Control device for an exhaust purge system of a NOx catalyst

Abstract

Control device (30) for an outlet flushing system, which comprises: a NOx catalyst (23) which is provided in an exhaust passage (20) of an internal combustion engine (10), wherein the NOx catalyst (23) serves as a NOx storage reduction catalyst which stores NOx contained in an exhaust gas and reduces the stored NOx with a reducing agent in order to purify the stored NOx accordingly; and an air-fuel ratio sensor (26) which detects an air-fuel ratio in a downstream section of the exhaust passage (20), wherein the downstream section of the exhaust passage (20) is arranged in the exhaust passage (20) downstream of the NOx catalyst (23), wherein the control device (30), which performs a purge control routine that directs the reducing agent into an upstream section of the outlet passage (20) to cause the NOx catalyst (23) to perform NOx reduction, wherein the upstream section of the outlet passage (20) is arranged upstream of the NOx catalyst (23) in the outlet passage (20), comprises: a sum calculation unit which calculates the sum (T) of values ​​of the reducing agent which have been supplied to the NOx catalyst (23) since the start of the purge control routine; a first unit of determination, which determines whether the sum (T) of the values ​​of the reducing agent calculated by the sum calculation unit is greater than or equal to a final threshold of determination (S); a second determining unit which determines whether the air-fuel ratio detected by the air-fuel ratio sensor (26) in the downstream section of the exhaust passage (20) after the start of the purge control routine is less than or equal to a predetermined value; and a purge termination unit which terminates the purge control routine in response to a previous determination consisting of a first confirmatory determination that the sum (T) of the reducing agent values ​​calculated by the sum calculation unit is greater than or equal to the final determination threshold (S), and a second confirmatory determination that the air-fuel ratio detected by the air-fuel ratio sensor (26) is less than or equal to the predetermined value after the purge control routine has started.

Description

Technical field

[0001] The present disclosure relates to a control device used for an outlet flushing system. General state of the art

[0002] A conventionally known exhaust scavenging device for an internal combustion engine is described in JP 4 759 496 B2. This exhaust scavenging device for an internal combustion engine comprises a three-way catalytic converter and a NOx catalyst located downstream of the three-way catalytic converter. The NOx catalyst stores, for example, the NOx emitted under lean conditions and reduces the stored NOx during rich operation, thus cleaning the exhaust gas.

[0003] In this exhaust scavenging device, a sum of the values ​​of the reducing agent supplied during a scavenging cycle, which reduces NOx under rich conditions, is calculated, and the scavenging cycle routine ends when the sum of the reducing agent values ​​is greater than or equal to a predetermined value. Specifically, a control device for the scavenging device calculates the values ​​of the reducing agent supplied to the scavenging device based on the output of an air-fuel ratio sensor located upstream of the three-way catalyst, predicts the values ​​of the reducing agent consumed in the three-way catalyst, and estimates the sum of the values ​​of the reducing agent supplied to the NOx catalyst. Furthermore, the scavenging cycle routine ends when the estimated sum of the values ​​of the reducing agent supplied to the NOx catalyst is greater than or equal to a predetermined value.

[0004] Further relevant state of the art can be found in the publications US 2008 / 0 131 346 A1, JP 2002 - 115 524 A and US 2013 / 0 047 589 A1. Summary of the invention: Technical problem

[0005] In the exhaust scavenging device of JP 4 759 496 B2, the final determination of the scavenging control routine is achieved solely by estimating the sum of the values ​​of the supplied reducing agent. However, it is known that the values ​​of the reducing agent actually used by the NOx catalyst for reduction vary due to various factors. If such a variation causes the estimated sum of the supplied reducing agent values ​​to exceed the values ​​of the reducing agent actually used, HC emissions and fuel consumption may worsen, which is undesirable.

[0006] The present disclosure was made in view of the aforementioned problems, and a main objective is to provide a control device for an exhaust gas purging system that is capable of suppressing deterioration of an exhaust gas while appropriately reducing NOx.

[0007] The aspect corresponds to a control device for an exhaust gas purging system, which comprises: a NOx catalyst which is provided in an exhaust gas or exhaust passage of an internal combustion engine, wherein the NOx catalyst serves as a NOx storage reduction catalyst which stores NOx contained in an exhaust gas and reduces the stored NOx with a reducing agent in order to clean the stored NOx accordingly;and an air-fuel ratio sensor which detects an air-fuel ratio in a downstream section of the exhaust passage, wherein the downstream section of the exhaust passage is located in the exhaust passage downstream of the NOx catalyst, wherein the control device which performs a scavenging control routine which introduces the reducing agent into an upstream section of the exhaust passage to cause the NOx catalyst to perform NOx reduction, wherein the upstream section of the exhaust passage is located in the exhaust passage upstream of the NOx catalyst, comprises: a summation calculation unit which calculates the sum of values ​​of the reducing agent which have been supplied to the NOx catalyst since the start of the scavenging control routine;a first unit of determination which determines whether the sum of the values ​​of the reducing agent calculated by the sum calculation unit is greater than or equal to a final threshold of determination; a second unit of determination which determines whether the air-fuel ratio detected by the air-fuel ratio sensor in the downstream section of the exhaust passage after the start of the purge control routine is less than or equal to a predetermined value;and a purge termination unit, which terminates the purge control routine in response to a previous determination consisting of a first confirmatory determination that the sum of the reducing agent values ​​calculated by the sum calculation unit is greater than or equal to the final determination threshold, and a second confirmatory determination that the air-fuel ratio detected by the air-fuel ratio sensor is less than or equal to the predetermined value after the purge control routine has started.

[0008] In the present setup, two determination conditions are provided as end conditions for the purge control routine, and the purge control routine terminates at the stage where the earlier of the two end conditions is met. Under the first end condition, the sum of the values ​​(amounts) of reducing agent added to the NOx catalyst since the start of the purge control routine is calculated, and the purge control routine terminates when the sum of the reducing agent values ​​is greater than or equal to a previously set end determination threshold.On the other hand, under the second end-determination condition, the purge control routine ends when the air-fuel ratio detected by the air-fuel ratio sensor provided in the downstream section of the exhaust passage is less than or equal to a predetermined value, that is, when the reducing agent component actually exits to the downstream section of the exhaust passage, the downstream section of the exhaust passage being located in the exhaust passage downstream of the NOx catalyst.

[0009] Here, it is usually more desirable for the purge control routine to terminate due to the first end-destination condition. This is because, under the second end-destination condition, the reducing agent component exits into the downstream section of the exhaust port, which is located downstream of the NOx catalyst. In contrast, under the first end-destination condition, it is possible to terminate the purge control routine without the reducing agent component exiting into the downstream section of the exhaust port, which is located downstream of the NOx catalyst.However, in the first final determination condition, the preset final determination threshold can sometimes be too high due to individual differences, deterioration of the NOx catalyst, setting errors, or similar factors. Therefore, if only the first final determination condition is provided, reducing agent is continuously added until the sum of the reducing agent values ​​is greater than or equal to the preset final determination threshold, even after the NOx catalyst has completed a reduction.

[0010] Therefore, the present device is equipped with a second end-destination condition, and under this condition, the purge control routine terminates when the air-fuel ratio, as detected by the air-fuel ratio sensor located in the downstream section of the exhaust port, is less than or equal to a predetermined value, the downstream section of the exhaust port being located downstream of the NOx catalyst. Consequently, even if the feed rate becomes excessive due to individual variations and the like, excessive discharge of the reducing agent component due to individual variations and the like can be suppressed by terminating the purge control routine when the air-fuel ratio is less than or equal to the predetermined value. Brief description of the illustrations

[0011] The aforementioned task, further tasks, features, and advantages of the present disclosure will become clearer from the following detailed description with reference to the accompanying illustrations. The accompanying illustrations show: Fig. 1 a schematic configuration diagram of an exhaust scavenging system of an internal combustion engine; Fig. 2. A time diagram during lean operation; Fig. 3. A flowchart of a process carried out during lean operation; Fig. 4 a diagram showing the relationship between an exhaust gas flow rate and a determined value; Fig. 5 a diagram showing the relationship between an exhaust gas flow rate and a basic feed rate during a purge control; Fig. 6 a diagram showing the relationship between NOx storage quantity and storage quantity amplification; Fig. 7 a diagram showing the relationship between an exhaust gas flow rate and a target air-fuel ratio during a purge control; Fig. 8 a time diagram during the execution of a flushing control; and Fig. 9 a time diagram showing a target air-fuel ratio during a purge control according to a further embodiment. Description of embodiments

[0012] In the present embodiment, an engine exhaust scavenging system is designed for use in a vehicle-integrated multi-cylinder four-stroke gasoline engine, which corresponds to an internal combustion engine. A schematic configuration diagram of such an engine exhaust scavenging system is shown in Fig. Figure 1 is shown. The diagram below shows only one of the multiple cylinders that comprise machine 10.

[0013] In each cylinder of the machine 10, a piston 11 is mounted in such a way that it can perform a reciprocating motion. Furthermore, a combustion chamber 12 is provided on the upper side of the piston 11 in each cylinder. The combustion chamber 12 is connected via an inlet opening 13 to an inlet passage 14 and via an outlet opening 21 to an outlet passage 20.

[0014] The inlet 14 and outlet 20 of the machine 10 are each equipped with an inlet valve 15, which closes the inlet opening 13, and an outlet valve 16, which closes the outlet opening 21. Opening of the inlet valve 15 causes air to flow from the inlet 14 into the combustion chamber 12, and opening of the outlet valve 16 directs the exhaust gas from the combustion chamber 12 into the outlet 20. The times at which the inlet valve 15 and the outlet valve 16 are opened and closed (valve timing) are variably controlled by a variable valve timing device.

[0015] The combustion chamber 12 is equipped with a spark plug 17. At a desired ignition time, a high voltage is applied to the spark plug 17 via an ignition device with an ignition coil or the like. As a result of the high voltage being applied to the spark plug 17, an arc discharge is generated between counter electrodes, which causes ignition of the air-fuel mixture inside the combustion chamber 12.

[0016] Each cylinder of the machine 10 is equipped with a fuel injection device 18, which supplies fuel directly into the combustion chamber 12. The fuel injection device 18 is connected to a fuel tank via a fuel line (not shown). The fuel in the fuel tank is supplied to the fuel injection device 18 of each cylinder and injected by the fuel injection device 18 into the combustion chamber 12.

[0017] In the exhaust port 20, a three-way catalyst 22, which cleans CO, HC, and NOx in the exhaust gas, and a NOx catalyst 23, which is a NOx storage-reduction type catalyst, are provided in series. The NOx catalyst 23 stores NOx when the air-fuel ratio of the exhaust gas is lean (oxidizing atmosphere) and reduces and cleans the stored NOx when the air-fuel ratio is rich (reducing atmosphere). The reducing agent of the NOx catalyst 23 is CO and / or HC, which are produced during rich combustion. It should be noted that instead of using CO and / or HC produced by rich combustion as the reducing agent, a fuel supply valve may be provided in the exhaust port 20, and the fuel supplied by the fuel supply valve may be used as the reducing agent.

[0018] In the exhaust port 20, an air-fuel ratio sensor 24, which detects the air-fuel ratio of an exhaust gas, is provided upstream of the three-way catalyst 22. A first compound sensor 25 is provided between the three-way catalyst 22 and the NOx catalyst 23, and a second compound sensor 26 is provided in the downstream section of the exhaust port, the downstream section of the exhaust port being located downstream of the NOx catalyst 23. The compound sensors 25 and 26 have the functions of a NOx sensor and an air-fuel ratio sensor (A / F sensor). An NOx sensor and an air-fuel ratio sensor can be provided individually instead of the compound sensors 25 and 26. Furthermore, the NOx sensor function is not required. In addition, various other sensors may be provided in the outlet passage 20.For example, an exhaust gas temperature sensor can be provided between the three-way catalyst 22 and the NOx catalyst 23.

[0019] The outputs of the various sensors are input to an ECU 30. The ECU 30 comprises a microcomputer with a CPU, ROM, RAM, and the like, and executes various control programs stored in the ROM to implement lean control, which sets the air-fuel ratio to a lean state at the time of combustion, and / or a scavenging control routine, which sets the air-fuel ratio to a rich state at the time of combustion and causes the NOx catalyst 23 to perform a reduction. The ECU 30 corresponds to a "control device".

[0020] The following describes the time at which the rinsing control routine is carried out. Fig. Figure 2 is a time diagram showing a state in which NOx is stored and reduced in lean operation, which includes lean control and a purge control routine. Fig. 2 LNT-in represents the NOx concentration in the exhaust gas flowing into the NOx catalyst 23, and LNT-out represents the NOx concentration in the exhaust gas flowing out of the NOx catalyst 23.

[0021] During lean-burn operation, NOx is stored in the NOx catalyst 23, and the amount of NOx stored in the NOx catalyst 23 increases. The amount of NOx stored is calculated using a known method. Specifically, the NOx concentration is calculated by the NOx sensor of the first compound sensor 25. Alternatively, a NOx exhaust concentration from the combustion chamber 12 is calculated based on the operating conditions of the machine 10, such as the machine speed and torque, and the NOx concentration after passing through the three-way catalyst 22 is calculated based on the calculated NOx exhaust concentration. Furthermore, the amount of NOx stored is calculated based on the NOx concentration calculated by these methods. Specifically, the amount of NOx stored is calculated according to the following formula. NOx storage quantity [g] = ∑(NOx concentration [ppm] × exhaust flow rate [mol / s] × NOx molar mass [g / mol] / 1,000,000)

[0022] When the NOx storage quantity reaches a predetermined upper storage limit, the purging control routine is started. Afterwards, the NOx stored in the NOx catalyst 23 is separated from and reduced, and the NOx storage quantity in the NOx catalyst 23 decreases.

[0023] Next, the control system used during lean operation will be described. Fig. 3 is a flowchart that is performed by the ECU 30 and is executed repeatedly by the ECU 30 in a predetermined cycle during lean operation.

[0024] In S10, the ECU 30 determines whether a purge flag is set to 1. The purge flag indicates whether the purge control routine is being executed. If the purge flag is set to 1, the purge control routine, which causes the NOx catalyst 23 to perform a reduction, is executed.

[0025] If the ECU 30 determines in S10 that the purge flag is not set to 1, meaning that lean operation is currently underway, the process continues with S11, and the ECU 30 determines whether a start condition of the purge control routine has been met. Specifically, if the NOx storage quantity in the NOx catalyst 23 reaches a predetermined value during lean operation of the machine 10, the ECU 30 determines that the start condition of the purge control routine has been met. If the ECU 30 determines in S11 that the start condition of the purge control routine has not been met, processing terminates.

[0026] The determined value of the NOx storage quantity at the beginning of the purging control routine can be constant or, as in Fig. As shown in Figure 4, the NOx storage capacity of the NOx catalyst 23 can be determined by a relationship between the exhaust gas flow rate and the temperature of the NOx catalyst. The higher the exhaust gas flow rate, the shorter the reaction time between the exhaust gas and the catalyst, and the lower the NOx storage capacity of the NOx catalyst 23. Consequently, with increasing exhaust gas flow rate during lean-burn operation, the initial NOx storage capacity set at the start of the purge control routine decreases.

[0027] Furthermore, the lower the temperature of the NOx catalyst 23, the lower the catalyst activity and the lower the storage capacity. Consequently, as the temperature of the NOx catalyst 23 decreases, the target value for the NOx storage quantity, which is set at the beginning of the purge control routine, is reduced. When the NOx storage quantity is then greater than or equal to the target value, the start condition of the purge control routine is met. The exhaust gas flow rate and / or the temperature of the NOx catalyst 23 used in calculating the target value can use the average value obtained during lean-burn operation, or they can use the exhaust gas flow rate and / or the temperature of the NOx catalyst 23 at a predetermined time.

[0028] If the ECU 30 determines in S11 that the start condition of the flushing control routine has been met, the flushing flag is set to 1 in S12. Then, an end-determination threshold S for the flushing control routine is set in S13. The end-determination threshold S can be constant or, as in Fig. 5 shown, according to the exhaust gas flow rate and / or the temperature of the NOx catalyst 23, or can be adjusted as shown in Fig. 6 shown, according to the NOx storage quantity, or the like. Fig. Figure 5 is a diagram showing the relationship between the exhaust gas flow rate and the baseline reducing agent feed rates during the purge control routine, and Fig. Figure 6 is a diagram showing the relationship between the NOx storage quantity and the storage quantity increase. The final detection threshold S is equal to the values ​​of the reducing agent added during the purge control routine.

[0029] As in Fig. As shown in Figure 5, for catalysts such as the three-way catalyst 22 and the NOx catalyst 23, the reaction time between the catalyst and the reducing agent is shorter and the reduction efficiency decreases at higher exhaust gas flow rates. When the reduction efficiency decreases, the values ​​of the reducing agent required to reduce the stored NOx increase. Furthermore, the oxygen storage capacity (OSC) of the catalyst decreases with decreasing catalyst temperature; therefore, the values ​​of the reducing agent required to reduce the stored NOx decrease. Consequently, the baseline reduction agent feed rates are determined based on the exhaust gas flow rate and the catalyst temperature using the relationship in Figure 5. Fig. 5. The exhaust gas flow rate and catalyst temperature used can correspond to the exhaust gas flow rate and catalyst temperature at the beginning of the purge control.

[0030] The final determination threshold S corresponds to a value calculated on the basis of the NOx storage quantity, whereby the storage quantity amplification is calculated on the basis of the NOx storage quantity, and the final determination threshold S is determined by multiplying a predetermined base value or one based on Fig. The storage gain is calculated using the base value calculated in section 5. The storage gain corresponds to a coefficient determined according to the fact that the amount of reducing material required in the NOx catalyst 23 increases with increasing NOx storage quantity, and is, in particular, as shown in Fig. 6 shown, set to a higher value when the NOx storage quantity in the NOx catalyst 23 increases.

[0031] Since the three-way catalyst 22 has an oxygen storage capacity (OSC), during the reduction of the oxygen stored in the three-way catalyst 22, the fatty component (HC, CO, and the like) required for the reduction is consumed in a reaction with the oxygen in the three-way catalyst 22, and the fatty component required for the reduction is not supplied to the side of the NOx catalyst 23. For this reason, the NOx catalyst 23 cannot reduce the stored NOx unless a certain amount of the reducing agent is supplied, regardless of the NOx storage capacity of the NOx catalyst 23. Therefore, as in Fig. As shown in Figure 6, the storage gain is shifted upwards (offset) by the amount of the oxygen storage capacity of the three-way catalyst 22. The storage gain is then calculated from the NOx storage quantity using the values ​​shown in Figure 6. Fig. The relationship shown in 6 is calculated.

[0032] Furthermore, in sections S14 to S18, the NOx storage rate of the NOx catalyst 23 is calculated based on the NOx inflow rate to the NOx catalyst 23 and the NOx leakage rate detected by the NOx sensor of the second composite sensor 26. The NOx storage rate can sometimes decrease as a result of an accumulation of unreduced NOx due to a lack of reducing agent. In particular, if a reduction agent deficiency with respect to the NOx storage rate occurred in a previous purge control due to individual differences in the NOx catalyst 23 or the like, unreduced NOx accumulates in the NOx catalyst 23, NOx storage cannot be carried out as planned, and NOx leakage occurs in the downstream section of the outlet passage, which is located in the outlet passage downstream of the NOx catalyst 23.

[0033] Therefore, in S14 to S18, the ECU 30 determines whether a deficiency of reducing agent has occurred in the NOx catalyst 23 with respect to the amount of stored NOx. It should be noted that the processing of S14 to S18 can be carried out at predetermined intervals during lean-burn operation, instead of when a purging condition in S11 is met.

[0034] In S14, the incoming NOx quantity is calculated by integrating the NOx quantity that flowed into the NOx catalyst 23 during lean operation. Specifically, the integration can be based on the NOx quantity detected by the first compound sensor 25, or the integration can be performed by calculating the NOx quantity discharged into the outlet 20 based on the operating state of the engine 10. The incoming NOx quantity can be reset at the end of the purge control routine.

[0035] In S15, the amount of NOx leakage is calculated by integrating the amount of NOx that escaped from the NOx catalyst 23 during lean-burn operation. The total amount of NOx that escaped to the downstream section of the exhaust port is calculated based on the amount of NOx detected by the second compound sensor 26, where the downstream section of the exhaust port is located downstream of the NOx catalyst 23. It should be noted that calculating the total amount of NOx that escaped from the NOx catalyst 23 requires that at least one NOx sensor be provided in the downstream section of the exhaust port, with the downstream section of the exhaust port being located downstream of the NOx catalyst 23. Furthermore, the amount of NOx leakage or NOx leakage quantity can be reset at the end of the purge control routine.

[0036] The NOx storage rate is calculated in S16. The NOx storage rate (%) can be calculated using the following formula. NOx storage rate (%) = (1 - NOx leakage amount / incoming NOx amount) × 100

[0037] If unreduced NOx accumulates, the amount of NOx that can be stored decreases and the amount of NOx leakage increases; therefore, the storage rate deteriorates, and the value representing the storage rate becomes smaller. Note that S14 to S16 correspond to the "storage rate calculation unit".

[0038] In S17, ECU 30 determines whether the storage rate calculated in S16 is less than a predetermined storage rate. If ECU 30 determines in S17 that the storage rate is greater than the predetermined storage rate, processing ends. Conversely, if ECU 30 determines in S17 that the storage rate is less than or equal to the predetermined storage rate, ECU 30 determines that unreduced NOx has accumulated, and in S18 the final determination threshold S is changed so that it becomes greater than the final determination threshold S set in S13, and processing ends. This means that the value of the reducing agent added during the purge control routine is changed so that it will be greater from the next purge control onward, and processing ends.In the procedure for changing the final destination threshold S, a new final destination threshold S is obtained by multiplying the final destination threshold S by a specific ratio greater than 1, or by increasing the final destination threshold S by a specific amount. The amount by which the final destination threshold S is increased is preferably a large value. This is because, rather than allowing an unreduced state to persist due to a lack of feed rate, it is more desirable to make a large change to the final destination threshold S all at once, resulting in an excessive feed rate. This allows the purge control routine to be completed with minimal leakage of the fatty component. Furthermore, the amount of the change can be variable based on the storage rate. S18 also corresponds to the "increase change unit".

[0039] On the other hand, in S10, if the ECU 30 determines that the purge flag is set to 1, the reducing agent is supplied to S20. In particular, fuel is injected by the fuel injection device 18 so that a predetermined rich air-fuel ratio is obtained. Fig. Figure 7 is an example of a map for setting a target air-fuel ratio when feedback control is performed during the purge control routine. For catalysts such as the three-way catalyst 22 and the NOx catalyst 23, the upper limit of the reducing agent values ​​that can be processed per unit time changes depending on the activity of the precious metal in the catalyst. Since, for example, the activity decreases with decreasing catalyst temperature, the reducing agent processing capacity decreases. Therefore, it is desirable that the richness level decreases with decreasing catalyst temperature. It is also desirable that the richness level decreases with increasing exhaust gas flow rate.It is desirable that the flow rate of the reducing agent flowing into the NOx catalyst 23 be kept constant per unit time by adjusting the target air-fuel ratio in this way. Furthermore, the target air-fuel ratio is determined from the exhaust gas flow rate and the catalyst temperature using the values ​​given in [reference]. Fig. The relationship shown in 7 is calculated.

[0040] In S21, the ECU 30 determines whether the air-fuel ratio detected by the second compound sensor 26 is less than or equal to a predetermined value. When the reduction of stored NOx by the NOx catalyst 23 ends, the rich component exits the downstream section of the exhaust port, which is located downstream of the NOx catalyst 23. In this case, the air-fuel ratio detected by the second compound sensor 26 becomes a value representing the rich side. That is, the ECU 30 determines whether the reduction of stored NOx by the NOx catalyst 23 ends and whether the air-fuel ratio detected by the second compound sensor 26 has become a value representing the rich side.If the air-fuel ratio is greater than the predetermined value, that is, if the air-fuel ratio detected by the second compound sensor 26 corresponds to a value close to stoichiometry, and a reduction of the stored NOx is carried out by the NOx catalyst 23, the processing continues with S22.

[0041] In S22, a sum T of values ​​of the added reducing agent since the start of the rinsing control routine is calculated. Specifically, the sum T(n) of reducing agent values ​​is calculated by adding the values ​​of the reducing agent added since the previous processing to the sum T(n-1) of reducing agent values ​​calculated in the previous processing.

[0042] The values ​​of the supplied reducing agent can be calculated from an excess fuel quantity, that is, a value obtained by subtracting the equivalent fuel quantity (the fuel quantity that creates a stoichiometric state with respect to the air quantity) from the fuel injection quantity, or they can be calculated from operating conditions, such as the output value of the air-fuel ratio sensor 24, the air quantity, and the like. Note that S22 corresponds to the "sum calculation unit".

[0043] In S23, the ECU 30 determines whether the sum T of the reducing agent values ​​calculated in S22 is greater than or equal to the final determination threshold S. If the ECU 30 determines that the sum T of the reducing agent values ​​is less than the final determination threshold S set in S13, the reduction of the stored NOx by the NOx catalyst 23 is assumed to be in progress, and the processing ends.

[0044] If the ECU 30 determines in S23 that the sum T of the reducing agent values ​​is greater than or equal to the preset final determination threshold S, the ECU 30 determines in S24 that the reduction of the stored NOx carried out by the NOx catalyst 23 is complete, and the purge control routine ends. Specifically, the purge flag is set to 0 to return to normal control, such as lean control. Additionally, the sum T of the reducing agent values ​​is set to 0. Note that S23 corresponds to the "first determination unit" and S24 corresponds to the "purge termination unit," which terminates the purge control routine based on the result from the first determination unit.

[0045] If, on the other hand, the air-fuel ratio in S21 is less than or equal to the predetermined value, that is, if the rich component is detected in the downstream section of the exhaust port, the process continues with S25, where the downstream section of the exhaust port is located in the exhaust port downstream of the NOx catalyst 23. In S25, the ECU 30 determines that the reduction of the stored NOx performed by the NOx catalyst 23 is complete, and the purge control routine ends. In particular, the purge flag is reset to 0 to return to normal control. Note that S21 corresponds to the "second determination unit" and S25 corresponds to the "purge termination unit," which terminates the purge control routine based on the result from the second determination unit.

[0046] If the purge control routine ends based on the result of the second determination unit rather than the result of the first, the set final determination threshold S is greater than the values ​​of the reducing agent actually used for reduction, and there is concern that the values ​​of the added reducing agent have become too high. If the final determination threshold S is too high, the rich component flows to the downstream section of the exhaust passage, which is undesirable, as the downstream section of the exhaust passage is located downstream of the NOx catalyst 23. Therefore, if the final determination threshold S is too high or excessive, it is necessary to adjust the final determination threshold S so that it is lower from the next purge control onwards.

[0047] Furthermore, if a high concentration (or a high flow rate) of the reducing agent is briefly or temporarily supplied during an execution of the purge control routine, the reducing agent processing capacity of the NOx catalyst 23 is exceeded, and a reducing agent escapes to the downstream section of the exhaust port in a phenomenon known as reducing agent blow-by, the downstream section of the exhaust port being located downstream of the NOx catalyst 23. If reducing agent blow-by occurs, the air-fuel ratio detected in the downstream section of the exhaust port is temporarily less than or equal to the predetermined value, the downstream section of the exhaust port being located downstream of the NOx catalyst 23.If, in such a case, the ECU 30 determines that the reducing agent values ​​are too high and the final determination threshold S is changed, a reducing agent shortage will occur from the rinsing control routine onwards. Therefore, in S26 and S27, the ECU 30 determines whether the feed rate is too high because the final determination threshold S is too high, or whether a blow-by of the reducing agent has occurred.

[0048] In S26, the sum T of reducing agent values ​​is obtained. Specifically, in the same manner as in S22, the sum T(n) of reducing agent values ​​is calculated by adding the reducing agent values ​​added since the previous processing to the sum T(n-1) of reducing agent values ​​calculated in the previous processing. The sum T of reducing agent values ​​is then added to the sum T of reducing agent values ​​from the start of the purge control routine until the endpoint. Furthermore, in S27, the ECU 30 determines whether the sum T of reducing agent values ​​obtained in S26 is greater than a prohibition value. The prohibition value corresponds to a value less than the final determination threshold S and is too small to equal the sum T of reducing agent values ​​if a reduction of the stored NOx carried out by the NOx catalyst 23 completes normally.In particular, the prohibition value corresponds to a value that is approximately half the final determination threshold S. In S27, after determining whether the sum T of reducing agent values ​​is greater than the prohibition value, the sum T of reducing agent values ​​is set to 0. Furthermore, instead of performing the process that sets the sum T of reducing agent values ​​to 0 after the rinsing control routine ends (S24 or S27), it is possible to perform a process that sets the previous sum T of reducing agent values ​​in S12, i.e., at the beginning of the rinsing control routine, to 0.

[0049] If the sum T of the reducing agent values ​​in S27 is less than the prohibition value, ECU 30 determines that a reducing agent blow-by has occurred, and processing ends. Conversely, if the sum T of the reducing agent values ​​in S23 is greater than the prohibition value, the final determination threshold S set in S13 is changed to a lower value in S28, and processing ends. This means that a change is made which reduces the values ​​of the added reducing agent during the purge control routine, starting with the next purge control, and processing ends. The procedure for changing the final determination threshold S involves obtaining a new final determination threshold S by multiplying the final determination threshold S by a specific ratio less than 1, or by reducing the final determination threshold S by a specific amount.The amount by which the final threshold S is reduced is preferably a small value. This is because if a large change is made all at once, there is a risk that the values ​​of the added reducing agent may become insufficient, leading to inadequate reduction. Furthermore, the amount of the change can be varied based on the difference between the final threshold S and the sum T of the reducing agent values, or based on the sum T of the reducing agent values ​​themselves. Additionally, S28 corresponds to the "reduction change unit".

[0050] Fig. Figure 8 is a time diagram during the execution of a flushing control. In particular, Fig. 8(a) a time diagram showing a case in which the rinsing control routine ends because the sum T of values ​​of the reducing agent is greater than or equal to the final determination threshold S, Fig. 8(b) is a time diagram showing a case in which the purge control routine ends because the air-fuel ratio is less than or equal to a predetermined value, and Fig. Figure 8(c) is a timing diagram showing a case where the purge control routine ends because reducing agent blow-by and the air-fuel ratio are less than or equal to a predetermined value. It should be noted that Fig. 8(a) is a time diagram showing a case in which the rinsing control routine ends due to the fulfillment of the determination condition of the first determination unit, and the Fig. Figures 8(b) and (c) are time diagrams showing cases where the rinsing control routine ends due to the fulfillment of the determination condition of the second determination unit.

[0051] First, using Fig. 8(a) describes a case in which the purge control routine terminates because the sum T of reducing agent values ​​is greater than or equal to the final determination threshold S. At time t11, after the purge flag is set to 1 and the purge control routine has started, the air-fuel ratio in the downstream section of the exhaust port, as detected by the second compound sensor 26, becomes a value representing a state close to stoichiometry, with the downstream section of the exhaust port being located downstream of the NOx catalyst 23. It should be noted that the air-fuel ratio does not immediately assume a value close to stoichiometry after the purge control routine has started. This is because it takes some time for the combustion gas to reach the second compound sensor 26 after switching to rich combustion (purge control).Furthermore, after the start of the rinsing control routine, the sum T of the reducing agent values ​​gradually increases.

[0052] While the reduction process at the NOx catalyst 23 continues, at time t12 the sum T of the reducing agent values ​​is greater than or equal to the final determination threshold S, and the purge control routine ends. The purge flag is set to 0, and the sum T of the reducing agent values ​​is also set to 0. Then, after the purge control routine ends and lean control begins, the air-fuel ratio detected by the second compound sensor 26 becomes a value representing the lean condition.

[0053] In many cases, the rinsing control routine ends because the sum T of the reducing agent values ​​is greater than or equal to the final determination threshold S, as in Fig. 8(a). In such a method for terminating the purge control routine, a reduction of the stored NOx, carried out by the NOx catalyst 23, is sufficiently achieved due to the supply of reducing agent up to the previously set final determination threshold S. Furthermore, in comparison to the final determination condition of the second determination unit, the rich component does not exit to the downstream section of the outlet passage, which is preferable, with the downstream section of the outlet passage being located downstream of the NOx catalyst 23.

[0054] The following will be based on Fig. 8(b) describes a case in which the purge control routine terminates because the air-fuel ratio is less than or equal to the predetermined value. At time t21, after the purge flag is set to 1 and the purge control routine has started, the air-fuel ratio in the downstream section of the exhaust port, as detected by the second compound sensor 26, becomes a value representing a state close to stoichiometry, with the downstream section of the exhaust port being located downstream of the NOx catalyst 23. Furthermore, after the purge control routine starts, the sum T of reducing agent values ​​gradually increases.

[0055] While the reduction process at the NOx catalyst 23 continues, at time t22, when the reduction of the stored NOx by the NOx catalyst 23 ends and the rich component flows to the downstream section of the exhaust port, the air-fuel ratio detected by the second compound sensor 26 is less than or equal to the predetermined value, with the downstream section of the exhaust port being located downstream of the NOx catalyst 23. When the air-fuel ratio is less than or equal to the predetermined value, the purge control routine ends. Then the purge flag is set to 0, and the sum T of reducing agent values ​​is set to 0. After the purge control routine ends and lean control begins, the air-fuel ratio detected by the second compound sensor 26 becomes a value representing the lean condition.

[0056] At the stage where the purge control routine is complete, the sum T of the reducing agent values ​​is greater than or equal to the prohibition value, but less than the final determination threshold S. Consequently, it is assumed that the final determination threshold S was too high and the reducing agent values ​​were too high. In this case, the final determination threshold S is adjusted so that it will be lower from the next purge control onwards.

[0057] As in Fig. As shown in Figure 8(b), if the preset final determination threshold S is too high and the reducing agent values ​​are too high, the rich component flows to the downstream section of the exhaust port, and the air-fuel ratio detected by the second compound sensor 26 is less than or equal to the predetermined value before the sum T of reducing agent values ​​is greater than or equal to the final determination threshold S, the downstream section of the exhaust port being located downstream of the NOx catalyst 23. As a result of the termination of the scavenging control routine when the rich component flows to the downstream section of the exhaust port, excessive discharge of the rich component can be suppressed, the downstream section of the exhaust port being located downstream of the NOx catalyst 23.

[0058] In this way, the flushing control routine ends when one of the two end-determination conditions of the flushing control routine, namely the one in Fig. 8(a) final determination condition shown and the one in Fig. The final determination condition shown in section 8(b) is met. Therefore, the purge control routine terminates when the final determination threshold S is correctly set, without the rich component exiting the downstream section of the exhaust port, which is located downstream of the NOx catalyst. Conversely, if the final determination threshold S is too high and the reducing agent values ​​are too high, the amount of rich component leakage can be minimized by terminating the purge control routine when the air-fuel ratio is less than or equal to the predetermined value.

[0059] A case in which the purge control routine terminates due to a blow-by of the reducing agent and the air-fuel ratio is less than or equal to the predetermined value is addressed using Fig. 8(c) described. At time t31, after the purge flag is set to 1 and the purge control routine is started, the air-fuel ratio in the downstream section of the exhaust port, as detected by the second compound sensor 26, becomes a value representing a state close to stoichiometry, with the downstream section of the exhaust port being located downstream of the NOx catalyst 23. Furthermore, after the purge control routine begins, the sum T of the reducing agent values ​​gradually increases.

[0060] At time t32, when the reducing agent concentration briefly increases and a blow-by of the reducing agent occurs, causing the rich component to flow to the downstream section of the exhaust port, the air-fuel ratio detected by the second compound sensor 26 is less than or equal to the predetermined value, with the downstream section of the exhaust port being located downstream of the NOx catalyst 23. When the air-fuel ratio is less than or equal to the predetermined value, the scavenging control routine ends. The scavenging flag is then set to 0, and the sum T of reducing agent values ​​is set to 0. After the scavenging control routine ends and lean control begins, the air-fuel ratio detected by the second compound sensor 26 becomes a value representing the lean condition.

[0061] At the stage where the purge control routine is complete because the sum T of the reducing agent values ​​is less than the prohibition value, it is assumed that a blow-by of the reducing agent has occurred. In this case, a correction that reduces the final determination threshold S from the next purge control is not carried out. This is because it is feared that if the final determination threshold S is reduced after a blow-by has occurred, a state could arise from the next purge control in which NOx is not reduced.

[0062] The present embodiment described above has the following effects.

[0063] In this embodiment, two determination conditions are provided as end conditions of the purge control routine, and the purge control routine terminates at the point where the earlier of the two end conditions is met. Under the first end condition, the sum T of reducing agent values ​​is calculated by integrating the values ​​of the reducing agent supplied to the NOx catalyst 23 since the start of the purge control routine, and the purge control routine terminates when the sum T of reducing agent values ​​is greater than or equal to the preset end threshold S.

[0064] On the other hand, under the second end-destination condition, the purge control routine continues until the air-fuel ratio detected by the second compound sensor 26, which is provided in the downstream section of the exhaust passage, is less than or equal to the predetermined value, that is, until the reducing agent component actually exits the downstream section of the exhaust passage, the downstream section of the exhaust passage being located downstream of the NOx catalyst 23. Therefore, it is usually more desirable for the purge control routine to terminate due to the first end-destination condition.This is because, in the second end-of-determination condition, the reducing agent component exits into the downstream section of the outlet passage, the downstream section of the outlet passage being located in the outlet passage downstream of the NOx catalyst 23, whereas in the first end-of-determination condition, it is possible to terminate the purge control routine without the reducing agent component exiting into the downstream section of the outlet passage, the downstream section of the outlet passage being located in the outlet passage downstream of the NOx catalyst 23.

[0065] However, under the first final determination condition, the preset final determination threshold S can cause the reducing agent values ​​to become too high due to individual differences, deterioration of the NOx catalyst 23, setting errors, or similar factors. Therefore, if only the first final determination condition is provided, the reducing agent is continuously supplied until the sum of the reducing agent values ​​is greater than or equal to the preset final determination threshold S, even after the reduction of the stored NOx by the NOx catalyst 23 has ended.Therefore, in the present embodiment, a second end-determination condition is provided, according to which the purge control routine is terminated when the air-fuel ratio detected by the second compound sensor 26, which is provided in the downstream section of the exhaust passage, is less than or equal to a predetermined value, wherein the downstream section of the exhaust passage is located in the exhaust passage downstream of the NOx catalyst 23. Consequently, even if the feed rate has become too high due to a setting error or the like, excessive discharge of the reducing agent component caused by the setting error or the like can be suppressed by terminating the purge control routine when the air-fuel ratio is less than or equal to the predetermined value.

[0066] If the purge control routine ends because the air-fuel ratio detected by the second compound sensor 26 is less than or equal to the predetermined value before the ECU 30 determines that the sum T of the reducing agent values ​​is greater than or equal to the set final determination threshold S, it is further assumed that a preset value of the final determination threshold S is too large and that the reducing agent values ​​are too high due to individual differences or deterioration of the NOx catalyst 23 or due to setting errors or the like.In this case, a reduction change occurs, modifying the preset value of the final determination threshold S used in the rinsing control routine to an updated value used in the next rinsing control routine. This updated value is lower than the preset value of the final determination threshold used in the previous rinsing control routine. Consequently, the final determination threshold S decreases from the next rinsing control onward, resulting in lower reducing agent levels and preventing them from becoming too high.

[0067] If the sum T of reducing agent values ​​is less than the final determination threshold S and also less than the prohibition value, there is a high probability that NOx reduction in the NOx catalyst 23 is incomplete, and it is assumed that the phenomenon where the air-fuel ratio is less than or equal to the predetermined value was caused by a brief period of high concentration output. If a correction is made in this case, the reduction amount will be insufficient during the next purge control. For this reason, if the sum T of reducing agent values ​​is less than the prohibition value, changes that reduce the final determination threshold S are prohibited. Consequently, a deficiency or shortfall in the reducing agent due to faulty learning can be prevented.

[0068] Furthermore, due to individual differences in the NOx catalyst 23 and the like, a deficiency of reducing agent may occur depending on the set final determination threshold S. If a deficiency of reducing agent occurs, unreduced NOx accumulates in the NOx catalyst 23, and NOx leakage occurs in the downstream section of the exhaust passage, which is located downstream of the NOx catalyst 23. Therefore, the second compound sensor 26, capable of detecting NOx, is provided in the downstream section of the exhaust passage, and the NOx leakage quantity is measured to calculate the NOx storage rate based on the NOx leakage quantity, with the downstream section of the exhaust passage being located downstream of the NOx catalyst 23.If the NOx storage rate is less than or equal to a predetermined storage rate, an increase is implemented, changing the preset value of the final destination threshold S used in the purge control routine to a controlled value used in the next purge control routine. This controlled value of the final destination threshold used in the next purge control routine is greater than the preset value of the final destination threshold used in the purge control routine. Consequently, the levels of the added reducing agent are increased from the next purge control onward, suppressing the accumulation of unreduced NOx and preventing a deterioration in NOx storage performance. <Weitere Ausführungsformen>

[0069] The present disclosure is not limited to the embodiment described above and may, for example, also be implemented as follows.

[0070] - Instead of being kept constant, the target air-fuel ratio can be set during the purge control routine so that the richness level becomes high until the NOx reduction carried out by the three-way catalyst 22 ends. Fig. Figure 9 is a time diagram showing the target air-fuel ratio during purge control.

[0071] After an initial control action, which causes the exhaust gas to be richer, is performed following the start of the purge control routine, the air-fuel ratio, as detected by the first compound sensor 25 located downstream of the three-way catalyst 22, changes from a value representing the lean state to a value representing the stoichiometric state. During the period in which the initial control action is performed and a certain amount of stored oxygen remains in the three-way catalyst 22, the rich component (HC, CO, and the like) in the exhaust gas is consumed in a reaction with the oxygen stored in the three-way catalyst 22. Consequently, the air-fuel ratio of the exhaust gas flowing from the three-way catalyst 22 becomes a value close to stoichiometry.Then, during the period in which the first control is carried out, and until the remaining amount of oxygen stored in the three-way catalyst 22 becomes small, the air-fuel ratio detected by the first compound sensor 25 is maintained at a value close to stoichiometry.

[0072] When the amount of oxygen stored in the three-way catalyst 22 decreases, the concentration of the rich component passing through the three-way catalyst 22 and being fed to the NOx catalyst 23 increases. As a result, the air-fuel ratio detected by the first compound sensor 25 changes to a value representing the rich side. When the first compound sensor 25 detects that the air-fuel ratio has changed to a value representing the rich side, a second control system switches over, and the richness level is reduced. Because the air-fuel ratio is switched over in this way, NOx reduction performed by the three-way catalyst 22 is carried out quickly, and the richness level continues to decrease even after the NOx reduction performed by the three-way catalyst 22 has finished.Consequently, the leakage rate of the rich component can be reduced if a leakage of the rich component from the NOx catalyst 23 occurs due to an excess of the reducing agent.

[0073] If the target air-fuel ratio is set in this way in two stages, furthermore, values ​​that can be achieved by multiplying the in Fig. The target air-fuel ratio obtained in section 7 can be determined using different coefficients for the first and second control stages, as the target air-fuel ratios are used. The coefficient of the first control stage is set higher than the coefficient of the second control stage. Furthermore, in addition to the one in Fig. In the case of two stages shown in 7, the target air-fuel ratio during the purge control routine can be implemented by dividing it into a plurality of stages, or it can be changed linearly.

[0074] In the case where the target air-fuel ratio is set in two stages, the processing of S20 is carried out. Fig. 3. The ECU 30 determines whether the first or second control is currently running, and then the target air-fuel ratio is set and the reducing agent is added. Furthermore, the sum T of the reducing agent values ​​in S26 is processed by... Fig. 8, and in S27, instead of determining whether the sum T of reducing agent values ​​is greater than or equal to the prohibition value, the ECU 30 can determine that reducing agent blow-by has occurred if the air-fuel ratio during the first control is less than or equal to the predetermined value. This is because reducing agent blow-by is likely to occur during the first control, resulting in the reducing agent being set to a high richness level. The procedure for determining whether blow-by has occurred, depending on whether the time when the air-fuel ratio is less than or equal to the predetermined value occurred during the first control, is described in detail. Instead of S26 and S27, the ECU 30 determines whether the first control is performed.If ECU 30 determines that the first control is performed, that is, if the air-fuel ratio during the first control is less than or equal to the predetermined value, ECU 30 determines that blow-by has occurred and the processing ends without performing a correction. If, on the other hand, ECU 30 determines that the first control is not performed, that is, if the air-fuel ratio during the second control is less than or equal to the predetermined value, a change is made in S28 that lowers the final determination threshold S. Such a determination can be made after determining whether the sum T of the reducing agent values ​​is greater than or equal to the prohibition value.

[0075] In this way, if the three-way catalyst 22 is positioned upstream of the NOx catalyst 23, almost no reducing agent is supplied to the NOx catalyst 23 after the purge control routine has started, unless the NOx reduction carried out by the three-way catalyst 22 has ended. Therefore, in order for the NOx reduction by the three-way catalyst 22 to be carried out quickly and at the same time suppress leakage of the rich component that would occur in the case of an excess, it is preferable for the fat content to be initially high and for the fat content to be low after the NOx reduction carried out by the three-way catalyst 22 has ended.Therefore, the first control, which causes the reducing agent's richness level to be set to a high value, is performed at the beginning of the purge control routine, and the second control, which causes the exhaust gas's richness level to be lower than the high reducing agent richness level, is performed afterward. Since reducing agent blow-by is likely to occur during the first control, which causes the reducing agent's richness level to be set to the high value, it can be assumed that reducing agent blow-by has occurred if the air-fuel ratio in the downstream section of the exhaust port is less than or equal to the predetermined value during the period in which the first control is performed, with the downstream section of the exhaust port being located in the exhaust port downstream of the NOx catalyst 23.Therefore, if the purge control routine ends as a result of the air-fuel ratio in the downstream section of the exhaust port being less than or equal to the predetermined value during the first control, a correction of the feed rate is prohibited, where the downstream section of the exhaust port is located in the exhaust port downstream of the NOx catalyst 23.

[0076] This can prevent a shortage of reducing agents from occurring due to faulty learning. - In the purge control routine, in addition to performing an air-fuel ratio feedback control that brings the actual air-fuel ratio into line with the target air-fuel ratio, a configuration can be used in which an open control sets a fuel quantity, which serves as a rich air-fuel ratio, as the fuel injection quantity. - When machine 10 is switched from lean operation to normal operation (operation in stoichiometric state), the ECU 30 can process Fig. 3. This can be carried out independently of the cycle, and in S11 the ECU 30 can determine that the start condition of the purge control routine has been met, and then carry out the purge control routine. In this case, when switching from lean operation to stoichiometric operation, the change can occur in the operating state after the NOx stored in the NOx catalyst 23 and the like have been processed in advance. - An oxidation catalyst can be used instead of the three-way catalyst 22. In addition, another catalyst can be provided alongside the three-way catalyst 22 and the NOx catalyst 23. - A turbocharger can be provided between the outlet passage 20 and the inlet passage 14, which compresses the air on the inlet side through the exhaust flow. - The internal combustion engine is not limited to a gasoline engine which injects fuel into the combustion chamber 12, and can be a gasoline engine which injects fuel into the intake passage 14, or a diesel engine. The control unit (control device) and the method as described in the present disclosure can be implemented by a dedicated computer provided by configuring a memory and a processor programmed to perform one or more functions embodied by a computer program. Alternatively, the control device and the method as described in the present disclosure can be implemented by a dedicated computer provided by configuring a processor with one or more dedicated hardware logic circuits.Furthermore, the control unit and the method as described in the present disclosure can be implemented by one or more dedicated computers configured by a combination of a memory and a processor programmed to perform one or more functions, and a processor configured by one or more hardware logic circuits. In addition, the computer program can be stored as instructions to be executed by a computer on a computer-readable, non-transitory, tangible recording medium.

[0077] The present disclosure has been described with reference to the embodiments; however, it is evident that the present disclosure is not limited to these embodiments and structures. The present disclosure also includes various modifications and equivalent variations. Furthermore, various combinations and configurations, and further combinations and configurations comprising only one additional element, two or more additional elements, or part of an additional element, also fall within the basic concept and scope of protection of the present disclosure.

Claims

[1] Control device (30) for an outlet flushing system, which comprises: a NOx catalyst (23) which is provided in an exhaust passage (20) of an internal combustion engine (10), wherein the NOx catalyst (23) serves as a NOx storage reduction catalyst which stores NOx contained in an exhaust gas and reduces the stored NOx with a reducing agent in order to purify the stored NOx accordingly; and an air-fuel ratio sensor (26) which detects an air-fuel ratio in a downstream section of the exhaust passage (20), wherein the downstream section of the exhaust passage (20) is arranged in the exhaust passage (20) downstream of the NOx catalyst (23), wherein the control device (30), which performs a purge control routine that directs the reducing agent into an upstream section of the outlet passage (20) to cause the NOx catalyst (23) to perform NOx reduction, wherein the upstream section of the outlet passage (20) is arranged upstream of the NOx catalyst (23) in the outlet passage (20), comprises: a sum calculation unit which calculates the sum (T) of values ​​of the reducing agent which have been supplied to the NOx catalyst (23) since the start of the purge control routine; a first unit of determination, which determines whether the sum (T) of the values ​​of the reducing agent calculated by the sum calculation unit is greater than or equal to a final threshold of determination (S); a second determining unit which determines whether the air-fuel ratio detected by the air-fuel ratio sensor (26) in the downstream section of the exhaust passage (20) after the start of the purge control routine is less than or equal to a predetermined value; and a purge termination unit which terminates the purge control routine in response to a previous determination consisting of a first confirmatory determination that the sum (T) of the reducing agent values ​​calculated by the sum calculation unit is greater than or equal to the final determination threshold (S), and a second confirmatory determination that the air-fuel ratio detected by the air-fuel ratio sensor (26) is less than or equal to the predetermined value after the purge control routine has started. [2] Control device (30) for the outlet flushing system according to claim 1, further comprising: a reduction change unit configured such that, when the purge control routine terminates in response to the second confirmatory determination that the air-fuel ratio detected by the air-fuel ratio sensor (26) is less than or equal to the predetermined value after the purge control routine has started, occurs earlier than the first confirmatory determination that the sum (T) of the reducing agent values ​​calculated by the sum calculation unit is greater than or equal to the final determination threshold (S), changes a preset final determination threshold (S) value used in the control terminal routine to an updated value used in a subsequent purge control routine. where the updated value of the final determination threshold (S) used in the next routine of the flushing control is less than the preset value of the final determination threshold (S) used in the routine of the flushing control. [3] Control device (30) for the outlet flushing system according to claim 2, wherein the reduction change unit is configured such that, when the purge control routine terminates in response to the second confirmatory determination that the air-fuel ratio detected by the air-fuel ratio sensor (26) is less than or equal to the predetermined value after the purge control routine has started, it prevents a change in the preset value of the final determination threshold (S) used in the purge control routine when it determines that the sum (T) of the reducing agent values ​​calculated by the sum calculation unit is less than a prohibition value. where the prohibition value is set to be less than the preset value of the final determination threshold (S). [4] Control device (30) for the outlet flushing system according to claim 2 or 3, wherein a three-way catalyst (22) is provided in the upstream section of the exhaust passage (20), wherein the upstream section of the exhaust passage (20) is arranged in the exhaust passage (20) upstream of the NOx catalyst (23); the reducing agent corresponds to a fuel; the routine of the flushing control exhibits: a first control routine which causes the exhaust gas to be set to a high richness level, so that a predetermined reduction operation is carried out by the three-way catalyst (22); and a second control routine, wherein the control device (30) executes the second control routine after the first control routine has ended, in order to cause a value of the exhaust gas richness to be lower than the high value of the exhaust gas richness; and the reduction change unit is configured such that when the purge control routine terminates in response to the second confirmatory determination that the air-fuel ratio detected by the air-fuel ratio sensor (26) during the first control routine is less than or equal to the predetermined value, it prevents a change in the preset value of the final determination threshold (S) used in the purge control routine. [5] Control device (30) for the outlet flushing system according to one of claims 1 to 4, wherein a NOx sensor (26) is provided in the downstream section of the exhaust passage (20), wherein the downstream section of the exhaust passage (20) is arranged in the exhaust passage (20) downstream of the NOx catalyst (23); and the control device (30) further comprises: a storage rate calculation unit which, during lean-burn operation in which the air-fuel ratio of the internal combustion engine (10) is lean, calculates a NOx storage rate of the NOx catalyst (23) based on a NOx inflow quantity to the NOx catalyst (23) and a NOx leakage quantity detected by the NOx sensor (26); and an increase-change unit configured such that, in response to the fact that the NOx storage rate calculated by the storage rate calculation unit is less than or equal to a predetermined storage rate, it changes the preset value of the final determination threshold (S) used in the purge control routine to a controlled value used in a subsequent purge control routine, where the controlled value of the final determination threshold (S) used in the next routine of the flushing control is greater than the preset value of the final determination threshold (S) used in the routine of the flushing control.

Images