JPH0417751A - Air-fuel ratio control system of internal combustion engine - Google Patents

Abstract

PURPOSE:To accelerate the progress of learning and enhance the accuracy of learning simultaneously by setting the learning correction value of the amount of correction of air-fuel ratio output from a lower air-fuel ratio sensor according to a uniform learning correction value for all of driving areas and a learning correction value for each of the subdivided driving areas. CONSTITUTION:A first air-fuel ratio correction amount setting means sets the first amount of correction of air-fuel ratio according to the detected value of a first air-fuel ratio sensor. A uniform learning correction value correcting means corrects and renews a uniform learning correction value according to a value obtained by addition of the value of a storage means to the average value of learning correction values for all of driving areas. A first correcting means for the learning correction value of each area corrects the learning correction value of each area according to the value of the storage means and the output of a second air-fuel ratio sensor, and a second correcting means for the learning correction value of each area decreases and renews the uniform learning correction value during learning of the uniform learning correction value. A second air-fuel ratio correction amount computing means computes the second amount of correction of air-fuel ratio according to the second sensor output, the uniform learning correction value, and the learning correction value of each area, and the final amount of correction of air-fuel ratio is obtained by an air-fuel ratio correction amount computing means.

Description

Detailed Description of the Invention <Industrial Application Field> The present invention relates to a device for controlling the air-fuel ratio of an internal combustion engine, and in particular, the present invention relates to an apparatus for controlling the air-fuel ratio of an internal combustion engine, and in particular, the present invention relates to an apparatus for controlling the air-fuel ratio of an internal combustion engine, and in particular, an air-fuel ratio sensor is provided on the upstream side and downstream side of an exhaust purification catalyst. The present invention relates to a device that highly accurately controls an air-fuel ratio based on a detected value of a fuel ratio sensor.

<Prior Art> A conventional general air-fuel ratio control device for an internal combustion engine is disclosed in, for example, Japanese Patent Application Laid-Open No. 60-240840.

To give an overview of this, it detects the intake air flow rate Q and rotational speed N of the engine and calculates the basic fuel supply amount TP (=K・Q/N; K is a constant) corresponding to the amount of air taken into the cylinder. Then, this basic fuel supply amount T, corrected by engine temperature, etc., is used as air-fuel ratio feedback, which is set by a signal from an air-fuel ratio sensor (oxygen sensor) that detects the air-fuel ratio of the air-fuel mixture by detecting the oxygen concentration in the exhaust gas. Feedback correction is performed using a correction coefficient (air-fuel ratio correction amount), and correction based on battery voltage is also performed to finally set the fuel supply amount T1.

Then, by outputting a drive pulse signal with a pulse width corresponding to the fuel supply amount T thus set to the fuel injection valve at a predetermined timing, a predetermined amount of fuel is injected and supplied to the engine. .

The air-fuel ratio feedback correction based on the signal from the air-fuel ratio sensor is performed to control the air-fuel ratio to around the target air-fuel ratio (stoichiometric air-fuel ratio). This is installed in the exhaust system and oxidizes Co and HC (hydrocarbons) in the exhaust, as well as NO
This is because the conversion efficiency (purification efficiency) of the exhaust purification catalyst (three-way catalyst) that reduces and purifies x is set to function effectively in the exhaust state during combustion at the stoichiometric air-fuel ratio.

As mentioned above, the electromotive force (output voltage) generated by the air-fuel ratio sensor has a characteristic that changes by 2 in the vicinity of the stoichiometric air-fuel ratio, and this output voltage o and the reference voltage (slice level) S corresponding to the stoichiometric air-fuel ratio
It is determined whether the air-fuel ratio of the air-fuel mixture is rich or lean with respect to the stoichiometric air-fuel ratio by comparing the air-fuel ratio with L. For example, when the air-fuel ratio is lean (rich), the feedback correction coefficient α multiplied by the basic fuel supply amount T2 is increased (decreased) by a large proportionality constant P the first time the basic fuel supply amount T2 is changed to lean (rich).
The air-fuel ratio is controlled to be close to the stoichiometric air-fuel ratio by gradually increasing (decreasing) the fuel supply amount T+ by a predetermined integral constant ■.

By the way, in the above-mentioned normal air-fuel ratio feedback control device, one air-fuel ratio sensor is used to improve responsiveness.
It is installed in the gathering part of the exhaust manifold as close as possible to the combustion chamber, but since the exhaust temperature in this part is high, the characteristics of the air-fuel ratio sensor are likely to change due to thermal effects and deterioration.
Because the exhaust gas from each cylinder is not sufficiently mixed, it is difficult to detect the average air-fuel ratio of all cylinders, and the air-fuel ratio detection accuracy is difficult.
This in turn worsened the accuracy of air-fuel ratio control.

In view of this, it has been proposed that an air-fuel ratio sensor is also provided on the downstream side of the exhaust purification catalyst, and the air-fuel ratio is feedback-controlled using the detected values of the two air-fuel ratio sensors (Japanese Patent Laid-Open No. 58-48756). (see official bulletin).

In other words, the air-fuel ratio sensor on the downstream side has difficulty in responsiveness because it is far from the combustion chamber, but since it is downstream of the exhaust purification catalyst, it is less sensitive to the influence of the exhaust component balance (CO; HC2NOx
, CO2, etc.), and the amount of poisoning by toxic components in the exhaust is small, making it less susceptible to changes in characteristics due to poisoning.Moreover, since the exhaust gas is in a good mixing state, the average air-fuel ratio of all cylinders can be detected. Highly accurate and stable detection performance can be obtained compared to the upstream air-fuel ratio sensor.

Therefore, it is possible to combine two air-fuel ratio feedback correction coefficients that are respectively set by calculations similar to those described above based on the detected values of the two air-fuel ratio sensors, or to combine the air-fuel ratio feedback correction coefficients that are set by the upstream air-fuel ratio sensor. The downstream air-fuel ratio sensor compensates for variations in the output characteristics of the upstream air-fuel ratio sensor by correcting the control constants (proportional and integral), comparison voltage and delay time of the output voltage of the upstream air-fuel ratio sensor. The system is designed to perform highly accurate air-fuel ratio feedback control.

However, in the air-fuel ratio control device using two air-fuel ratio sensors as described above, the required level for air-fuel ratio correction during feedback control may be significantly different from that during non-feedback control, and especially during non-feedback control. The following problem occurs at the start of feedback control when transitioning to feedback control.

That is, in the above case, the feedback control speed by the downstream air-fuel ratio sensor is usually set smaller than the feedback control speed by the upstream air-fuel ratio sensor, so the feedback control speed by the downstream air-fuel ratio sensor is not used. air-fuel ratio correction amount (for example, the correction amount of the proportional portion of the air-fuel ratio feed hack correction coefficient by the upstream air-fuel ratio sensor)
It takes time for the air-fuel ratio to reach the required value, and in turn, it takes time to reach the target air-fuel ratio, resulting in deterioration of fuel efficiency, drivability, exhaust emissions, etc.

Furthermore, even during air-fuel ratio feedback control, when the operating state of the engine changes to a different range, the air-fuel ratio may deviate significantly from the target air-fuel ratio, which also causes deterioration of fuel efficiency, drivability, and exhaust emissions. .

Therefore, by sequentially calculating the average value of the second air-fuel ratio correction amount as a learning correction value and storing it for each driving region, and correcting and setting the fuel supply amount using the learning correction value, A system that allows stable air-fuel ratio control at all times has been proposed (see Japanese Patent Application Laid-Open No. 63-97851, etc.).
.

<Problems to be Solved by the Invention> By the way, since the second air-fuel ratio correction amount is for long-term correction of the deviation of the first air-fuel ratio correction amount, the control period is the same as that of the first air-fuel ratio correction amount. It is very long compared to the control cycle, so if the operating range in which the learning correction value is stored is made smaller, the time spent in each range becomes shorter and learning does not progress.

On the other hand, the required value of the learning correction value varies greatly depending on driving conditions (e.g., presence or absence of EGR), the value of the proportional component (in vehicles equipped with manual transmission, the proportional component in a certain area is made particularly small to avoid surges), etc. Therefore, if the operating range in which the learning correction value is stored is increased, the accuracy of learning will be impaired.

Therefore, conventionally, the driving range in which the learning correction value is stored is set by compromising the two goals of promoting the progress of learning and improving the accuracy of learning, but it is difficult to achieve both of these goals. This resulted in deterioration of drivability due to deterioration of exhaust emission characteristics and variations in air-fuel ratio.

The present invention was developed in view of these conventional problems, and it matches uniform learning over a wide driving range with area-specific learning for each subdivided driving range in order to maintain improvement in the accuracy of learning. An object of the present invention is to provide an air-fuel ratio control device for an internal combustion engine that simultaneously promotes the progress of learning and improves the accuracy of learning.

Means for Solving the Problems> For this reason, as shown in FIG. calculating a first air-fuel ratio correction amount according to the output value of first and second air-fuel ratio sensors whose output value changes depending on the concentration ratio of the middle specific gas component; and the first air-fuel ratio sensor. a first air-fuel ratio correction amount calculation means; a second air-fuel ratio correction amount calculation means for calculating a second air-fuel ratio correction amount based on the output of the second air-fuel ratio sensor and the learning correction value; an air-fuel ratio correction amount calculation means for calculating a final air-fuel ratio correction amount based on the first air-fuel ratio correction amount and the second air-fuel ratio correction amount; In the control device, a rewritable uniform learning correction value storage means storing a uniform learning correction value for uniformly correcting the second air-fuel ratio correction amount in the entire N range of operation; a rewritable area-specific learning correction value storage means that stores area-specific learning correction values for correcting the area-specific learning correction value for each of a plurality of divided driving regions; and an area-specific learning correction value retrieved from the area-specific learning correction value storage means. a first correcting area-based learning correction value for a corresponding operating region based on the value and the output of the second air-fuel ratio sensor;
area-specific learning correction value correction means, and uniform learning correction for correcting the uniform learning correction value by adding a value obtained by averaging the area-specific learning correction values to the uniform learning correction value retrieved from the uniform learning correction value storage means. a uniform learning correction value updating means for rewriting the uniform learning correction value of the uniform learning correction value storage means with the uniform learning correction value modified by the uniform learning correction value modifying means; a second area-specific learning correction value correction means for correcting the area-specific learning correction values by subtracting the added correction amount from the area-specific learning correction values of all driving regions; and the first area-specific learning correction value correction means. and an area-specific learning correction value updating means for rewriting the area-specific learning correction value of the corresponding driving area N with the area-specific learning correction value corrected by the second area-specific learning correction value correcting means; The learning correction value for calculating the second air-fuel ratio correction amount is set using the learning correction value and the area-specific learning correction value.

Function> The first air-fuel ratio correction amount setting means sets the first air-fuel ratio correction amount based on the detected value from the first air-fuel ratio sensor.

Further, the uniform learning correction value correction means performs learning for modifying the uniform learning correction value with a value obtained by adding a value calculated by averaging the learning correction values for each area to the uniform learning correction value retrieved from the uniform learning correction value storage means. The uniform learning correction value stored in the uniform learning correction value storage means is rewritten and updated with the modified uniform learning correction value by the uniform learning correction value updating means.

On the other hand, the area-specific learning correction value is corrected by the first area-specific learning correction value correction means based on the search value from the area-specific learning correction value storage means and the output of the second air-fuel ratio sensor;
When learning the uniform learning correction value, the uniform learning correction value is further corrected by decreasing the correction amount by the second area-based learning correction value correcting means, and the area-based learning correction value updating means corrects the corrected area-based learning correction value. The area-specific learning correction value stored in the area-specific learning correction value storage means is rewritten and updated with the value.

Then, the second air-fuel ratio correction amount calculation means calculates the second air-fuel ratio correction amount based on the output from the second air-fuel ratio sensor, the uniform learning correction value and the area-specific learning correction value, and The final air-fuel ratio correction amount is calculated by the air-fuel ratio correction amount calculation means based on the first air-fuel ratio correction amount and the second air-fuel ratio correction amount.

<Example> Embodiments of the present invention will be described below based on the drawings.

In FIG. 2 showing the configuration of one embodiment, an air flow meter 13 for detecting the intake air flow rate Q and a throttle valve 14 for controlling the intake air flow rate Q in conjunction with the accelerator pedal are provided on the intake road 12 of the engine 11. An electromagnetic fuel injection valve 15 serving as a fuel supply means is provided for each cylinder in the downstream manifold portion.

The fuel injection valve 15 is driven to open by an injection pulse signal from a control 22-16 having a built-in microcomputer, and injects fuel that is pressure-fed from a fuel pump (not shown) and controlled to a predetermined pressure by a pressure regulator. do. Further, a water temperature sensor 17 is provided to detect the temperature Tw of cooling water in the cooling jacket of the engine 11. On the other hand, the exhaust passage I8 is provided with a first air-fuel ratio sensor 19 that detects the air-fuel ratio of the intake air-fuel mixture by detecting the oxygen concentration in the exhaust gas in the manifold gathering part, and the A three-way catalyst 20 is provided as an exhaust purification catalyst that performs oxidation of Co and HC and reduction of NOx for purification, and furthermore, a second air-fuel ratio sensor having the same function as the first air-fuel ratio sensor is provided downstream of the three-way catalyst 20. An air-fuel ratio sensor 21 is provided.

Further, the distributor (not shown in FIG. 2) has a built-in crank angle sensor 22, and a crank angle signal outputted from the crank angle sensor 22 in synchronization with the engine rotation is counted for a certain period of time, or The engine rotation speed N is detected by measuring the period of the crank reference angle signal.

Next, the air-fuel ratio control routine by the control unit 16 will be explained according to the flowcharts of FIGS. 3 and 4. FIG. 3 shows a fuel injection amount setting routine, and this routine is performed at predetermined intervals (for example, every 10 m5).

In step 1 (denoted as S in the figure), the amount of intake air per unit rotation is determined based on the intake air flow rate Q detected by the air flow meter 13 and the engine rotation speed N calculated based on the signal from the crank angle sensor 24. The basic fuel injection amount T, which corresponds to T, is calculated using the following equation.

TP = KxQ/N (K is a constant) In step 2, various correction coefficients C0EF are set based on the cooling water temperature Tw etc. detected by the water temperature sensor 17.

In step 3, a feedback correction coefficient α set by a feedback correction coefficient setting routine to be described later is read.

In step 4, a voltage correction numerator is set based on the battery voltage value. This is to correct changes in the injection flow rate of the fuel injection valve 15 due to battery voltage fluctuations.

In step 5, the final fuel injection amount (fuel supply amount) T
I is calculated according to the following formula.

T r = T P X COE F X α + T.

In step 6, the calculated fuel injection valve T1 is set in the output register.

As a result, at a predetermined fuel injection timing synchronized with the engine rotation, a drive pulse signal having a pulse width of the calculated fuel injection amount T1 is applied to the fuel injection valve 15 to perform fuel injection.

Next, the air-fuel ratio feedback correction coefficient setting routine will be explained with reference to FIG. This routine is executed in synchronization with engine rotation.

In step 11, the operating conditions for performing feedback control of the air-fuel ratio (matching the operating conditions for learning the uniform learning correction value P HOSM and the area-specific learning correction value P HO5SX, which will be described later), however, the learning is performed in consideration of steady conditions. It is also possible to improve the accuracy by doing so.

If the operating conditions are not satisfied, this routine is terminated. In this case, the feedback correction coefficient α is clamped to the value at the end of the previous feedback control or a constant reference value, and the feedback control is stopped.

In step 12, the signal voltage V from the first air-fuel ratio sensor 19 is determined. 2 and the signal voltage V'o2 from the second air-fuel ratio sensor 21 are input.

In step 13, the signal voltage V of the first air-fuel ratio sensor 19 input in step 11 is determined. 2 and a reference value SL corresponding to the target air-fuel ratio (stoichiometric air-fuel ratio) to determine whether the air-fuel ratio is changing from lean to rich or from rich to lean.

When it is determined that it is the time of reversal, the process proceeds to step 14, where a uniform learning correction value P805M for learning and correcting the proportional correction amount PHOS of the air-fuel ratio feedback correction coefficient α, which is the second air-fuel ratio correction amount, is stored. The engine speed N and basic fuel injection amount T are searched from the uniform learning correction value MANOBU (stored in the RAM of the microcomputer built into the control unit 16).

The area-specific learning correction value P stored in the corresponding operating area from the area-specific learning correction value map (also stored in RAM) in which the area-specific learning correction value of the proportional correction amount P HOS is stored based on Search for HO5SX.

Incidentally, as shown in FIG. 5, the uniform learning correction value Matsubu includes one uniform learning correction value P8 for all operating ranges in which learning is performed.
05M is stored, and the area-based learning correction value map includes the engine rotational speed N and the basic fuel injection amount T.

Area-based learning correction values are stored in each of the driving areas, which are divided into three areas and divided into a total of nine areas.

Here, the RAM storing the uniform learning correction value P HOSM and the area-specific learning correction value P HOSSX is the uniform learning correction value P HOSM storage means and the area-specific learning correction value P H
Configure OSSX.

In step 15, the signal voltage V' from the second air-fuel ratio sensor 21 is determined. 2 and a reference value SL corresponding to the target air-fuel ratio (stoichiometric air-fuel ratio) to determine whether the air-fuel ratio is changing from lean to rich or from rich to lean.

When it is determined that it is the time of reversal, the process proceeds to step 16, where the area-specific learning correction value PH08SX searched in step 14 is set as the current value PHO3P, and then step 1
Proceed to step 7 and uniformly increase the learning correction value P HOSM correction amount)
IO5P is calculated using the following formula.

DPHOSP=M (PHO5P0+PH05P-1)
/ 2 Here, P HO5P-, is the area-specific learning correction value P Boss, when the output v', , of the second air-fuel ratio sensor 21 was reversed last time, and M is a positive constant (<l). be.

In other words, the correction amount DPHO5P is set as a bone value of a predetermined percentage of the value obtained by averaging the area-based learning correction value PHOSS every time the area is reversed.

In step 18, the uniform learning correction value P HOSM is determined by adding the correction amount DPHOSP calculated in step 17 to the uniform learning correction value P HOSM retrieved in step 14.
M is corrected, and the uniform learning correction value P HOSM stored in the RAM is updated with the corrected value.

Therefore, the portions of steps 17 and 18 correspond to uniform learning correction value modifying means, and the portion of step 18 also corresponds to uniform learning correction value updating means.

Next, in step 19, the area-by-area learning correction value PHOSSX for all driving ranges of the area-by-area learning correction value map is corrected by the value obtained by subtracting the correction amount DPHOSP.

That is, this step 19 corresponds to the second area-based learning correction value modification means.

In step 20, the area-based learning correction value P HOSSX calculated in step 19 is set as P HO5P-+ for the next calculation in step 17, and then the process proceeds to step 21.

When it is determined in step 15 that it is not inverted, step 1
Jump from step 6 to step 20 and proceed to step 21.

In step 21, the output of the second air-fuel ratio sensor 21 is determined. 2 to the standard value SL to find the air-fuel ratio rich.

Determine lean.

When the air-fuel ratio is determined to be rich (■"., >SL), the process proceeds to step 22, and a predetermined value [DP] is determined from the area-specific learning correction value PH08SX retrieved in step 14.
Area-specific learning correction value P HO by subtracting IO3R
Perform corrective calculations on SSX. Also, the air-fuel ratio is lean (■“.

When it is determined that z < S L ), step 23
Proceed to the searched area-specific learning correction value P HOSSX
The area-specific learning correction value PHOSSX is corrected using the value obtained by adding a predetermined value DPHO5L to DPHO5L. That is, step 22
and step 23 corresponds to the first area-based learning correction value modification means.

In step 24, the area-specific learning correction value PHOSS stored in the corresponding driving region of the area-specific learning correction value map is rewritten and updated with the area-specific learning correction value PH08Sx corrected in step 22 or 23. That is, this step 24 corresponds to area-based learning correction value updating means.

In step 25, the uniform learning correction value P HOSM and the area-specific learning correction value P H updated as described above are updated.
OSSX is added to calculate the proportional correction amount P)IQs as the second air-fuel ratio correction amount.

Next, the process proceeds to step 26, where the first air-fuel ratio sensor 19 makes a rich/lean determination, and when the lean-rich state is reversed, the process proceeds to step 27, where the decreasing direction given at the time of lean reversal is used to set the air-fuel ratio feedback correction coefficient α. The proportional bone PR is changed from the reference value PIIO to the second air-fuel ratio correction amount P.
Update HO5 with the decreased value. Then step 28
The air-fuel ratio feedback correction coefficient α is updated to a value obtained by subtracting the proportional bone PR from the current value.

Also, when the transition from rich to lean is reversed, proceed to step 29,
The increasing proportion Pc given at the time of lean inversion for setting the air-fuel ratio feedback correction coefficient α is updated with the value obtained by adding the second air-fuel ratio correction amount PHO5 to the reference value PLO. Next, in step 30, the air-fuel ratio feedback correction coefficient α is updated with a value obtained by adding the proportional amount PL to the current value.

Further, when it is determined in step 13 that the output of the first air-fuel ratio sensor 19 is not in the inversion state, the process proceeds to step 31 to perform a rich/lean determination, and if the output is rich, step 32
The process proceeds to step 33, where the air-fuel ratio feedback correction coefficient α is updated with a value obtained by subtracting the integral IR from the current value, and when lean, the process proceeds to step 33, where it is updated with a value obtained by adding the integral It.

Here, steps 26 to 33 are replaced by step 27. The function of setting the air-fuel ratio feedback correction coefficient α, excluding the correction in step 29, corresponds to the first air-fuel ratio correction amount setting means by the first air-fuel ratio sensor 19, and step 27. The portions from step 26 to step 33 including step 29 correspond to air-fuel ratio correction amount setting means.

With such a configuration, it is possible to promote convergence to the reference value by promoting learning in all driving regions where learning is performed using the uniform learning correction value P HOSM, and to promote convergence to the reference value.
HO5SX allows for highly accurate learning that varies by area.

In addition, FIGS. 6 and 7 show the uniform learning correction value P HO
It shows how the SM and area-specific learning correction value PHO5Sx are updated.

In this embodiment, while air-fuel ratio feedback control is based on the detected value of the first air-fuel ratio sensor 19, the proportional portion of the air-fuel ratio feedback correction coefficient is corrected based on the detected value of the second air-fuel ratio sensor. Although we have shown an example in which the air-fuel ratio feedback correction coefficient is set by each air-fuel ratio sensor and the air-fuel ratio feedback correction coefficient obtained by combining both values is used, the application is not limited to this. The present invention can also be applied to a system in which the air-fuel ratio feedback control is performed using the first air-fuel ratio sensor while correcting the reference value SL for rich/lean determination or the output delay time using the detection by the second air-fuel ratio sensor.

<Effects of the Invention> As explained above, according to the present invention, air-fuel ratio sensors are provided on the upstream and downstream sides of the exhaust purification catalyst, and air-fuel ratio feedback control is performed based on the detected values of these rain air-fuel ratio sensors. , the learning correction value of the air-fuel ratio correction amount that is set based on the output of the downstream air-fuel ratio sensor is divided into a uniform learning correction value that is learned and corrected uniformly for all operating regions, and a learning correction value that is learned and corrected for each subdivided operating region. Since the configuration is set using area-specific learning correction values, it is possible to simultaneously promote the progress of learning and improve the accuracy of learning, and maintain good exhaust emission characteristics and driving performance.

[Brief explanation of drawings]

FIG. 1 is a block diagram showing the configuration of the present invention, FIG. 2 is a diagram showing the configuration of an embodiment of the present invention, FIG. 3 is a flowchart showing the fuel injection amount setting routine of the same embodiment, and FIG. Similarly, FIGS. 5A and 5B are flowcharts showing the air-fuel ratio feedback correction coefficient setting routine, and FIGS. 2 is a diagram showing how the uniform learning correction value and the area-by-area learning correction value are updated, respectively. 11... Internal combustion engine 12... Intake passage injection valve 16.
...Air-fuel ratio sensor of control unit 20...Three-way catalyst air-fuel ratio sensor 15...Fuel injection 19...1st 21...2nd patent applicant Japan Electronics Co., Ltd. Agent Patent attorney Sasa Shima Fujio Figure 2 Figure 3 Figure 4 Shira Figure 5 Figure (Δ) Figure (B) Hunting station rotation 7ffN

Claims (1)

[Claims] The catalyst is provided on the upstream and downstream sides of an exhaust purification catalyst provided in the exhaust passage of an engine, and its output value changes in response to the concentration ratio of a specific gas component in the exhaust gas, which changes depending on the air-fuel ratio. first and second air-fuel ratio sensors; first air-fuel ratio correction amount calculation means for calculating a first air-fuel ratio correction amount according to the output value of the first air-fuel ratio sensor; a second air-fuel ratio correction amount calculating means for calculating a second air-fuel ratio correction amount based on the output of the fuel ratio sensor and the learning correction value; the first air-fuel ratio correction amount and the second air-fuel ratio correction amount; and an air-fuel ratio correction amount calculating means for calculating a final air-fuel ratio correction amount based on the following: In an air-fuel ratio control device for an internal combustion engine, the second air-fuel ratio correction amount is applied to the entire operating range. a rewritable uniform learning correction value storage unit storing a uniform learning correction value for uniformly correcting the second air-fuel ratio correction amount; and area-specific learning for correcting the second air-fuel ratio correction amount for each of the plurality of operating regions. a rewritable area-specific learning correction value storage means that stores correction values; and a corresponding operating region based on the area-specific learning correction value retrieved from the area-specific learning correction value storage means and the output of the second air-fuel ratio sensor. a first area-specific learning correction value correction means for correcting the area-specific learning correction value; and adding a value obtained by averaging the area-specific learning correction values to the uniform learning correction value retrieved from the uniform learning correction value storage means. uniform learning correction value correction means for modifying the uniform learning correction value by using the uniform learning correction value modifying means; and uniform learning correction value updating for rewriting the uniform learning correction value in the uniform learning correction value storage means with the uniform learning correction value modified by the uniform learning correction value modifying means. and second area-specific learning correction value correction means for correcting the area-specific learning correction values by subtracting the correction amount added by the uniform learning correction value correction means from the area-specific learning correction values of all driving regions. , Area-specific learning correction for rewriting the area-specific learning correction value of the corresponding driving region with the area-specific learning correction value corrected by the first area-specific learning correction value correction means and the second area-specific learning correction value correction means. An air-fuel ratio control system for an internal combustion engine, characterized in that the learning correction value for calculating the second air-fuel ratio correction amount is set using the uniform learning correction value and the area-specific learning correction value. Fuel ratio control device.