through the LNT increases, as a result, the NOx removal rate of the engine underside catalyst decreases. The set point Vcmd (m) of the O ₂ sensor that is to compensate for this reduction in the NOx removal rate becomes, as in 18 The larger the exhaust gas volume increases, the more the exhaust gas volume increases, so that the generation of reducing agents such as CO, H ₂ , NH ₃ and the like increases at the engine underside catalyst.

Returning to 16 , in S77, the target air-fuel ratio AFcmd (m) is determined using a known feedback algorithm such that a deviation E_v (m) between the output Vout (m) of the O ₂ sensor after the catalyst and its set value Vcmd (m ) (see equation (7-1) below) becomes "0", and a transition is made to S78. As an example of the calculation in S77, a calculation formula is shown, and in the equations (7-1) to (7-3) below, the target air-fuel ratio AFcmd (m) is determined using a sliding mode algorithm. In equation (7-2), "Pole_af" is the switching function parameter and is set to a value greater than -1 and less than 0 (for example, -0.85). In Equation (7-3), the two feedback gains "Krch_v" and "Kadp_v" are set to a negative value. [Equation 7]

In S78, a later reference is made to 22 described adaptive calculation for the three-way catalyst characteristic executed, with which this processing ends. As described above, in the air-fuel feedback mode after the catalyst, based on the output of the O ₂ sensor after the catalyst, the target air-fuel ratio AFcmd is determined with respect to the output of the LAF sensor before the catalyst. Therefore, the target air-fuel ratio AFcmd reflects the state of the engine bottom catalyst. In the adaptive calculation for the three-way catalyst characteristic, by using this characteristic, by performing statistical processing on the target air-fuel ratio AFcmd set in the air-fuel feedback mode after the catalyst calculates adaptive catalyst three-way characteristic correction value Dtw, which corresponds to the deterioration factor of the HC oxidation performance of the engine bottom catalyst. The adaptive calculation for the three-way catalyst characteristic of S78 is omitted when in S53 10 the oxidation characteristic correction coefficient Kmod is determined in a method other than Type 3.

19 11 is a time chart showing a concrete example of the change of the output Afact of the LAF sensor before the catalyst and the output Vout of the O ₂ sensor after the catalyst in determining the target air-fuel ratio AFcmd by the feedback calculation for the air-fuel Ratio according to the catalyst. 19 FIG. 12 shows a case where the stoichiometric operation mode flag F_Stoic_mode changes from "0" to "1" at a time t1.

As with reference to 16 is described, the target air-fuel ratio AFcmd immediately after the start of stoichiometric driving (in 19 Time t1) is set to low-fat (see p75) 16 ), whereby the oxygen stored in the engine bottom catalyst is released and used for the oxidation of the reducing agent, which is supplied to low-fat due to the control. At a time point t2, when it is judged that the reversal of the output Vout of the O ₂ sensor has taken place, the reduction treatment end flag F_CRD_Done is switched from "0" to "1" (see FIG 20 below), and the low-fat operating mode (tilting toward rich) is canceled. After the time t2, the target air-fuel ratio AFcmd is determined such that the output Vout of the O ₂ sensor after the catalyst reaches the predetermined target value Vcmd in accordance with the running mode.

20 FIG. 12 is a flowchart showing the concrete procedure of processing for the engine bottom catalyst reduction characteristic judgment for updating the reduction treatment end flag F_CRD_Done. In the 20 The processing shown is executed by the ECU in the same control cycle tm (10 to 50 ms) as the post-catalyst air-fuel ratio feedback calculation 16 executed. In the engine bottom catalyst reduction characteristic judgment processing 20 is estimated by the amount of reducing agent, which is supplied to the engine underside catalyst, the reduction treatment end flag F_CRD_Done updated. Hereinafter, the concrete procedure of processing for the engine underside catalyst reduction characteristic judgment will be made with reference to the timing chart 21 described.

In S81, it is judged whether or not the stoichiometric operation mode flag F_Stoic_mode (m) is set to "1". When the judgment in S81 is "No", that is, when the lean driving mode is present, a transition is made to S82, and when the judgment in S81 is "Yes", that is, when stoichiometric driving is performed, a transition is made to S86.

In S82, an estimated value Rd_hat (m) of the amount of reducing agent supplied to the engine bottom side catalyst during the control cycle tm (hereinafter, "current reductant amount estimated value"), a provisional value Rd_hat_tmp (m) of the current reductant amount estimated value, and the product Crd_hat (m) of the current reductant amount estimated value (hereinafter "Reducing agent supply amount estimated value") are reset to 0, respectively, and a transition is made to S83. In lean driving operation, the engine bottom catalyst is fed virtually no reducing agent. In S83, a catalyst reduction characteristic update completion flag F_CrdAdp_done (m) is set to 0, and a transition is made to S84. This flag F_crdAdp_done (m) serves to indicate that an updating of a parameter of the reduction characteristic of the engine side catalyst has been completed by an adaptive calculation of the catalyst reduction characteristic described later in S92. In S84, the reduction-treatment-end flag F_CRD_Done (m) is reset to 0, thus ending this processing.

In S86, based on the output AFact (m) of the LAF sensor in front of the catalyst, the current reducing agent amount estimated value Rd_hat (m) is calculated, and a transition is made to S87. More specifically, when the output value AFact (m) of the LAF sensor before the catalyst falls below the standard value AFcmd_bs, a value obtained by multiplying this excess by the exhaust gas volume estimated value Gex_hat (m) is obtained as the current reducing agent amount estimated value Rd_h (m ). Concretely, the provisional value Rd_hat_tmp (m) is introduced as shown in Equations (8-1) and (8-2) below. [Equation 8]

In S87, it is judged whether or not the output value Vout (m) of the O ₂ sensor after the catalyst is smaller than the inverse judgment threshold value V1n. As with reference to 16 to 19 is described, immediately after the start of the stoichiometric driving operation, the air-fuel ratio is set to low-rich, and the stored oxygen in the engine lower side catalyst is released downstream. Whether the reduction treatment is completed or not can thus be judged from whether or not the output value Vout (m) of the O ₂ sensor after the catalyst has exceeded the inversion judgment threshold value V1n. When the judgment in S87 is "Yes", a transition is made to S88, and by adding the current reducing agent amount estimation value Rd_hat (m), the reducing agent supply amount estimation value Crd_hat (m) is calculated (see Equations (9) and 21 below), with which this processing ends.

[Equation 9]

• Crd_hat (m) = Crd_hat (m - 1) + Rd_hat (m) (9)

When the judgment in S87 is "No", the reduction treatment end flag F_CRD_Done (m) indicating that the reduction treatment has been completed immediately after the start of the stoichiometric driving operation is set to 1 (S90), and a transition is made to S91. In this way, in the feedback calculation for the air-fuel ratio after the catalyst, it is changed from the weak rich operation mode to the catalyst air-fuel ratio feedback operation mode (see S74 in FIG 16 ).

In S91, it is judged whether or not the catalyst reduction characteristic update completion flag F_CrdAdp_done (m) is "1". If the judgment in S91 is "No", the later will be referred to 24 described adaptive calculation of the catalyst reduction characteristic is performed (S92), with which this processing ends. If the judgment in S91 is "Yes" and the adaptive calculation of the catalyst reduction characteristic has been completed, this processing ends immediately. In this way, the air-fuel ratio is controlled to low-rich immediately after the start of stoichiometric driving until the output of the O ₂ sensor reverses after the catalyst. Meanwhile, therefore, the condition of the engine underside catalyst is reflected in the estimated value Crd_hat of the amount of reducing agent supplied to the engine underside catalyst. In the adaptive calculation of the catalyst reduction characteristic, using this characteristic, by performing statistical processing on the reduced-agent supply amount estimated value Crd_hat calculated in the low-rich operation mode, an adaptive catalyst reduction characteristic coefficient Krd corresponding to the deterioration factor of the HC oxidation performance of the engine bottom catalyst is calculated. The adaptive calculation for the catalyst reduction characteristic from S92 is omitted when in S53 10 the oxidation characteristic correction coefficient Kmod is determined in a method other than Type 2.

22 FIG. 12 is a flowchart showing the concrete procedure of calculating the adaptive catalyst three-way characteristic correction value Dtw. The processing off 22 is used by the ECU as a subroutine of the 16 shown feedback calculation after the catalyst in the air-fuel feedback mode after the catalyst in the cycle tm.

In the adaptive calculation for the three-way catalyst characteristic, by performing statistical processing on the target air-fuel ratio AFcmd, which is in the stoichiometric feedback operation mode by means of the post-catalyst feedback calculation (see FIG 16 ) which calculates adaptive catalyst three-way characteristic correction value Dtw corresponding to the deterioration factor. This correction value Dtw corresponds to the deviation from the standard value AFcmd_bs of the target air-fuel ratio AFcmd to maintain the output value Vout of the O ₂ sensor after the catalyst in the stoichiometric feedback operation mode at the target value Vcmd. A state in which the output value Vout of the O ₂ sensor after the catalyst is maintained at the target value Vcmd that is greater than the inverse judgment threshold value V1n corresponds to a state in which downstream of the engine underside catalyst a slight slip of reducing agent corresponding to the target value Vcmd is present. At this time, if the oxidizing ability or reducibility of the engine bottom catalyst decreases, the amount of reducing agent to be supplied to the engine bottom catalyst also decreases to maintain the output Vout at the target value Vcmd, thus calculating the target air-fuel ratio AFcmd calculated by the feedback calculation is shifted towards lean.

However, the displacement amount of the target air-fuel ratio (AFcmd-AFcmd_bs) may be different depending on, for example, the carrier temperature of the engine bottom catalyst. In other words, even if the degree of deterioration of the engine underside catalyst remains the same in the case that the carrier temperature of the engine bottom catalyst is high, the displacement amount is higher than that in the case where it is low. In addition, it is by no means the case that the displacement amount uniformly decreases over the entire temperature range in accordance with the degree of deterioration of the engine bottom catalyst. Thus, instead of using the displacement amount as a deterioration factor that is directly proportional to the degree of deterioration, the temperature dependency is preferably eliminated by performing the statistical processing described below. In this calculation, for eliminating the temperature dependency of the displacement amount, a plurality of weighting functions Wtw_i defined on a one-dimensional line with the carrier temperature as a basis and a local adaptation coefficient Dtw_i following the individual weighting functions are applied, with the aid of which weighted statistical processing is performed calculate the adaptive catalyst three-way characteristic correction value Dtw.

In S151, the estimated value Tcc_hat (m) of the carrier temperature of the engine side catalyst is obtained, and from this estimated value Tcc_hat (m), by searching a predetermined map, the individual weighting functions for the three-way catalyst characteristic value Wtw_i (m) (where i is a positive integer ) and a transition is made to S152.

23 FIG. 12 is a view showing an example of a map for calculating the three-way catalyst characteristic value, that is, the shape of a weighting function for the three-way catalyst characteristic Wtw_i. Hereinafter, the description will be made for the case i = 1, 2, 3, that is, for three weighting functions, but the present invention is not limited thereto. Also, a number of four weighting functions or more can be easily generalized.

As in 23 1, the first to third weighting functions are defined with respect to the estimated value Tcc_hat of the carrier temperature of the engine bottom catalyst. The domain of the individual weighting functions (that is, the range in which the weighting function values are not 0) is defined as the first to third ranges, respectively. In the example off 23 For example, the first region is about 300 ° C or more, the second region is about 150 to 450 ° C or more, and the third region is about 300 ° C or more. The individual areas are set so that they overlap each other. In the example off 23 Overlay the first area and the second area at 150 to 300 ° C and the second area and the third area at about 300 to 450 ° C.

The magnitude of the weighting functions Wtw_i is normalized such that the sum of all weighting function values at each temperature is always 1. This is achieved by setting an area that does not overlap any other area as 1, while defining areas that overlap another area with a monotone increasing function or a monotone decreasing function. In the example off 23 is the first weighting function Wtw_1 at about 150 ° C or less 1, and is defined between about 150 ° C to about 300 ° C as a monotonically decreasing function from 1 to 0. The second weighting function Wtw_2 is defined between about 150 ° C to about 300 ° C as a monotonically increasing function from 0 to 1 and between about 300 ° C and about 450 ° C as a monotonically decreasing function from 1 to 0. The third weighting function Wtw_3, in turn, is defined between about 300 ° C and about 450 ° C as a monotonically increasing function from 0 to 1 and above about 300 ° C and above as 1.

Returning to 22 at S152, as shown in Equation (10) below, the sum of the products of the weighting function values Wtw_i (m) and the local adaptation coefficients Dtw_i (m) calculated in S151 is calculated and calculated as the three-way adaptive characteristic correction value Dtw (m ). The local adaptation coefficient Dtw_i has as output value 0 and is calculated, for example, by means of integral calculus, such that an adaptive error signal E_adp '(m) described later becomes 0. [Equation 10]

By calculating the standard value AFcmd_bs and the adaptive catalyst three-way characteristic correction value Dtw (m), the adaptive air-fuel target ratio AFcmd_adp (m) is calculated in S153 (see Equation (11) below).

[Equation 11]

• AFcmd_adp (m) = AFcmd_bs + Dtw (m) (11)

By subtracting the target adaptive air-fuel ratio AFcmd_adp (m) from the target air-fuel ratio AF_cmd (m) in S154, the adaptive error signal Eadp '(m) is calculated (see Equation (12-1) below) and the adaptive error signal Eadp '(m) is distributed to the individual areas, a local adaptive error signal E_adp'_i (m) is also calculated. More specifically, a value resulting from multiplying the adaptive error signal Eadp '(m) by the individual weighting function values Wtw_i (m) is taken as the local adaptive error signal E_adp'_i (m) (see Equation (12-2) below). ,

[Equation 12]

• E_adp '(m) = AF_cmd (m) - AFcmd_adp (m) (12-1) E_adp'_i (m) = Wtw_i (m) * E_adp '(m) (12-2)

In S155, the adaptive local catalyst three-way characteristic correction value Dtw_i (m) is calculated by, as shown in, for example, equation (13) below, a value obtained by substituting the local adaptive error signal E_adp'_i (m ) is multiplied by a negative adaptive gain Kadp_t, such that the local adaptive error signal E_adp'_i (m) calculated for the individual regions becomes 0.

[Equation 13]

• Dtw_i (m) = Dtw_i (m - 1) + Kadp_t · E_adp'_i (m) (13)

In the adaptive calculation for the three-way catalyst characteristic, first to third regions are defined superimposed on the one-dimensional straight line with the carrier temperature of the engine bottom catalyst, and for the individual regions, the weighting functions Wtw_i are defined, but the present invention not limited to this. For example, a plurality of regions superimposed in a two-dimensional plane with the base temperature of the engine bottom catalyst may be defined, and weighting functions Wtw_ij may be defined for the regions in the two-dimensional plane.

24 FIG. 13 is a flowchart showing the concrete procedure of the catalyst reduction characteristic adaptive calculation for calculating an adaptive catalyst reduction characteristic coefficient Krd. The processing off 24 is output from the ECU as a subroutine of the engine bottom catalyst reduction characteristic judgment processing 20 executed in the control cycle tm, whenever it is changed from the weak rich operating mode in the air-fuel feedback mode after the catalyst.

In the calculation of the adaptive catalyst reduction characteristic coefficient, an adaptive catalyst reduction characteristic correction coefficient Krd is calculated from the estimated value Crd_hat (m) of the reducing agent amount supplied to the engine underside catalytic converter from the start of the stoichiometric driving operation until the reversal of the output of the O ₂ sensor (m), which corresponds to the deterioration factor. For example, it is conceivable that at the same time decreases the storage function of the engine underside catalyst with decreasing oxidizing ability or reducing ability of the engine underside catalyst. Therefore, it is considered that, until the output of the O ₂ sensor after the catalyst is reversed, the amount of reducing agent supplied to the engine bottom catalyst decreases along with the oxidizing ability or reducing ability.

However, this reducing agent supply amount is as described with reference to FIG 25 described, for example, depending on the carrier temperature of the engine bottom catalyst differently. In addition, it is by no means the case that the reducing agent supply amount decreases uniformly over the entire temperature range in accordance with the degree of deterioration of the engine bottom catalyst. Thus, instead of taking the reducing agent supply amount estimation value Crd_hat as a deterioration factor that is directly proportional to the deterioration degree, the temperature dependency is preferably eliminated by performing the statistical processing described below. In this calculation, as well as in referring to 22 described adaptive calculation for the three-way catalyst characteristic for eliminating the temperature dependence of the estimated value Crd_hat has a plurality of weighting functions Wrd_i, which are defined on a one-dimensional line with the carrier temperature as a basis, and a following the individual weighting functions local adaptation coefficient Krd_i introduced by their statistical Processing with weighting is performed to calculate the adaptive catalyst reduction characteristic correction value Krd.

In S101, the estimated value Tcc_hat (m) of the carrier temperature of the engine side catalyst is calculated, and based on the estimated value Tcc_hat (m), a standard reducing agent supply amount Crd_bs (m) which is a comparison object of the reducing agent supply amount estimated value Crd_hat (m) is calculated by searching a predetermined map.

25 FIG. 16 shows an example of a map for determining the standard reducing agent supply amount Crd_bs (m). As in 25 As shown in FIG. 4, the standard reducing agent supply amount Crd_bs (m) is set to increase as the carrier temperature of the engine bottom catalyst increases. The reason for this is that as the support temperature increases, the oxidizing ability or reducing capability of the engine bottom catalyst increases, thereby increasing the amount of reductant required until the output of the O ₂ sensor after the catalyst is reversed.

Returning to 24 At S102, by calculating a predetermined map, the individual reducing agent supply amount weighting function values Wrd_i (m) are calculated based on the estimated value Tcc_hat (m) of the carrier temperature of the engine side catalyst, and a flow goes to S103.

26 11 is a view showing an example of a map for calculating the reducing agent supply amount weighting function value, that is, the shape of the reducing agent supply amount weighting function Wrd_i. The shape of the reducing agent supply amount weighting function Wrd_i is the same as that of referring to FIG 23 described weighting function Wtw_i, which is why a detailed description is omitted.

Returning to 24 At S103, as shown in Equation (14) below, the sum of the products of the weighting function values Wrd_i (m) and the local adaptation coefficients Krd_i (m) calculated in S102 is calculated and used as the adaptive catalyst reduction characteristic coefficient Krd (m). The output value of the local adaptation coefficient Krd_i is 1 [Equation 14]

In S104, by multiplying the standard reducing agent supply amount Crd_bs (m) and the adaptive catalyst reduction characteristic coefficient Krd (m) calculated in S101, an adaptive reducing agent supply amount Crd_adp (m) is calculated (see Equation (15) below).

[Equation 15]

• Crd_adp (m) = Krd (m) · Crd_bs (m) (15)

In S105, from that in S88 20 calculated adaptive value Crd_adp (m), the adaptive error signal Eadp (m) is calculated (see equation (16-1) below), and the adaptive error signal Eadp (m) is distributed among the individual regions In addition, a local adaptive error signal E_adp_i (m) is calculated (see equation (16-2) below).

[Equation 16]

• E_adp (m) = Crd_hat (m) - Crd_adp (m) (16-1) E_adp_i (m) = Wrd_i (m) * E_adp (m) (16-2)

In S106, the local adaptation coefficient Krd_i (m) is calculated by multiplying, as shown in the equation (17) below, for example, a value obtained by multiplying the local adaptive error signal E_adp_i (m) by a negative adaptive gain Kadp_c. is integrated, such that the local adaptive error signal E_adp_i (m) calculated for the individual regions becomes 0.

[Equation 17]

• Krd_i (m) = Krd_i (m-1) + Kadp_c · E_adp_i (m) (17)

In S107, it is judged whether the update of the reduction characteristic has been completed. More specifically, in the case where the local adaptive error signal E_adp_i has become smaller than a certain threshold, or if since the start of the processing 24 has passed a certain time, judged that the update processing of the local adaptation coefficient Krd_i indicative of the reduction characteristic of the engine bottom catalyst is completed. When the judgment in S107 is "Yes", the flag F_CrdAdp_done (m) is set to "1" (S108), thus ending this processing. If the judgment in S107 is "No", this processing ends, while the flag F_CrdAdp_done (m) remains at "0", so that the processing is off 24 continues to run.

In the adaptive calculation for the catalyst reduction characteristic, first to third regions superimposed on the one-dimensional straight line with the carrier temperature of the engine bottom catalyst are defined, and for the individual regions, the weighting functions Wrd_i are defined, but the present invention is not limited thereto. For example, a plurality of regions superimposed in a two-dimensional plane with the base temperature of the engine bottom catalyst may be defined, and weighting functions Wrd_ij may be defined for the regions in the two-dimensional plane.

27 FIG. 12 is a flowchart showing the concrete procedure of the sequential identification calculation of a heat model of the engine bottom catalyst. In this sequential identification calculation, several model parameters included in the heat model of the engine bottom catalyst (see equation (18) below) are identified. In the 27 The processing shown is executed by the ECU in a certain control cycle tn (200 to 500 ms). In the following, values which are updated or sampled in the cycle tm are provided with an "n" in brackets. The processing off 27 not applicable if in S53 off 10 the oxidation characteristic correction coefficient Kmod is determined in a method other than Type 1.

Equation (18) is a heat model of the engine bottom catalyst. In other words, the equation (18) is an equation for calculating an estimated value Tds_hat (n) of the exhaust gas temperature downstream of the engine underside catalyst based on a known value. The carrier temperature of the engine side catalyst in the inside of the exhaust pipe (and the temperature of the exhaust gas downstream thereof) changes, besides the heat exchange with the exhaust gas flowing in the exhaust pipe, also by heat exchange with the outside air outside the exhaust pipe. When HC is contained in the exhaust gas flowing into the engine bottom side catalyst, heat is generated at the engine bottom side catalyst by oxidizing the HC.

[Equation 18]

• Tds_hat (n) = Tds_hat (n-1) + a * (Tds_hat (n-1) - Ta (n)) + b (n - 1) · Gex_hat (n) · Tup (n) + c (n - 1) · Gfuel_aft_tm (n) (18)

In Equation (18) below, the first term on the right side is the temperature estimate Tds_hat (n-1) before the control cycle tn, while the second through fourth terms correspond to a respective increase in the temperature from the time before the control cycle tn to the present time , More specifically, the second term on the right side is a heat release term, that is, a term indicating the contribution to the movement of heat between the engine bottom catalyst and the outside air, which is proportional to the difference between the estimated value Tds_hat (n-1). of the Carrier temperature of the engine bottom catalyst and the outside temperature Ta (n) is. The proportionality coefficient a of the second term is a fixed value determined by a system identification executed in advance, and may also be a value which is set and determined according to the output of the downstream temperature sensor.

The third term on the right is a heat transfer term, that is a term that indicates the contribution to the movement of heat between the engine underside catalyst and the outside air and that is proportional to the product of the exhaust gas volume estimate Gex_hat (n) and the output Tup (n) of the upstream temperature sensor. The value of the proportionality coefficient b (n-1) of this heat transfer term is determined by the processing of S123 in FIG 27 updated for every cycle tn.

The fourth term on the right side is a heat generation term, that is, a term indicating the contribution of the combustion of HC in the exhaust gas flowing into the engine bottom side catalyst in the engine bottom side catalyst. By increasing the post-injection amount as described above, the amount of HC in the exhaust gas also increases. Therefore, the heat generation term is proportional to the fuel amount Gfuel_aft_tn (n) supplied by post injection during the cycle tn. The value of the proportionality coefficient c (n-1) of this heat generation term is determined by the processing of S124 in FIG 27 updated for every cycle tn. In addition, the heat generation term increases with increasing post-injection amount, and also increases with increasing HC oxidation performance of the engine bottom catalyst. Therefore, the coefficient c is a parameter showing the contribution of the post-injection fuel to the temperature increase of the engine bottom catalyst, and is a parameter indicating the HC oxidation performance of the engine bottom catalyst. Hereinafter, this coefficient is called a heat generation coefficient. Hereinafter, a concrete procedure for updating the coefficient b and the heat generation coefficient c will be described.

First, in S121, it is judged whether the upstream temperature sensor and the downstream temperature sensor necessary for identifying the coefficients b and c are normal. If the judgment in S121 was "Yes", it goes to S122, and if it is "No", the processing ends 27 immediately.

In S122, it is judged whether or not post-injection has been performed during the cycle tn, that is, whether the post-injection amount Gfuel_aft_tn (n) is 0 or not. When the judgment in S122 is "Yes", a transition is made to S123, and the value of the coefficient b (n) is updated in accordance with equations (19) and (20) below, thus ending this processing. When the judgment in S122 is "No", according to equations (21) and (22) below, the value of the heat generation coefficient c (n) is updated (see S124), the catalytic oxidation characteristic adaptive calculation for calculating the catalytic adaptive correction value Oxidizing characteristic Kox (n) is carried out (see S125 and the later described 28 ), with which this processing ends. That is, if the post injection is not performed, the value of the coefficient b (n) is updated, and only when the post injection is executed, the value of the heat generation coefficient c (n) is updated.

In S123, using a specific parameter identification algorithm on equation (18), the value of the coefficient b (n) is updated while keeping the value of the heat generation coefficient c (n) at the previous value. The concrete procedure of an example of this will be described below. First, based on the output Tds (n) of the downstream temperature sensor, the outside air temperature sensor output Ta (n), the coefficient b (n-1), the exhaust gas volume estimated value Gex_hat (n), and the upstream temperature sensor output Tup (n) a virtual output W (n) and its estimated value W_hat (n) defined in equations (19-1) to (19-3) below.

[Equation 19]

• W (n) = Tds (n) - Tds (n-1) -a (Tds (n-1) + Ta (n)) (19-1) W_hat (n) = b (n-1) Gex_hat (n) · Tup (n) = b (n-1) · ς (n) (19-2) ς (n) = Gex_hat (n) · Tup (n) (19-3)

In this case, when the judgment in S122 is "Yes", the post-injection amount Gfuel_aft_tn (n) = 0, therefore, in the execution of S123, the fourth term on the right side of Equation (18) is 0. The estimated value W_hat (n) defined in equation (19-2) therefore corresponds to one below, as in equation (20) Estimated Virtual Output W (n). Updating the value of the coefficient b such that the difference between the virtual output W (n) defined in equation (19-1) and the estimated value W_hat (n) defined in equation (19-2) is as small as possible corresponds to that Updating the value of the coefficient b so that the difference between the output Tds (n) of the downstream temperature sensor and its estimated value Tds_hat (n) is as small as possible.

[Equation 20]

• W_hat (n) = Tds_hat (n) - Tds_hat (n - 1) - a (Tds_hat (n - 1) + Ta (n)) (20)

The coefficient b (n) enabling this is calculated by, for example, as shown in equation (21-1) below, a value that results by taking the variable gain KP1 (n) calculated according to equation (21-2). 2) is recursively updated, multiplied by the difference of W (n) and W_hat (n-1). In Equation (21-2), the coefficient P1 is a certain identification gain. Equations (21-1) and (21-2) are algorithms among the parameter identification algorithms, which are generalized as so-called RLS algorithms, to algorithms called fixed gain algorithms. [Equation 21]

In S123, using a certain parameter identification algorithm on equation (18), the value of the coefficient c is updated while keeping the value of the coefficient b at the previous value. The concrete procedure of an example of this will be described below. First, based on the output Tds of the downstream temperature sensor, the output Ta of the outside air temperature sensor, the coefficients b, c, the estimated value Gex_hat of the exhaust gas volume, and the output Tup of the upstream temperature sensor, one of equations (22-1) to (22-2) calculated below virtual output W (n) and its estimated value W_hat (n).

[Equation 22]

• R (n) = Tds (n) - Tds (n-1) -a (Tds (n-1) + Ta (n)) -b (n-1) Gex_hat (n) Tup (n) (22-1 ) R_hat (n) = c (n-1) Gfuel_aft_tm (n) (22-2)

Equation (23) is derived by modifying equation (22-2) from equation (18), and therefore the estimate R_hat (n) defined in equation (22-2) corresponds to the estimate of the virtual output R (n). Updating the value of the coefficient c such that the difference between the virtual output R (n) defined in equation (22-1) and the estimated value R_hat (n) defined in equation (22-2) is as small as possible corresponds to that Updating the value of the coefficient c such that the difference between the output Tds (n) of the downstream temperature sensor and its estimated value Tds_hat (n) is as small as possible.

[Equation 23]

• R_hat (n) = Tds_hat (n) - Tds_hat (n - 1) - a (Tds_hat (n - 1) + Ta (n)) - b (n - 1) Gex_hat (s) Tup (n) (23)

The coefficient c (n) permitting this is calculated by, for example, as shown in equation (24-1) below, a value that results by taking the variable gain KP2 (n) calculated according to equation (24). 2) is recursively updated, multiplied by the difference of R (n) and R_hat (n-1). In Equation (24-2), the coefficient P2 is a certain identification gain. [Equation 24]

28 FIG. 10 is a flowchart showing the concrete procedure of the catalyst oxidation characteristic adaptive calculation for calculating an adaptive catalyst oxidation characteristic coefficient Kox. The processing off 28 is issued by the ECU as a subroutine of the recursive identification calculation of the heat model 27 executed exclusively in execution of a post-injection in the control cycle tn.

As described above, the heat generation coefficient c (n) updated according to the equations (18) to (24-2) is a characteristic that increases with increasing HC oxidation performance of the engine bottom catalyst. A deviation from a certain standard value C_bs of the heat generation coefficient c (n) can thus be used as a deterioration factor of the engine bottom catalyst. However, the heat generation coefficient c (n) is different depending on the support temperature of the engine bottom catalyst. In addition, it is by no means such that the heat generation coefficient c (n) decreases uniformly over the entire temperature range in accordance with the degree of deterioration of the engine bottom catalyst. Therefore, in this adaptive calculation of the catalyst oxidation characteristic as well as in reference to FIG 22 described adaptive calculation for the three-way catalyst characteristic for eliminating the temperature dependence of the heat generation coefficient c a plurality of weighting functions Wox_i, which are defined on a one-dimensional straight line with the support temperature of the engine bottom catalyst as a basis, and a the individual weighting functions following local adaptation coefficient Kox_i introduced with their help weighted statistical processing is performed to calculate the catalyst oxidation characteristic application correction value Kox corresponding to the deterioration factor of the engine bottom catalyst.

In S141, the estimated value Tcc_hat (m) of the carrier temperature of the engine bottom catalyst is calculated, and from this estimated value Tcc_hat (m), by searching a predetermined map, the individual weighting functions for the oxidation characteristic value Wox_i (n) (where i is a positive integer) are computed, and There is a transition to S142.

29 FIG. 12 is a view showing an example of a map for calculating the oxidation characteristic value, that is, the shape of a weighting function for the oxidation characteristic Wox_i. The form of the weighting function Wox_i is the same as that of referring to FIG 23 described weighting function Wtw_i, which is why a detailed description is omitted.

Returning to 28 At S142, as shown in Equation (25) below, the sum of the products of the weighting function values Wox_i (n) and the local adaptation coefficients Kox_i calculated in S141 is calculated and used as the adaptive catalyst oxidation characteristic correction value Kox (n). The output value of the local adaptation coefficient Kox_i is 1. [Equation 25]

In S143, a predetermined standard heat generation coefficient c_bs is multiplied by the catalyst oxidation characteristic adaptive correction value Kox (n), whereby an adaptive heat generation coefficient value C_adp (n) is calculated (see Equation (26) below). The standard heat generation coefficient C_bs corresponds to the heat generation coefficient of a standard engine bottom catalyst without deterioration, and is determined in advance by system identification. Hereinafter, the standard heat generation coefficient C_bs will be described as a fixed value without depending on the temperature, but this is not necessarily the case. The standard heat generation coefficient C_bs may also be determined from the carrier temperature of the engine bottom catalyst by searching a predetermined map.

[Equation 26]

• C_adp (m) = Kox (m) · C_bs (26)

By subtracting the adaptive value C_adp (n) from the heat generation coefficient c (n) in S144, the adaptive error signal E_adp '' (n) is calculated (see equation (27-1) below) and by the adaptive error signal E_adp '( m) is distributed to the individual areas, a local adaptive error signal E_adp '' _ i (n) is also calculated (see equation (27-2) below). Further, in S145, the adaptive local characteristic correction value Kox_i (n) is calculated by, as shown in Equation (27-3) below, a value obtained by adding the local adaptive error signal E_adp "_ i (n) A negative adaptive gain Kadp_o is multiplied, such that the local adaptive error signal E_adp '' _ i (n) calculated for the individual regions becomes 0.

[Equation 27]

• E_adp '' (n) = c (n) - C_adp (n) (27-1) E_adp_i '' (n) = Wox_i (n) * E_adp '' (m) (27-2) Kox_i (n) = Kox_i (n-1) + Kadp_o * Eadp '' _i (n) (27-3)

In the catalyst oxidation characteristic adaptive calculation, first to third regions superimposed on the one-dimensional straight line with the base temperature of the engine base catalyst are defined, and for the individual regions, the weighting functions Wox_i are defined, but the present invention is not limited thereto. For example, a plurality of regions superimposed in a two-dimensional plane with the support temperature of the engine bottom catalyst may be defined, and for the regions in the two-dimensional plane, weighting functions Wox_ij may be defined.

In addition, in the above embodiment, the example of an exhaust gas purification system ( 1 ) using the DeNOx function of an LNT, but the present invention is not limited thereto. For example, the present invention may also be applied to an exhaust gas purification system 2A be applied as it is in 30 which utilizes the DeNOx function of a selective NH3 reduction catalyst. In 30 Only those elements are provided with other reference numerals, which differ from the exhaust gas purification system 2 differ.

In the exhaust duct 11 the emission control system 2A starting from the upstream side of the exhaust gas, an oxidation catalyst (hereinafter "DOC") 61 , a DPF 62 , a reducing agent injection device 63 and a selective reduction catalyst (hereinafter, "SCR catalyst") 64 provided in this order.

The oxidation catalyst 61 has at least two functions, namely an oxidation function and a three-way cleaning function. The oxidation catalyst 61 can also be replaced by an LNT, which has three functions, namely an oxidation function, a DeNOx function and a three-way cleaning function.

The SCR catalyst 64 has a DeNO x function, whereby NOx contained in exhaust gas in an atmosphere in the presence of a reducing agent such as NH _{3 is} selectively reduced. More specifically, when injecting carbamide water through the reducing agent injection device 63 the carbamide water is thermally decomposed or digested by the exhaust heat, producing NH ₃ . The generated NH ₃ becomes the SCR catalyst 64 supplied, and by the NH ₃ , the NOx in the exhaust gas is selectively reduced. As in 30 shown are the oxidation catalyst 61 and the DPF 62 provided in the engine base area during the SCR 64 is provided in the bottom surface area.

The carbamide water injection device 63 has a carbamide water tank 631 and a carbamide water injection nozzle 633 on. The carbamide water tank 631 stores carbamide water and is via a carbamide water pump 635 with the carbamide water injection nozzle 633 connected. The carbamide water injection nozzle 633 is via an unillustrated drive device with the ECU 3A connected. The ECU 3A determines, by means of a carbamide water control (not shown), the carbamide water injection amount and the carbamide water injection timing of the carbamide water injection nozzle 633 , and the drive device actuates the carbamide water injection nozzle 633 to achieve the carbamide water injection condition.

In this way, in the exhaust gas purification system described above 2A the lean NOx in the exhaust gas using the DeNOx function of the SCR catalyst 64 in stoichiometric operation using the three-way DOC cleaning function 61 away. That is, in the system off 1 takes over the LNT in lean driving 41 the DeNOx function, and off at the system 30 takes over the SCR catalytic converter during lean engine operation 64 the deNOx function while the systems off 1 and 30 otherwise identical. The referring to 2 to 29 Therefore, processing can also be described by the emission control system 2A out 30 be executed. However, in this case, since another device takes over the DeNOx function, S23 is out of processing for judging the stoichiometric operation conditions 6 amended as follows. Off in S23 6 It is judged whether there is a condition in which the DeNOx function can be sufficiently achieved. More specific expression tone is expressed in the system 30 in S23 off 6 judged that the SCR catalyst 64 the DeNOx function can sufficiently achieve, when the estimated value of the SCR catalyst carrier temperature is equal to or greater than a certain cleaning temperature, for example, set at 180 to 200 ° C. When the SCR catalyst 64 Below this cleaning temperature, the digestion can be carried out using the carbamide water injection nozzle 633 not well drained carbamide water drain.

LIST OF REFERENCE NUMBERS

1

Engine (internal combustion engine)

11

exhaust duct

13

Fuel injection valve

2, 2A

emission Control system

41

LNT (HC oxidation catalyst)

3, 3A

ECU (fuel injection control means, temperature detecting means, oxidizing ability estimating means, NOx surplus judging means)

53

Temperature sensor in front of the catalytic converter (temperature measuring device)

54

Temperature sensor downstream of the catalytic converter (temperature sensing device)

61

DOC (HC oxidation catalyst)

64

SCR catalyst (DeNOx catalyst)

Claims (8)

An exhaust purification system for an internal combustion engine, comprising: an HC oxidation catalyst provided in an exhaust passage of the internal combustion engine and having a function of oxidizing HC in the exhaust gas, a fuel injection valve capable of performing a main injection and a post-injection after the main injection, which is a fuel injection in a region in which the particulate matter amount in the exhaust pipe increases with increasing distance from the main injection, and a fuel injection control means that controls the fuel injection amount and the fuel injection timing of the fuel injection valve, characterized by a temperature detecting means detecting the temperature of the HC oxidation catalyst, an oxidation ability estimating means which estimates, based on the temperature of the HC oxidation catalyst, the HC oxidation ability of the HC oxidation catalyst, wherein the higher the HC oxidizing ability, the more the fuel injection control means increases the distance between the main injection timing post-injection timing. An exhaust gas purification system for an internal combustion engine according to claim 1, characterized in that in the exhaust passage also a DeNOx catalyst is provided which has a DeNOx function by removing NOx from the exhaust gas in the presence of a reducing agent added to the exhaust gas channel, the HC oxidation catalyst is a three Path cleaning function by removing HC, CO and NOx from the exhaust gas in a stoichiometric air-fuel mixture, and also providing NOx surplus judgment means which judges according to one or both of the engine condition and the condition in the exhaust passage; whether there is an excess NOx state, with a proportion of the amount of NOx flowing into the NOx removal catalyst increasing in proportion to the amount of NOx removable by the DeNOx function, wherein, in the event that there is no excess NOx state, the fuel injection control means controls the air-fuel ratio to lean rather than stoichiometric and, in the event that there is a NOx excess state, stoichiometrically controls the air-fuel ratio. An exhaust purification system for an internal combustion engine according to claim 1, characterized in that the HC oxidation catalyst further comprises a three-way purifying function wherein, in the stoichiometric air-fuel mixture, HC, CO and NOx are removed from the exhaust gas and has a DeNOx function, wherein, in the presence of a reducing agent added to the exhaust passage, NOx is removed from the exhaust gas, and there is further provided a NOx surplus judgment means which judges whether or not there is a NOx surplus condition according to one or both of the engine condition and the condition in the exhaust passage increases the amount of NOx flowing into the NOx removal catalyst in proportion to the amount of NOx removable by the DeNOx function, the fuel injection control means controlling the air-fuel ratio to lean rather than stoichiometric in the event of no NOx excess condition , and in the event that a NOx over The fuel-to-fuel ratio is stoichiometric. An exhaust purification system for an internal combustion engine, comprising: an HC oxidation catalyst having an HC oxidation function, wherein HC is removed from the exhaust gas, and a three-way cleaning function, wherein in stoichiometric air-fuel mixture HC, CO and NOx are removed from the exhaust gas, a fuel injection valve capable of performing a main injection and a post-injection after the main injection, which is a fuel injection in a region in which the particulate matter amount in the exhaust pipe increases with increasing distance from the main injection, and a fuel injection control means that controls the fuel injection amount and the fuel injection timing of the fuel injection valve, characterized by a temperature detecting means detecting the temperature of the HC oxidation catalyst, an oxidation ability estimating means that estimates the HC oxidation ability of the HC oxidation catalyst based on the temperature of the HC oxidation catalyst, and an NOx surplus judgment means that judges whether there is a NOx surplus condition according to one or both of the engine condition and the condition in the exhaust passage, wherein a proportion of the NOx amount flowing into the NOx removal catalyst is proportional to that by the DeNOx function removeable amount of NOx increases, wherein the fuel injection control means stoichiometrically regulates the air-fuel ratio in the case where it is judged that there is a NOx surplus state, while the post-injection amount is increased so as to increase the HC oxidation ability. An exhaust purification system for an internal combustion engine according to any one of claims 1 to 4, characterized in that the HC oxidation ability is represented by numbers, namely a base factor depending on the temperature of the HC oxidation catalyst and the exhaust gas volume, and deterioration factors depending on the degree of deterioration of the HC oxidation catalyst. An exhaust purification system for an internal combustion engine according to any one of claims 1 to 5, characterized in that the oxidation ability estimation means estimates the HC oxidation ability from a heat generation coefficient of the HC oxidation catalyst indicating the contribution of the fuel supplied by the post injection to the temperature increase of the HC oxidation catalyst. An exhaust purification system for an internal combustion engine according to any one of claims 1 to 5, characterized in that the oxidizing ability estimating means estimates the HC oxidizing ability from the amount of reducing agent needed to separate the oxygen stored in the HC oxidizing catalyst or from the reducing agent amount used to reduce the amount of the oxidizing agent HC oxidation catalyst stored NOx is needed. An exhaust purification system for an internal combustion engine according to any one of claims 1 to 5, characterized by further comprising an O ₂ sensor detecting the oxygen concentration downstream of the HC oxidation catalyst, the oxidation ability estimating means detecting the HC Estimates oxidation capability based on an air-fuel ratio setpoint that is set such that an output of the O ₂ sensor is maintained at a particular setpoint.

Images