JPH0364643A - Idle air-fuel ratio learning and reflecting method - Google Patents

Abstract

PURPOSE:To prevent an engine stall by using no idle learning value in the feedback control of the air-fuel ratio in the idling such as an engine heating, but controlling to reflect a preset lower limit guard value to the effective injection time. CONSTITUTION:When the fuel injection time of fuel injection valves 58 provided in a suction maniford 54 responding to the cylinders is controlled by a microcomputer 41, the fuel injection time is feedback-controlled to be a specific value after the engine heating is completed. And in this case, a learning value to be the correction value in respect of the air-fuel ratio of the fuel injection time is stored in a memory. And when the fuel injection time is open-loop- cotrolled in a cold time or in an engine heating time, the stored learning value is reflected to the fuel injection time. And in this case, in a cold time and in an idling time, the learning value read from the memory is compared with the specific value, and when the learning value is smaller than the specific value, the learning value is converted to a lower limit guard value, and reflected to the fuel injection time.

Description

DETAILED DESCRIPTION OF THE INVENTION (Field of Industrial Application) The present invention relates to an idle air-fuel ratio learning reflection method, and particularly relates to an idle air-fuel ratio learning reflection method that reflects a learned value of the idle air-fuel ratio to the fuel injection time during idling operation of an internal combustion engine. Concerning learning reflection methods.

[Conventional technology]

In a fuel-injected internal combustion engine, after warming up, the air-fuel ratio (
A/F) to the target air-fuel ratio, the engine load (intake pipe pressure PM or intake air per engine revolution ff1Q/N
Basic injection time tT based on E ) and engine speed NE
P is determined, and in addition to this, the intake temperature correction coefficient FTHA, the air-fuel ratio feedback correction coefficient FAF, and other various coefficients and A
An effective injection time TAU is calculated by multiplying /F learning value (learning control correction coefficient) KG, etc., and feedback control is performed to open the fuel injection valve for a time corresponding to this effective injection time TAU. During idling after completion of warm-up, the effective injection time TAU is calculated in the same manner as described above, and feedback control is performed.

However, when the engine is cold or warmed up, the oxygen concentration sensor detects the oxygen concentration in the exhaust gas and feeds back the detected signal (air-fuel ratio signal) to the computer, allowing the computer to calculate the air-fuel ratio. is inert,
Since a correct oxygen concentration detection result cannot be obtained, the above feedback control is not possible, and open loop control is performed.

Conventionally, during this cold time and during open loop control during warm-up, the A/F learning value learned during feedback control after complete warm-up is read from the memory and reflected when calculating the effective injection time. Improves air-fuel ratio control accuracy during open-loop control. (For example, Japanese Patent Application Laid-open No. 134741/1983).

[Problem to be solved by the invention]

However, the engine load, engine speed, and other conditions do not match when the engine is idling when it is cold or warmed up and when it is idling after being completely warmed up. /F Store the learned value in memory and use it during cold time or OJ.
! In the conventional method of directly reflecting the A/F learning value when the aircraft is idling, even if the idle A/F learning value is normally obtained after complete warm-up, depending on the value, the In some cases, A/F lean may cause rough idle or engine stall.

For example, if the fuel injection time is controlled by feedback control after warm-up is completed, and the engine enters an idling state after decelerating, some of the fuel injected from the fuel injection valve may adhere to the intake boat or intake pipe. Since the amount of liquid fuel in the fuel tank increases, the air-fuel mixture becomes richer. Therefore, the A/F learning value during idling is changed to a value for controlling the mixture to the lean side (a value smaller than the median value 1.0) in order to bring the mixture to the target air-fuel ratio, and is updated and stored in the memory. Ru.

When the engine is subsequently started, the idle switch is always on, so when the engine is cold or idling while warming up, the conventional method uses an A/F learning value that is smaller than the median value stored in the memory above. In order to directly reflect this in the effective injection time TAU, the effective injection time TALI is calculated based on a value on the lean side than the A/F learning value during idle that is originally required during cold and @ mid-aircraft. It becomes lean and causes rough idle and engine stall.

Especially when fast idle or the air conditioner is turned on when it is cold or when it is idling while warming up, the engine load increases, so it is necessary to increase the amount of fuel injection 121. Therefore, from the memory, the A/F learning value is set to a value smaller than the central ignition value. If this is read out and reflected in the effective injection time TAU, it is likely to cause the rough idle or engine stall described above.

The present invention has been made in view of the above points, and by setting a lower limit guard value to the learned value of the idle air-fuel ratio when cold or during warm-up, A/F The purpose of this invention is to provide a method for reflecting idle air-fuel ratio learning that prevents the occurrence of rough idle and engine stall due to lean conditions.

[Means to solve the problem]

In order to achieve the above object, the idle air-fuel ratio learning reflection method according to the present invention includes: after the internal combustion engine has been warmed up, the fuel injection time by the fuel injection valve is feedback-controlled so that the air-fuel ratio becomes a predetermined value; A learned value, which is a correction value related to the air-fuel ratio for fuel injection time, is stored in memory for each operating region, and the fuel injection time is controlled in an open loop during cold periods and warm-up.
In the system for reading out the learning value stored in the memory and reflecting it in the fuel injection time, the idle learning value read out from the memory at cold time, warm-up, and idling time is compared with a predetermined value;
When the idle learning value is smaller than the predetermined value, the idle learning value is changed to a predetermined lower limit guard value and reflected in the fuel injection time, and when the idle learning value is greater than the predetermined value, the idle learning value is changed to a predetermined lower limit guard value and reflected in the fuel injection time. is directly reflected in the fuel injection time.

[Effect]

In the present invention, when the idle air-fuel ratio learning value read from the memory, that is, the idle learning value, is smaller than a predetermined value when the internal combustion engine is cold, warmed up, and idling, the idle learning value is smaller than a predetermined value. When the learned value is leaner than the predetermined value, it is changed to a lower limit card value that is larger than the read value, and this is reflected in the fuel injection time.

Therefore, in the present invention, if the learned value during idling after warm-up is stored in the memory as a value on the leaner side than the predetermined value, and it is used as is when idling when cold or during warm-up, If the value is too lean than the required value, it will be changed to the lower limit card value, so the reflected idle learning value will not be too lean, and the idle learning value will not be too lean.
It can be set to the idle learning value originally required during the flight or a value close to it.

〔Example〕

Next, embodiments of the method of the present invention will be described with reference to the drawings. This example is applied to an electronically controlled fuel injection system (EFI-D) that detects the amount of air taken into an automobile engine per cycle based on the pressure (negative pressure) of the intake pipe. The idle learning value is reflected in the fuel injection time according to the flowchart shown in FIG.

Here, the operation according to the flowcharts shown in FIGS. 1 and 2 is performed by the engine shown in the schematic configuration diagram in FIG. 3 and the microcomputer 4 in FIG. 3 shown in detail in FIG. 4.
1, so FIGS. 3 and 4 will be explained first.

The engine shown in FIG. 3 includes an automatic transmission 42, and a microcomputer 41 controls fuel injection. In FIG. 3, 43 is a battery.

44 is a lamp for self-diagnosis, 45 is an air conditioner, and these are connected to the microcomputer 41. Also,
46 is an air cleaner, and a surge tank 48 is provided downstream of the air cleaner via a throttle valve 47. An intake temperature sensor 49 for detecting the intake temperature is installed near the air cleaner 46, and the throttle valve 47 has a
A slot position sensor 50 is attached that is turned on when the throttle valve 47 is fully open. Further, a diaphragm type pressure sensor 51 is attached to the surge tank 48.

Further, a bypass passage 52 is provided that bypasses the throttle valve 47 and communicates the upstream side and the downstream side of the throttle valve 47, and an idle speed control whose opening degree is controlled by a solenoid is provided in the middle of the bypass passage 52. - Control valve (ISCV) 53 is installed. The duty ratio of the current flowing through this l5CV53 is controlled to control the valve opening degree, and thereby the bypass passage 52
By adjusting the amount of air flowing through the engine, the idling rotation speed is controlled to the target rotation speed.

The surge tank 48 is the intake manifold (intake pipe) 5
4 and the combustion chamber 5 of the engine 56 via the intake boat 55
It is connected to 7. A fuel injection valve 58 is arranged for each cylinder so that a portion thereof protrudes into the intake manifold 54. Fuel in a fuel tank 59 is fed to this fuel injection valve 58 by a fuel pump 60 that is driven while the ignition switch 73 is on.
The fuel injection valve 58 injects this fuel into the airflow passing through the intake manifold 54 for an effective injection time TAU, which will be described later, according to instructions from the microcomputer 41.

The combustion chamber 57 is communicated with a catalyst device (not shown) via an exhaust boat 62 and an exhaust manifold (exhaust pipe) 63. Further, 64 is a spark plug, which is provided so that a part thereof protrudes into the combustion chamber 57. Further, 65 is a piston that reciprocates in the vertical direction in the figure.

The rotation angle sensor 66 detects the rotation of the shaft of the distributor and outputs an engine rotation signal to the microcomputer 41, for example, every 30' CA.

A water temperature sensor 67 is provided so as to penetrate through the engine block 68 and partially protrude into the water tank 17, and detects the temperature of the engine cooling water and outputs a water temperature sensor signal (THW). . Furthermore, 6 are oxygen a
The oxygen concentration detection sensor (02 sensor) is arranged so that a part thereof protrudes through the exhaust manifold 63, and detects the oxygen concentration in the exhaust gas before it enters the catalyst device.

70 is a vehicle speed sensor, and magnet 1 is rotated by the rotation of the speedometer cable and gears of the transmission.
~ drives the reed switch to generate a vehicle speed signal,
This is input to the microcomputer 41.

Furthermore, 72 is a start time injection switch attached to the water jacket 68.

73 is a cold start injector attached to the center of the surge tank 48. The cold start injector 73 is operated by the ignition switch 71 and the start injector time switch 72, completely apart from the control of the microcomputer 41, and injects fuel only when starting when the engine cooling water temperature is below a certain temperature, improving startability at low temperatures. make things better.

Next, the hardware configuration of the microcomputer 41 will be explained with reference to FIG. 4. In the figure, the same components as those in FIG. 3 are designated by the same reference numerals, and their explanations will be omitted. In FIG. 4, the microcomputer 41 is a central processing unit (
CPU)81. Read-only memory (ROM) that stores processing programs 82. Random Access Memory (RAM) 83. used as a work area. Backup RAM 84, C that retains data even after the engine is stopped
It has a clock generator 85 that supplies a master clock to the PU 81, connects these to each other via a bidirectional pass line 86, and has an input/output port 87. The configuration is such that the input boat 8 is connected to the output boats 89 to 91, respectively.

The microcomputer 41 also receives a pressure detection signal from the pressure sensor 51 taken out through a filter 92 and a buffer 93 in series, an intake temperature detection signal from the intake temperature sensor 49 taken out through a buffer 94, and a buffer 93.
The multiplexer 96 selects and outputs the water temperature sensor signal (THW) taken out via the A/D converter 9.
After converting the digital signal into a digital signal at step 7, the signal is sent to the pass line 86 via the input/output port 87 together with the vehicle speed signal from the vehicle speed sensor 70. Note that the filter 92 described above is a filter for removing the pulsating component of the intake pipe pressure contained in the output detection signal of the pressure sensor 51.

As a result, each input detection signal of the multiplexer 96 is
Under the control of the PU 81, the output is sequentially selected from the multiplexer 96, converted into a digital signal by the A/D converter 97, and then stored in the RAM 831. :Remembered.

Further, the microcomputer 41 inputs the oxygen concentration detection signal from the 02 sensor 69 to the comparator 99 via the buffer 798, where it is waveform-shaped and supplied to the input port 88. A signal obtained by waveform shaping the detection signal from
(not shown) and the output signal of the throttle position sensor 50 are respectively supplied to the input boat 88.

Furthermore, the microcomputer 41 has drive circuits 101 to 1.
03, and supplies the signal from the output boat 89 to the igniter 80 (not shown in FIG. 3) via the drive circuit 101. Further, a signal from the output boat 90 is supplied to the fuel injection valve 58 via a drive circuit 102 equipped with a down counter, and a signal from the output boat 91 is further supplied to the l5CV 53 via a drive circuit 103.

The microcomputer 21 having such a hardware configuration realizes the operations according to the flowcharts of FIGS. 1 and 2 of this embodiment. FIG. 1 is a flowchart for explaining the operation of an embodiment of the present invention, which is executed by interrupt processing. In FIG. 1, it is first determined whether or not it is the timing to calculate the effective injection time TAU of the fuel injection valve 58 based on the engine rotational speed signal from the rotation angle sensor 66 described above (step 11), and if it is not the calculation timing, this process is performed. When the routine is finished and it is time to calculate, proceed to the next step 1.
Proceed to step 2 to calculate the injection time TAU'.

This injection time TAU' is a value obtained by multiplying the basic injection time by various increase coefficients, and is calculated by the following formula.

TAU' =FAF-FTHA-FFC・(FIDL
+FOTP) ・ (FASE+FWL+1) ・ (tTP+tTPAEW) (1) In the above formula, F
AF is air-fuel ratio feedback correction coefficient, FTHA is intake temperature correction coefficient, FFC is reduction coefficient when returning from fuel cut, FIDL is idle stabilization increase/decrease L, FOTP is 0T
Pjlffi correction coefficient.

FASE is the temperature after starting m, FWL is the basic warm-up increase coefficient, t
TP is the basic injection time, and tTPAEW is the warm-up acceleration correction injection time.

Here, the basic injection time tTP is calculated from the intake pipe pressure and the engine rotation speed, and the basic injection time tTP is calculated from the intake pipe pressure and the engine rotation speed.
Values determined in advance through experiments are stored in the ROM 82 so that the air-fuel mixture supplied into the engine cylinders has a predetermined air-fuel ratio when fuel is injected from the control valve 58 for the basic injection time tTP.

In addition, the warm-up acceleration correction injection time tTPAEW is the time delay from the start of calculation of the effective injection time TAU until the actual fuel injection and the intake air in transient operating conditions such as during acceleration or deceleration. Correction for correcting the deviation of the air-fuel ratio of the air-fuel mixture from the target air-fuel ratio due to the time delay until the liquid fuel adhering to the boat 55 etc. flows into the combustion chamber 57. It takes a negative value during deceleration operation, and becomes zero during steady operation.

In addition, during feedback control after warm-up is completed, the air-fuel ratio feedback correction coefficient FAF increases when the air-fuel mixture is lean and decreases when the mixture is rich, depending on the 'fli elementary concentration detection signal from the 02 sensor 69. This value is calculated by a routine different from that shown in FIG. 1, and is set to 1.0 during open loop control during cold and warm-up.

The intake temperature correction coefficient FTHA is a coefficient calculated in a routine different from that shown in FIG. 1 based on the intake temperature detection signal from the intake temperature sensor 49, and is set to a value of 1.0 or close to 10 after warm-up is completed. When it is cold or in a ti machine, it is set to a value greater than 1.0. Furthermore, in formula (1), each remaining coefficient FFC, FWL
, FASE, FOTP.

FIDL is determined depending on when the engine 56 is in a predetermined state.

Next, in step 13 of FIG. 1, the CPU 81 receives the throttle valve 4 from the throttle position sensor 50
Based on the signal according to the opening degree of 7, whether it is in the idle state or not (
In other words, it is determined whether the throttle valve 47 is fully open or not, and if it is not in the idling state (off idling), the process proceeds to step 14 and selects and distinguishes the type of learning term, that is, the learning control correction coefficient KG, based on the rotation and load. . That is, there are seven learning control correction coefficients KG during off-idling, KG1 to KG7, as shown on the vest, depending on the intake pipe pressure PM and the engine speed NE.

Here, the values of KG1 to KG7 are tKG determined by interpolation from the map in FIG. 5 based on the intake pipe pressure PM.
is used.

In this way, KGi obtained from the map in Figure 5
(However, i=1 to 7) is multiplied by the injection time TAtJ' in step 15 of FIG. 1 to calculate the effective injection time TAU. The CPU 81 injects fuel from the fuel injection valve 58 once per engine rotation during the effective injection time TAU.

By the way, the learning control correction coefficient KG mentioned above is a learning value that is calculated by the character spacing routine shown in FIG. 2 during feedback control of the injection time, and is updated and stored in the backup RAM 84 every skip. Therefore, to explain the academic language routine shown in FIG. 2, first, in step 21, it is determined whether or not learning conditions are satisfied. This learning condition is, for example, when air-fuel ratio feedback control is in progress and the engine cooling water temperature is 80° C. or higher.

When it is determined that the learning condition is satisfied, it is determined in step 22 whether or not the air-fuel ratio feedback correction coefficient FAF has been skipped. As shown in FIG. 6, the air-fuel ratio feedback correction coefficient FAF is calculated from six O2 sensors, and as shown in FIG.
It is done like this.

When it is determined in step 22 that the air-fuel ratio feedback correction coefficient FAF should be skipped, in the next step 23, the values of the air-fuel ratio feedback correction coefficient FAF at the time of skipping (indicated by A, B, C, . . . in FIG. 6) are determined. and step 2
Two consecutive air-fuel ratio feedback correction coefficients FA of 4
The arithmetic mean of the values of F 1imFAFAV, i.e. (A+B
/2°(B+C)/2. (C+D)/2... is calculated.

Subsequently, in step 25, the arithmetic average value FAFAV is set to the upper limit value 1 of a predetermined range including a value corresponding to the target air-fuel ratio during feedback control (1.0 when the target air-fuel ratio is the stoichiometric air-fuel ratio).
．． 01 is exceeded, and if it is, the process proceeds to step 26, where it is determined whether or not the throttle position sensor 50 is on, and when it is on (that is, when idling), the idling learning control correction coefficient ( This is a learned value of the idle air-fuel ratio, and hereinafter referred to as "idle learned value j".
The value of KGO is changed to the value obtained by adding 0.005 to the previous value (Step 27), and when it is off, the learning control correction coefficient KG during feedback control (however, i = 1 to
7> is changed to a value obtained by adding o and oos to the previous value (step 28).

If it is determined in step 25 that the arithmetic average value FAFAV is less than the upper limit value 1.1, the process proceeds to step 29 and it is determined whether the arithmetic average value FAFAV is less than the lower limit value 0.95. If it is less than the lower limit, the process proceeds to step 30 to determine whether or not the throttle position sensor 50 is on, and if it is on, the idle learning value KGO is changed to a value obtained by subtracting 0.005 from the previous value (step 31).
Feedback when off! ! The value of the learning υ control correction coefficient KGi (hereinafter also referred to as "off-idle academic value") during control is changed to a value obtained by subtracting 0.005 from the previous value (step 32).Step 27゜28
．． Learning v obtained by calculating either 31.32
J correction coefficients KGO, KGi (these are collectively called K
G) is stored in the backup RAM 84.

Further, in step 29, when it is determined that the value of the arithmetic average value FAFAV is 0.95 or more, that is, 0.
．． When 99≦FAFAV≦1.01, the learning control correction coefficient KG is not updated; similarly, when it is determined in step 21 that the learning condition is not satisfied, and in step 2
Even when it is determined in step 2 that the air-fuel ratio feedback correction coefficient FAF has not been skipped, the learning control correction coefficient KG is not updated and this learning routine is ended.

In this way, when the air-fuel ratio is lower than the target air-fuel ratio, the arithmetic average value FAFAV exceeds the upper limit value, so the learning control correction coefficient KG is increased, while the air-fuel ratio is lower than the target air-fuel ratio. When the fuel ratio is richer than the fuel ratio, the arithmetic average value FAFAV is less than the lower limit value, so the learning control correction coefficient KG is reduced, and this learning control correction coefficient KG is reflected in the effective injection time TAU in step 15 of FIG. By doing so, the arithmetic average value FAFAV is 0.99 or more.1.
Control is performed such that the air-fuel ratio falls within a range of 0.01 or less, that is, the air-fuel ratio falls within a predetermined range centered on the target air-fuel ratio.

The above is the explanation during feedback control, but next, the operation effect during idling will be explained. When the CPU 81 determines the idle state in step 13 of FIG.
+ (e.g. 0°C) or above, the temperature t2 (e.g. 50°C)
to 60° C.).

When the engine cooling water temperature is not within this temperature range (feedback control after warm-up is completed, or during open loop control when the engine cooling water temperature is extremely low), the CPU 81 is locked out by the character spacing routine shown in FIG. The injection time T calculated in step 12 using the idle learning value KGO read from the backup RAM 84 as is.
AU' is multiplied to calculate the effective injection time TAU (step 17).

On the other hand, when the engine cooling water temperature is in the temperature range above t1 and below t2, that is, during open loop control during cold time and warm-up, the CPU 81 uses the backup RAM.
It is determined whether the value of the idle learning value KGO read from 84 is less than the median value of 1.0 (step 18), and if it is a rich value of 1.0 or more, the process proceeds to step 17 and backs up. The idle learning value KGO read from the up RAM 84 is used as is.

On the other hand, when the idle learning value KGO read from the backup RAM 84 is less than 1.0, which is the value on the lean side, the CPU 81 proceeds to step 19, does not use the idle learning value KGO, and sets the lower limit guard in advance. The idle learning value KGOB set as the value is set as the injection time TAt.
The effective injection time is calculated by multiplying J'. Here, the idle learning value KG as the lower limit guard value
The value of OB is set to the median value of KGO of 1.0 or a value near 1.0.

In this embodiment, when the engine coolant temperature is low (less than temperature t1), the lower limit guard value KGOB is not used and the idle learning value KGO is used as is, but this is different from the injection time TA.
In equation (1) above for calculating U', the late start amount F
There are ASE, basic warm-up increase coefficient FWL, etc., and these are designed to increase the amount of fuel injection at temperatures below t1, so using the lower limit guard value KGOB will correct the mixture to be excessively rich. This is because there is a possibility that the

On the other hand, when the engine coolant temperature is 11 or higher, the temperature is 4.
This is because when it is less than 2, the amount of fuel injected by the injection time TAU' is not increased so much, and the effective injection time TAU is greatly influenced by the value of the idle learning value KGO.

In this way, according to this embodiment, when the idle learning +1tlK GO is on the lean side when the idle time is cold or during warm-up, the lower limit guard value KG○B is reflected in the effective injection time TAU. (Step 19)
The engine will not become too lean, and therefore rough idle and engine stall can be prevented.

In addition, when the engine cooling water temperature is outside the predetermined temperature range, and even if it is within the predetermined temperature range, the idle learning value KGO
is 1.0 or more, the idle learning value obtained by learning during feedback control is used as is for the effective injection time T.
It is reflected in ALI, and in this example, idle learning 1
It! We are using KGO as effectively as possible.

It should be noted that the present invention is not limited to the above-mentioned embodiments, but can also be applied to an electronically controlled fuel injection system that uses an air flow meter to directly measure the amount of air taken into the engine. .

〔Effect of the invention〕

As described above, according to the present invention, when idling when cold or warming up, the preset lower limit guard value is used as the effective injection time without using the idling learning value learned during air-fuel ratio feedback control. This has the advantage of being able to prevent rough idle and engine stall due to excessive lean.

[Brief explanation of drawings]

FIG. 1 is a flowchart 1 for explaining the operation of an embodiment of the present invention.
~, M2 is a flow chart showing an example of a learning routine for the learning control correction coefficient KG used in FIG. 1, FIG. 3 is a schematic diagram of an engine to which the present invention can be applied MI4 Fig. 5 shows a map between the intake pipe pressure and the value of the learning control correction coefficient, and Fig. 6 shows the relationship between the air-fuel ratio signal and the air-fuel ratio feedback correction coefficient FAF. FIG. 11-19...Fuel injection time calculation processing step, 2
4-32...Learning! lI control correction coefficient/F learning 1f
i) calculation processing step, 41...microcomputer, 50...throttle position sensor, 51.
... Pressure sensor, 58 ... Fuel injection valve, 66 ... Rotation angle sensor, 67 ... Water temperature sensor, 69 ... Oxygen concentration detection sensor (02 sensor). Fig. 1 Fig. tKG Fig. 00 00 00 00 00 00 Intake pipe pressure (mmHg) Fig. Hi

Claims (1)

[Claims] After the internal combustion engine has been warmed up, the fuel injection time by the fuel injection valve is feedback-controlled so that the air-fuel ratio becomes a predetermined value, and a learning value that is a correction value for the air-fuel ratio of the fuel injection time is controlled. is stored in memory for each operating region, and the fuel injection time is controlled in an open loop during cold and warm-up periods.
In the system for reading out the learning value stored in the memory and reflecting it in the fuel injection time, the idle learning value read out from the memory at cold time, warm-up, and idling time is compared with a predetermined value; When the idle learning value is smaller than the predetermined value, the idle learning value is changed to a predetermined lower limit guard value and reflected in the fuel injection time, and when the idle learning value is greater than the predetermined value, the idle learning value is changed to a predetermined lower limit guard value and reflected in the fuel injection time. An idle air-fuel ratio learning reflection method characterized in that an idle learning value is directly reflected in the fuel injection time.