JP2007192088A - Fuel injection control device of internal combustion engine - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To properly perform learning correction of a deposit ratio and a residual ratio by using a dual injector system. <P>SOLUTION: From a state of only a cylinder injection of fuel such as a (k-1) cycle, port injection such as a k cycle is performed. Since part of fuel by the port injection adheres on an intake port wall face or the like, an exhaust air-fuel ratio AFRs is changed to a side leaner than a stoichiometric air-fuel ratio. The port deposition ratio Rp is calculated from the change of the exhaust air-fuel ratio AFRs to the leaner side. Then in the (k+1) cycle, if the state is returned to the state of only the cylinder injection of fuel again, the exhaust air-fuel ratio AFRs is changed to a side richer than the stoichiometric air-fuel ratio. The port residual ratio Pp is calculated from the change of the exhaust air-fuel ratio AFRs to the richer side. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to a fuel injection control device for an internal combustion engine, and more particularly to a fuel injection control device for a dual injector system.

An apparatus that performs learning correction of a fuel adhesion model based on an exhaust air-fuel ratio at the time of transition is known (see, for example, Patent Document 1). According to this apparatus, the fuel adhesion rate is corrected based on the peak value of the deviation amount of the exhaust air-fuel ratio with respect to the target air-fuel ratio, and the evaporation time constant is corrected based on the integral value of the deviation amount. Therefore, it is possible to perform air-fuel ratio control corresponding to variations in internal combustion engines, changes with time, and changes in use environment.
Further, a dual injector system including a port injector and an in-cylinder injector is known (see, for example, Patent Document 4).

Japanese Patent Application Laid-Open No. 11-229931 JP-A-11-200909 Japanese Patent Laid-Open No. 10-122017 JP 2005-48730 A

  By the way, the apparatus by the said patent document 1 is provided only with the port injector, and is not provided with the in-cylinder injector. Therefore, even if this apparatus is applied to a dual injector system, there is a possibility that the adhesion rate and the residual rate, which are model parameters, cannot be corrected appropriately.

  The present invention has been made to solve the above-described problems, and an object of the present invention is to appropriately perform learning correction of the adhesion rate and the residual rate using a dual injector system.

In order to achieve the above object, a first invention is a fuel injection control device for an internal combustion engine having a port injector for injecting fuel into an intake port and an in-cylinder injector for injecting fuel into a cylinder.
A fuel injection amount calculating means for calculating a fuel injection amount from the port injector and a fuel injection amount from the in-cylinder injector using a dynamic behavior model in consideration of fuel adhesion in the vicinity of the intake port;
An injection rate changing means capable of changing an injection rate of the port injector and the in-cylinder injector;
An air-fuel ratio detecting means for detecting an air-fuel ratio that changes due to the injection rate being changed by at least one cycle by the injection rate changing means;
And an adhesion rate correction unit that corrects the adhesion rate of the injected fuel, which is a parameter of the dynamic behavior model, based on the changed air-fuel ratio detected by the air-fuel ratio detection unit.

In a second aspect based on the first aspect, the injection rate changing means changes the injection rate of the port injector from zero to a predetermined value in a predetermined cycle.
The air-fuel ratio detection means detects an air-fuel ratio that changes in the predetermined cycle,
The adhesion rate correction means includes learning value calculation means for calculating a learning value using an air-fuel ratio that has changed in the predetermined cycle, and corrects the adhesion rate based on the learning value calculated by the learning value calculation means. It is characterized by being.

  According to a third aspect, in the second aspect, the injection rate changing means changes the injection rate of the in-cylinder injector to zero in the predetermined cycle.

In a fourth aspect based on the second or third aspect, the injection rate changing means changes the injection rate of the in-cylinder injector from the predetermined value to zero in a cycle after the predetermined cycle. Is what
The air-fuel ratio detection means detects an air-fuel ratio that changes in the subsequent cycle,
A residual rate correcting means for correcting the residual rate of attached fuel, which is a parameter of the dynamic behavior model, is further provided based on the detected air-fuel ratio.

Further, a fifth invention is the first invention according to the first invention, a target fuel amount calculating means for calculating a target fuel amount from the target air-fuel ratio;
Actual fuel amount calculating means for calculating an actual fuel amount from the air-fuel ratio detected by the air-fuel ratio detecting means;
Integrating means for integrating the difference between the target fuel amount calculated by the target fuel amount calculating means and the actual fuel amount calculated by the actual fuel amount calculating means;
It further comprises an adhesion amount correction means for correcting the fuel adhesion amount, which is a parameter of the dynamic behavior model, based on the difference accumulated by the accumulation means.

  According to a sixth aspect of the present invention, in the fifth aspect, when the injection rate of the in-cylinder injector is changed from zero to a predetermined value by the injection rate changing means, the target air-fuel ratio is made fuel leaner than the stoichiometric ratio. When the injection rate of the port injector is changed from zero to a predetermined value by the injection rate changing means, target air-fuel ratio changing means for changing the target air-fuel ratio to a fuel rich side from stoichiometry is provided. It is further provided with the feature.

  According to the first aspect, the adhesion rate, which is a parameter of the dynamic behavior model, is corrected based on the air-fuel ratio that changes due to the change in the injection rate by the injection rate changing means. By using variable control of the injection rate unique to the dual injector system, it is possible to appropriately correct the adhesion rate. Therefore, the fuel injection amount can be calculated with high accuracy, and the controllability of the air-fuel ratio can be improved.

  According to the second aspect of the invention, the learning value is calculated from the air-fuel ratio that changes due to the port injection from the state of only the in-cylinder injection, and the correction of the adhesion rate is appropriately executed based on the learning value. be able to.

  According to the third aspect of the invention, the change amount of the air-fuel ratio can be increased by changing from the state of only in-cylinder injection to the state of only port injection, so that the adhesion rate can be easily corrected. it can.

  According to the fourth aspect of the invention, the residual ratio is corrected based on the air-fuel ratio that changes due to the change from the state in which port injection is performed to the state in which only in-cylinder injection is performed. Using the variable control of the injection rate unique to the dual injector system, it is possible to appropriately correct the residual rate.

  According to the fifth aspect of the present invention, based on the integrated value of the difference between the actual fuel amount calculated from the air-fuel ratio that changes due to the change in the injection rate and the target fuel amount calculated from the target air-fuel ratio, In addition, the correction of the fuel adhesion amount can be executed. Using the variable control of the injection rate unique to the dual injector system, it is possible to appropriately correct the fuel adhesion amount.

  According to the sixth aspect of the invention, the amount of deviation of the exhaust air / fuel ratio from the stoichiometry due to the change in the injection rate can be reduced. Therefore, drivability and emissions can be improved when correcting the fuel adhesion amount.

  Embodiments of the present invention will be described below with reference to the drawings. In addition, the same code | symbol is attached | subjected to the element which is common in each figure, and the overlapping description is abbreviate | omitted.

Embodiment 1 FIG.
[Description of system configuration]
FIG. 1 is a schematic diagram showing a system configuration according to Embodiment 1 of the present invention. The system shown in FIG. 1 includes an internal combustion engine 1. Although the internal combustion engine 1 has a plurality of cylinders, only one of them is shown in FIG. The internal combustion engine 1 includes a cylinder block 4 having a piston 2 therein. The cylinder block 4 is provided with a water temperature sensor 6. The water temperature sensor 6 is configured to detect the cooling water temperature Tw.
The piston 2 is connected to the crankshaft 8 via a crank mechanism. A crank angle sensor 10 is provided in the vicinity of the crankshaft 8. The crank angle sensor 10 is configured to detect the rotation angle (crank angle CA) of the crankshaft 8.

  A cylinder head 12 is assembled to the upper part of the cylinder block 4. A space from the upper surface of the piston 2 to the cylinder head 12 forms a combustion chamber 14. The cylinder head 12 is provided with a spark plug 16 that ignites the air-fuel mixture in the combustion chamber 14.

  The cylinder head 12 includes an intake port 18 that communicates with the combustion chamber 14. An intake valve 20 is provided at a connection portion between the intake port 18 and the combustion chamber 14. The intake valve 20 is driven by a variable valve mechanism 21. The variable valve mechanism 21 is configured to be able to change the valve timing and valve lift amount of the intake valve 20.

  An intake passage 22 is connected to the intake port 18. A surge tank 24 is provided in the middle of the intake passage 22. A throttle valve 26 is provided upstream of the surge tank 24. The throttle valve 26 is an electronically controlled valve that is driven by a throttle motor 28. The throttle valve 26 is driven based on the accelerator opening AA detected by the accelerator opening sensor 32. In the vicinity of the throttle valve 26, a throttle opening sensor 30 is provided. The throttle opening sensor 30 is configured to detect the throttle opening TA.

  An air flow meter 34 is provided upstream of the throttle valve 28 in the intake passage 22. The air flow meter 34 is configured to detect the intake air amount Ga. An air cleaner 36 is provided upstream of the air flow meter 34.

  The cylinder head 12 includes an exhaust port 40 that communicates with the combustion chamber 14. An exhaust valve 42 is provided at a connection portion between the exhaust port 40 and the combustion chamber 14. The exhaust valve 42 is driven by a variable valve mechanism 43. The variable valve mechanism 43 is configured to be able to change the valve timing and the valve lift amount of the exhaust valve 42.

  An exhaust passage 44 is connected to the exhaust port 40. A catalyst 46 is provided in the exhaust passage 44. The catalyst 46 is configured to purify exhaust gas discharged from the combustion chamber 14. An air-fuel ratio sensor 48 is provided upstream of the catalyst 46 in the exhaust passage 44. The air-fuel ratio sensor 48 is configured to detect air-fuel ratio (hereinafter referred to as “exhaust air-fuel ratio”) AFRs of exhaust gas discharged from the combustion chamber 14.

  The system of the present embodiment is a so-called dual injector system that includes a port injector 50 and an in-cylinder injector 52 for each cylinder. The port injector 50 is provided in the vicinity of the intake port 18 and is configured to inject fuel in the vicinity of the intake port 18. On the other hand, the in-cylinder injector 52 is provided in the cylinder head 12 and configured to inject fuel directly into the combustion chamber 14.

Further, the system of the present embodiment includes an ECU (Electronic Control Unit) 60 as a control device. On the output side of the ECU 60, the ignition plug 16, the variable valve mechanisms 21, 43, the throttle motor 28, the port injector 50, the in-cylinder injector 52, and the like are connected. Connected to the input side of the ECU 60 are a water temperature sensor 6, a crank angle sensor 10, a throttle opening sensor 30, an accelerator opening sensor 32, an air flow meter 34, an air-fuel ratio sensor 48, and the like.
The ECU 60 calculates the engine speed NE based on the crank angle CA. Further, the ECU 60 calculates the engine load KL based on the throttle opening degree TA and the like. The ECU 60 performs fuel injection control based on the output of each sensor. Hereinafter, the fuel injection control will be described in detail.

[Features of Embodiment 1]
The fuel injection control in the first embodiment will be described with reference to FIG. FIG. 2 is a diagram schematically illustrating the behavior of fuel in the internal combustion engine 1.
In FIG. 2, the symbol “fip” represents the amount of fuel injected from the port injector 50 (hereinafter referred to as “port injection amount fip”). All of the fuel injected from the port injector 50 is not sucked into the combustion chamber 14 (hereinafter also referred to as “in-cylinder”), but a part of the fuel is injected into the wall surface of the intake port 18 or the intake valve 20 ( Hereinafter referred to as “port wall surface”). The symbol “fwp” represents the amount of fuel that adheres to the wall of the port or the like and is not sucked into the combustion chamber 14 (hereinafter referred to as “deposition amount fwp”). The fuel adhering to the port wall surface is vaporized and is sucked into the combustion chamber 14. Accordingly, the portion of the fuel injected from the port injector 50 that has not adhered to the port wall surface and the portion that has vaporized the fuel that has adhered to the port wall surface are sucked into the combustion chamber 14 from the intake port 18. .

In FIG. 2, the symbol “fid” represents the amount of fuel injected from the in-cylinder injector 52 (hereinafter referred to as “in-cylinder injection amount fid”). Before the internal combustion engine 1 is completely warmed up, not all of the fuel injected from the in-cylinder injector 52 exists in an atomized state in the combustion chamber 14, but a part of the fuel adheres to the wall surface of the combustion chamber 14. To do.
Here, according to the knowledge of the present inventors, the amount of fuel adhering to the wall surface of the combustion chamber 14 after the internal combustion engine 1 is completely warmed up (hereinafter referred to as “in-cylinder adhesion amount”) is the in-cylinder adhesion amount. This is considered to be an extremely small amount that does not require correction of the in-cylinder injection amount fid. Therefore, in the first embodiment, it is assumed that all of the fuel injected from the in-cylinder injector 52 exists as an atomized state in the fuel chamber 14.

  In FIG. 2, the symbol “fc” represents the amount of fuel supplied into the combustion chamber 14 (hereinafter referred to as “in-cylinder fuel amount fc”). The fuel is mixed with new gas sucked into the combustion chamber 14 from the intake port 18 and exists in the state of an air-fuel mixture. The symbol “Ga” represents the amount of new gas flowing into the combustion chamber 14 (hereinafter referred to as “intake air amount Ga”). This intake air amount Ga can be detected by the air flow meter 34 as described above.

The in-cylinder fuel amount fc is a fuel amount (in-cylinder required fuel amount) required to be present in the combustion chamber 14 in accordance with the operating state of the internal combustion engine 1. As a method of calculating the in-cylinder fuel amount fc, a method of calculating by dividing the measured value of the intake air amount Ga by the target air-fuel ratio AFR is generally used. This is based on the premise that the exhaust air-fuel ratio matches the air-fuel ratio AFRc of the new gas calculated by the following equation (1).
AFRc = Ga / fc (1)

However, not only the new gas sucked in the intake stroke of the current cycle (hereinafter also referred to as “k cycle”) but also the previous cycle (hereinafter also referred to as “(k−1) cycle”) in the combustion chamber 14. There are residual gas that has not been exhausted in the exhaust stroke, and blow-back gas that has been blown back to the intake port 18 in the (k-1) cycle and sucked again in the intake stroke of the k cycle. For example, as in transient operation, the air-fuel ratio of residual gas or blown-back gas may not match the air-fuel ratio AFRc of the new gas. For this reason, the exhaust air-fuel ratio cannot necessarily be matched with the target air-fuel ratio AFR by the above general method.
Therefore, in the system of the first embodiment, in consideration of the influence of the cylinder residual gas and the blowback gas, more strictly, the amount of fuel remaining in the cylinder and the amount of fuel blown back to the intake port 18 are considered. In principle, the fuel injection amounts fip and fid from the injectors 50 and 52 are calculated so that the exhaust air-fuel ratio AFRe becomes the target air-fuel ratio AFR.

  In FIG. 2, the symbol “Gr” represents the amount of gas remaining in the combustion chamber 14 at the start of intake (hereinafter referred to as “in-cylinder residual gas amount Gr”). The symbol “fr” represents the amount of fuel (burned fuel) contained in the residual gas (hereinafter referred to as “cylinder residual fuel amount fr”). The symbol “Gb” represents the amount of gas blown back from the combustion chamber 14 to the intake port 18 (hereinafter referred to as “blowback gas amount Gb”). The symbol “fb” represents the amount of fuel (burned fuel) contained in the blown-back gas (hereinafter referred to as “blow-back fuel amount fb”). The in-cylinder residual gas amount Gr, the in-cylinder residual fuel amount fr, the blow-back gas amount Gb, and the blow-back fuel amount fb all have a correlation with the operation state of the internal combustion engine 1. For this reason, by grasping the relationship between them and the operating state of the internal combustion engine 1 in advance and creating a map, the cylinder residual gas amount Gr, the cylinder residual fuel amount fr, the blowback gas amount Gb, and the blowback fuel Each value of the quantity fb can be estimated on the vehicle.

In FIG. 2, the symbol “Gex” represents the amount of exhaust gas discharged from the combustion chamber 14 during one cycle operation (hereinafter referred to as “exhaust gas amount Gex”). The exhaust gas amount Gex can be expressed using the intake air amount Ga, the in-cylinder residual gas amount Gr, and the blown back gas amount Gb.
Specifically, at the start of the k cycle, an amount of residual gas represented by Gr (k-1) remaining in the exhaust stroke of the (k-1) cycle exists in the cylinder, and the (k-1) cycle The amount of blown-back gas represented by Gb (k-1) blown back in the intake stroke or the exhaust stroke is present in the intake port 18. Furthermore, if the amount of new gas represented by Ga (k) and the amount of blown back gas represented by Gb (k-1) are sucked into the cylinder during the intake stroke of K cycle, The in-cylinder total gas amount at the end of the intake stroke is an amount represented by Ga (k) + Gr (k-1) + Gb (k-1). In the exhaust stroke of the internal combustion engine 1, a part of the in-cylinder total gas amount Ga (k) + Gr (k-1) + Gb (k-1) is discharged as an exhaust gas amount, and other in-cylinder total gas amount A part of Gb (k) is blown back to the intake port 18, and the remaining part becomes the cylinder residual gas amount Gr (k) remaining in the cylinder at the start of the next cycle. Therefore, the exhaust gas amount Gex (k) of k cycles can be expressed as the following equation (2).
Gex (k) = Ga (k) + Gb (k-1) + Gr (k-1) -Gb (k) -Gr (k) (2)

The symbol “fex” represents the amount of fuel (burned fuel) included in the exhaust gas and discharged from the combustion chamber 14 (hereinafter referred to as “exhaust fuel amount fex”). The discharged fuel amount fex can be expressed by using the in-cylinder fuel amount fc, the in-cylinder residual fuel amount fr, and the blowback fuel amount fb.
Specifically, the in-cylinder residual fuel amount at the start of the k cycle is fr (k), and the in-cylinder residual fuel amount at the end of the k cycle (that is, the in-cylinder residual fuel amount in the next cycle) is fr (k). If the amount of blown-back fuel at the end of the k cycle is fb (k) and the amount of in-cylinder fuel in the k cycle is fc (k), the exhausted fuel amount fex (k) in the k cycle is Can be expressed as the following equation (3).
fex (k) = fc (k) + fb (k-1) + fr (k-1) -fb (k) -fr (k) ... (3)

Further, the exhaust air-fuel ratio AFRe can be expressed by the following equation (4) using the exhaust gas amount Gex and the exhaust fuel amount fex.
AFRe = Gex / fex ... (4)
Therefore, from the above equations (2) to (4), the condition for matching the exhaust air-fuel ratio AFRe (k) in the k cycle with the target air-fuel ratio AFR (k) can be expressed by the following equation (5). it can.
AFR (k) = Gex (k) / fex (k)
= {Ga (k) + Gb (k-1) + Gr (k-1) -Gb (k) -Gr (k)} / {fc (k) + fb (k-1) + fr (k-1 ) -fb (k)
-fr (k)} ... (5)

In other words, the condition for setting the exhaust air-fuel ratio AFRe (k) in the k cycle to the target air-fuel ratio AFR (k) is that the fuel amount fc (k) flowing into the cylinder during the k cycle is expressed by the following equation (6). Value.
fc (k) = {Ga (k) + Gb (k-1) + Gr (k-1) -Gb (k) -Gr (k)} / AFR (k) -fb (k-1) -fr ( k-1) + fb (k)
+ fr (k) ... (6)
In other words, if the in-cylinder fuel amount fc (k) in the k cycle is calculated according to the above equation (6), the exhaust air / fuel ratio AFRe (k) in the k cycle can be matched with the target air / fuel ratio AFR (k). Become.

Next, an example of a method for calculating the fuel injection amounts fip and fid from the injectors 50 and 52 based on the in-cylinder fuel amount fc calculated as described above will be described.
FIG. 3 is a diagram for explaining a fuel behavior model representing the relationship between the fuel injection amounts fip and fid from the injectors 50 and 52 and the in-cylinder fuel amount fc. The system of the first embodiment calculates the fuel injection amounts fip and fid corresponding to the in-cylinder fuel amount fc on the premise of this fuel behavior model.

  Part of the fuel injected from the port injector 50 adheres to the port wall surface and the remainder flows into the combustion chamber 14. Assuming that the ratio of the fuel injected from the port injector 50 to the port wall surface or the like is “port adhesion rate Rp”, the amount of the fuel adhering to the port wall surface among the newly injected fuel is “Rp × fip”. It is represented by On the other hand, of the injected fuel, the amount of fuel taken into the combustion chamber 14 is represented by “(1-Rp) × fip”.

  In addition to the fuel “(1-Rp) × fip” supplied directly from the port injector 50, vaporized fuel generated by vaporization of the fuel adhering to the wall surface of the port flows into the combustion chamber 14. Assuming that the ratio of the fuel adhering to the wall of the port, etc. remaining in the adhering state is “port residual ratio Pp”, “Pp × fwp” of the port adhering amount fwp generated in the (k-1) cycle is The amount shown remains adhering to the port wall surface or the like, while the amount of fuel expressed by “(1-Pp) × fwp” is vaporized and sucked into the combustion chamber 14.

Therefore, the port adhesion amount at the start of the k cycle (for example, at the start of the intake stroke) is fwp (k), and the port adhesion rate and the port residual rate in the k cycle are Rp (k) and Pp (k), respectively. Yes, when the port injection amount in the k cycle is fip (k), the port adhesion amount fwp (k + 1) in the (k + 1) cycle can be expressed as the following equation (7).
fwp (k + 1) = Pp (k) × fwp (k) + Rp (k) × fip (k) (7)
Further, the fuel amount fcp (k) from the port injector 50 flowing into the combustion chamber 14 from the intake port 18 in the k cycle can be expressed as the following equation (8).
fcp (k) = (1-Rp (k)) × fip (k) + (1-Pp (k)) × fwp (k) (8)

On the other hand, as described above, the in-cylinder adhesion amount is not considered after the internal combustion engine 1 is completely warmed up. That is, it is assumed that the fuel injected from the in-cylinder injector 52 does not adhere to the wall surface of the combustion chamber 14 and exists in a vaporized (or atomized) state in the combustion chamber 14. Therefore, the fuel amount fcd (k) from the in-cylinder injector 52 existing in the vaporized (or atomized) state in the combustion chamber 14 in the k cycle can be expressed as the following equation (9).
fcd (k) = fid (k) ... (9)

From the above equations (8) and (9), the total amount of fuel (in-cylinder fuel amount) fc (k) supplied into the combustion chamber 14 in the k cycle can be expressed as the following equation (10). it can.
fc (k) = fcp (k) + fcd (k)
= (1-Rp (k)) × fip (k) + (1-Pp (k)) × fwp (k) + fid (k) (10)

As shown in the above equation (10), the in-cylinder fuel amount fc (k) is expressed as a function of the port injection amount fip (k) and the in-cylinder injection amount fid (k). During normal operation, in order to determine the port injection amount fip (k) and in-cylinder injection amount fid (k) that satisfy the above equation (10), as shown in the following equation (11), the port injection amount fip (k) And the injection ratio γ (k) of the in-cylinder injection amount fid (k). That is, the ECU 60 stores in advance a map that defines the relationship between the injection ratio γ (k) and the operating state of the internal combustion engine 1. From this map, the injection ratio γ (k) corresponding to the operating state of the internal combustion engine is read for each cycle.
fip (k) × (1-γ (k)) = fid (k) × γ (k) (11)

According to the above equations (6), (10), and (11), the port injection amount fip (k) and the in-cylinder injection amount necessary for supplying the fuel of fc (k) into the combustion chamber 14 fid (k) can be expressed by the following equations (12) and (13), respectively.
fip (k) = γ (k) / {(1-γ (k)) + (1-Rp (k)) × γ (k)}
× [{Ga (k) + Gb (k-1) + Gr (k-1) -Gb (k) -Gr (k)} / AFR (k)-(1-Pp (k)) × fwp (k )
-fb (k-1) -fr (k-1) + fb (k) + fr (k)] ... (12)
fid (k) = (1-γ (k)) / {(1-γ (k)) + (1-Rp (k)) × γ (k)}
× [{Ga (k) + Gb (k-1) + Gr (k-1) -Gb (k) -Gr (k)} / AFR (k)-(1-Pp (k)) × fwp (k )
-fb (k-1) -fr (k-1) + fb (k) + fr (k)] ... (13)

  The port adhesion rate Rp of the injected fuel and the port residual rate Pp of the adhered fuel have a correlation with the operation state of the internal combustion engine 1 (for example, the engine speed NE, the engine load KL, the cooling water temperature Tw, etc., and so on). is doing. For this reason, if the relationship between them and the operation state of the internal combustion engine 1 is grasped in advance and a map is created, both the port adhesion rate Rp and the port residual rate Pp can be estimated on the vehicle. is there. If they are estimated, the port injection amount fip and the in-cylinder injection amount fid for generating the in-cylinder fuel amount fc can be calculated according to the above equations (12) and (13).

  By the way, the port adhesion rate Rp and the port residual rate Pp, which are parameters of the fuel behavior model, may deviate from appropriate values due to a change with time, machine difference, and the like. In such a case, the port injection amount fip and the in-cylinder injection amount fid cannot be calculated with high accuracy, and thus learning correction of the port adhesion rate Rp and the port residual rate Pp is necessary. According to the apparatus disclosed in Patent Document 1, the learning correction of the adhesion rate and the residual rate is executed based on the deviation amount of the exhaust air / fuel ratio with respect to the target air / fuel ratio.

  In the dual injector system of the first embodiment, the injection rate of the in-cylinder injector 52 (hereinafter referred to as “in-cylinder injection rate”) and the injection rate of the port injector 50 (hereinafter referred to as “port injection rate”) are arbitrarily changed. It has the property of being able to. The present inventor has found a new learning correction method for the port adhesion rate Rp and the port residual rate Pp by paying attention to this characteristic. Hereinafter, a new learning correction method for the port adhesion rate Rp and the port residual rate Pp according to the first embodiment will be described.

  FIG. 4 is a timing chart for illustrating fuel injection control at the time of learning correction of the port adhesion rate Rp and the port residual rate Pp in the first embodiment. More specifically, FIG. 4 (A) is a diagram showing an in-cylinder injection rate, FIG. 4 (B) is a diagram showing a port injection rate, and FIG. 4 (C) is an exhaust air-fuel ratio AFRs. It is a figure which shows the value of.

  As shown in FIGS. 4A and 4B, the in-cylinder injection rate is 100% and the port injection rate is 0% before k cycles (for example, (k-1) cycles). That is, prior to the k cycle, only in-cylinder injection is performed, so that fuel is not attached to the port wall surface or the like. As shown in FIG. 4C, before the k cycle, the target air-fuel ratio AFR and the exhaust air-fuel ratio AFRs are values near the stoichiometric range. Therefore, during steady operation, the injection rate for realizing the exhaust air-fuel ratio AFRs in the vicinity of stoichiometry is 100% (the same applies to Embodiment 2-5 described later).

  Next, in the k cycle, the in-cylinder injection rate is changed to A%, and the port injection rate is changed to (100-A)%. Here, the total fuel injection amount {fip (k) + fid (k)} in the k cycle is the total fuel injection amount {fip (k-1) + fid (k-1) in the (k-1) cycle. Same as}. That is, only the injection ratios of the two injectors 50 and 52 are different between the k cycle and the (k-1) cycle. In this k cycle, part of the fuel injected from the port injector 50 adheres to the port wall surface and the like. As a result, although the total fuel injection amount has not changed, the fuel amount fc sucked into the combustion chamber 14 decreases. Therefore, as shown in FIG. 4C, the exhaust air-fuel ratio AFRs (k) in the k cycle changes to a value on the lean side with respect to the stoichiometry. In this way, by performing fuel injection from the port injector 50 only during k cycles (ie, one injection) during steady operation, the exhaust air-fuel ratio AFRe can be shifted to the lean side.

Next, a method for calculating the port adhesion rate Rp based on the above-described change in the exhaust air-fuel ratio AFRs from the vicinity of the stoichiometry to the lean side will be described.
In the k cycle, of the fuel injection amount fip from the port injector 50, the amount of fuel adhering to the port wall surface is “Rp × fip”, and the amount of fuel sucked into the combustion chamber 14 is (1-Rp) × fip. Further, considering the above-described blown back fuel amount fb and in-cylinder residual fuel amount fr, the fuel amount fex (k) discharged from the combustion chamber 14 in the k cycle can be expressed as the following equation (14).
fex (k) = (1-Rp (k)) × fip (k) + fid (k) + fb (k-1) + fr (k-1) -fb (k) -fr (k) ... (14)
Further, in the k cycle, the air amount Gex (k) discharged from the combustion chamber 14 is expressed by the above equation (2). Here, Gb (k-1), Gb (k), Gr (k-1), and Gr (k) in the above equation (2) are determined in relation to the operating state of the internal combustion engine 1 as described above. By referring to the obtained map, it can be estimated on the vehicle. Further, since the intake air amount Ga can be detected by the air flow meter 34, Gex (k) can be obtained. Further, since the internal combustion engine 1 is in steady operation, it can be assumed that the amount of exhaust air has not changed and Gex (k) = Gex. Therefore, the exhaust air / fuel ratio AFRe (k) in the k cycle calculated as a model is expressed by the following equation (15).
AFRe (k) = Gex (k) / fex (k)
= Gex / {(1-Rp (k)) × fip (k) + fid (k) + fb (k-1) + fr (k-1) -fb (k)
-fr (k)} ... (15)

  Here, it is impossible to simply compare the exhaust air-fuel ratio AFRe (k) based on the model calculation represented by the above equation (15) and the exhaust air-fuel ratio AFRs actually detected by the air-fuel ratio sensor 48. This is because, in the exhaust stroke of the k cycle, there is a dead time τ and a response delay T described later until the air-fuel ratio sensor 48 accurately detects the exhaust air-fuel ratio AFRs after the exhaust valve 42 is opened.

FIG. 5 is a diagram for explaining the dead time τ and the response delay T that are generated when the air-fuel ratio sensor 48 detects the exhaust air-fuel ratio AFRs.
When the exhaust valve 42 is opened at time t1, the exhaust air-fuel ratio AFRe (k) by the model calculation changes. Thereafter, the output AFRs of the air-fuel ratio sensor 48 starts to rise from a time t2 when a predetermined time has elapsed from the time t1. Thereafter, at time t3 when a predetermined time has elapsed from time t2, the exhaust air / fuel ratio AFRs detected by the air / fuel ratio sensor 48 reaches a value of 63% of the exhaust air / fuel ratio AFRe (k) calculated by the model. The time from time t1 to time t2 is the dead time τ, and the time from time t2 to time t3 is the response delay T. In the example shown in FIG. 5, the response delay T is the time constant of the 63% response of the air-fuel ratio sensor 48. The dead time τ has a correlation with the operation state and the origin (the length from the exhaust valve 42 to the air-fuel ratio sensor 48) of the internal combustion engine 1. Further, the response delay T has a correlation with the operation state of the internal combustion engine 1. Specifically, the response delay T is influenced by the amount of air and tends to be larger as the load is lighter and smaller as the load is higher. For this reason, the dead time τ and the response delay T can be estimated on the vehicle by grasping the relationship between them and the operating state of the internal combustion engine 1 in advance and creating a map.

The inventor has found that there is a relationship expressed by the following equation (16) between the exhaust air-fuel ratio AFRe obtained by model calculation and the exhaust air-fuel ratio AFRs detected by the air-fuel ratio sensor 48. It was.
T × {d (AFRs) / dt} + AFRs = AFRe ・ ・ ・ (16)
When the above equation (16) is discretized so that it can be calculated by the ECU 60, the following equation (17) is obtained.
AFRs (k + 1) = (1-Δt / T) × AFRs (k) + (Δt / T) × AFRe (k) (17)
Here, the exhaust air-fuel ratio AFRs (k) detected by the air-fuel ratio sensor 48 in the k cycle can be transformed as the following equation (18). In the following equation (18), “AFRsτ (t)” represents the exhaust air / fuel ratio AFRs detected by the air / fuel ratio sensor 48 at the calculation time t shifted by τ, and “AFRs (t + τ)” represents the air / fuel ratio. The exhaust air-fuel ratio AFRs detected by the sensor 48 is shifted by time (t + τ).
AFRs (k) = AFRsτ (t) = AFRs (t + τ) (18)
When the above equation (15) and the above equation (18) are applied to the above equation (17), the following equation (19) is obtained.
AFRs (t + τ) = (1-Δt / T) × AFRs (k-1) + (Δt / T) × [Gex / {(1-Rp (k)) × fip (k) + fid (k)
+ fb (k-1) + fr (k-1) -fb (k) -fr (k)}] ... (19)

When the above equation (19) is solved for Rp (k), the following equation (20) is obtained.
Rp (k) = 1- (1 / fip (k)) × [1 / {AFRs (t + τ)-(1-Δt / T) × AFRs (k-1)} × {Gex × (Δt / T )}
-fid (k) + fb (k-1) + fr (k-1) -fb (k) -fr (k)] ... (20)
Here, since only the injection rates of the two injectors 50 and 52 are different between the (k-1) cycle and the k cycle, the port injection amount fip (k) and the in-cylinder injection amount fid (k) in the k cycle. The value of (k-1) cycles can be used. In addition, the dead time τ and the response delay T can be estimated with reference to a map defined in relation to the operating state of the internal combustion engine 1. Of Gex, Ga (k) can be detected by the air flow meter 34, and the blow-back gas amounts Gb (k-1), Gb (k) and the cylinder residual gas amounts Gr (k-1), Gr (k) can be estimated with reference to a map defined in relation to the operating state of the internal combustion engine 1. Further, the blowback fuel amounts fb (k-1) and fb (k) and the in-cylinder residual fuel amounts fr (k-1) and fr (k) are also maps determined in relation to the operating state of the internal combustion engine 1. It can be estimated by reference. Further, the exhaust air / fuel ratio AFRs (t + τ), AFRs (k−1) can be detected by the air / fuel ratio sensor 48.
If they are estimated and detected, the port adhesion rate Rp (k) of k cycles can be calculated by calculating with the calculation period Δt according to the above equation (20). Further, since the internal combustion engine 1 is in steady operation, the operating conditions do not change greatly. Therefore, the port adhesion rate Rp (k) of k cycles obtained by the above equation (20) can be constant, that is, Rp = Rp (k) = Rp (k−1).

Returning to FIG. 4, in the (k + 1) cycle, the in-cylinder injection rate is changed to 100% again, and the port injection rate is changed to 0%. That is, after fuel injection is performed from the port injector 50 only for k cycles (only one injection), the state is returned to the state of only in-cylinder injection again. In the (k + 1) cycle, part of the fuel adhering to the port wall surface in the k cycle evaporates and is sucked into the combustion chamber 14. As a result, the fuel amount in the combustion chamber 14 increases although the total fuel injection amount has not changed. For this reason, as shown in FIG. 4C, the exhaust air-fuel ratio AFRe (k + 1) in the (k + 1) cycle changes to a richer value than the stoichiometric value.
In the subsequent (k + 2) cycle and (k + 3) cycle, the amount of fuel evaporated from the port wall surface and the like and sucked into the combustion chamber 14 gradually decreases. As a result, as shown in FIG. 4C, the exhaust air-fuel ratio AFRs gradually approaches stoichiometry and converges to the target air-fuel ratio AFR in the (k + 4) cycle.

Next, a method for calculating the port residual ratio Pp based on the above-described change of the exhaust air-fuel ratio AFRs to the rich side will be described.
When the port injection is performed in k cycles as in the example shown in FIG. 4, the port residual ratio Pp is obtained after the subsequent (k + 1) cycles.
In the (k + 1) cycle, the fuel amount fex (k + 1) discharged from the combustion chamber 14 can be expressed as the following equation (21).
fex (k + 1) = Rp × fip (k) × (1-Pp) + fid (k + 1) + fb (k) + fr (k) -fb (k + 1) -fr (k + 1) ··(twenty one)
Further, since the internal combustion engine 1 is in steady operation, the air amount Gex (k + 1) discharged from the combustion chamber 14 in the (k + 1) cycle is the same as the air amount Gex (k) in the k cycle. . That is, Gex (k + 1) = Gex (k) = Gex. Therefore, the exhaust air / fuel ratio AFRe (k + 1) in the (k + 1) cycle by the model calculation is expressed by the following equation (22).
AFRe (k + 1) = Gex (k + 1) / fex (k + 1)
= Gex / {Rp × fip (k) × (1-Pp) + fid (k + 1) + fb (k) + fr (k) -fb (k + 1)
-fr (k + 1)} ... (22)

The port residual ratio Pp in the (k + 1) cycle can be expressed as in the following equation (23) by the same method as that for calculating Rp.
Pp (k + 1) = 1- {1 / (Rp × fip (k))} × [[1 / {AFRs (t + τ)-(1-Δt / T) × AFRs (k)}]
× (Gex × Δt / T) -fid (k + 1) + fb (k) + fr (k) -fb (k + 1) -fr (k + 1)] ... (23)
Here, since the port injection amount and the in-cylinder injection amount are the same in the (k + 1) cycle and the (k-1) cycle, the port injection amount fip (k + 1) and the in-cylinder amount in the (k + 1) cycle As the injection amount fid (k + 1), a value of (k-1) cycles can be used. In addition, the dead time τ and the response delay T can be estimated with reference to a map defined in relation to the operating state of the internal combustion engine 1. For Gex, the value obtained when calculating the port adhesion rate Rp can be used. Further, the blowback fuel amounts fb (k), fb (k + 1) and the in-cylinder residual fuel amounts fr (k), fr (k + 1) are also maps determined in relation to the operating state of the internal combustion engine 1. It can be estimated by reference. The exhaust air / fuel ratio AFRs (t + τ) and AFRs (k) can be detected by the air / fuel ratio sensor 48.
If they are estimated and detected, the port residual ratio (k + 1) of (k + 1) cycles can be calculated by calculating with the calculation cycle Δt according to the above equation (23).

By reflecting the port adhesion rate Rp obtained by the above equation (20) and the port residual rate Pp obtained by the above equation (23) in the map, the learning correction of the port adhesion rate Rp and the port residual rate Pp is performed. Done.
Note that the port adhesion rate Rp is obtained only in k cycles in which port injection is performed. On the other hand, the port residual ratio Pp can be calculated in a plurality of cycles until the exhaust air-fuel ratio AFRs settles in a steady state. In the example shown in FIG. 4, the port residual ratio Pp can be calculated up to (k + 3) cycles. Thus, by calculating the port residual rate Pp in a plurality of cycles, the accuracy of learning correction of the port residual rate Pp can be improved.

[Specific Processing in Embodiment 1]
In the system according to the first embodiment, the routine shown in FIG. 6 is executed in learning correction of the port adhesion rate Rp and the port residual rate Pp. FIG. 6 is a flowchart of the learning correction control of the port adhesion rate Rp and the port residual rate Pp executed by the ECU 60 in the first embodiment.

In the routine shown in FIG. 6, first, it is determined whether or not the internal combustion engine 1 is in steady operation (step 100). If it is determined in step 100 that the engine is not in steady operation, the intake air amount Ga changes and it is determined that the port adhesion rate Rp and the port residual rate Pp cannot be obtained with high accuracy. Is temporarily terminated.
If it is determined in step 100 that the engine is in steady operation, it is determined whether or not the in-cylinder injection rate is 100%, that is, whether or not only in-cylinder injection is in effect (step 102). . If it is determined in step 102 that the in-cylinder injection rate is not 100%, it is determined that the fuel has already adhered to the port wall surface and the port adhesion rate Rp cannot be calculated. Exit once.

  If it is determined in step 102 that the in-cylinder injection rate is 100%, the injection rates of the injectors 50 and 52 are changed (step 104). In step 104, the in-cylinder injection rate is changed to A%, and the port injection rate is changed to (100-A)%. In the example shown in FIG. 4, the in-cylinder injection rate and the port injection rate are changed as in step 104 in k cycles. Thereby, in addition to the fuel injection from the in-cylinder injector 52, the fuel injection from the port injector 50 is performed.

  Next, the exhaust air-fuel ratio AFRs is detected by reading the output of the air-fuel ratio sensor 48 (step 106). In this step 106, the exhaust air-fuel ratio AFRs shifted to the fuel lean side due to the change in the injection rate in the above step 104 is detected. Then, using the exhaust air-fuel ratio AFRs detected in step 106, a port adhesion rate Rp, which is a learning value, is calculated according to the above equation (21) (step 108).

  Thereafter, in the next cycle, the injection rate of the injectors 50 and 52 is restored (step 110). That is, in this step 110, the in-cylinder injection rate is changed to 100%, and the port injection rate is changed to 0%. Then, the exhaust air-fuel ratio AFRs is detected in the same manner as in step 106 (step 112). In step 112, the exhaust air-fuel ratio AFRs shifted to the fuel rich side due to returning the injection rate in step 110 is detected.

Next, it is determined whether or not the exhaust air-fuel ratio AFRs detected in step 112 has settled steady (step 114). In step 114, it is determined whether or not the exhaust air-fuel ratio AFRs is the same as the previously detected exhaust air-fuel ratio AFRs.
If it is determined in step 114 that the exhaust air-fuel ratio AFRs is not steady, the port residual ratio Pp, which is a learning value, is calculated according to the above equation (24) (step 116).

  Thereafter, the process returns to step 112, and the exhaust air-fuel ratio AFRs in the next cycle is detected. Then, the determination in step 114 is performed. Thus, the port residual ratio Pp is calculated in step 116 until the detected exhaust air-fuel ratio AFRs settles steady.

  On the other hand, if it is determined in step 114 that the exhaust air-fuel ratio AFRs has settled in a steady state, the port adhesion rate Rp calculated in step 108 and the port residual rate Rp calculated in step 116 are reflected in the map. (Step 118). Thereby, learning correction of the port adhesion rate Rp and the port residual rate Rp is executed.

  As described above, according to the routine shown in FIG. 6, during the steady operation of the internal combustion engine 1, the exhaust air-fuel ratio AFRs shifted to the lean side by performing only one port injection from the state of only in-cylinder injection. Based on this, the port adhesion rate Rp is calculated. Further, the port residual ratio Pp is calculated based on the exhaust air-fuel ratio AFRs shifted to the fuel rich side by returning to the state of only the in-cylinder injection again. By reflecting the calculated port adhesion rate Rp and port residual rate Pp on the map, the port adhesion rate Rp and port residual rate Pp, which are parameters of the fuel behavior model, can be learned and corrected with high accuracy. Therefore, the fuel injection amounts fip and fid can be calculated with high accuracy, and the controllability of the air-fuel ratio can be improved.

  Incidentally, in the first embodiment, fuel is injected from the two injectors 50 and 52 in the k cycle, but fuel may be injected only from the port injector 50. That is, in the example shown in FIG. 4, the in-cylinder injection rate in the k cycle may be changed to 0%. In this case, since the amount of shift of the exhaust air-fuel ratio AFRs from the stoichiometry increases, the learned values of the port adhesion rate Rp and the port residual rate Pp can be easily calculated.

  In the first embodiment, the exhaust air-fuel ratio AFRs detected by the air-fuel ratio sensor 48 is shifted by the time τ and applied to the above equations (20) and (23). The exhaust air-fuel ratio AFRe may be applied while being shifted by the time τ.

  In the first embodiment, when the ECU 60 executes the process of step 104, the “injection rate changing means” in the first to third inventions executes the process of step 106, thereby The “adhesion rate correcting means” in the first and second inventions is realized by executing the processing of steps 108 and 118 in the “air-fuel ratio detecting means” in the second invention. Further, when the ECU 60 executes the process of step 110, the “injection rate changing means” according to the fourth aspect of the invention is executed. When the process of step 116 is executed, the “air-fuel ratio detecting means” of the fourth aspect of the invention is changed to the step. By executing the process 118, the “residual rate correction means” in the fourth invention is realized.

Embodiment 2. FIG.
Next, a second embodiment of the present invention will be described with reference to FIGS. The system of the second embodiment can be realized by causing the ECU 60 to execute a routine shown in FIG. 8 described later using the hardware configuration shown in FIG.

[Features of Embodiment 2]
In the first embodiment, the attached amount fwp, which is the amount of fuel attached to the port wall surface or the like, is calculated according to the above equation (7). For example, the adhesion amount fwp (k + 2) of (k + 2) cycles can be calculated according to the following equation (24).
fwp (k + 2) = Pp × fwp (k + 1) + Rp × fip (k + 1)
= Pp × (1-Rp) × fip (k + 1) + Rp × fip (k + 1) (24)

Incidentally, the adhesion amount fwp has a correlation with the operation state of the internal combustion engine 1. Therefore, it is possible to estimate the adhesion amount fwp on the vehicle by referring to a map determined in relation to the operating state of the internal combustion engine 1. Thus, the calculation load of ECU60 can be reduced by calculating | requiring the adhesion amount fwp using a map.
However, in this case, as with the port adhesion rate Rp and the port residual rate Pp, the adhesion amount fwp, which is a parameter of the fuel behavior model, may deviate from an appropriate value due to changes over time, machine differences, and the like. In such a case as well, learning correction of the adhesion amount fwp is necessary to accurately calculate the injection amounts fip and fid. Therefore, in the second embodiment, a learning method for the adhesion amount fwp will be described.

FIG. 7 is a timing chart for explaining fuel injection control at the time of learning correction of the adhesion amount fwp in the second embodiment. More specifically, FIG. 7 (A) is a diagram showing an in-cylinder injection rate, FIG. 7 (B) is a diagram showing a port injection rate, and FIG. 7 (C) is an exhaust air-fuel ratio AFRs. It is a figure which shows the value of.
As shown in FIGS. 7A and 7B, before the k cycle, the in-cylinder injection rate is 0% and the port injection rate is 100%. That is, prior to the k cycle, only port injection is performed, so that fuel is constantly attached to the port wall surface and the like. Further, as shown in FIG. 7C, before the k cycle, the target air-fuel ratio AFR and the exhaust air-fuel ratio AFRs are values near the stoichiometric range.

  Next, in the k cycle, the in-cylinder injection rate is changed to 100%, and the port injection rate is changed to 0%. Here, the fuel injection amount fid (k) in the k cycle is the same as the fuel injection amount fip (k-1) in the (k-1) cycle. That is, the total fuel injection amount is the same in the k cycle and the (k−1) cycle, and the injector for injecting fuel is changed from the port injector 50 to the in-cylinder injector 52. In this k cycle, part of the fuel fwp (k-1) adhering to the port wall surface during the (k-1) cycle evaporates and is sucked into the combustion chamber 14. As a result, the in-cylinder fuel amount fc (k) increases in the k cycle. Therefore, as shown in FIG. 7C, the exhaust air-fuel ratio AFRs (k) in the k cycle changes to a value on the rich side with respect to the stoichiometry. In this way, the exhaust air-fuel ratio AFRs can be shifted to the rich side by changing from the port injection only state to the in-cylinder injection only state in the k cycle during the steady operation.

  In the subsequent (k + 1) cycle and (k + 2) cycle, the amount of fuel that evaporates from the port wall surface and the like and is sucked into the combustion chamber 14 gradually decreases. As a result, as shown in FIG. 7C, the exhaust air-fuel ratio AFRs gradually approaches stoichiometry and converges to the target air-fuel ratio AFR in the (k + 3) cycle. The fuel amount corresponding to the deviation of the exhaust air-fuel ratio AFRs indicated by oblique lines in FIG. 7C can be considered as the adhesion amount fwp.

Next, a method for calculating the adhesion amount fwp based on the above-described change of the exhaust air-fuel ratio AFRs to the rich side will be described.
Assuming that the target air-fuel ratio is AFR and the exhaust gas amount from the combustion chamber 14 is Gex, the target fuel amount fref can be expressed by the following equation (25).
fref = Gex / AFR ... (25)
The exhaust gas amount Gex can be calculated according to the above equation (2) using the same method as in the first embodiment. Further, the target air-fuel ratio AFR is set according to the operating state of the internal combustion engine 1.

Further, assuming that the exhaust air-fuel ratio detected by the exhaust sensor 48 is AFRs and the exhaust gas amount from the combustion chamber 14 is Gex, the actual fuel amount fs discharged from the combustion chamber 14 is expressed by the following equation (26). It can be expressed as
fs = Gex / AFRs ... (26)

It is considered that the difference between the target fuel amount fref and the actual fuel amount fs adheres to the port wall surface or the like. Therefore, the adhesion amount fwp is considered to be obtained by integrating the difference between the target fuel amount fref and the actual fuel amount fs, and can be expressed as the following equation (27).
fwp = ∫ _k | fref (k) -fs (k) | dk (27)
The learning correction of the adhesion amount fwp is performed by reflecting the adhesion amount fwp obtained by the above equation (27) in the map.

[Specific Processing in Second Embodiment]
In the system of the second embodiment, the routine shown in FIG. 8 is executed when learning and correcting the adhesion amount fwp. FIG. 8 is a flowchart of the learning correction control of the adhesion amount fwp executed by the ECU 60 in the second embodiment.

  In the routine shown in FIG. 8, first, as in the first embodiment, it is determined whether or not the internal combustion engine 1 is in steady operation (step 100). If it is determined in step 100 that the routine operation is not being performed, it is determined that the intake air amount Ga and the exhaust air amount Gex change and the adhesion amount fwp cannot be obtained with high accuracy. Exit once.

If it is determined in step 100 that the vehicle is in steady operation, it is determined whether or not the port injection rate is 100%, that is, whether or not only the port injection is in effect (step 120). If it is determined in step 120 that the port injection rate is not 100%, this routine is temporarily terminated.
On the other hand, if it is determined in step 120 that the port injection rate is 100%, the target fuel amount fref is calculated according to the above equation (25) (step 122).

Next, the injection rate is changed (step 124). In step 124, the in-cylinder injection rate is changed to 100%, and the port injection rate is changed to 0%. In the example shown in FIG. 7, the injection rate is changed as in step 124 in k cycles. Thereby, it changes from the state of only port injection to the state of only in-cylinder injection.
Then, the exhaust air-fuel ratio AFRs is detected by reading the output of the air-fuel ratio sensor 48 (step 126). In step 126, the exhaust air-fuel ratio AFRs shifted to the fuel rich side due to the change in the injection rate in step 124 is detected. Thereafter, the actual fuel amount fs is calculated according to the above equation (26) using the exhaust air-fuel ratio AFRs detected in step 126 (step 128).

  Next, according to the above equation (27), the difference (absolute value) between the target fuel amount fref calculated in step 122 and the actual fuel amount fs calculated in step 128 is integrated (step 130).

  Then, the exhaust air-fuel ratio AFRs in the next cycle is detected in the same manner as in step 126 (step 132). Thereafter, it is determined whether or not the exhaust air-fuel ratio AFRs detected in the above step 132 has settled steady (step 134). If it is determined in step 134 that the exhaust air-fuel ratio AFRs is not steady, it is determined that the calculation of the adhesion amount has not ended. In this case, the actual fuel amount fs is calculated according to the above equation (27) using the exhaust air-fuel ratio AFRs calculated in step 132 (step 136). Thereafter, the process returns to step 130, and the difference between the target fuel amount fref calculated in step 122 and the actual fuel amount fs calculated in step 136 is integrated.

  Next, the processes in steps 132 and 134 are sequentially executed. If it is determined in step 134 that the exhaust air-fuel ratio AFRs has settled down steadily, it is determined that the adhesion rate calculation has been completed. In this case, the integrated value of the difference between the target fuel amount fref and the actual fuel amount fs accumulated in step 130 is reflected on the map as the adhesion amount fwp (step 138). Thereby, learning correction of the adhesion amount fwp is executed.

  As described above, according to the routine shown in FIG. 8, the exhaust air-fuel ratio shifted to the rich side by changing from the port injection only state to the in-cylinder injection only state during the steady operation of the internal combustion engine 1. AFRs are calculated. Then, the adhesion amount fwp is calculated by integrating the difference between the target fuel amount fref obtained from the target air-fuel ratio AFR and the actual fuel amount fs obtained from the exhaust air-fuel ratio AFRs. By reflecting the calculated adhesion amount fwp on the map, the adhesion amount fwp, which is a parameter of the fuel behavior model, can be learned and corrected with high accuracy. Therefore, the fuel injection amounts fip and fid can be calculated with high accuracy, and the controllability of the air-fuel ratio can be improved.

  By the way, in the second embodiment, the exhaust air / fuel ratio AFRs is largely shifted to the rich side when calculating the adhesion amount fwp, so that emission and drivability may be deteriorated. However, by performing learning correction of the adhesion amount fwp according to the second embodiment at the time of executing the rich spike of the NOx catalyst, the influence of the emission deterioration can be suppressed. Furthermore, the deterioration of drivability can be suppressed by retarding the output torque by retarding the ignition timing by the spark plug 16 from the normal timing.

  In the second embodiment, the example in which the injection rate is changed when the port injection rate is 100% before the k cycle has been described. However, the port injection rate is not necessarily 100%. The present invention can be applied when the port injection rate is such that a state in which fuel is steadily adhered to a port wall surface or the like is obtained.

  In the second embodiment, the ECU 60 executes the process of step 122, so that the “target fuel amount calculating means” in the fifth aspect of the invention executes the process of step 128 or 136 to execute the fifth. When the “actual fuel amount calculating means” in the invention executes the process of step 130, the “accumulating means” in the fifth invention executes the process of step 138, thereby “attachment amount correcting means in the fifth invention” "Is realized.

Embodiment 3 FIG.
Next, Embodiment 3 of the present invention will be described with reference to FIG. 9 and FIG. The system of the third embodiment can be realized by causing the ECU 60 to execute a routine shown in FIG. 10 described later using the hardware configuration shown in FIG.

[Features of Embodiment 3]
In the second embodiment, as shown in FIG. 7, the shift amount of the exhaust air-fuel ratio AFRs to the rich side due to the change in the injection rate is large. For this reason, it is possible to easily learn the adhesion amount fwp, which is a parameter of the fuel model, but there is a possibility that emission and drivability are deteriorated.
Therefore, in the third embodiment, a method for reducing the deterioration of emission and drivability by reducing the shift amount of the exhaust air-fuel ratio AFRs to the rich side will be described.

FIG. 9 is a timing chart for explaining fuel injection control at the time of learning correction of the adhesion amount fwp in the third embodiment. More specifically, FIG. 9 (A) is a diagram showing an in-cylinder injection rate, FIG. 9 (B) is a diagram showing a port injection rate, and FIG. 9 (C) is an exhaust air-fuel ratio AFRs. It is a figure which shows the value of.
Prior to the K cycle, as in the second embodiment, the in-cylinder injection rate is 0% and the port injection rate is 100%. As shown in FIG. 9C, before the k cycle, the target air-fuel ratio AFR and the exhaust air-fuel ratio AFRs are values near the stoichiometric range.

  Next, in the k cycle, the port injection rate is changed to 0% as in the second embodiment. However, in the third embodiment, the in-cylinder injection rate is not changed to 100% as in the second embodiment, but is changed to B% smaller than 100%. As a result, the total fuel injection amount is reduced as compared with the (k-1) cycle. In other words, the target air-fuel ratio AFR for k cycles is set to a value that is leaner than the stoichiometric value. Further, in this k cycle, part of the fuel fwp (k-1) adhering to the port wall surface during the (k-1) cycle evaporates and is sucked into the combustion chamber 14. However, by reducing the fuel injection amount in the k cycle as described above, the increase amount of the in-cylinder fuel amount fc (k) in the k cycle can be reduced as compared with the second embodiment. As a result, the shift amount of the exhaust air-fuel ratio AFRs to the rich side is made smaller than in the second embodiment.

  Thereafter, in the (k + 1) cycle, the port injection rate remains 0%, and the in-cylinder injection rate is further increased by B% and changed to 2B%. Thereby, the fuel injection amount is increased as compared with the k cycle. That is, in the (k + 1) cycle, the target air-fuel ratio AFR is set to a richer side than the k cycle. Further, although the amount of fuel evaporated from the port wall surface or the like and sucked into the combustion chamber 14 is reduced as compared with the k cycle, the exhaust air-fuel ratio AFRs in the (k + 1) cycle approaches the stoichiometry.

  Thereafter, in the (k + 2) cycle, the port injection rate remains 0% and the in-cylinder injection rate is changed to 100%. As a result, the fuel injection amount is further increased than in the (k + 1) cycle. That is, in the (k + 2) cycle, the target air-fuel ratio AFR is set to a richer side than in the (k + 1) cycle. Although the amount of fuel that evaporates from the port wall surface and is sucked into the cylinder is reduced as compared to the (k + 1) cycle, the exhaust air-fuel ratio AFRs in the (k + 2) cycle converges to stoichiometry.

  Note that the method of calculating the adhesion amount fwp based on the above-described change of the exhaust air-fuel ratio AFRs to the rich side is the same as that in the second embodiment, and thus the description thereof is omitted in the third embodiment.

[Specific Processing in Embodiment 3]
In the system of the third embodiment, a routine shown in FIG. 10 is executed when learning and correcting the adhesion amount fwp. FIG. 10 is a flowchart of the learning correction control of the adhesion amount fwp executed by the ECU 60 in the third embodiment.

In the routine shown in FIG. 10, first, the processes of steps 100, 120, and 122 are sequentially executed as in the second embodiment.
Subsequently, the injection rate is changed (step 140). In step 140, the in-cylinder injection rate is changed to B%, and the port injection rate is changed to 0%. In the example shown in FIG. 9, the injection rate is changed as in step 140 in k cycles. As a result, the state of only the port injection is changed to the state of only the in-cylinder injection, and the total fuel injection amount is reduced.

  Next, similarly to the second embodiment, the processes of steps 126, 128, and 130 are sequentially executed. Thereafter, in the next cycle, only the in-cylinder injection rate is increased by B% (step 142). In the example shown in FIG. 9, the in-cylinder injection rate is increased as in step 142 in the (k + 1) cycle. In step 142, the in-cylinder injection rate is increased by B% until the in-cylinder injection rate reaches 100%. Then, the exhaust air-fuel ratio AFRs changed due to the increase in the in-cylinder injection rate in step 142 is detected (step 132).

Next, it is determined whether or not the exhaust air-fuel ratio AFRs detected in step 132 has settled steady (step 134). If it is determined in step 134 that the exhaust air-fuel ratio AFRs is not steady, the actual fuel amount fs is calculated using the exhaust air-fuel ratio AFRs calculated in step 132 (step 136). The difference between the target fuel amount fref and the actual fuel amount fs is integrated (step 130). Thereafter, the processes in steps 142 and 132 are executed again.
On the other hand, if it is determined in step 134 that the exhaust air-fuel ratio AFRs has settled in a steady state, the integrated value integrated in step 130 is reflected on the map as the adhesion amount fwp (step 138). Thereby, learning correction of the adhesion amount fwp is executed.

  As described above, according to the routine shown in FIG. 10, the exhaust air-fuel ratio shifted to the rich side by changing from the port injection only state to the in-cylinder injection state during the steady operation of the internal combustion engine 1. AFRs are detected. The adhesion amount fwp is calculated by integrating the difference between the target fuel amount fref obtained from the target air-fuel ratio AFR and the actual fuel amount fs obtained from the exhaust air-fuel ratio AFRs. Then, by reflecting the calculated adhesion amount fwp on the map, the adhesion amount fwp, which is a parameter of the fuel behavior model, can be learned and corrected with high accuracy. Here, when switching from port injection to in-cylinder injection, by reducing the total fuel injection amount, that is, by setting the target air-fuel ratio AFR to be leaner than stoichiometric, the rich side of the exhaust air-fuel ratio AFRs Can be suppressed. For this reason, at the time of learning correction of the adhesion amount fwp, deterioration of emission and drivability can be suppressed.

  In the third embodiment, the in-cylinder injection rate is increased stepwise from 0% to B%, but the increase amount of the in-cylinder injection rate may be different. For example, the increase amount of the in-cylinder injection rate may be gradually increased. Also in this case, when calculating the adhesion amount fwp, the shift amount on the rich side of the exhaust air-fuel ratio AFRs can be suppressed, and the exhaust air-fuel ratio AFRs can be settled quickly and stably.

  In the third embodiment, the “target air-fuel ratio changing means” according to the sixth aspect of the present invention is realized by the ECU 60 executing the process of step 140.

Embodiment 4 FIG.
Next, a fourth embodiment of the present invention will be described with reference to FIG. 11 and FIG. The system of the fourth embodiment can be realized by causing the ECU 60 to execute a routine shown in FIG. 12 described later using the hardware configuration shown in FIG.

[Features of Embodiment 4]
In the second embodiment, the exhaust air-fuel ratio AFRs is shifted to a richer side than the stoichiometric state by changing from the state of only port injection to the state of only in-cylinder injection.
In the fourth embodiment, a case will be described in which the exhaust air-fuel ratio AFRs is shifted to the lean side by changing from the in-cylinder injection only state to the port injection only state, and the attached amount fwp is learned and corrected.

FIG. 11 is a timing chart for explaining fuel injection control at the time of learning correction of the adhesion amount fwp in the fourth embodiment. More specifically, FIG. 11 (A) is a diagram showing an in-cylinder injection rate, FIG. 11 (B) is a diagram showing a port injection rate, and FIG. 11 (C) is an exhaust air-fuel ratio AFRs. It is a figure which shows the value of.
As shown in FIGS. 11A and 11B, before the k cycle, the in-cylinder injection rate is 100% and the port injection rate is 0%. That is, prior to the k cycle, only in-cylinder injection is performed, so that fuel is not attached to the port wall surface or the like. As shown in FIG. 11C, before the k cycle, the target air-fuel ratio AFR and the exhaust air-fuel ratio AFRs are values near the stoichiometric range.

  Next, in the k cycle, the in-cylinder injection rate is changed to 0%, and the port injection rate is changed to 100%. Here, the fuel injection amount fip (k) in the k cycle is the same as the fuel injection amount fid (k-1) in the (k-1) cycle. That is, in the (k-1) cycle and the k cycle, the total fuel injection amount is the same, and the injector for injecting fuel is changed from the in-cylinder injector 52 to the port injector 50. In this k cycle, part of the fuel injected from the port injector 50 adheres to the port wall surface and the like, so the amount of fuel sucked into the combustion chamber 14 is reduced compared to the (k-1) cycle. As a result, as shown in FIG. 11C, the exhaust air-fuel ratio AFRs (k) in the k cycle changes to a value on the lean side with respect to the stoichiometry. As described above, in the k cycle during the steady operation, the exhaust air-fuel ratio AFRs can be shifted to the lean side by changing from the in-cylinder injection only state to the port injection only state.

  Thereafter, in the (k + 1) cycle, part of the fuel adhering to the port wall surface in the k cycle evaporates and is sucked into the fuel chamber 14. As a result, the amount of fuel sucked into the fuel chamber 14 increases as compared with the k cycle. For this reason, as shown in FIG. 11C, the exhaust air-fuel ratio AFRs of (k + 1) cycles approaches the stoichiometric state rather than k cycles.

  Further, in the (k + 2) cycle, the fuel adhering to the port wall surface in the k cycle and (k + 1) cycle evaporates and is sucked into the fuel chamber 14. For this reason, the amount of fuel sucked into the combustion chamber 14 from the intake port 18 further increases compared to (k + 1) cycles. As a result, as shown in FIG. 11C, the exhaust air-fuel ratio AFRs gradually approaches stoichiometry and converges to the target air-fuel ratio AFR in the (k + 3) cycle. Similar to the second embodiment, in FIG. 11C, the fuel amount corresponding to the deviation of the exhaust air-fuel ratio indicated by the oblique line can be considered as the adhesion amount fwp.

  In the fourth embodiment, as in the second embodiment, the difference between the target fuel amount fref and the actual fuel amount fs is integrated based on the above-described change of the exhaust air-fuel ratio AFRs to the lean side. Then, the adhesion amount fwp is calculated. Therefore, in the fourth embodiment, a detailed description of the method for calculating the adhesion amount fwp is omitted.

[Specific Processing in Embodiment 4]
In the system of the fourth embodiment, the routine shown in FIG. 12 is executed when learning and correcting the adhesion amount fwp. FIG. 12 is a flowchart of the learning correction control of the adhesion amount fwp executed by the ECU 60 in the third embodiment.

  According to the routine shown in FIG. 12, first, the processing of steps 100 and 102 is sequentially executed as in the first embodiment. Next, similarly to the second embodiment, the target fuel amount fref is calculated according to the above equation (26) (step 122).

Subsequently, the injection rate is changed (step 146). In step 146, the in-cylinder injection amount is changed to 0%, and the port injection rate is changed to 100%. In the example shown in FIG. 11, the injection rate is changed as in step 146 in k cycles. As a result, the state of only in-cylinder injection is changed to the state of only port injection.
Next, as in the second embodiment, the exhaust air-fuel ratio AFRs is detected (step 126), the actual fuel amount fs is calculated according to the above equation (27) (step 128), and the above equation (28) is obtained. Therefore, the difference (absolute value) between the target fuel amount fref and the actual fuel amount fs is integrated (step 130).

Thereafter, in the same manner as in the second embodiment, the exhaust air / fuel ratio AFRs in the next cycle is detected (step 132), and it is determined whether or not the exhaust air / fuel ratio AFRs detected in step 132 has settled in a steady state (step 132). Step 134). If it is determined in step 134 that the exhaust air-fuel ratio AFRs is not steady, the actual fuel amount fs is calculated from the exhaust air-fuel ratio AFRs detected in step 132 according to the above equation (26) (step 136). Thereafter, the process returns to step 130.
On the other hand, if it is determined in step 134 that the exhaust air-fuel ratio AFRs has settled steadily, the integrated value of the difference between the target fuel amount fref and the actual fuel amount fs accumulated in step 130 is mapped as the adhesion amount fwp. (Step 138). Thereby, learning correction of the adhesion amount fwp is executed.

  As described above, according to the routine shown in FIG. 12, the exhaust air-fuel ratio shifted to the lean side by changing from the in-cylinder injection only state to the port injection only state during the steady operation of the internal combustion engine 1. AFRs are detected. Then, the adhesion amount fwp is calculated by integrating the difference between the target fuel amount fref obtained from the target air-fuel ratio AFR and the actual fuel amount fs obtained from the exhaust air-fuel ratio AFRs. By reflecting the calculated adhesion amount fwp on the map, the adhesion amount fwp, which is a parameter of the fuel behavior model, can be learned and corrected with high accuracy. Therefore, the fuel injection amounts fip and fid can be calculated with high accuracy, and the controllability of the air-fuel ratio can be improved.

  By the way, in the fourth embodiment, when the adhesion amount fwp is calculated, the exhaust air-fuel ratio AFRs is largely shifted to the lean side, so that the emission and drivability may be deteriorated. However, it is possible to suppress the deterioration of the emission by performing learning correction of the adhesion amount fwp according to the fourth embodiment at the timing when the target air-fuel ratio AFR is intentionally made lean. Furthermore, the deterioration of drivability can be suppressed by increasing the output torque by advancing the ignition timing by the spark plug 16 more than usual.

  In the fourth embodiment, only the adhesion amount fwp is calculated. At the same time as the calculation of the adhesion amount fwp, the port adhesion rate Rp and the port residual rate Pp are calculated by the same method as in the first embodiment. It can also be calculated. In the example shown in FIG. 11, the port adhesion rate Rp can be calculated in k cycles in which the injection rate is changed, and the port residual rate Pp can be calculated in (k + 1) cycles and (k + 2) cycles. Yes (the same applies to Embodiment 5 described later). Then, since the adhesion amount fwp, the port adhesion rate Rp, and the port residual rate Pp can be learned and corrected at a time, the calculation accuracy of the fuel injection amounts fip and fid can be further improved.

  In the fourth embodiment, when the ECU 60 executes the process of step 122, the “target fuel amount calculating means” according to the fifth aspect of the invention executes the process of step 128 or 136 to execute the fifth step. When the “actual fuel amount calculating means” in the invention executes the process of step 130, the “accumulating means” in the fifth invention executes the process of step 138, thereby “attachment amount correcting means in the fifth invention” "Is realized.

Embodiment 5 FIG.
Next, a fifth embodiment of the present invention will be described with reference to FIGS. The system of the fifth embodiment can be realized by causing the ECU 60 to execute a routine shown in FIG. 14 described later using the hardware configuration shown in FIG.

[Features of Embodiment 5]
In the fourth embodiment, as shown in FIG. 11, the amount of shift of the exhaust air-fuel ratio AFRs due to the change in the injection rate to the lean side is large. For this reason, it is possible to easily learn the adhesion amount fwp, which is a parameter of the fuel model, but there is a possibility that emission and drivability are deteriorated.
Therefore, in the fifth embodiment, a method for reducing the deterioration of emission and drivability by reducing the shift amount of the exhaust air-fuel ratio AFRs to the lean side will be described.

FIG. 13 is a timing chart for illustrating fuel injection control at the time of learning correction of the adhesion amount fwp in the fifth embodiment. More specifically, FIG. 13 (A) is a diagram showing an in-cylinder injection rate, FIG. 13 (B) is a diagram showing a port injection rate, and FIG. 13 (C) is an exhaust air-fuel ratio AFRs. It is a figure which shows the value of.
Prior to the K cycle, as in the fourth embodiment, the in-cylinder injection rate is 100% and the port injection rate is 0%. As shown in FIG. 13C, before the k cycle, the target air-fuel ratio AFR and the exhaust air-fuel ratio AFRs are values near the stoichiometric range.

  Next, in the k cycle, as in the fourth embodiment, the in-cylinder injection rate is changed to 0%. However, in the fifth embodiment, the port injection rate is not changed to 100% as in the fourth embodiment, but is changed to (100 + C)%. As a result, the total fuel injection amount is increased as compared with the (k-1) cycle. That is, the target air-fuel ratio AFR of k cycles is set to a richer value than stoichiometric. In this k cycle, part of the fuel injected from the port injector 50 adheres to the port wall surface and the like, and the amount of fuel sucked into the combustion chamber 14 decreases. However, by increasing the fuel injection amount in the k cycle as described above, the reduction amount of the in-cylinder fuel amount fc (k) in the k cycle can be reduced as compared with the fourth embodiment. As a result, the shift amount of the exhaust air-fuel ratio AFRs to the lean side is made smaller than in the fourth embodiment. In FIG. 13C, the fuel amount corresponding to the deviation of the exhaust air-fuel ratio AFRs indicated by the oblique lines can be considered as the adhesion amount fwp.

  Thereafter, in the (k + 1) cycle, the in-cylinder injection rate remains 0% and the port injection rate is reduced by D%. As a result, the total fuel injection amount is reduced compared to the k cycle. Therefore, the exhaust air-fuel ratio AFRs approaches the stoichiometric state rather than the k cycle.

  Thereafter, in the (k + 2) cycle, the in-cylinder injection rate remains 0%, and the port injection rate is further reduced by D%. As a result, the total fuel injection amount is reduced as compared with the (k + 1) cycle. Further, in the (k + 3) cycle, the total fuel injection amount is further reduced, and the exhaust air-fuel ratio AFRs converges to stoichiometry.

  Note that the method for calculating the adhesion amount fwp based on the above-described change of the exhaust air-fuel ratio AFRs to the lean side is the same as that in the fourth embodiment, and thus the description thereof is omitted in the fifth embodiment.

[Specific Processing in Embodiment 5]
In the system of the fifth embodiment, a routine shown in FIG. 14 is executed when learning and correcting the adhesion amount fwp. FIG. 14 is a flowchart of the learning correction control of the adhesion amount fwp executed by the ECU 60 in the fifth embodiment.
In the routine shown in FIG. 14, first, the processing of steps 100, 102, and 122 is sequentially executed as in the fourth embodiment.

  Subsequently, the injection rate is changed (step 148). In step 148, the in-cylinder injection rate is changed to 0%, and the port injection rate is changed to (100 + C)%. In the example shown in FIG. 13, the injection rate is changed as in step 148 in k cycles. As a result, the in-cylinder injection only state is changed to the port injection only state, and the total fuel injection amount is increased.

  Next, similarly to the fourth embodiment, the processes of steps 126, 128, and 130 are sequentially executed. Thereafter, in the next cycle, only the port injection rate is reduced by D% (step 150). In the example shown in FIG. 13, in the (k + 1) cycle, the port injection rate is decreased as in step 150. In this step 150, the port injection rate is decreased by D% until the port injection rate reaches 100%. Then, the exhaust air-fuel ratio AFRs changed due to the decrease in the port injection rate in step 150 is detected (step 132).

  Next, similarly to the fourth embodiment, the process of step 134 is executed. If it is determined in step 134 that the exhaust air-fuel ratio AFRs is not steady, the actual fuel amount fs is calculated using the exhaust air-fuel ratio AFRs calculated in step 132 (step 136). Thereafter, the difference between the target fuel amount fref and the actual fuel amount fs is integrated (step 130). Thereafter, the process proceeds to step 142.

  On the other hand, if it is determined in step 134 that the exhaust air-fuel ratio AFRs has settled in a steady state, the integrated value integrated in step 130 is reflected on the map as the adhesion amount fwp (step 138). Thereby, learning correction of the adhesion amount fwp is executed.

  As described above, according to the routine shown in FIG. 14, the exhaust air-fuel ratio shifted to the lean side by changing from the in-cylinder injection only state to the port injection only state during the steady operation of the internal combustion engine 1. AFRs are detected. The adhesion amount fwp is calculated by integrating the difference between the target fuel amount fref obtained from the target air-fuel ratio AFR and the actual fuel amount fs obtained from the exhaust air-fuel ratio AFRs. Then, by reflecting the calculated adhesion amount fwp on the map, the adhesion amount fwp, which is a parameter of the fuel behavior model, can be learned and corrected with high accuracy. Here, when switching from in-cylinder injection to port injection, it is possible to suppress the lean side shift amount of the exhaust air-fuel ratio AFRs by increasing the total fuel injection amount, so at the time of learning correction of the adhesion amount fwp, Deterioration of emissions and drivability can be suppressed.

  In the fifth embodiment, the port injection rate is gradually decreased from (100 + C)% by D%, but the amount of decrease in the port injection rate may be different. For example, the reduction amount of the port injection rate may be gradually increased. In this case as well, when calculating the adhesion amount fwp, the shift amount on the lean side of the exhaust air-fuel ratio AFRs can be suppressed, and the exhaust air-fuel ratio AFRs can be settled quickly and stably.

  In the fifth embodiment, the “target air-fuel ratio changing means” according to the sixth aspect of the present invention is realized by the ECU 60 executing the process of step 148.

It is the schematic which shows the system configuration | structure by Embodiment 1 of this invention. 2 is a diagram schematically showing the behavior of fuel in the internal combustion engine 1. FIG. It is a figure for demonstrating the fuel behavior model showing the relationship between the fuel injection quantity fip, fid from each injector 50, 52, and the cylinder fuel quantity fc. 4 is a timing chart for explaining fuel injection control at the time of learning correction of the port adhesion rate Rp and the port residual rate Pp in the first embodiment of the present invention. FIG. 6 is a diagram for explaining a dead time τ and a response delay T that are generated when an air-fuel ratio sensor 48 detects exhaust air-fuel ratio AFRs. 4 is a flowchart of learning correction control of a port adhesion rate Rp and a port residual rate Pp executed by an ECU 60 in the first embodiment of the present invention. 9 is a timing chart for explaining fuel injection control at the time of learning correction of the adhesion amount fwp in the second embodiment of the present invention. In Embodiment 2 of this invention, it is a flowchart of the learning correction control of the adhesion amount fwp performed by ECU60. 10 is a timing chart for explaining fuel injection control at the time of learning correction of the adhesion amount fwp in Embodiment 3 of the present invention. In Embodiment 3 of this invention, it is a flowchart of the learning correction | amendment control of the adhesion amount fwp performed by ECU60. 10 is a timing chart for explaining fuel injection control at the time of learning correction of the adhesion amount fwp in Embodiment 4 of the present invention. In Embodiment 4 of this invention, it is a flowchart of learning correction | amendment control of the adhesion amount fwp performed by ECU60. In Embodiment 5 of this invention, it is a timing chart for demonstrating the fuel-injection control at the time of learning correction of the adhesion amount fwp. In Embodiment 5 of this invention, it is a flowchart of the learning correction | amendment control of the adhesion amount fwp performed by ECU60.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Internal combustion engine 6 Water temperature sensor 10 Crank angle sensor 14 Combustion chamber 16 Spark plug 18 Intake port 20 Intake valve 22 Intake passage 26 Throttle valve 30 Slot opening sensor 32 Accelerator opening sensor 34 Air flow meter 40 Exhaust port 42 Exhaust valve 44 Exhaust passage 46 Catalyst 48 Air-fuel ratio sensor 50 Port injector 52 In-cylinder injector 60 ECU

Claims (6)

A fuel injection control device for an internal combustion engine having a port injector for injecting fuel into an intake port and an in-cylinder injector for injecting fuel into a cylinder,
A fuel injection amount calculating means for calculating a fuel injection amount from the port injector and a fuel injection amount from the in-cylinder injector using a dynamic behavior model in consideration of fuel adhesion in the vicinity of the intake port;
An injection rate changing means capable of changing an injection rate of the port injector and the in-cylinder injector;
An air-fuel ratio detecting means for detecting an air-fuel ratio that changes due to the injection rate being changed by at least one cycle by the injection rate changing means;
An internal combustion engine comprising: an adhesion rate correcting means for correcting an adhesion rate of the injected fuel that is a parameter of the dynamic behavior model based on the changed air-fuel ratio detected by the air-fuel ratio detection means; Fuel injection control device. The fuel injection control device for an internal combustion engine according to claim 1,
The injection rate changing means changes the injection rate of the port injector from zero to a predetermined value in a predetermined cycle,
The air-fuel ratio detection means detects an air-fuel ratio that changes in the predetermined cycle,
The adhesion rate correction means includes learning value calculation means for calculating a learning value using an air-fuel ratio that has changed in the predetermined cycle, and corrects the adhesion rate based on the learning value calculated by the learning value calculation means. What is claimed is: 1. A fuel injection control device for an internal combustion engine. The fuel injection control device for an internal combustion engine according to claim 2,
The fuel injection control device for an internal combustion engine, wherein the injection rate changing means changes the injection rate of the in-cylinder injector to zero in the predetermined cycle. The fuel injection control device for an internal combustion engine according to claim 2 or 3,
The injection rate changing means changes the injection rate of the port injector from the predetermined value to zero in a cycle after the predetermined cycle,
The air-fuel ratio detection means detects an air-fuel ratio that changes in the subsequent cycle,
A fuel injection control device for an internal combustion engine, further comprising: a residual rate correcting means for correcting the residual rate of attached fuel, which is a parameter of the dynamic behavior model, based on the detected air-fuel ratio. The fuel injection control device for an internal combustion engine according to claim 1,
Target fuel amount calculating means for calculating a target fuel amount from the target air-fuel ratio;
Actual fuel amount calculating means for calculating an actual fuel amount from the air-fuel ratio detected by the air-fuel ratio detecting means;
Integrating means for integrating the difference between the target fuel amount calculated by the target fuel amount calculating means and the actual fuel amount calculated by the actual fuel amount calculating means;
A fuel injection control device for an internal combustion engine, further comprising: an adhesion amount correction unit that corrects an amount of fuel adhesion that is a parameter of the dynamic behavior model based on the difference accumulated by the accumulation unit. The fuel injection control device for an internal combustion engine according to claim 5,
When the injection rate of the in-cylinder injector is changed from zero to a predetermined value by the injection rate changing means, the target air-fuel ratio is changed to the fuel lean side with respect to the stoichiometry, and the port injector is changed by the injection rate changing means. The fuel injection control for an internal combustion engine, further comprising target air-fuel ratio changing means for changing the target air-fuel ratio to a fuel rich side with respect to stoichiometry when the injection rate of the engine is changed from zero to a predetermined value apparatus.

Images