DE102010043983B4 - Exhaust gas cleaning device for a machine - Google Patents

Abstract

An exhaust purification device for an engine (10) comprising: a catalyst (40, 41) disposed in an exhaust passage (32) of the engine (10), the catalyst (40, 41) storing NOx in a lean atmosphere and the stored NOx reduced in a rich atmosphere; a reduction treatment performing section (S15) for causing the catalyst (40, 41) to undergo reduction treatment; a NOx purification rate estimation section (S14, S24) for estimating a NOx purification rate (η) which is a difference between corresponds to an amount of NOx in an exhaust gas before passing through the catalyst (40, 41) and an amount of NOx in an exhaust gas after passing through the catalyst (40, 41), the NOx purification rate (η) before performing the reduction treatment a pre-reduction rate (ηbefore) and a post-reduction rate (ηafter) after performing the reduction treatment; change calculating means (S25) for calculating a purity change rate (Δη) of the purification rate (η) generated by the reduction treatment based on the pre-reduction rate (ηpre) and the post-reduction rate (ηafter); a NOx reduction amount estimation section (S21) for estimating a NOx reduction amount (ΔNOx) obtained in the reduction treatment is reduced; and a deterioration determination section (S29) for determining a deteriorated state of the catalyst (40, 41) based on a relationship between the purification rate change (Δη) and the NOx reduction amount (ΔNOx), characterized in that the deterioration determination section (S29) has a threshold value setting section (S28) for setting a threshold value (Δηmax) for the purification rate change (Δη) based on the NOx reduction amount (ΔNOx) for determining whether the catalyst (40, 41) is normal or deteriorated, and a comparison section (S29) for comparing the purification rate change (Δη) with the threshold value (Δηmax).

Description

The present invention relates to an exhaust gas purification device for an engine.

A catalyst for storing and reducing nitrogen oxides (NOx) is disposed in an exhaust passage of a system that includes an internal combustion engine such as a diesel engine. The catalyst is, for example, a lean NOx trap (LNT) catalyst and purifies NOx contained in the exhaust gas emitted from the engine. The LNT stores NOx in a lean atmosphere. Further, in a rich atmosphere, the LNT reduces the stored NOx and emits the reduced NOx. Because the diesel engine has the substantially lean atmosphere, NOx contained in the exhaust gas is stored in the LNT so that the exhaust gas can be purified. However, when an amount of NOx stored in the LNT is increased, NOx purifying performance is decreased. Therefore, the rich atmosphere is periodically introduced into the system having the NOx catalyst, and a rich purge corresponding to a reduction treatment is performed to reduce the NOx stored in the LNT.

If a sulfur component (S) contained in the fuel is accumulated in the LNT, the NOx purification performance of the LNT will be deteriorated. This deterioration represents S poisoning. When an accumulated S component amount is increased, it is necessary to cause the LNT to recover from S poisoning to an S release state. For example, the S component is discharged at a temperature equal to or higher than 700 ° C. and at an air-fuel ratio (A / F ratio) equal to or higher than 14.5.

However, the recovery process may decrease fuel mileage or cause a catalyst component of the LNT to have heat deterioration because the LNT is made to have a high temperature. The heat deterioration means a reduction in surface area generated when a catalyst such as platinum gradually combines with its surroundings because the LNT is exposed to the high temperature, for example equal to or higher than 650 ° C. As a result, it becomes impossible for the LNT to purify the exhaust gas.

Therefore, it is desirable to perform the restoration process at the minimum level in accordance with a deterioration degree of the LNT such that it is necessary that the deterioration degree be detected more accurately. If the heat deterioration has progressed, the exhaust gas can have values that exceed legal requirements. In this case, it is necessary to determine that the LNT is in an abnormal condition when it is not possible to restore the LNT. If the degree of deterioration of the LNT is accurately determined, the determination of an abnormal condition can be accurately performed.

The LNT deterioration refers to both S poisoning deterioration and heat deterioration.

The JP 2000 - 34 946 A or the JP 2008-19 790 A discloses a method for determining a degree of degradation of an LNT. An amount of NOx discharged from an engine is estimated based on an operating state of the engine such as a load or a rotational speed. An amount of NOx stored in the LNT is estimated based on the estimate of the discharged amount of NOx. Further, an amount of NOx actually stored in the LNT is calculated based on an amount of a reducing agent consumed in the LNT when a single rich purge is performed. The degree of deterioration of the LNT is determined based on a difference between the estimated value and the actual value of the stored amount of NOx.

However, it is necessary to continue the estimation of the discharged amount of NOx in order to estimate the stored amount of NOx. The discharged amount of NOx is changed by a factor such as a characteristic difference in an exhaust gas recirculation (EGR) system and an injection system even when the engine is approximately in the same operating condition. Furthermore, it is necessary to calculate a ratio of the stored amount of NOx to the discharged amount of NOx in order to obtain the estimated value of the stored amount of NOx. Therefore, it is also necessary to continue estimating NOx purification rate characteristics of the LNT that indicate the ratio of the stored amount of NOx to the discharged amount of NOx.

Further, the NOx purification rate is changed by a temperature of the LNT. An error may be generated in the estimation because the discharged amount of NOx is changed by the above factor and because the NOx purification rate is changed by the temperature of the LNT. Furthermore, the error becomes large in that it is accumulated when the estimated value is integrated by continuing to estimate the discharged amount of NOx and the NOx purification rate. In this case, the detected degree of deterioration of the LNT may become inaccurate.

The US 2007/0 144 143 A1 discloses an exhaust gas purification device according to the preamble of claim 1. The DE 10 2008 002 366 A1 , the JP H10-259 713 A and JP 2005-36653 A disclose further prior art.

In view of the foregoing problem and other problems, it is an object of the present invention to provide an exhaust gas purification device which is improved over the prior art.

The object is achieved by an exhaust gas cleaning device according to claim 1. According to the present invention, an exhaust gas purification device has a catalyst disposed in an exhaust passage of an engine. In a lean atmosphere, the catalyst stores NOx and reduces the stored NOx in a rich atmosphere. A reduction treatment performing section performs reduction treatment for the catalyst. A NOx purification rate estimation section estimates a purification rate (η) of NOx determined by a difference between an amount of NOx in the exhaust gas before passing through the catalyst and a purification rate before reduction (ηpre) before the reduction treatment is performed, and a rate after reduction (ηafter) after the reduction treatment was performed. A change calculator calculates a change (Δη) in the cleaning rate, which is defined by a difference between the rate before a reduction (η before) and the rate after a reduction (η after). A NOx reduction amount estimation section estimates a NOx reduction amount (ΔNOx) that is reduced in the reduction treatment. A deterioration determination section determines a deterioration state of the catalyst based on a relationship between the purification rate change (Δη) and the NOx reduction amount (ΔNOx), and the deterioration determination section includes a threshold setting section for setting a threshold (Δηmax) for the purification rate change (Δη) based on the NOx reduction amount (ΔNOX) for determining whether the catalyst is normal or deteriorated, and has a comparison section for comparing the purification rate change (Δη) with the threshold value (Δηmax).

As a result, the deteriorated state of the catalytic converter can be accurately determined.

• 1 ist eine schematische Ansicht, die eine Abgasreinigungsvorrichtung gemäß einem ersten Ausführungsbeispiel zeigt;
• 2 ist eine schematische Ansicht, die eine Abwandlung der Abgasreinigungsvorrichtung zeigt;
• 3 ist eine graphische Darstellung, die die Kennlinien einer NOx-Reinigungsrate relativ zu einer NOx-Speichermenge zeigt;
• 4 ist eine graphische Darstellung, die eine Veränderung der NOx-Reinigungsrate relativ zu einer NOx-Reduktionsmenge zeigt;
• 5 ist ein Flussdiagramm, das eine Hauptroutine einer Verschlechterungsbestimmungsverarbeitung zeigt;
• 6 ist ein Flussdiagramm, das eine Nebenroutine von S26 von 5 zeigt;
• 7 ist ein Blockdiagramm, das eine Berechnung einer NOx-Abgabemenge und eine Berechnung der NOx-Reinigungsrate zeigt;
• 8 ist eine graphische Darstellung, die eine Veränderung eines Luft-Kraftstoff-Verhältnisses mit der Zeit zeigt;
• 9 ist eine graphische Darstellung, die ein Verfahren zum Bestimmen eines Beendens einer NOx-Reduktion zeigt;
• 10 ist eine graphische Darstellung, die eine Beziehung zwischen einer Reduktionsmittelverbrauchsmenge und der NOx-Reduktionsmenge zeigt;
• 11 ist eine Ansicht, die ein Konzept zum Berechnen einer Veränderung der NOx-Reinigungsrate darstellt;
• 12 ist eine graphische Darstellung, die eine Beziehung zwischen einer Temperatur der LNT und der NOx-Reinigungsrate zeigt;
• 13 ist eine Ansicht, die ein Konzept zum Berechnen eines Schwellenwerts für die Veränderung der NOx-Reinigungsrate zeigt;
• 14 ist ein Blockdiagramm, das eine Berechnung eines Schwellenwerts für eine Veränderung einer NOx-Reinigungsrate gemäß einem zweiten Ausführungsbeispiel zeigt;
• 15 ist eine graphische Darstellung, die Beispiele eines Korrekturkoeffizienten für eine Temperatur einer LNT des zweiten Ausführungsbeispiels zeigt;
• 16 ist eine graphische Darstellung, die Kennlinien einer NOx-Reinigungsrate relativ zu einer NOx-Speichermenge gemäß einem dritten Ausführungsbeispiel zeigt; und
• 17 ist eine Darstellung, die Beispiele eines Temperaturbereichs zeigt, der eingestellt ist, um einen Verschlechterungsgrad einer LNT des dritten Ausführungsbeispiels zu bestimmen.

The above object and other objects, features and advantages of the present invention will become apparent from the following detailed description made with reference to the accompanying drawings. In the drawings:

• 1 Fig. 13 is a schematic view showing an exhaust gas purification device according to a first embodiment;
• 2 Fig. 13 is a schematic view showing a modification of the exhaust gas purification device;
• 3 Fig. 13 is a graph showing the characteristics of a NOx purification rate relative to a NOx storage amount;
• 4th Fig. 13 is a graph showing a change in NOx purification rate relative to an NOx reduction amount;
• 5 Fig. 13 is a flowchart showing a main routine of deterioration determination processing;
• 6th FIG. 13 is a flowchart showing a subroutine of FIG S26 of 5 shows;
• 7th Fig. 13 is a block diagram showing calculation of NOx discharge amount and calculation of NOx purification rate;
• 8th Fig. 13 is a graph showing a change in an air-fuel ratio with time;
• 9 Fig. 13 is a diagram showing a method of determining completion of NOx reduction;
• 10 Fig. 13 is a graph showing a relationship between a reducing agent consumption amount and the NOx reducing amount;
• 11 Fig. 13 is a view showing a concept for calculating a change in the NOx purification rate;
• 12th Fig. 13 is a graph showing a relationship between a temperature of the LNT and the NOx purifying rate;
• 13th Fig. 13 is a view showing a concept for calculating a threshold value for the change in the NOx purification rate;
• 14th Fig. 13 is a block diagram showing calculation of a threshold value for a change in a NOx purifying rate according to a second embodiment;
• 15th Fig. 13 is a graph showing examples of a correction coefficient for a temperature of an LNT of the second embodiment;
• 16 FIG. 13 is a graph showing characteristics of a NOx purification rate relative to FIG Fig. 11 shows a NOx storage amount according to a third embodiment; and
• 17th Fig. 13 is a diagram showing examples of a temperature range set to determine a degree of deterioration of an LNT of the third embodiment.

(First embodiment)

Like it in 1 Shown is an exhaust gas purification device on an engine system 1 mounted that a diesel engine 10 and an electronic control unit 70 (ECU).

Air comes out of an inlet passage 31 into the diesel engine 10 fed. An air flow meter 21st and an inlet pressure sensor 22nd are in the inlet passage 31 arranged. An intake air amount Fa is determined by the air flow meter 21st is measured and an intake air pressure Pm is determined by the intake pressure sensor 22nd measured. The measured values are stored in the ECU 70 entered. An intake throttle valve (not shown) is downstream of the air flow meter 21st arranged. A lot of in the machine 10 Ingested air is increased / decreased by changing an opening of the throttle valve.

The machine 10 has a cylinder (not shown), a piston (not shown), a common rail (not shown) and an injector 11 . While the machine 10 is operated, high pressure fuel is supplied from the common rail through a branch pipe, and the supplied fuel is made from the injector 11 injected into the cylinder. The fuel may be light oil with an injection pressure equal to or greater than 101.3 MPa (1000 atmospheres).

When an intake valve (not shown) is opened, intake air is released from the intake passage 31 inserted into the cylinder. The air introduced is with that of the injector 11 injected fuel mixed. The air-fuel mixture is compressed in the cylinder by the piston and auto-ignition and combustion are caused. When an exhaust valve (not shown) is opened, the burned air is released into the exhaust passage 32 submitted.

A speed sensor 23 is with the machine 10 connected and detects an engine speed NE. The speed sensor 23 can for example be a crank angle sensor for measuring a rotation angle of a crankshaft (not shown) associated with the engine 10 connected is. A crank angle θ measured by the crank angle sensor is entered in the ECU 70 entered.

An in-cylinder pressure sensor 24 is in the machine 10 is arranged and detects an internal pressure P in the cylinder.

The sensor 24 may be integrated with a glow plug that corresponds to an ignition assist section. A combustion state in the cylinder is detected by the sensor 24 won. That is, an ignition timing and a combustion temperature are determined by using the output of the sensor 24 estimated. In addition, knocking, a combustion pressure peak position, or misfire can be detected. The internal pressure P generated by the sensor 24 is detected is entered into the ECU 70 entered.

A catalyst 40 with lean NOx trap (LNT) is in the exhaust passage 32 arranged. A support layer of the LNT 40 is arranged on a base material made of, for example, ceramic, and a storage means and a catalyst are supported on the support layer. For example, when gamma-alumina is used as the backing layer, an amount of a storage agent and a catalyst can be supported on asperities of the backing layer having a large surface area. The storage means can be made of barium, lithium or potassium. The catalyst can be made of platinum, for example.

NOx contained in the exhaust gas is stored in the storage means of the LNT in a lean atmosphere in which the fuel content is leaner than a theoretical air-fuel ratio 40 saved. Usually, in the lean atmosphere, a value of an air-fuel ratio (A / F ratio) is equal to or greater than 17.

When the fuel is excessive at the theoretical air-fuel ratio, a rich atmosphere is defined with the A / F value equal to or less than 14.5. When the rich atmosphere is created and a predetermined temperature condition is satisfied, NOx stored in the storage means is reduced to harmless nitrogen by a reducing agent such as CO or HC contained in the fuel, and the nitrogen is released.

An oxidation catalyst 50 is upstream of the LNT 40 in the exhaust passage 32 arranged. The oxidation catalyst 50 burns hydrocarbons (HC) that enter the exhaust gas passage 32 is supplied by a catalytic reaction, whereby a temperature of the exhaust gas is increased. Therefore the temperature of the LNT 40 increased to a catalyst reactivity of the LNT 40 to improve. That is, the NOx cleaning ability of the LNT 40 is funded.

In the event that the machine system 1 a diesel particulate filter (DPF) for collecting solids in the exhaust gas, such as soot, can then, if the DPF has an oxidation catalyst, the oxidation catalyst 50 replaced by the oxidation catalytic converter of the DPF. If a cleaning power of the LNT 40 is high enough, the oxidation catalyst can 50 be omitted because the oxidation catalyst 50 only to improve the catalyst reactivity of the LNT 40 is used. In this case, an area surrounded by a broken line 100 of the machine system 1 of 1 by an area surrounded by a dashed line 101 of 2 be replaced. Like it in 2 shown can be the LNT 40 and the oxidation catalyst 50 through an LNT 41 be replaced. The LNT 40 and the oxidation catalyst 50 can be placed in a housing of the LNT 41 be integrated.

An exhaust air temperature sensor 25th and a first air-fuel ratio sensor 26th are between the LNT 40 and the oxidation catalyst 50 in the exhaust passage 32 arranged. The sensor 25th measures a temperature Tforward of exhaust gas entering the LNT 40 flows. The first air-fuel ratio sensor 26th measures an air-fuel ratio λ before the exhaust gas entering the LNT 40 flows. The air-fuel ratio λvοrn is an air-fuel ratio of an exhaust gas upstream of the LNT 40 . The sensor values Tvorn, λvorn generated by the sensors 25th , 26th are measured in the ECU 70 entered. The first air-fuel ratio sensor 26th can be located upstream of the oxidation catalyst 50 are located.

A second air-fuel ratio sensor 27 is downstream of the LNT 40 in the exhaust passage 32 arranged. The sensor 27 measures an air-fuel ratio λ after the exhaust gas downstream of the LNT 40 . The air-fuel ratio λ nach is an air-fuel ratio of the exhaust gas downstream of the LNT 40 . The sensor value λnach obtained by the second air-fuel ratio sensor 27 is measured is entered in the ECU 70 entered.

An O ₂ sensor for measuring an O ₂ concentration can be used in place of the second air-fuel ratio sensor 27 be used. In this case, the air-fuel ratio λ nach is calculated based on the measured O ₂ concentration. If a NOx sensor 28 is able _{to measure an O 2} concentration and an air-fuel ratio λ, the second air-fuel ratio sensor 27 through the NOx sensor 28 be replaced.

The NOx sensor 28 is downstream of the LNT 40 in the exhaust passage 32 arranged. The sensor 28 measures a NOx concentration V _{NOx of} the exhaust gas downstream of the LNT 40 . The NOx sensor 28 measures a NOx concentration, for example, by using an oxygen ion conductivity of zirconia (ZrO2). That is, NOx contained in the exhaust gas is _{split into O 2} and N ₂ , and a NOx concentration is measured based on an electromotive force generated when an O ion moves within zirconia. The NOx sensor may be able _{to measure an O 2} concentration and an air-fuel ratio λ based on this principle. The one through the NOx sensor 28 measured sensor value V _Nox is entered in the ECU 70 entered.

An accelerator sensor 29 is with the ECU 70 and outputs an electrical signal corresponding to an offset amount of an accelerator (not shown). The accelerator may correspond to an operating part for transmitting a torque requested by an occupant of a vehicle. The sensor value of the accelerator sensor 29 gets into the ECU 70 entered.

The ECU 70 can be a normal computer and has a CPU (not shown) to do calculations and memory 71 to store a variety of information. The ECU 70 detects an operating state of the machine 10 based on the sensor values output from the various sensors and calculates optimal values for fuel injection amount, fuel injection time and fuel injection pressure. That is, the ECU 70 controls fuel injection of the injector 11 the machine 10 .

The ECU 70 controls the machine 10 basically in the lean atmosphere and periodically creates the rich atmosphere. When a rich purge corresponding to a reduction treatment is performed in the rich atmosphere, it becomes in the LNT 40 stored NOx reduced.

The rich purge can be performed using a rich burning method in which the fuel is briefly out of the injector 11 is injected after a main injection has been performed. Alternatively, a post injection method can be used as the rich purging in which injection is performed after a long period of time.

The ECU 70 may correspond to an exhaust gas purification device for an engine of the present invention, and evaluates the degree of deterioration of the LNT 40 . The evaluation method is explained below.

A relationship between a NOx purification rate η and a NOx storage amount Y that is in the LNT 40 is stored is by reference on 3 explained. 3 shows characteristics of the NOx purification rate η and the NOx purification rate η is determined by a change in the NOx storage amount Y, which is in the LNT 40 is saved, changed. A solid line 211 of 3 represents the NOx purification rate η when the LNT 40 is in the normal state. Dashed line 221 of 3 represents the NOx purification rate η when the LNT 40 has deteriorated.

The NOx purification rate η is a characteristic of the LNT 40 and represents a rate of an amount of NOx produced by the LNT 40 stored and purified, to an amount of NOx released by the engine 10 is delivered. The NOx purification rate η is defined as 100% when the NOx storage amount Y is zero. That is, it is determined that all of the discharged NOx is purified when the NOx storage amount Y is zero.

Like it in the line 211 of 3 is shown, if the LNT 40 is normal, namely when the LNT 40 does not have so much deterioration, the NOx purification rate η gradually decreases as the NOx storage amount Y is increased. Therefore, when the NOx storage amount Y is increased, it is necessary to perform a rich purge in order to reduce the NOx storage amount Y.

When the NOx purification rate η becomes η10 of one point C11 on the line 211 is reduced, a rich rinse is carried out. At this time, a NOx reduction amount ΔNOx reduced by the rich purge is defined as ΔNOx1. The NOx purification rate η becomes from the point C11 to a point C21 on the line 211 increased by the NOx reduction amount ΔNOx1. That is, the NOx cleaning ability of the LNT 40 can be restored. When the NOx purification rate η at the point C21 is defined as η21, a cleaning rate change Δη has a value of Δη1 (= η21-η10). The cleaning rate change Δη indicates a change in the cleaning rate η which is generated by the rich rinsing.

When the LNT 40 has deteriorated due to S poisoning or heat, as was the case with the lineage 221 of 3 as shown, the NOx purification rate η is decreased as the NOx storage amount Y is increased. A slope of the line 221 that indicates the deterioration time is stronger than a slope of the line 211 , which indicates the standard time. Therefore, if the LNT 40 has deteriorated compared with the normal time, a degree of decrease in the NOx purification rate η becomes large.

When the NOx purification rate η to η10 from one point C12 on the line 221 that the point C11 on the line 211 is reduced, a rich rinse is carried out. At this time, since a NOx reduction amount ΔNOx reduced by the rich purge is ΔNOx1, the NOx purification rate becomes from the point C12 on one point C22 on the line 221 increased by the NOx reduction amount ΔNOx1.

Because the slope of the line 221 stronger than the slope of the line 211 is, a NOx purification rate becomes η22 from the point C22 of the deterioration time is higher than the NOx purification rate η21 of the normal time. That is, the purification rate change Δη2 (= η22-η10) from the point C12 to the point C22 is greater than the change in the cleaning rate Δη1 (= η21-η10) from normal time.

The purification rate change Δη produced by the rich purge becomes larger in the deteriorating time than in the normal time when the NOx reduction amount ΔNOx is the same.

In addition, the purification rate change Δη is changed when the NOx reduction amount ΔNOx generated by a single rich purge is different even if the LNT 40 has the same degree of deterioration. That is, the purification rate change Δη becomes larger as the NOx reduction amount ΔNOx is at one of the line 211 , which indicates the standard time, and the line 221 indicating the deterioration time becomes larger.

4th FIG. 13 shows a change in the purification rate change Δη relative to the NOx reduction amount ΔNOx generated by the single rich purge. A line 311 represents standard time and a line 321 represents the deterioration time. In 4th a NOx purification rate prior to purging η is specified. The NOx purification rate before purging ηvor indicates a NOx purification rate before the rich purge is carried out.

Like it in 4th as shown, the purification rate change Δη gradually becomes large as the NOx reduction amount ΔNOx comes to one of the line 311 , which indicates the standard time, and the line 321 becomes large, which indicates the deterioration time. A degree of change in the cleaning rate change Δη is on the line 321 bigger than the line 311 because the slope of the line 221 greater than the slope of the line 211 is.

Therefore, as it is in 4th shown may be a threshold line 331 , which has a threshold value Δηmax, for the cleaning rate change Δη between the line 311 and the line 321 be set so that it can be determined that the LNT 40 normal or worsened. The threshold Δηmax of the line 331 becomes progressive larger as the NOx reduction amount ΔNOx becomes larger.

Further is the line 211 , 221 of 3 not straight forward. If the pre-purge NOx purification rate ηpre is different, the purification rate change Δη will be changed even if the NOx reduction amount ΔNOx is the same. That is, the line 311 , 321 of 4th will be changed if the NOx purification rate before purging ηvor is different.

The degree of deterioration of the LNT 40 is determined using a change in cleaning rate change Δη produced by the rich flush. The purification rate change Δη is changed based on the deterioration degree, the NOx reduction amount ΔNOx, and the NOx purification rate before purging ηpre.

More specifically, it becomes a map of the threshold value Δηmax of the cleaning rate change Δη in the memory 71 saved in advance. The map has parameters of the NOx reduction amount ΔNOx and the NOx purification rate before purging η.

The degree of deterioration of the LNT 40 for example, by using the flowcharts of the 5 and 6th rated. The processing of determining the degree of deterioration of the LNT 40 made by the ECU 70 is performed with reference to the 5 and 6th described. 5 shows a main routine of deterioration determination and 6th shows a subroutine of S26 of 5 . This deterioration determination is started, for example, when the machine 10 is started.

At S11 of 5 it is determined whether or not a rich flush needs to be requested. More specifically, if the NOx storage amount that is in the LNT 40 is stored the same as a predetermined amount, it is determined that the rich flush needs to be requested. Alternatively, when a predetermined time has passed since the last time a rich flush was performed, it is determined that the rich flush needs to be requested. Alternatively, when the integration value of the discharge amount of NOx discharged from the engine becomes equal to a predetermined value, it is determined that the rich purging needs to be requested. Alternatively, based on the sensor value V _{NOx of} the NOx sensor 28 downstream of the LNT 40 be determined that the rich flush must be requested. When the sensor _{value V NOx} becomes large, it is determined that the NOx purifying ability of the LNT 40 has decreased because it is determined that the NOx storage amount is large. An edition of S11 is NO when it is determined that the rich flush need not be requested. S11 is repeated before the fat flush is requested. An edition of S11 is YES when it is determined that the rich flush must be requested, and S12 is running.

At S12 becomes a NOx discharge amount prior to a purge Xe1 coming out of the engine 10 is omitted, calculated before the rich flush is performed. 7th is used to explain a calculation method of the NOx discharge amount before purge Xe1. Each step performed in the calculation process is shown as a block diagram. 7th In addition to the steps for the NOx discharge amount before purging, Xe1 also shows each step for calculating the NOx purification rate η. The calculation method of the pre-purge NOx discharge amount Xe1 will be described with reference to FIG 7th explained.

At S101 becomes a base NOx discharge amount Xe0 generated by the engine 10 based on the operational status of the machine 10 calculated. Specifically, the operating state corresponds to an engine speed NE and a torque Tq. An adaptation value of the NOx discharge amount caused with the engine speed value NE and the torque Tq is stored in the memory 71 saved as a card in advance. The adaptation value is set as the basic discharge NOx amount Xe0 by referring to the map stored in the memory 71 is stored. The engine speed NE is obtained based on the crank angle θ obtained from the speed sensor 23 is entered. Alternatively, the engine speed NE is obtained based on an offset amount of the accelerator obtained from the accelerator sensor 29 is entered. A fuel injection amount Q given by the injector 11 can be used instead of the moment Tq.

Because the basic NOx discharge amount Xe0 calculated at S101 is the adaptation value that is used for each operating state of the engine 10 Being predetermined may be an actual exhaust NOx amount actually produced by the engine 10 may differ from the basic NOx discharge amount Xe0 due to, for example, the O ₂ concentration of the intake air or the combustion center. A process of correcting the basic NOx discharge amount Xe0 is performed based on the O ₂ concentration of the intake air or the center of combustion as follows.

At S102 becomes the O ₂ concentration of the intake air based on at least one value of the intake air amount Fa obtained from the air flow meter 21st is input, the intake air pressure Pm obtained from the intake pressure sensor 22nd is input, and the air-fuel ratio λforward obtained from the first air-fuel ratio sensor 26th is entered is calculated. Specifically, the O ₂ concentration is high when the intake air amount Fa is great because the O ₂ concentration is calculated from the intake air amount Fa. Therefore, the O ₂ concentration calculated based on a predetermined relationship between the intake air amount Fa and the O ₂ concentration.

The O ₂ concentration becomes high when the inlet pressure Pm becomes high because the O ₂ concentration is calculated from the inlet pressure Pm. Therefore, the O ₂ concentration calculated based on a predetermined relationship between the intake pressure Pm and the O ₂ concentration.

The O ₂ concentration becomes high when the air-fuel ratio λforward becomes large because the O ₂ concentration is calculated from the air-fuel ratio λvοrn. Therefore, the O ₂ concentration is calculated based on a predetermined relationship between the air-fuel ratio λvοrn and the O ₂ concentration.

The O ₂ concentration can be calculated by using a plurality of values of the intake air amount Fa, the intake pressure Pm, and the air-fuel ratio λforward. Thus, at S102, the O ₂ concentration is calculated from the sensor value Fa, Pm, λvοrn without using an O ₂ sensor.

At S103 becomes a correction factor K1 calculated for the O ₂ concentration of the inlet air. The correction factor K1 represents an increase / decrease in the basic NOx discharge amount Xe0 based on a difference between the O ₂ concentration calculated at S102 and the adaptation value of the O ₂ concentration obtained from the operating state of the engine 10 The adaptation value corresponds to a setpoint value of the O ₂ concentration of the intake air and the calculated value corresponds to an estimated value of the O ₂ concentration of the intake air.

Specifically, a map of the _{target O 2} concentration is made in the memory 71 in advance relative to each operating state of the machine 10 stored, which corresponds to the engine speed NE and the torque Tq. There is also a map of the correction factor K1 in the memory 71 ₂ concentration stored in advance relative to each deviation amount between the O ₂ -Sollkonzentration and the estimated O. The correction factor K1 is determined by referring to the map of O ₂ to correspond -Sollkonzentration and the estimated O ₂ concentration.

At S104 becomes the combustion center (the center of combustion) based on the in-cylinder pressure P obtained from the pressure sensor 24 is input and the crank angle θ obtained from the speed sensor 23 is entered is calculated. Specifically, a waveform of the in-cylinder pressure P is defined as a parameter indicating a combustion state relative to the crank angle θ, and a center (center of gravity) of the waveform of the in-cylinder pressure P is calculated as the combustion center.

At S105 becomes a correction factor K2 calculated for the burn center. The correction factor K2 represents an increase / decrease in the basic NOx discharge amount Xe0 based on a difference between the combustion center calculated at S104 and the adaptation value of the combustion center derived from the operating state of the engine 10 is determined. The adaptation value corresponds to a nominal value of the combustion center.

Specifically, a map of the target incineration center is kept in memory 71 in advance relative to each operating state of the machine 10 stored according to the engine speed NE and the torque Tq. There is also a map of the correction factor K2 in the memory 71 stored in advance relative to each deviation amount between the target value and the estimated value. The correction factor K2 is set for the combustion center by referring to the map to correspond to the current target value and the estimated value.

At S106 the basic NOx discharge amount Xe0 is corrected to increase / decrease based on the correction factor K1 the O ₂ concentration and the correction factor K2 of the burn center. Therefore, the discharge amount of NOx _{can be calculated in consideration of the O 2} concentration and the combustion center. The corrected amount of NOx is defined as the amount of NOx released over time before a purge X'e1.

The NOx purification rate η is determined by using the NOx discharge amount obtained from the engine 10 is omitted, and a NOx permeation amount after passing through the LNT 40 calculated. However, there is a time lag before the NOx is released from the engine 10 is omitted by the NOx sensor 28 is detected while the NOx permeation amount by using the NOx sensor 28 is calculated. Therefore, it is necessary to make a timing of the NOx discharge amount and a timing of the NOx passage amount correspond to each other.

At S107 a time delay Ln is calculated, which is generated until that by the machine 10 released NOx the NOx sensor 28 reached. The time lag Ln has a value corresponding to a discharge amount Ga. The discharge amount Ga may correspond to the intake air amount Fa. A map of the time delay Ln is in the Storage 71 stored relative to each intake air amount Fa. The time lag Ln is set based on the map so as to correspond to the current intake air amount Fa.

Furthermore, the NOx sensor 28 have their own response delay. That is why at S108 a time constant Tn is calculated, which is the response delay of the NOx sensor 28 corresponds to. Because the time constant Tn corresponds to the intake air amount Fa, a map of the time constant Tn becomes in the memory 71 stored relative to the intake air amount Fa. The time constant Tn is set based on the map to correspond to the current intake air amount Fa.

At S109 becomes the NOx discharge temporal amount before purging X'e1 with delay time processing based on the temporal delay Ln, which at S107 is calculated, and the time constant Tn, which at S108 is calculated, filtered. Therefore, the temporal NOx release amount before purging X'e1 can correspond to a NOx permeation amount which is at S13 is to be calculated. The NOx discharge amount that has passed the delay time processing is taken as a NOx discharge amount before purge Xe1 of S12 used. Thus, the amount of NOx released before purging corresponds to Xe1 that at S12 is calculated, the amount of NOx permeation that occurs at S13 is to be calculated. The ECU 70 , the S107 to S109 may correspond to a time processing section of the present invention.

S110 from 7th is processing for calculating the NOx purification rate η and corresponds to S14 or S24 which will be mentioned later.

Like it in S13 of 5 is shown, the NOx permeation amount before purging X11 after going through the LNT 40 before subjecting it to the rich purge based on the NOx concentration V _NOx calculated by the NOx sensor 28 is entered. For example, if the NOx permeation amount before purge Xl1 is a total amount of NOx that has passed through the LNT 40 passes through, and when the NOx concentration V _{NOx detected} by the NOx sensor 28 is inputted is an amount of NOx per unit volume, the total NOx permeation amount is calculated _{from the NOx concentration V NOx.}

At S14 of 5 , the S110 from 7th corresponds to, a NOx purification rate before purging η before rich purging by substituting the NOx discharge amount before purging Xe1 and the NOx passing amount before purging X11 is calculated into the following formula (1). That is, a ratio of the amount of NOx before going through the LNT 40 to the amount of NOx after going through the LNT 40 is defined as the pre-purge NOx purification rate ηpre. The ECU 70 , the S14 may correspond to a pre-reduction NOx purification rate calculating section for calculating the NOx purification rate before purge η prior to the present invention. $$\mathrm{\eta }\text{in front}=\text{X11 / Xe1}$$

A fat rinse will be at S15 began. Specifically, for example, rich combustion is performed to remove fuel from the injector 11 to be injected with a short time interval or a post-injection is carried out with a long interval, namely after the main injection. Because the exhaust gas becomes rich, it becomes so in the storage means of the LNT 40 stored NOx is reduced by the reducing agent such as CO or HC in the fuel, and harmless nitrogen corresponding to the reduced NOx is emitted.

At S16 becomes an upstream air-fuel ratio λ in front of upstream of the LNT 40 by the first air-fuel ratio sensor 26th is obtained and a downstream air-fuel ratio λ behind the downstream of the LNT 40 by the second air-fuel ratio sensor 27 receive.

The discharge amount Ga is calculated at S17 based on the intake air amount Fa obtained from the air flow meter, for example 21st is entered.

At S18 For example, a reducing agent amount Δra consumed in reducing the NOx is calculated at the present time after the rich purging is started. 8th is used to explain a method of calculating the reducing agent amount Δra. A solid line from 8th shows a change in the upstream air-fuel ratio λvοrn and a broken line of FIG 8th FIG. 13 shows a change in the downstream air-fuel ratio λ rear over time.

When the rich purging is started at time t1, the upstream air-fuel ratio λforward moves in a rich range during the rich purging. In contrast, the downstream air-fuel ratio λ rear has an approximately stoichiometric value (λ = 1) during that in the LNT 40 stored NOx is reduced because fuel is in the LNT 40 is consumed as a reducing agent. Thus, the upstream air-fuel ratio λ front and the downstream air-fuel ratio λ rear can be used as an index indicating a situation of reducing agent consumption in the LNT 40 indicates.

The reducing agent amount Δra is calculated by substituting the upstream air-fuel ratio λ front, the downstream air-fuel ratio λ rear, and the discharge amount Ga into the following formula (2). The reducing agent amount Δra corresponds to a value of a multiplication of a hatched area of 8th with the delivery amount Ga. $$\Delta \text{ra}=\sum \left(\mathrm{\lambda }\text{back}-\mathrm{\lambda }\text{front}\right)\times \text{Ga}$$

At S19 of 5 it is determined that the reduction treatment of NOx in the LNT 40 to end or not. 9 is used to explain the determination procedure. 9 is a revision of 8th showing the downstream air-fuel ratio λ rear over time. The downstream air-fuel ratio X rear has approximately the stoichiometric value (λ = 1) while NOx is reduced. However, the downstream air-fuel ratio λ rear has a value on the rich side with the passage of time because fuel corresponding to the reducing agent is passed through the LNT 40 will pass if the reducing treatment of NOx is sufficient. That is why at S19 a judgment value λ1 is set as the value on the rich side, and the completion of the reduction treatment is judged by whether the downstream air-fuel ratio λ behind becomes richer than the judgment value λ1.

As long as the reduction treatment is not finished, there is an output of S19 NO and will S16 to S18 is repeated such that the reducing agent amount Δra is updated. When the reduction treatment is finished, there is an output of S19 YES and will flush the fat at S20 ended by ending rich combustion or post-injection. Therefore, the rich scavenging can be prevented from continuing when the reducing treatment is finished. As a result, exhaust gas containing a lot of CO and HC can be prevented from being emitted. The ECU 70 , the S11 , S15 and S19 may correspond to a reduction treatment performing section of the present invention.

At S21 a NOx reduction amount ΔNOx, which is reduced by the rich purging, is calculated. Because the reducing agent amount Δra used in the reducing treatment corresponds to the NOx reduction amount ΔNOx, the present NOx reduction amount ΔNOx is determined by referring to a map of FIG 10 calculated that in the memory 71 has been saved in advance. The ECU 70 which executes S121 corresponds to a NOx reduction amount estimation section for estimating the NOx reduction amount ΔNOx of the present invention.

At S22 For example, a discharge amount of NOx after a purge Xe2 is calculated after the engine has performed the rich purge 10 is left out. This calculation is made by approximately the same procedure as the NOx discharge amount before a purge Xe1 of 7th carried out. The ECU 70 , the S12 and S22 corresponds to a discharged NOx amount estimating section for estimating the discharged NOx amount of the present invention.

At S23 we are a NOx permeation amount after a purge X12 by the LNT 40 passes through after the rich purge occurs based on the NOx concentration V _NOx calculated by the NOx sensor 28 is entered. The ECU 70 , the S13 and executes S23 corresponds to a NOx permeation amount estimating section for estimating the NOx permeation amount of the present invention.

At S24 becomes a NOx purification rate after purging η after having the rich purge by replacing the NOx discharge amount after purging Xe2 and the NOx permeating amount after purging X12 is calculated into the following formula (3). The ECU 70 , the S12 to S14 and S22 to S24 corresponds to a NOx purification rate estimating section for estimating the NOx purification rate η of the present invention. The ECU 70 , the S24 corresponds to a post-purge NOx purification rate calculating section for calculating the post-purge NOx purification rate η according to the present invention. $$\mathrm{\eta }\text{to}=\text{X12 / Xe2}$$

At S25 as it is in 11 As shown in FIG. 1, a purification rate change Δη indicating a degree of change in the purification rate η is calculated by using the NOx purification rate before purge η and the NOx purification rate after purge η after. Specifically, the calculation is carried out by using the following formula (4). $$\Delta \mathrm{\eta \; =\; \eta }\text{to}-\mathrm{\eta }\text{in front}$$

That is, a difference between the NOx purification rate after purge η after and the NOx purification rate before purge η before is defined as a purification rate change Δη. Therefore, even if the sensor such as the NOx sensor 28 , which is used for calculating the NOx purification rate η after, η before has an offset error that is smoothly increased or decreased, the error can be corrected such that the cleaning rate change Δη is precisely calculated.

$$\Delta \mathrm{\eta =}\Delta \mathrm{\eta }\text{'}/\mathrm{\eta }\text{vor}$$

The following formulas (5) and (6) can be used to calculate the cleaning rate change Δη in place of the formula (4). $$\Delta \mathrm{\eta }\text{'}\mathrm{=\; \eta }\text{to}-\mathrm{\eta }\text{in front}$$

$$\Delta \mathrm{\eta \; =}\Delta \mathrm{\eta }\text{'}/\mathrm{\eta }\text{in front}$$

The cleaning rate change Δη is calculated by dividing a difference Δη 'between the cleaning rate after a rinse ηafter and the cleaning rate before a rinse ηvor by the pre-rinse cleaning rate ηvor. Therefore, the cleaning rate change Δη is calculated to correspond to a relative change between the post-wash cleaning rate ηafter and the pre-wash cleaning rate η before.

Even if the sensor, such as the NOx sensor 28 , which is used to calculate the NOx purification rate η post, η pre has a gain error that is proportionally offset relative to the output, and even if the calculated NOx discharge amount has an error, the error can be canceled so that the purification rate change Δη is calculated exactly. The ECU 70 , the S25 corresponds to a change calculation section of the present invention.

At S26 becomes a processing for determining a temperature TLNT of the LNT 40 carried out. During processing, the temperature TLNT becomes the LNT 40 determined to determine whether the deterioration level of the LNT 40 is detected or not.

Temperature properties of the LNT 40 are explained before processing is explained. 12th shows an example in which the NOx purification rate η is relative to the temperature TLNT of the LNT 40 will be changed. In 12th is the condition such as a deterioration degree of the LNT 40 or the NOx storage amount is the same for each temperature TLNT of the LNT 40 .

As it is in ( 1 ) of 12th as shown, when the temperature TLNT is low, for example, equal to or less than 200 ° C., the NOx purification rate η is small because the catalyst activity of the LNT 40 is low.

As it is in ( 2 ) of 12th is shown, when the temperature TLNT is increased, for example, in a range from 200 ° C to 350 ° C, the NOx purification rate η is increased.

The NOx purification rate η is not stable when the temperature T _{LNT is} low, as shown in ( 1 ) and ( 2 ) of 12th is shown.

As it is in ( 3 ) of 12th as shown, when the temperature T _{LNT is} raised, for example, in a range from 350 ° C. to 450 ° C., the NOx purification rate η becomes high. The NOx purification rate η in ( 3 ) of 12th has less variation even if the temperature TLNT is changed.

As it is in ( 4th ) of 12th is shown, when the temperature T _{LNT is} increased, for example, in a range from 450 ° C. to 550 ° C., the NOx purification rate η is decreased as the temperature T _{LNT is} increased.

As it is in ( 5 ) of 12th as shown, when the temperature T _{LNT is} raised, for example, to a temperature equal to or higher than 550 ° C., the NOx purification rate η is decreased. The NOx purification rate η is not stable when the temperature TLNT is high as shown in ( 4th ) and ( 5 ) of 12th is shown.

Because the LNT 40 has the temperature characteristics, the NOx purification rate η is changed by changing the temperature T _LNT . Therefore, if the temperature TLNT is not taken into account, the result of the deterioration determination may become inaccurate.

According to the first embodiment, execution of the deterioration determination is permitted only when the temperature TLNT is in a predetermined range that is ( 3 ) of 12th in which the NOx purification rate η is stable.

6th FIG. 13 is a flowchart showing the processing for determining the temperature TLNT of S26 which is used to determine whether or not the temperature TLNT is in the predetermined range.

At S261 becomes the temperature Tfront of the exhaust gas upstream of the LNT 40 through the sensor 25th of 1 receive.

At S262 the temperature TLNT is estimated based on the exhaust gas temperature Tvorn. For example, a map of the temperature TLNT relative to the exhaust gas temperature Tforward is stored in the memory in advance 71 is stored and the current temperature TLNT is determined by referring to the map.

At S263 it is determined whether or not the temperature TLNT is in a predetermined temperature range (T _{LNT_min} -T _{LNT_max} ). This temperature range (T _{LNT_min} -T _{LNT_max} ) corresponds to the range ( 3 ) of 12th . For example, the temperature range can be all or part of the range ( 3 ) of 12th be constructed.

When the temperature TLNT is in this temperature range (T _{LNT_min} -T _{LNT_max} ), an output of S263 is YES and a signal indicating YES becomes at S264 issued. When the temperature T _LNT is not within this temperature range (T _{LNT LNT_min} -T _ _max), an output of S263 is NO, and a signal that indicates NO, wherein S265 issued. The processing of the flowchart of 6th will terminate after S264 or S265 is carried out. The ECU 70 , the S26 corresponds to a temperature judgment section of the present invention.

Like it in 5 shown is at S26 the signal output is set so that this at S27 Is YES or NO. An edition of S27 is NO if the signal output is at S26 No is. At this time, because the temperature TLNT is not in the predetermined temperature range (T _{LNT_min} -T _{LNT_max} ), processing returns S11 back. In this case, the degree of deterioration becomes the LNT 40 not assessed, as a wrong assessment can be carried out. When the signal output is at S26 YES is an issue of S27 YES and will S28 is carried out to make the determination of the degree of deterioration.

At S28 a threshold value Δηmax is set for the current cleaning rate change Δη to determine that the LNT 40 normal or worsened. The purification rate change Δη is changed by the NOx reduction amount ΔNOx and the NOx purification rate before purging ηbefore even if the degree of deterioration is the same.

Like it in 13th shown is at S28 the current threshold value Δηmax is set by referring to a map of the threshold value Δηmax of the purification rate change Δη having a parameter of the pre-purge NOx purification rate ηpre. The ECU 70 , the S28 corresponds to a threshold value setting section of the present invention.

At S29 it is determined whether or not the cleaning rate change Δη is larger than the threshold value Δηmax. When the cleaning rate change Δη is equal to or smaller than the threshold value Δηmax, an output of S29 NO. Therefore it is determined that the LNT 40 is normal and reverses processing to S11 back. In this case, the degree of deterioration becomes the LNT 40 reassessed after the next rich flush is done.

When the cleaning rate change Δη is larger than the threshold value Δηmax, an output of S29 YES.

That is why at S30 determines that the LNT 40 is degraded, and the flowchart of FIG 5 completed. In this case, a process is carried out to regenerate from the S poisoning or an alarm is issued so that the LNT 40 is checked or replaced. The ECU 70 , the S27 to S30 corresponds to a deterioration determining section of the present invention. The ECU 70 , the S29 and S30 executes corresponds to a comparison section of the present invention.

According to the first embodiment, the deterioration of the LNT becomes 40 is determined based on the purification rate change Δη and not based on the NOx storage amount. Therefore, the deterioration determination can be made accurately.

(Second embodiment)

In the first embodiment, the deterioration degree becomes the LNT 40 only determined if the temperature TLNT of the LNT 40 is in the predetermined temperature range (T _{LNT_min} -T _{LNT_max} ) in which the NOx purification rate η has a stable value.

In a second embodiment, the threshold value Δηmax is set based on the NOx reduction amount ΔNOx, the NOx purification rate before purging ηpre and the temperature T _LNT such that the degree of deterioration of the LNT 40 is determined even when the NOx purification rate η is unstable. Aspects of the second embodiment that are different from the first embodiment will be described below.

The construction of the second embodiment is similar to that of the first embodiment of FIG 1 . The degree of deterioration of the LNT 40 is based on the flowchart of 5 determined similar to the first embodiment. The procedure for setting the threshold value Δηmax at S28 in the second embodiment is different from that of the first embodiment. Each step that is performed in the method of setting the threshold value Δηmax is shown in FIG 14th shown in a block diagram.

At S281 a basic threshold value Δηmax 'is preset based on the NOx reduction amount ΔNOx and the NOx purification rate before purge η. The basic threshold value is Δηmax ' equivalent to the threshold value Δηmax in the first embodiment. The current basic threshold Δηmax 'is set by referring to a map of the basic threshold Δηmax' having a parameter of the NOx reduction amount ΔNOx and the NOx purification rate before purge ηpre.

At S282 becomes a correction factor K _T based on the temperature TLNT of the LNT 40 is calculated to correct the basic threshold Δηmax 'set at S281. 15th includes examples showing the correction factor K _T relative to the temperature TLNT of the LNT 40 shows.

In 15 (a) the correction factor K _{T is} constant (K _T = 1) in the temperature range (T _{LNT_min} -T _{LNT_max} ). That is, at S282 the correction factor K _{T is} set to be constant regardless of the temperature T _LNT. At S283 the threshold value Δηmax is set by multiplying the basic threshold value Δηmax 'by the correction factor K _T. Therefore, in the example of 15 (a) the threshold value Δηmax is equal to the basic threshold value Δηmax ', without depending on the temperature TLNT of the LNT 40 to be influenced, similar to the first embodiment.

The lower limit value T _{LNT_min} and the upper limit value T _{LNT_max} of 15 (a) are set in such a manner that the NOx purification rate η shows a stable value as the temperature range of ( 3 ) of 12th .

In 15 (b) the correction factor K _{T is} gradually reduced as the temperature TLNT increases. Therefore, as the temperature TLNT increases, the correction factor K _{T is} gradually reduced, namely at S282 , and will be at S283 the threshold value Δηmax is gradually reduced.

The lower limit value T _{LNT_min} and the upper limit value T _{LNT_max} of 15 (b) are set in such a way that the correction factor K _{T is} the smallest value in the temperature range ( 3 ) of 12th becomes. That is, the correction factor K _T is in the temperature range ( 3 ) of 12th in which the NOx purification rate η shows the stable value is made the smallest value. The correction factor K _T is used in the low temperature range ( 1 ), ( 2 ) of 12th in which the NOx purification rate η shows an unstable value is made larger.

Therefore, the determination of the degree of deterioration can be made in the low temperature region showing the unstable value. Thus, the determination can be made frequently. Further, it can be prevented that the LNT 40 is deteriorated when the LNT 40 is not deteriorated.

In 15 (c) the correction factor K _{T is} gradually reduced as the temperature TLNT becomes higher. Then the correction factor K _{T is} gradually increased as the temperature T _LNT becomes higher.

The lower limit value T _{LNT_min} and the upper limit value T _{LNT_max} of 15 (c) are set in such a way that the correction factor K _{T is} in the temperature range ( 3 ) of 12th becomes smallest. Therefore, the degree of deterioration in the low temperature region can be ( 1 ), ( 2 ) of 12th and according to the high temperature range ( 4th ), ( 5 ) of 12th can be determined in which the NOx purification rate η is not stable. Thus, the determination can be made more frequently. Further, it can be prevented that the LNT 40 has deteriorated when the LNT 40 has not deteriorated.

(Third embodiment)

In a third embodiment, if the degree of deterioration of the LNT 40 is detected, the NOx purification rate η is controlled. Aspects of the third embodiment that are different from the first and second embodiments will be described below.

A construction of the third embodiment is similar to that of the first embodiment of FIG 1 . The degree of deterioration of the LNT 40 is based on the flowchart of 5 determined similar to the first embodiment.

The degree of deterioration of the LNT 40 is determined based on the purification rate change Δη when the NOx purification rate η of the LNT 40 is in a predetermined range.

A solid line 211 of 16 is the same line as the line 211 of 3 showing the standard time and a dashed line 221 of 16 is the same line as the line 221 of 3 showing the deterioration time.

Like it in 16 as shown, when the NOx purification rate η is in a range from 40% to 90%, the NOx purification rate η becomes the line 211 , 221 is changed almost linearly relative to the NOx storage amount Y. The range from 40% to 90% represents a linear range of the line 211 , 221 In contrast, the NOx purification rate η is less linear in a range other than 40% to 90%. When the NOx purification rate η is equal to or greater than 90%, the line becomes 211 , 221 convex upward and the NOx purification rate η has a small variation even if the NOx storage amount Y is changed.

Therefore, when the deterioration determination is made when the NOx purification rate η is equal to or greater than 90%, the purification rate change Δη is small even if the LNT is used 40 is deteriorating. In this case, wrong judgment may be made.

In the third embodiment, the deterioration degree becomes the LNT 40 is determined in the linear range of the NOx purification rate η from 40% to 90%. In particular, the temperature range (T _{LNT_min} -T _{LNT_max} ) of S263 is from 6th is set in a manner that the NOx purification rate η changes in a range from 40% to 90%, for example.

If the rich purging is performed when the LNT temperature T _{LNT is} high, reduction is promoted. Therefore, the NOx storage amount Y, which is in the LNT 40 remains decreased when the rich flush is stopped. Like it in 16 As shown, the NOx purification rate η can become equal to or greater than 90% as the NOx storage amount Y is reduced. Therefore, as it is in 17 (c) , 17 (d) is shown, the temperature range (T _{LNT_min} -T _{LNT_max} ) is set to be comparatively low, for example less than or equal to 300 ° C. In this case, the reduction reaction is limited compared with a case where the temperature T _LNT in 17 (a) , 17 (b) is high. When the rich purging is completed, a predetermined amount of NOx can be caused to be in the LNT 40 remains and the NOx purification rate η can be made equal to or smaller than 90%. Thus, the degree of deterioration of the LNT 40 in the linear region of the line 211 , 221 be rated.

If the NOx purification rate η before a purge, which at S14 is calculated has an error, or if the NOx purification rate after a purge η after that at S24 is calculated has an error, an error in the cleaning rate change Δη can be compensated for within the linear range. Therefore, the degree of deterioration of the LNT 40 can be determined more precisely.

The temperature range (T _{LNT_min} -T _{LNT_max} ) from 17 (a) , 17 (b) is suitable for the second embodiment in which the determination is required to be made frequently.

The line 211 , 221 is out of the linear range when the pre-purge NOx purification rate η is equal to or less than 40%. In this case, the deterioration determination can be prevented from being performed. For example, after S27 of 5 a new step is added in which it is judged whether the pre-purge NOx purification rate η is equal to or less than 40%. If it is determined that the pre-purge NOx purification rate η is equal to or less than 40%, it is prevented from occurring S28 is carried out.

The present invention is not limited to the above embodiments, and the embodiment can be modified within the scope of the present invention.

A deterioration determining section ( S29 ) an exhaust gas purification device determines a deteriorated state of a catalytic converter ( 40 , 41 ), which is in an exhaust gas passage ( 32 ) a machine ( 10 ) is arranged based on a relationship between a change (Δη) of a purification rate (η) and a NOx reduction amount (ΔNOx). The catalyst has the NOx reduction amount (ΔNOx) through reduction treatment. The change (Δη) in the purification rate (η) corresponds to a difference between a pre-reduction rate (ηbefore) before the reduction treatment is carried out and a post-reduction rate (ηafter) after the reduction treatment is carried out.

Claims (11)

An exhaust purification device for an engine (10) comprising: a catalyst (40, 41) disposed in an exhaust passage (32) of the engine (10), the catalyst (40, 41) storing NOx in a lean atmosphere and the stored NOx reduced in a rich atmosphere; a reduction treatment performing section (S15) for causing the catalyst (40, 41) to undergo reduction treatment; a NOx purification rate estimation section (S14, S24) for estimating a NOx purification rate (η) which is a difference between an amount of NOx in an exhaust gas before passing through the catalyst (40, 41) and an amount of NOx in an exhaust gas after a Passing through the catalytic converter (40, 41), the NOx purification rate (η) having a pre-reduction rate (ηbefore) before performing the reduction treatment and a post-reduction rate (ηafter) after performing the reduction treatment; change calculating means (S25) for calculating a cleaning rate change (Δη) of the cleaning rate (η) generated by the reduction treatment based on the pre-reduction rate (ηbefore) and the post-reduction rate (ηafter); a NOx reduction amount estimation section (S21) for estimating a NOx reduction amount (ΔNOx) reduced in the reduction treatment; and a deterioration determination section (S29) for determining a deterioration state of the catalyst (40, 41) based on a relationship between the purification rate change (Δη) and the NOx reduction amount (ΔNOx), characterized in that the deterioration determination section (S29) includes a threshold value setting section (S28) for setting a threshold value (Δηmax) for the purification rate change (Δη) based on the NOx reduction amount (ΔNOx), for determining whether the catalyst (40, 41) is normal or deteriorated, and a comparison section (S29) for comparing the purification rate change ( Δη) with the threshold value (Δηmax). Exhaust gas cleaning device according to Claim 1 further comprising: a NOx sensor (28) disposed downstream of the catalyst (40, 41) to detect a NOx concentration in the exhaust passage (32), wherein the NOx purification rate estimation section (S14, S24) comprises a NOx discharge amount estimating section (S12, S22) for estimating a NOx discharge amount (Xe1, Xe2) discharged from the engine (10), a NOx passage amount estimating section (S13, S23) for estimating a NOx passage amount (XI1, XI2) after passing through the catalyst (40, 41) by using the NOx sensor (28), a pre-reduction NOx purification rate calculator for calculating the pre-reduction rate (ηvor) based on the NOx discharge amount (Xe1) and the NOx passage amount (XII) before the reduction treatment is performed, and post-reduction NOx purification rate calculating means for calculating the post-reduction rate (ηnach) based on the NOx discharge amount (Xe2) and the NOx permeation amount (XI2) after which the reduction treatment is performed. Exhaust gas cleaning device according to Claim 1 wherein the threshold value setting section (S28) sets the threshold value (Δηmax) for the purification rate change (Δη) further based on the pre-reduction NOx purification rate (ηpre). Exhaust gas cleaning device according to Claim 3 wherein the threshold value setting section (S28) corrects the threshold value (Δηmax) for the purification rate change (Δη) based on a temperature (TLNT) of the catalyst (40, 41), and the threshold value setting section (S28) sets the corrected value as the threshold value (Δηmax) . Exhaust gas cleaning device according to Claim 4 wherein the threshold value setting section (S28) corrects the threshold value (Δηmax) to become higher as the temperature (T _LNT ) of the catalyst (40, 41) is changed in one direction to decrease the NOx purification rate (η). Exhaust gas purification device according to one of the Claims 1 to 3 further comprising: a temperature determination section (S26) for determining whether or not a temperature (T _LNT ) of the catalytic converter (40, 41) is in a predetermined temperature range, the deterioration determination section (S29) determining the deteriorated state of the catalytic converter (40 , 41) is determined based on the purification rate change (Δη) when the temperature determination section (S26) determines that the temperature (T _LNT ) of the catalyst (40, 41) is in the predetermined temperature range. Exhaust gas cleaning device according to Claim 6 wherein the NOx purification rate (η) has a characteristic (211, 221) relative to a NOx storage amount (Y) stored in the catalyst (40, 41), and the deterioration determination section (S29) shows the deteriorated state of the catalyst ( 40, 41) is determined based on the cleaning rate change (Δη) calculated based on the pre-reduction rate (ηpre) and the post-reduction rate (ηafter) when the characteristic curve (211, 221) is in a linear range in which the characteristic curve is linear. Exhaust gas cleaning device according to Claim 7 , wherein the reduction treatment performing section (S20) stops the reduction treatment when an air-fuel ratio downstream of the catalyst (40, 41) is richer than a predetermined value, and the predetermined temperature range is set in a manner that a predetermined NOx- Storage amount remains in the linear range when the reduction treatment is stopped. Exhaust gas cleaning device according to Claim 2 wherein the NOx discharge amount estimating section (S12, S22) estimates the NOx discharge amount (Xe1, Xe2) based on a basic NOx discharge amount (Xe0) predetermined for each operating state of the engine (10). Exhaust gas cleaning device according to Claim 2 or 9 further comprising: a time processing section (S107, S108, S109) for causing a time behavior of the NOx passage amount (XI1, XI2) and a time behavior of the NOx discharge amount (Xe1, Xe2) correspond to each other by taking into account a time lag until the NOx discharged from the engine (10) is detected by the NOx sensor (28), the NOx purification rate calculating portion calculating the NOx purification rate (η) based on the NOx - The passage amount (XI1, XI2) and the NOx discharge amount (Xe1, Xe2), which correspond to each other, are calculated. Exhaust gas purification device according to one of the Claims 1 to 10 further comprising: an upstream sensor (26) disposed upstream of the catalyst (40, 41) to detect an air-fuel ratio or an O ₂ concentration; and a downstream sensor (27) disposed downstream of the catalyst (40, 41) to detect an air-fuel ratio or an O ₂ concentration, the NOx reduction amount estimating section (S21) the NOx reduction amount (ΔNOx ) based on detection values detected by the upstream and downstream sensors (26, 27).

Images