Combustion control system of internal combustion engine

Abstract

In an internal combustion engine equipped with a fuel injection system and an exhaust purifying device, such as NOx trap catalyst and/or diesel particulate filter (DPF), a control unit is provided. The control unit controls the fuel injection system to permit the engine to have a predetermined combustion mode in which, under a predetermined condition of the exhaust purifying device, the fuel injection system causes the engine to carry out a main combustion to produce a torque and at least one preliminary combustion prior to the main combustion. The at least one preliminary combustion is effected at a timing in the vicinity of top dead center of compression stroke, and the main combustion is effected at a first timing after completion of the preliminary combustion. The control unit further controls the fuel injection system in such a manner that upon switching of the engine operation from a previous combustion mode to the predetermined combustion mode, the main combustion that takes place after completion of the preliminary combustion is effected at a second timing that is retarded as compared with the first timing.

Description

BACKGROUND OF THE INVENTION

1. Field of the Invention

The present invention relates in general to a control system for controlling operation of an internal combustion engine, and more particularly to a combustion control system for controlling combustion of an internal combustion engine that is equipped at an exhaust system thereof with an exhaust purifying device.

2. Description of the Related Art

In order to clarify the task of the present invention, a known combustion control system of an internal combustion engine will be briefly discussed prior to describing the detail of the present invention, with the aid of the disclosure of Laid-open Japanese Patent Application (Tokkai) 2000-320386.

Disclosed by the publication is a combustion control system of a diesel engine that has an exhaust purifying device installed in an exhaust system thereof. Upon need of increasing the temperature of the exhaust purifying device, the combustion control system controls a fuel injection device in such a manner that each fuel injector injects a reference quantity of fuel, that meets the torque needed by the engine, into a corresponding cylinder at a timing in the vicinity of TDC (viz., top dead center) of compression stroke while splitting the fuel injection by three, that is, first, second and third fuel injection shots.

SUMMARY OF THE INVENTION

In the combustion control system of the publication, the split fuel injection is so controlled as to keep the combustion of fuel that has been injected into the cylinder in the fuel splitting manner. In such control, the second or third fuel shot is directed to a flame of the combustion of fuel that has been injected by the first or second fuel shot, and thus the fuel injected by the second and third shots is subjected to a combustion that includes mainly a diffusion combustion. If, under this combustion condition, the exhaust air/fuel ratio is made richer, deterioration of emission characteristic, particularly smoke characteristic, becomes severe.

Accordingly, the present invention is provided by taking the above-mentioned fact into consideration and aims to provide a combustion control system of an internal combustion engine equipped at an exhaust system with an exhaust purifying device, that can suppress or at least minimize the deterioration of emission characteristic even when the exhaust air/fuel ratio is turned rich for increasing the temperature of the exhaust purifying device and that can, upon switching of a combustion mode of the engine, induce a desired exhaust air/fuel ratio without inducing undesirable torque fluctuation.

According to the present invention, there is provided a combustion control system of an internal combustion engine equipped at an exhaust system with an exhaust purifying device, that, upon sensing a given condition of the exhaust purifying device, starts a predetermined combustion control for controlling the fuel injection to the engine in a manner to cause the engine to carry out both at least one preliminary combustion and a main combustion that follows the preliminary combustion to produce a main torque, wherein the at least one preliminary combustion is carried out at a timing in the vicinity of TDC (viz., top dead center) of compression stroke and the main combustion is forced to start after completion of the preliminary combustion, and wherein upon switching to the predetermined combustion control from a previous combustion control, the switching to the main combustion after switching of the preliminary combustion is retarded by a certain degree.

Because the main combustion is forced to start after completion of the preliminary combustion, the main combustion can include mainly a premixed combustion, so that deterioration of emission characteristic, that would be caused by the enrichment of exhaust air/fuel ratio, can be suppressed or at least minimized.

Furthermore, since the incylinder temperature is increased by means of the preliminary combustion, start timing of the main combustion can be retarded and thus the exhaust temperature can be increased.

Accordingly, by switching the combustion mode as mentioned hereinabove, the regeneration of the exhaust purifying device, that would be achieved by enrichment of the exhaust air/fuel ratio and/or increase of the exhaust temperature, is assuredly carried out without inducing deterioration of emission characteristic of the engine.

If, upon switching to the predetermined combustion control due to occurrence of the given condition of the exhaust purifying device, a certain fluctuation is shown in the intake system, the preliminary combustion that is effected at the timing in the vicinity of TDC of compression stroke is hardly affected by such fluctuation, and thus, the exhaust air/fuel ratio can be switched to a target ratio instantly. While, since the main combustion that is effected after the compression stroke is easily affected by such intake system fluctuation, instant switching to the target air/fuel ratio tends to induce unstable ignition and thus intends to induce incomplete combustion of the mixture in the cylinder.

Accordingly, in the present invention, at first, the preliminary combustion is carried out to produce a compression end temperature that enables the main combustion, and then, switching to the main combustion is carried out with a certain retardation. With this, the switching to the main combustion is effected after the intake system fluctuation has been reduced, and thus, the main combustion can be stably effected establishing a smoothed switching of the combustion control.

In accordance with a first aspect of the present invention, there is provided a combustion control system of an internal combustion engine. The combustion control system comprises a fuel injection system provided an intake system of the engine; an exhaust purifying device provided at an exhaust system of the engine; and a control unit that controls the fuel injection system to permit the engine to have a predetermined combustion mode, the predetermined combustion mode being a mode in which, under a predetermined condition of the exhaust purifying device, the fuel injection system causes the engine to carry out a main combustion to produce a torque and at least one preliminary combustion prior to the main combustion, the at least one preliminary combustion being effected at a timing in the vicinity of top dead center of compression stroke, the main combustion being effected at a first timing after completion of the preliminary combustion, wherein the control unit further controls the fuel injection system in such a manner that upon switching of the engine operation from a previous combustion mode to the predetermined combustion mode, the main combustion that takes place after completion of the preliminary combustion is effected at a second timing that is retarded as compared with the first timing.

In accordance with a second aspect of the present invention, there is provided a method for controlling an internal combustion engine that is equipped at an intake system with a fuel injection system and at an exhaust system with an exhaust purifying device. The method comprises controlling the fuel injection system to permit the engine to have a predetermined combustion mode in which, under a predetermined condition of the exhaust purifying catalyst, the fuel injection system causes the engine to carry out a main combustion to produce a torque and at least one preliminary combustion prior to the main combustion; controlling the fuel injection system in such a manner that the at least one preliminary combustion is effected at a timing in the vicinity of top dead center of compression stroke, and the main combustion is effected at a first timing after completion of the preliminary combustion; and further controlling the fuel injection system in s such a manner that upon switching of the engine operation from a previous combustion mode to the predetermined combustion mode, the main combustion that takes place after completion of the preliminary combustion is effected at a second timing that is retarded as compared with the first timing.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1 is a schematic diagram of a combustion control system of an internal combustion engine, according to the present invention;

FIG. 2 is a general flowchart showing the entire operation steps executed by a control unit employed in the present invention for effecting a combustion control;

FIG. 3 is a flowchart of a first branch extending from the general flowchart, showing operation steps of “DPF regeneration mode” executed by the control unit;

FIG. 4 is a flowchart of a second branch extending from the general flowchart, showing operation steps of “sulfur poisoning recovery mode” executed by the control unit;

FIG. 5 is a flowchart of a third branch extending from the general flowchart, showing operation steps of “rich spike mode” executed by the control unit;

FIG. 6 is a flowchart of a fourth branch extending from the general flowchart, showing operation steps of “melt down suppression mode” executed by the control unit;

FIG. 7 is a flowchart of a fifth branch extending from the general flowchart, showing operation steps executed by the control unit for determining the order of priority of regeneration in case of presence of request for DPF regeneration;

FIG. 8 is a flowchart of a sixth branch extending from the general flowchart, showing operation steps executed by the control unit for determining the order of priority of regeneration in case of presence of request for sulfur poisoning recovery;

FIG. 9 is a flowchart of a seventh branch extending from the general flowchart, showing operation steps executed by the control unit for turning “rq-DPF-flag” to 1;

FIG. 10 is a flowchart of an eighth branch extending from the general flowchart, showing operation steps executed by the control unit for turning “rq-desul-flag” to 1;

FIG. 11 is a flowchart of a ninth branch extending from the general flowchart, showing operation steps executed by the control unit for turning “rq-sp-flag” to 1;

FIG. 12 is a flowchart of a tenth branch extending from the general flowchart, showing operation steps of “catalyst activation promotion mode” executed by the control unit;

FIG. 13 is a time chart showing a combustion condition exhibited by a first reference (1);

FIG. 14 is a time chart showing a combustion condition exhibited by a second reference (2);

FIG. 15 is a time chart showing a combustion condition exhibited by the present invention (3);

FIG. 16 is a graph showing a comparison of emission characteristics effected by the first and second references ((1), (2)) and the present invention (3);

FIGS. 17A, 17B, 17C and 17D are graphs showing a relation between a main fuel injection timing and the emission characteristic;

FIG. 18 is a map showing a target fuel injection timing for a preliminary combustion;

FIG. 19 is a map showing a target fuel injection quantity for the preliminary combustion;

FIG. 20 is a graph showing a target fuel injection timing for a main combustion;

FIG. 21 is a time chart similar to FIG. 15, but showing another combustion condition exhibited by the present invention;

FIGS. 22A and 22B are flowcharts of two types of switching, each showing operation steps executed by the control unit for switching to a split retard combustion mode (SRCM);

FIG. 23 is a graph showing a relation between PM accumulation quantity and a target “λ” during regeneration;

FIG. 24 is a map used for looking up a target intake air quantity for an engine operation of “λ=1”;

FIG. 25 is a graph showing a relation between a main fuel injection timing and a torque correction factor “K1”;

FIG. 26 is a graph showing a relation between the target “λ” and a torque correction factor “K2”;

FIG. 27 is a map used for looking up a target intake air quantity for a rich spike operation; and

FIG. 28 is a graph showing an area where both DPF regeneration and sulfur poisoning recovery are possible.

DETAILED DESCRIPTION OF THE INVENTION

In the following, the present invention will be described in detail with reference to the accompanying drawings.

Referring to FIG. 1, there is schematically shown a combustion control system of a diesel engine 1 as an internal combustion engine, which is an embodiment of the present invention.

As shown, an intake passage 2 of engine 1 has at its upstream part a compressor 3a of a turbo-charger 3. Air led into intake passage 2 is super-compressed by compressor 3a, cooled by an intercooler 4, flow-controlled by an intake throttle valve 5 and fed to respective combustion chambers through a collector 6. Fuel fed into a fuel pipe is compressed by a fuel injection pump 7 and led to a common rail 8 and directly injected to the respective combustion chambers from corresponding fuel injection valves 9. The air and the fuel thus fed to each combustion chamber are subjected to a combustion caused by a compression ignition. Exhaust gas produced as a result of the combustion is led to an exhaust passage 10.

Fuel injection pump 7, common rail 8 and fuel injection valves 9 thus constitute a common rail type fuel injection device.

Part of the exhaust gas led to exhaust passage 10 is returned back to the intake system as EGR gas through an EGR passage 11 and an EGR valve 12. The most part of the exhaust gas drives an exhaust turbine 3b of turbocharger 3.

At a part of exhaust passage 10 downstream of exhaust turbine 3b, there is disposed a NOx trap catalyst 13 for purifying the exhaust gas. That is, NOx trap catalyst 13 functions to trap NOx in the exhaust gas when the exhaust air/fuel ratio is lean and to release the trapped NOx (viz., desorption of the trapped NOx) therefrom when the exhaust air/fuel ratio is rich. To NOx trap catalyst 13, there is attached an oxidation catalyst (viz., precious metal catalyst) for oxidizing HC and CO in the exhaust gas.

At a part of exhaust passage 10 downstream of NOx trap catalyst 13, there is disposed a diesel particulate filter (or DPF) 14 for cleaning the exhaust gas. That is, DPF 14 functions to collect particulate matter (or PM) in the exhaust gas. Also to DPF 14, there is attached an oxidation catalyst (viz., precious metal catalyst) for oxidizing HC and CO in the exhaust gas.

If desired, NOx trap catalyst 13 and DPF 14 may be arranged at mutually reversed positions. Furthermore, if desired, to DPF 14, there may be attached NOx trap catalyst 13 to constitute an integrated exhaust gas purifying unit.

Denoted by numeral 20 is a control unit constructed by a microcomputer, that comprises CPU, RAM, ROM and input and output interfaces. To control unit 20, there are fed information signals from an engine speed sensor 21 that senses an engine speed “Ne”, an accelerator position sensor 22 that senses an accelerator position “APO”, an air flow meter 23 that senses an intake air quantity “Qac” and a water temperature sensor 24 that senses the temperature “Tw” of an engine cooling water.

To control unit 20, there are further fed information signals from a catalyst temperature sensor 25 that senses the temperature of NOx trap catalyst 13, an exhaust pressure sensor 26 that senses a pressure of the exhaust gas at an inlet side of DPF 14, a DPF temperature sensor 27 that senses the temperature of DPF 14, and an exhaust air/fuel ratio sensor 28 that senses an exhaust air/fuel ratio of the exhaust gas at an outlet side of DPF 14.

The exhaust air/fuel ratio will be referred to as “exhaust-λ” and the numeral value of the ratio represents an excess rate of air.

If desired, the temperature of NOx trap catalyst 13 and that of DPF 14 may be indirectly derived from the temperature of the exhaust gas that flows downstream of such device 13 and/or 14.

By processing the information signals fed thereto, control unit 20 issues instruction signals to fuel injection valves 9 for controlling the fuel injection quantity and fuel injection timing, to intake throttle valve 5 for controlling an open degree of the same and to EGR valve 12 for controlling an open degree of the same.

As will be described in detail hereinafter, in addition to the above-mentioned control, control unit 20 carries out an exhaust gas purifying control that includes a control for regeneration of DPF 14 by burning collected PM on DPF 14, a control for desorption or release of trapped NOx from NOx trap catalyst 13 and a control for recovery of sulfur poisoning of NOx trap catalyst 13.

For carrying out the exhaust gas purifying control, programmed operation steps are executed by control unit 20, which are shown in flowcharts in FIGS. 2 to 12.

The flowchart shown in FIG. 2 is a general flowchart that shows the entire of operation steps executed by control unit 20.

At step S1, the information signals from the various sensors are read. That is, the engine speed “Ne”, the accelerator position “APO”, the intake air quantity “Qac”, the temperature of NOx trap catalyst 13, the pressure of the exhaust gas at an inlet side of DPF 14, the temperature of DPF 14 and the exhaust-λ at the outlet side of DPF 14 are read.

At step S2, judgment is carried out as to whether NOx trap catalyst 13 is under activation (or hot) condition or inactivation (or cool) condition. If the temperature of the catalyst 13 is lower than “T5” that is the lowest activation temperature of the catalyst 13, the operation flow goes to an after-mentioned “catalyst activation promotion mode” of FIG. 12 judging that the catalyst 13 is still under the inactivation condition. While, if the temperature of the catalyst 13 is equal to or higher than “T5”, the operation flow goes to step S3 judging that the catalyst 13 is under the activation condition (viz., the activation condition achieved after completion of an after-mentioned catalyst activation promotion mode).

If desired, the activation condition judgment of NOx trap catalyst 13 may be carried out based on the concentration of at least one of HC and CO detected at an outlet side of the catalyst 13. That is, if the concentration of HC or CO is higher than a threshold value, judgment is so made that the catalyst 13 is under the inactivation condition. The concentration judgment of HC or CO may be carried out based on an output level of the air/fuel ratio sensor 28. The estimation is so made that the concentration of HC or CO increases as the rich degree of the exhaust air/fuel ratio increases.

At step S3, the quantity of NOx trapped by and accumulated on NOx trap catalyst 13 (viz., NOx accumulation quantity) is calculated. As is described in column 8 of U.S. Pat. No. 5,473,887, the NOx accumulation quantity may be estimated from an integrated value of engine speed, or a travel distance of the vehicle. In case of using the integrated value of engine speed for the estimation, the integrated value should be reset to zero at the time when the desorption of trapped NOx from NOx trap catalyst 13 is completed and/or when the sulfur poisoning recovery is completed together with the desorption of trapped NOx.

At step S4, the quantity of sulfur (viz., sulfur accumulation quantity) accumulated on NOx trap catalyst 13 due to the sulfur poisoning is calculated. Like in the above-mentioned NOx accumulation quantity, the sulfur accumulation quantity may be estimated from the integrated value of engine speed or the travel distance of the vehicle. In case of using the integrated value of engine speed for the estimation, the integrated value should be reset to zero at the time when the sulfur poisoning recovery is completed.

At step S5, the quantity of PM (viz., PM accumulation quantity) trapped by and accumulated on DPF 14 is calculated. The method of this calculation is as follows.

In view of the fact that the exhaust pressure at the inlet side of DPF 14 increases with increase of the PM accumulation quantity, the PM accumulation quantity is estimated by comparing the current exhaust pressure at the inlet side of DPF 14 that is detected by exhaust pressure sensor 26 with a reference exhaust pressure that has been previously derived in a corresponding operation condition of the engine. The current operation condition is derived from the engine speed and engine load. If desired, the PM accumulation quantity may be derived from a combination of the integrated value of engine speed (or the travel distance of the vehicle) and the exhaust pressure.

At step S6, judgment is carried out as to whether “reg-flag” set for indicating the state of “DPF regeneration mode” has been raised or not. If “reg-flag=1” has been established, the operation flow goes to an after-mentioned control for the “DPF regeneration mode” of FIG. 3.

At step S7, judgment is carried out as to whether “desul-flag” set for indicating the state of “sulfur poisoning recovery mode” of NOx trap catalyst 13 has been raised or not. If “desul-flag=1” has been established, the operation flow goes to an after-mentioned control for the “sulfur poisoning recovery mode” of FIG. 4.

At step S8, judgment is carried out as to whether “sp-flag” set for indicating the state of “rich spike mode” for NOx desorption of NOx trap catalyst 13 has been raised or not. If “sp-flag=1” has been established, the operation flow goes to an after-mentioned control for the “rich spike mode” of FIG. 5.

At step S9, judgment is carried out as to whether “rec-flag” set for indicating the state of “melt down suppression mode” after the DPF regeneration and sulfur poisoning recovery has been raised or not. If “rec-flag=1” has been established, the operation flow goes to an after-mentioned control for the “melt down suppression mode” of FIG. 6.

At step S10, judgment is carried out as to whether “rq-DPF-flag” set for indicating presence/absence of request for DPF regeneration has been raised or not. If “rq-DPF-flag=1” has been established due to presence of such request, the operation flow goes to the flowchart of FIG. 7 to determine the order of priority of regeneration in case of presence of request for DPF regeneration.

At step S11, judgment is carried out as to whether“req-desul” set for indicating presence/absence of request for “sulfur poisoning recovery” has been raised or not. If “req-desul=1” has been established due to presence of such request, the operation flow goes to the flowchart of FIG. 8 to determine the order of priority of regeneration in case of presence of request for sulfur poisoning recovery.

At step S12, judgment is carried out as to whether the PM accumulation quantity calculated at step S5 has reached a predetermined quantity PM1 or not, that is, whether it is the time for DPF regeneration or not.

If “PM accumulation quantity>PM1” is established judging that it is the time for DPF regeneration, the operation flow goes to the flowchart of FIG. 9. As is seen from the flowchart of this drawing, at step S701, the “rq-DPF-flag” is turned to 1 to issue the request for DPF regeneration.

At step S13, judgment is carried out as to whether the sulfur accumulation quantity calculated at step S4 has reached a predetermined quantity S1 or not, that is, whether it is the time for the sulfur poisoning recovery or not.

If “sulfur accumulation quantity>S1” is established judging that it is the time for the sulfur poisoning recovery, the operation flow goes to the flowchart of FIG. 10. As is seen from the flowchart of this drawing, at step S801, the “rq-desul-flag” is turned to 1 to issue the request for the sulfur poisoning recovery.

At step S14, judgment is carried out as to whether the NOx accumulation quantity on NOx trap catalyst 13 calculated at step S3 has reached a predetermined quantity NOx1 or not, that is, whether it is the time for the NOx desorption from the catalyst 13 or not.

If “NOx accumulation quantity>NOx1” is established judging that it is the time for the NOx desorption, the operation flow goes to the flowchart of FIG. 11. As is seen from the flowchart of this drawing, at step S901, the “rq-sp-flag” is turned to 1 to issue the request for NOx desorption, viz., the request for rich spike operation.

In the following, the “DPF regeneration mode” will be described in detail with reference to the flowchart of FIG. 3.

That is, as is seen from the flowchart of FIG. 2, when the PM accumulation quantity has reached the predetermined amount PM1 and thus, the “reg-flag” is turned to 1, the flowchart of FIG. 3 is started.

At step S101, for actually carrying out the DPF regeneration, the engine combustion is switched from a normal lean combustion mode to a split retard combustion mode (SRCM) according to the present invention.

In the following, the split retard combustion mode (SRCM) will be described. It is to be noted that in addition to the DPF regeneration, the split retard combustion mode (SRCM) is used for the sulfur poisoning recovery, NOx desorption (or rich spike) and catalyst activation promotion.

For carrying out the DPF regeneration, it is necessary to control the “exhaust λ” within a range from 1.0 to 1.4 and control the temperature of DPF 14 over 600° C. For carrying out the sulfur poisoning recovery of NOx trap catalyst 13, it is necessary to control the “exhaust λ” smaller than 1.0 and control the exhaust temperature over 600° C.

In a normal operation of the internal combustion engine at a lean air/fuel ratio, a pilot fuel injection is usually carried out. The pilot fuel injection timing is set at 40° to 10° BTDC (before top dead center), the pilot fuel injection quantity is set at 1 to 3 mm³/stroke, the main fuel injection timing is set at 10° to −20° BTDC, and the interval between the pilot fuel injection and the main fuel injection is within a range from 10° to 30° in CA (crank angle).

For actualizing the lower λ and the higher temperature exhaust for the DPF regeneration and sulfur poisoning recovery from a normal operation of the engine, it is necessary to reduce the intake air quantity. However, when the quantity of intake air is reduced, the compression end temperature in each cylinder is inevitably lowered, which affects the combustion characteristic of the air/fuel mixture. Thus, as will be understood from the time chart of FIG. 13, that shows a combustion characteristic of a first reference, if the pilot fuel injection is set like in the fuel injection of the normal lean combustion, it is necessary to advance the injection timing of the main fuel injection. In such setting of fuel injection quantity and fuel injection timing, retarding the fuel injection timing for the purpose of increasing the exhaust temperature brings about an unstable combustion, and thus, the retarding has a limit of its own and thus actualization of the lower λ and higher temperature exhaust is considerably difficult.

In view of the above, the above-mentioned Laid-open Japanese Patent Application (Tokkai) 2000-320386 employs a measure to split the main fuel injection. With this splitting, a retarding limit of the fuel injection timing is expanded to actualize the lower λ and higher temperature exhaust. This is depicted by the time chart of FIG. 14 that shows the combustion characteristic of a second reference.

However, in case of the '386 publication, during a brisk combustion of previously injected fuel, a subsequent fuel injection is carried out. Thus, as is seen from the time chart of FIG. 14, a continuous combustion is inevitably carried out. That is, since a split fuel for the main combustion is injected to a flame of the combustion of fuel that has been previously injected, the split fuel is forced to start its combustion as soon as it is injected. Thus, percentage of diffusive combustion is increased, so that a partial equivalent ratio becomes very rich causing a severe deterioration of emission characteristic (viz., smoke).

Accordingly, in the present invention, as is seen from the time chart of FIG. 15, the fuel injections ((a) and (b)) are so controlled that the main combustion for producing a torque and a preliminary combustion effected before the main combustion are both carried out, and the preliminary combustion is carried out at a timing in the vicinity of TDC (viz., top dead center) of compression stroke and the main combustion is started after completion of the preliminary combustion.

That is, under the compression stroke, fuel is injected to effect the preliminary combustion to increase the incylinder temperature (viz., compression end temperature) in the vicinity of TDC, as is indicated by reference “a”. As is known, the fuel injection quantity bringing about a heat release of the preliminary combustion depends on the operation condition of the engine.

However, in the invention, the quantity of fuel injected for the preliminary combustion is so determined as to realize a certain heat release from the preliminary combustion and the incylinder temperature obtained at the time just before the fuel injection for the main combustion shows a level that is higher than a self ignition temperature that enables a self ignition of the fuel in the combustion chamber. Furthermore, in the invention, the fuel injection quantity and fuel injection timing for the preliminary combustion are varied in accordance with the compression end temperature that is estimated in each operation condition of the engine. With this, stability of the preliminary combustion is improved.

After completion of the preliminary combustion, the fuel for the main combustion is injected at a timing after TDC (viz., top dead center) as is indicated by reference “b”.

That is, by increasing the incylinder temperature by effecting the preliminary combustion, the retarding limit of the main combustion is expanded to improve a controllability for obtaining a target or desired temperature. Furthermore, by injecting fuel for the main combustion after the preliminary combustion is completed, an ignition retarding period for the main combustion is obtained, so that the rate of premixed combustion for the main combustion is increased for suppressing or at least minimizing emission of the smoke.

Although depending on the engine speed, the period (viz., interval period) from start timing of the preliminary combustion to that of the main combustion should be controlled larger than 20° in CA (crank angle). If not, the completion of the preliminary combustion is not obtained, that is, the heat release by the preliminary combustion is not sufficiently carried out. With such interval period, deterioration of the main combustion characteristic is suppressed and thus deterioration of the smoke characteristic is suppressed. Furthermore, since the main combustion is forced to start in the expansion stroke, the burning speed of the fuel is very late, and thus, the main combustion is completed at a timing of over 50° ATDC (after top dead center) of compression stroke. Because the completion of the main combustion is delayed, the main combustion becomes slow and thus combustion noise is minimized.

As is understood from the graph of FIG. 16, in the split retard combustion mode (SRCM) according to the present invention, such a combustion as to feature a high exhaust temperature and low smoke exhaust is actualized even under richer condition in the air/fuel mixture. In the graph of FIG. 16, the emission characteristic exhibited by the present invention is shown by bars indicated by numeral (3), and the emission characteristic exhibited by the first and second references are shown by bars indicated by numerals (1) and (2) respectively. As is seen from this graph, in the present invention, also the concentration of HC in the exhaust gas is very small.

Due to employment of the preliminary combustion, the retarding limit of the main combustion is expanded. Thus, even when the fuel injection timing for the main combustion is retarded, the combustion at lower λ is stable, which enables the high exhaust temperature.

As is understood from the graph of FIG. 17, when the timing of the main combustion is retarded, the premixed rate for the main combustion is increased. Thus, even when the value of λ is small, undesired smoke emission is suppressed by retarding the main combustion. Furthermore, when the timing of the main combustion is retarded, higher exhaust temperature is realized. Thus, by varying the fuel injection timing for the main combustion, the exhaust temperature can be controlled.

FIG. 18 is a map showing a target fuel injection timing for the preliminary combustion using the engine speed Ne and the engine load Q as parameters.

FIG. 19 is a map showing a target fuel injection quantity for the preliminary combustion using the engine speed Ne and the engine load Q as parameters.

FIG. 20 is a map showing a target fuel injection timing for the main combustion to achieve a target exhaust temperature, using the engine speed Ne and the engine load Q as parameters.

When, under a low load condition of the engine, a target exhaust temperature is required, the combustion timing of the main combustion is largely retarded. Thus, in such condition, it tends to occur that even when the preliminary combustion is effected, the incylinder temperature is not kept sufficiently high to the time when the fuel injection for the main combustion is started. That is, there may be such a case that only one preliminary combustion fails to keep the incylinder temperature sufficiently high to the timing for the fuel injection for the main combustion.

In view of such anxiety, as is seen from FIG. 21, in the present invention, a plurality of preliminary combustions are effected before the main combustion in such a manner that the heat releases by the respective preliminary combustions are not overlapped. With this measure, both lower smoke emission and high exhaust temperature are achieved even under the low load condition of the engine.

Thus, when the lower λ and high exhaust temperature are needed for effecting the DPF regeneration and sulfur poisoning recovery, switching to the split retard combustion mode (SRCM) is carried out in the present invention.

Two types (viz., type-1 and type-2) of such switching will be described in detail with reference to the flowcharts of FIGS. 22A and 22B.

In one type (viz., type-1) of the switching shown in FIG. 22A, at step S1101, switching to a fuel injection control for the preliminary combustion is effected. That is, the fuel injection is carried out at a fuel injection timing for a preliminary combustion (see FIG. 18) with a fuel injection quantity for the preliminary combustion (see FIG. 19).

At step S1102, the fuel injection for the main combustion is carried out at a retarded fuel injection timing. That is, although the fuel injection timing for the preliminary combustion is instantly switched, switching of the fuel injection timing for the main combustion is carried out with a retard. More specifically, switching is gradually made by repeating several cycles for obtaining a target main fuel injection timing as shown in FIG. 20. Since the torque produced is reduced as the fuel injection timing is retarded, the fuel injection quantity for the main injection is gradually increased to compensate for the torque reduction by the retarded fuel injection timing, for keeping the torque produced.

That is, when the switching from a normal combustion mode to the split retard combustion mode (SRCM) is carried out for effecting the regeneration of DPF 14 and NOx trap catalyst 13, fresh air quantity needed by the engine and the EGR quantity are varied due to change of the target λ, which induces a fluctuation of the intake system of the engine. As has been mentioned hereinabove, since the preliminary combustion effected at the timing in the vicinity of the top dead center of compression stroke is hardly affected by the fluctuation of the intake system, switching to the target value can be instantly made. However, the main combustion after the compression stroke is easily affected by the fluctuation of the intake system, and thus, if the switching to the target value for the main combustion is instantly made, unstable ignition is induced causing a misfiring of the main combustion.

The above description will be easily understood from the following.

That is, the fresh air quantity and the behavior of the EGR gas, that change in accordance with the change of the target λ induced by the combustion mode switching, affect the main combustion. That is, although the throttle opening can instantly take a target throttle opening, it takes a certain time lag that a differential pressure between upstream and downstream portions of throttle valve 5 is developed to a degree corresponding to the target throttle opening, which tends to induce a severe lack of fresh air. In such case, the compression end temperature becomes excessively reduced and thus in a worst case, the main combustion is subjected to a misfiring. Like the above, since the differential pressure between upstream and downstream portions of the EGR valve 12 is unstable at the transient time, there is such a case that the EGR gas is excessively increased. In this case, an ignition retard becomes marked and thus in a worst case the main combustion is subjected to a misfiring. Furthermore, due to occurrence of such lack of fresh air and the excessive increase of the EGR gas, the misfiring tendency of the main combustion tends to be much increased.

Accordingly, in the present invention, as has been mentioned hereinabove, switching to the preliminary combustion is effected at first for the purpose of obtaining a compression end temperature that enables the main combustion, and then, switching to the main combustion is gradually carried out, and when the fluctuation of the intake system becomes sufficiently small, the switching to the main combustion is actually carried out.

Thus, in the present invention, the main combustion is stably established, and thus smoothed switching of the combustion control is achieved.

The switching speed for the main combustion is set in accordance with retard factors of the intake system (which are the collector capacitor, 2 to 3 repetition cycles and the like), and a real λ (which is determined based on the fresh air quantity, the fluctuation of the EGR system and the like). For example, the switching to the main combustion may be so made that ⅓ of an entire retard is effected for one cycle and the entire retard is established by repeating 3 cycles.

In the other type (viz., type-2) of the switching shown in FIG. 22B, at step S1111, a retardation timing needed for subsidence of the transient fluctuation (viz., fresh air quantity and EGR gas quantity) of the intake system is determined in accordance with a target λ for the combustion after the combustion switching. This may be achieved by looking up a suitable map.

At step S1112, like the above-mentioned step S1101, switching to the fuel injection control for a preliminary combustion is effected. That is, the fuel injection is carried out at a timing for the preliminary combustion (see FIG. 18) with a fuel injection quantity for the preliminary combustion (see FIG. 19).

At step S1113, judgment is carried out as to whether a time when a transient fluctuation of the fresh intake air and EGR gas induced by operation of the intake system subsides has passed or not. The time may be previously derived through experiments. If YES, that is, when it is judged that the time has passed, the operation flow goes to step S1114.

At step S1114, switching to the fuel injection for the main combustion is carried out.

As is seen from the above, in the other type of switching, when the transient fluctuation of the fresh air intake air and EGR gas subsides and thus when the differential pressure between the upstream and downstream portions of the throttle valve 5 becomes stable, switching to the main combustion is actually effected. Thus, stable main combustion is achieved, and thus, smoothed switching of the combustion control is effected.

Referring back to FIG. 3, at step S101, for carrying out the DPF regeneration, the engine combustion is switched from the normal lean combustion mode to the split retard combustion mode (SRCM) according to the present invention.

Then, at step S102, the exhaust λ is controlled to a target value. In the regeneration of DPF 14, the target value of the exhaust λ depends on the PM accumulation quantity. Accordingly, by comparing the exhaust pressure at the inlet side of DPF 14 with a reference exhaust pressure that has been previously derived in a corresponding operation condition of the engine, the PM accumulation quantity is estimated, and then by using the map of FIG. 23, the target λ corresponding to the estimated PM accumulation quantity is looked up.

After the switching to the split retard combustion mode (SRCM) is effected at step S101, the fresh air quantity is controlled by intake throttle valve 5 and EGR valve 12 for achieving the target air/fuel ratio. More specifically, at first, the intake throttle valve 5 is so controlled as to achieve a target intake air quantity (viz., target intake air quantity for an operation on λ=1) that is provided by multiplying the target λ by a value looked up from the map of FIG. 24, and if the air/fuel ratio differs from the target value, intake throttle valve 5 and/or EGR valve 12 is controlled to bring the air/fuel ratio to the target value.

When the switching to the split retard combustion mode (SRCM) is actually made, the fuel injection timing is greatly retarded. Thus, in the invention, in addition to the above-mentioned control for the intake air quantity, the following control for suppressing a torque fluctuation, that would be induced at the switching, is carried out. That is, the target intake air quantity and fuel injection quantity provided based on the map of FIG. 24 are corrected by means of a torque correction factor K1 looked up from the map of FIG. 25 that shows a relation between the target fuel injection timing for the main fuel injection and the correction factor K1.

When the target air/fuel ratio is reduced to or near a stoichiometric value, undesirable pumping loss tends to occur. Thus, in the invention, the target intake air quantity and the fuel injection quantity for the main combustion are corrected by means of a correction factor K2 that is looked up from the map of FIG. 26 that shows a relation between the target λ and the correction factor K2.

Referring back to FIG. 3, at step S103, judgment is carried out as to whether the temperature of DPF 14 has exceeded a target upper limit T22 under regeneration of DPF 14 or not.

If “DPF temperature>T22” is made, the operation flow goes to step S111 judging that the temperature of DPF 14 has exceeded the upper limit during the regeneration of DPF 14. At step S111, the fuel injection timing for the main combustion is advanced to lower the exhaust temperature, and then, at step S112, a torque correction (viz., reduction of the fuel injection quantity for the main combustion) is carried out for compensating a torque change (viz., increase) caused by the advancing of the fuel injection timing.

If, at step S103, “DPF temperature<T22” is made, the operation flow goes to step S104 to judge whether the temperature of DPF 14 has become below a target lower limit T21 under regeneration of DPF 14 or not.

If “DPF temperature<T21” is made, the operation flow goes to step S109 judging that the temperature of DPF 14 has become below the lower limit T21 during the regeneration of DPF 14. At step S109, the fuel injection timing for the main combustion is retarded to increase the exhaust temperature, and then at step S110, a torque correction (viz., increase of the fuel injection quantity for the main combustion) is carried out for compensating a torque drop caused by the retardation of the fuel injection timing.

If, at step S104, “DPF temperature≧T21” is made, the operation flow goes to step S105. At this step S105, judgment is carried out as to whether a predetermined time “tDPFreg” has passed from the start of the DPF regeneration or not. If YES, the operation flow goes to step S106 judging that the PM accumulated on DPF 14 has completely burnt.

At step S106, switching from the split retard combustion mode to the normal combustion mode is made because the regeneration of DPF 14 has been completed. With this switching, heating of DPF 14 is stopped.

At step S107, the reg-flag is turned to 0 due to completion of the DPF regeneration.

Then, at step S108, the rec-flag is turned to 1 for preparation of an after-mentioned “melt-down suppression mode”. That is, if there is cinder (viz., burnable rest) of particulate matter (PM) on DPF 14 even when the DPF regeneration has been finished, a rapid increase of the exhaust λ causes a sudden burning of the cinder, which has a possibility of inducing melting of DPF 14. Thus, it is necessary to prepare such melt-down suppression mode.

In the following, control for the sulfur poisoning recovery mode will be described in detail with reference to the flowchart of FIG. 4.

As is seen from the flowchart of FIG. 2, when the sulfur accumulation quantity of NOx trap catalyst 13 reaches the predetermined quantity S1 to induce “rq-desul-flag=1” inducing “desul-flag=1” in an after-mentioned flowchart of FIG. 8, the control for the sulfur poisoning recovery starts like the manner as shown in the flowchart of FIG. 4.

At step S201, for carrying out the sulfur poisoning recovery of NOx trap catalyst 13, the engine combustion is switched from the normal combustion mode to the split retard combustion mode (SRCM) according to the present invention.

At step S202, the exhaust λ is controlled to a stoichiometric value. This control is carried out by setting the target λ to a stoichiometric ratio.

At step S203, judgment is carried out as to whether the temperature of NOx trap catalyst 13 is higher than a predetermined temperature T4 or not. If the catalyst 13 is produced basically by Ba (viz., barium), it is necessary to keep the temperature of the catalyst 13 over 600° C. in rich-stoichiometric atmosphere for normal activation of the same. Thus, in such catalyst 13, the predetermined temperature T4 is set at 600° C.

If the catalyst temperature is lower than the predetermined value T4, the operation flow goes to step S210 to retard the fuel injection timing for the main combustion thereby to increase the exhaust temperature, and then at step S211, a torque correction is carried out for compensating a torque change caused by the retardation of the fuel injection timing.

At step S204, judgment is carried out as to whether a predetermined time “tdesul” has passed from the start of the sulfur poisoning recovery mode or not. If YES, the operation flow goes to step S205 judging that the sulfur poisoning has been recovered or removed.

At step S205, because of completion of the sulfur poisoning recovery, the engine combustion is switched from the split retard combustion mode (SRCM) to the normal combustion mode stopping heating of NOx trap catalyst 13. At the same time, the stoichiometric air/fuel ratio operation is cancelled.

At step S206, because of completion of the sulfur poisoning recovery, the “desul-flag” is turned to 0.

At step S207, the rec-flag is turned to 1 for preparation of after-mentioned “melt-down suppression mode”. That is, if a certain amount of particulate matter (PM) is left on DPF 14 after completion of the sulfur poisoning recovery, a rapid increase of the exhaust λ causes a sudden burning of the particulate matter (PM), which has a possibility of inducing melting of DPF 14. Thus, it is necessary to prepare such melt-down suppression mode.

At step S208, the rq-sp-flag is turned to 0 to disable the request for rich spike operation. That is, under the sulfur poisoning recovery mode, NOx trap catalyst 13 is kept exposed to a stoichiometric exhaust atmosphere for a certain time, and thus, the NOx desorption is carried out at the same time. Thus, if a request for the NOx desorption (viz., request for rich spike operation) has been issued, it is necessary to cancel such request.

In the following, control for the rich spike mode (viz., NOx desorption mode) will be described in detail with reference to the flowchart of FIG. 5.

As is seen from the flowchart of FIG. 2, when the NOx accumulation quantity on NOx trap catalyst 13 reaches the predetermined quantity NOx1 to induce “rq-sp-flag=1 inducing “sp-flag=1” in an after-mentioned flowchart of FIG. 7 or 8, the control for the rich spike mode starts like the manner shown in the flowchart of FIG. 5.

At step S301, for carrying out the NOx desorption, the engine combustion is switched from a normal combustion mode to the split retard combustion mode (SRCM) according to the present invention.

At step S302, the exhaust λ is controlled to a rich value. That is, the control is carried out by controlling intake throttle valve 5 for achieving a target intake air quantity for the rich spike operation, and then, like in the above-mentioned case of DPF regeneration, by controlling the fresh intake air by intake throttle valve 5 and/or EGR valve 12.

At step S303, judgment is carried out as to whether a predetermined time “tspike” has passed from the start of the rich spike mode or not. If YES, the operation flow goes to step S304 judging that the NOx desorption has completely effected.

At step S304, because of completion of the NOx desorption, the engine combustion is switched from the split retard combustion mode (SRCM) to the normal combustion mode. At the same time, the rich air/fuel ratio operation is cancelled.

At step S305, because of completion of the NOx desorption, the sp-flag is turned to 0.

In the following, control for the melt-down suppression mode will be described in detail with reference to the flowchart of FIG. 6.

As is seen from the flowchart of FIG. 2, when the DPF regeneration or sulfur poisoning recovery is finished to induce “rec-flag=1” in the flowchart of FIG. 3 or 4, the control for the metal-down suppression mode starts in such a manner as shown in the flowchart of FIG. 6.

At step S401, the exhaust λ is controlled to a target value, for example, “λ≦1.4”. That is, just after the DPF regeneration for example, the DPF 14 is still highly heated. If, under such condition, the exhaust λ is instantly controlled to a lean level, the cinder (viz., burnable rest) of particulate matter (PM) on DPF 14 is suddenly burnt, which has a possibility of inducing melting of DPF 14. Thus, it is necessary to provide the exhaust λ with such upper limit target value. Since, in the melt-down suppression mode, a lower exhaust temperature is preferable, the split retard combustion mode is not used. That is, the exhaust λ is controlled to the target value under the normal combustion mode.

At step S402, judgment is carried out as to whether or not the DPF temperature has become lower than a predetermined temperature T3 (for example 500° C.) that is the lowest level for enabling a rapid oxidation of particulate matter (PM). If NO, that is, when the DPF temperature is higher than T3, the control for the exhaust λ is continued. If YES, that is, when the DPF temperature is lower than T3, the operation flow goes to step S403 judging that the undesired melting of DPF 14 may be avoided even if the oxygen concentration in the exhaust gas becomes to the level of the atmospheric air.

At step S403, the control for the exhaust λ is stopped because of disappearance of fear of the melt-down.

At step S404, because of completion of the melt-down suppression mode, the rec-flag is turned to 0.

In the following, control-1 for determining the order of priority of regeneration will be described with reference to the flowchart of FIG. 7.

This control-1 is carried out when request for the DPF regeneration and one of request for the sulfur poisoning recovery and request for the NOx desorption are issued at the same time.

As is seen from the flowchart of FIG. 2, when the request for DPF regeneration is issued to induce “rq-DPF-flag=1”, the control-1 for determining the order starts in such a manner as shown in the flowchart of FIG. 7.

At step S501, judgment is carried out as to whether, after issuance of request for the DPF regeneration, the sulfur accumulation quantity has reached the predetermined quantity S1 or not, that is, whether the timing for the sulfur poisoning recovery has come or not, like the case of the above-mentioned step S13.

If YES, the operation flow goes to step S801 of the flowchart of FIG. 10 judging that the timing for the sulfur poisoning recovery has come. At the step S801, the rq-desul-flag is turned to 1 to issue a request for the sulfur-poisoning recovery. In this case, the order of priority is decided by the flowchart of FIG. 8, as will be described hereinafter.

If NO at step S501, that is, when it is judged that the timing for the sulfur poisoning recovery has not come yet, the operation flow goes to step S502.

At step S502, judgment is carried out as to whether “rq-sp-flag=1” has been established or not, that is, whether the request for the NOx desorption has been issued or not. If NO, the operation flow goes to step S503.

At step S503, judgment is carried out as to whether, after issuance of request for the DPF regeneration, the NOx accumulation quantity has reached the predetermined quantity NOx1 or not, that is, whether the timing for the NOx desorption has come or not, like the case of the above-mentioned step S14.

If YES, that is, when the NOx accumulation quantity is judged greater than the predetermined quantity NOx1, the operation flow goes to step S901 of the flowchart of FIG. 11. At the step S901, the rq-sp-flag is turned to 1 to issue a request for the NOx desorption (viz., request for rich spike operation).

If NO at step S503, that is, when the NOx accumulation quantity is judged smaller than the predetermined quantity NOx1, the operation flow goes to step S504. That is, in such case, only the request for DPF regeneration is issued.

At step S504, judgment is carried out as to whether or not the current engine operation is under a possible condition enabling the DPF regeneration and sulfur poisoning recovery with reference to the map of FIG. 28. As is seen from this map, the possible condition is the condition wherein the engine speed and engine load are not low, irregularity of temperature increase is relatively small and deterioration degree of the exhaust performance does not exceed an allowable value.

If YES, that is, when the current engine operation condition is judged to be under the possible condition, the operation flow goes to step S505 where the reg-flag is turned to 1 for preparation of the DPF regeneration.

If YES at step S502, that is, when “rq-sp-flag=1” has been established, the operation flow goes to step S506 judging that the request for the DPF regeneration and the request for NOx desorption have been issued at the same time.

At step S506, judgment is carried out as to whether or not the current engine operation is under a low NOx condition that enables reduction in NOx in the exhaust gas (for example, under a stationary condition). If YES, that is, when it is judged that the current engine operation is under the low NOx condition, the operation flow goes to step S507 inferring that under such low NOx condition, deterioration of the exhaust property at the tail pipe is hardly seen even when the regeneration of the NOx trap catalyst 13 is delayed, and thus it is preferable to give priority to the DPF regeneration that would affect the operation of the engine.

If NO at step S506, that is, when it is judged that the current engine operation is not under the low NOx condition (that is, under high NOx condition, such as under vehicle acceleration condition), the operation flow goes to step S508 inferring that under such high NOx condition, it is preferable to give priority to the regeneration of NOx trap catalyst 13 for suppressing the deterioration of the exhaust property at the tail pipe. At step S508, the sp-flag is turned to 1 for preparation of the NOx desorption (viz., rich spike operation).

At step S507, judgment is carried out as to whether or not the temperature of DPF 14 is higher than an activation temperature T6 of the oxidation catalyst carried on DPF 14. If, before starting operation for increasing the exhaust temperature, the temperature of the oxidation catalyst of DPF 14 is lower than the activation temperature T6, it takes a longer time to reach the temperature that enables the regeneration of DPF 14, and thus, there is a fear of deterioration of NOx characteristic in the tail pipe of the exhaust system during the temperature increase. Thus, in such case, it is preferable to give priority to the regeneration of NOx trap catalyst 13. Accordingly, also in this case, the operation flow goes to step S508 inducing “sp-flag=1” for preparation of NOx desorption (viz., rich spike operation).

If YES at step S507, that is, when it is judged that the temperature of DPF 14 is higher than T6, the operation flow goes to the above-mentioned step 5504 intending to give priority to the regeneration of DPF 14.

In the following, control-2 for determining the order of priority of regeneration will be described with reference to the flowchart of FIG. 8.

This control-2 is carried out when request for the sulfur poisoning recovery and request for the NOx desorption are issued at the same time.

As is seen from the flowchart of FIG. 2, when the request for sulfur poisoning recovery is issued to induce “rq-desul-flag=1”, the control-2 for determining the order starts in such a manner as shown in the flowchart of FIG. 8.

At step S601, judgment is carried out as to whether, after issuance of request for the sulfur poisoning recovery, the PM accumulation quantity has reached the predetermined quantity PM1 or not, that is, whether the timing for the regeneration of DPF 14 has come or not, like the case of the above-mentioned step S12.

If YES, the operation flow goes to step S701 of the flowchart of FIG. 9 judging that the timing for the regeneration of DPF 14 has come. At step S701, the rq-DPF-flag is turned to 1 to issue a request for the regeneration of DPF 14. In this case, the order of priority is decided by the flowchart of FIG. 7.

If NO at step S601, that is, when it is judged that the timing for the regeneration of DPF 14 has not come yet, the operation flow goes to step S602.

At step S602, judgment is carried out as to whether the temperature of the catalyst 13 is higher than a predetermined temperature T1 or not. If YES, the operation flow goes to step S603.

At step S603, judgment is carried out as to whether or not the current engine operation is under a possible condition that enables the DPF regeneration and sulfur poisoning recovery with reference to the map of FIG. 28.

If the current engine operation condition is judged to be the condition (viz., possible condition) that enables the sulfur poisoning recovery, the operation flow goes to step S604 to induce “desul-flag=1” for preparation of sulfur poisoning recovery.

If, at step S602, it is judged that the temperature of the catalyst 13 is lower than T1, the operation flow goes to step S605. That is, even if, under such lower temperature of the catalyst, the operation for increasing the exhaust temperature is started, it takes a longer time to reach the temperature that enables the sulfur poisoning recovery, and thus, there is a fear of deterioration of NOx characteristic in the tail pipe of the exhaust system during the temperature increase. Thus, in such case, it is preferable to give priority to the NOx desorption. Accordingly, the operation flow goes to step S605.

At step S605, judgment is carried out as to whether “rq-sp-flag=1” has been established or not, that is, whether the request for the NOx desorption has been issued or not. If YES, the operation flow goes to step S607 to induce “sp-flag=1” for preparation of the NOx desorption (viz., rich spike operation).

If NO at step S605, that is, when it is judged that “rq-sp-flag=1” has not been established, the operation flow goes to step S606.

At step S606, judgment is carried out as to whether after issuance of the request for the sulfur poisoning recovery, the NOx accumulation quantity has reached the predetermined quantity NOx1 or not, that is, whether the timing for the NOx adsorption has come or not, like the case of the above-mentioned step S14.

If YES, that is, when it is judged that the NOx accumulation quantity is larger than the predetermined quantity NOx1, the operation flow goes to step S901 of the flowchart of FIG. 11 to induce “rq-sp-flag=1”.

In the following, control for carrying out the “catalyst activation promotion mode” will be described with reference to the flowchart of FIG. 12.

As is seen from the flowchart of FIG. 2, when the temperature of NOx trap catalyst 13 is lower than a predetermined temperature T5 that is the lowest activation temperature of the catalyst 13, the operation flow goes to step S1001 of the flowchart of FIG. 12.

At step S1001, judgment is carried out as to whether or not the current engine operation is under a possible condition that enables the catalyst activation promotion. Since the catalyst activation promotion is carried out by the split retard combustion mode (SRCM) according to the invention, such judgment means a judgment as to whether the current engine operation enables the split retard combustion or not. Specifically, such judgment is carried out with reference to the map of FIG. 28 with the possible zone of the DPF regeneration and sulfur poisoning recovery replaced with a possible zone of the catalyst activation promotion. If YES, that is, when it is judged that the current engine operation enables the catalyst activation promotion, the operation flow goes to step S1002.

At step S1002, for achieving the catalyst activation promotion, the engine combustion is switched from the normal lean combustion mode to the split retard combustion mode (SRCM) according to the present invention. With this switching, the exhaust temperature is increased and thus, heating of the catalyst is promoted. Also in this case, the control is effected by setting the target λ. That is, for the control to the target λ, a torque correction is made in view of a torque reduction caused by the retard combustion.

At step S1003, judgment is carried out as to whether the temperature of NOx trap catalyst 13 has become higher than T5 or not. If YES, that is, when it is judged that the temperature has exceeded T5, the operation flow goes to step S1004 to switch the engine combustion from the split retard combustion mode to the normal lean combustion mode.

The entire contents of Japanese Patent Application 2003-284311 filed Jul. 31, 2003 are incorporated herein by reference.

Although the invention has been described above with reference to the embodiment of the invention, the invention is not limited to such embodiment as described above. Various modifications and variations of such embodiment may be carried out by those skilled in the art, in light of the above description.

Claims

1. A combustion control system of an internal combustion engine, comprising:

a fuel injection system provided at an intake system of the engine;

an exhaust purifying device provided at an exhaust system of the engine; and

a control unit that controls the fuel injection system to permit the engine to have a predetermined combustion mode, the predetermined combustion mode being a mode in which, under a predetermined condition of the exhaust purifying catalyst, the fuel injection system causes the engine to carry out a main combustion to produce a torque and at least one preliminary combustion prior to the main combustion, the at least one preliminary combustion being effected at a timing in the vicinity of top dead center of compression stroke, the main combustion being effected at a first timing after completion of the preliminary combustion,

wherein the control unit further controls the fuel injection system in such a manner that upon switching of the engine operation from a previous combustion mode to the predetermined combustion mode, the main combustion that takes place after completion of the preliminary combustion is effected at a second timing that is retarded as compared with the first timing.

2. A combustion control system as claimed in claim 1, in which the second timing for the main combustion is controlled to have a value that gradually reaches a target value.

3. A combustion control system as claimed in claim 2, in which a fuel injection timing for the main combustion is gradually retarded to have a target fuel injection timing.

4. A combustion control system as claimed in claim 1, in which the main combustion after completion of the preliminary combustion is carried out after a fluctuation of the intake system, that is induced by the switching to the predetermined combustion mode from the previous combustion mode, subsides.

5. A combustion control system as claimed in claim 1, in which a fuel injection quantity for the preliminary combustion is determined to a quantity that permits an incylinder temperature at the second timing of fuel injection for the main combustion to be higher than a temperature that enables a self ignition of the main combustion.

6. A combustion control system as claimed in claim 1, in which a combustion start timing of the main combustion is retarded from a combustion start timing of the preliminary combustion by over 20 degrees in crank angle.

7. A combustion control system as claimed in claim 1, in which the main combustion is completed at a timing of over 50 degrees in crank angle after top dead center of compression stroke.

8. A combustion control system as claimed in claim 1, in which the exhaust temperature of the engine is controlled by controlling a fuel injection timing for the main combustion.

9. A combustion control system as claimed in claim 1, in which the exhaust purifying device is a PM filter that collects particulate matter in the exhaust gas, and in which under a predetermined condition of the filter, the control unit controls the fuel injection system in a manner to increase the exhaust temperature for burning the particulate matter accumulated on the filter.

10. A combustion control system as claimed in claim 1, in which the exhaust purifying device is a NOx trap catalyst that traps NOx in the exhaust gas when the exhaust air/fuel ratio is lean, and in which under a predetermined condition of the NOx trap catalyst, the control unit controls the fuel injection system in a manner to make the exhaust air/fuel ratio richer thereby to release the trapped NOx from the catalyst.

11. A combustion control system as claimed in claim 1, in which the exhaust purifying device is a NOx trap catalyst that traps NOx in the exhaust gas when the exhaust air/fuel ratio is lean, and in which under a predetermined condition of the NOx trap catalyst, the control unit controls the fuel injection system in a manner to increase the exhaust temperature to release a sulfur poisoning from the catalyst.

12. A combustion control system as claimed in claim 1, in which the predetermined condition of the exhaust purifying device is a condition wherein the exhaust purifying device is under inactivation condition.

13. A combustion control system as claimed in claim 1, in which the combustion of the previous combustion mode contains mainly a diffusive combustion that is produced when a main fuel injection is effected during a combustion induced by a pilot fuel injection.

14. In an internal combustion engine that is equipped at an intake system with a fuel injection system and at an exhaust system with an exhaust purifying device,

a method for controlling the engine, comprising:

controlling the fuel injection system to permit the engine to have a predetermined combustion mode in which, under a predetermined condition of the exhaust purifying device, the fuel injection system causes the engine to carry out a main combustion to produce a torque and at least one preliminary combustion prior to the main combustion;

controlling the fuel injection system in such a manner that the at least one preliminary combustion is effected at a timing in the vicinity of top dead center of compression stroke, and the main combustion is effected at a first timing after completion of the preliminary combustion; and

controlling the fuel injection system in such a manner that upon switching of the engine operation from a previous combustion mode to the predetermined combustion mode, the main combustion that takes place after completion of the preliminary combustion is effected at a second timing that is retarded as compared with the first timing.