JPS6338652A - Air-fuel ratio control device for internal combustion engine - Google Patents

Abstract

PURPOSE:To improve control precision during the transient state, of an air-fuel ratio by a method wherein, from an amount or air passing a fuel injection valve part and a target air-fuel ratio, a fundamental fuel injection amount is computed, and learning correction is made on fuel delay due to a deviation between an actual air-fuel ratio and a target air-fuel ratio and learning correction on a fundamental injection amount. CONSTITUTION:Detecting values from a throttle valve opening sensor 25, a crank angle sensor 32, a water temperature sensor 33, an oxygen sensor 34 and the like are inputted to a control device 35. From the opening of a throttle valve and the number of revolutions, an intake air amount is determined, and a fundamental fuel injection amount is computed so that an air-fuel ratio is adjusted to a target value. An adhesion amount under the steady running conditions of a suction system and a change in an adhesion amount changed due to a primary delay are computed from an air amount and a cooling water temperature to determine a correction amount in relation to a fuel delay. Based on a deviation between a target air-fuel ratio and an actual air-fuel ratio from an oxygen sensor 34, the learning correction factor of a fuel delay and the learning correction factor of a fundamental injection amount are computed, and a fundamental fuel injection amount is corrected by means of a correction regarding to a fuel delay and a learning correction factor.

Description

[Detailed description of the invention] (Industrial application field) The present invention relates to an air-fuel ratio control device for an internal combustion engine.

(Prior art) Electronically controlled fuel injection engines are actually widely adopted due to their high fuel metering accuracy, and when controlling the injection amount supplied from the injection valve to the engine intake system, the engine load (e.g. intake air The basic fuel injection amount (basic pulse width) Tp (= -Qa/N, where K is a constant) based on the amount Q,) and the engine speed N is corrected according to other operating variables. Based on the following equation (1), the injection amount (injection pulse width) Ti
is calculated (for example, see "Automotive Engineering", Vol. 34, No. 11, p. 28, published by Railway Former Headquarters, November 1985).

T i ” T p X COE F X L A M
B D , A + T s...(1) However, Go E F: Each! 111 Total trol of correction coefficient L A M B D A: Air-fuel ratio correction coefficient Ts: Invalid pulse width (52 light ljr question to be solved, -γ) By the way, the fuel injection valve is Ba! In a device in which one or more injection valves are installed at a gathering point in the intake path far from the injection valve (hereinafter referred to as the "l5PI device"), errors associated with metering the amount of intake air and fuel delay are required for injection pulse width control. The two factors, the error associated with

This is because the air-fuel ratio is uniquely determined by the amount of air passing through the injection valve and the amount of fuel injected, so it is necessary to supply the amount of fuel commensurate with the amount of air in the injection valve. This is because the amount of air in the area was not measured.

For example, the amount of air QA I NJ passing through the injection valve section and the amount of air QCYL flowing into the engine cylinder do not correspond during a transient period. Now, the change in intake pipe pressure PB dPB/dt can be expressed by the following formula (2), so dPn/dt=Q A + N J-Qcyt,
-(2) When this equation is transformed, QAINJ is finally expressed by the following equation (3).

QA I N J =Q CY L +c
-dQ CY L /dt...(3) However, C is the change in dPo/dL and QCYL (dQcy
These are the coefficients when approximated assuming that L/dt) are approximately equal.

As can be seen from the opening formula (3), in steady state, QA
INJ and QCYL match, but fi
A deviation occurs by an amount of air corresponding to the S2 term.

FIG. 9 is a waveform diagram showing this situation. Therefore, S
In a PI device, QA I N J must be used as the air amount in order to improve accuracy during transient times.

In addition, with the SPI device, some of the injected fuel adheres to the inner wall surface of the intake pipe or intake boat before it reaches the shilling, or the amount of fuel that is floating in the intake pipe without being inhaled (these fuel amounts are (hereinafter collectively referred to as "adhesion amount") greatly affects the control accuracy of the air-fuel ratio.

For this reason, various correction methods have been proposed for fuel delay, but none of these directly deals with the amount of fuel adhering to the intake system, but instead detects and corrects the factors that affect the amount of adhering fuel. Therefore, there are inevitably many conditions for determining whether correction is necessary or not.

Therefore, we will try to match the air amount measurement error and the fuel delay system error at the same time. Since it is not possible to determine which error is due to fIJ, an attempt to improve drivability will increase the number of matching steps.Furthermore, if accuracy is compromised to a certain extent, drivability will inevitably deteriorate. On the other hand, if the allowance for matching is increased in order to prevent poor drivability, this will result in poor fuel efficiency and exhaust emissions at the expense of drivability.

Therefore, in order to improve the control accuracy of the air-fuel ratio in SPI devices and the like, it is necessary to separate the measurement error associated with air amount measurement from the error associated with fuel delay. Unless they are separated, the content of the correction will become ambiguous.

This invention was made by focusing on the conventional problem α, and it is possible to control the target air-fuel ratio by independently calculating the amount of air passing through the injector and the correction amount due to fuel delay. It is an object of the present invention to provide an air-fuel ratio control device which performs the following and also provides a learning function for errors after initial setting to accurately control the target air-fuel ratio.

(Means for Solving the Problem) In the present invention, as shown in FIG. , a means 3 for calculating the target air-fuel ratio TFBYA, and a basic fuel injection amount Tp from these air ff1QA+NJ and the target air-fuel ratio TFBYA.
means 4 for calculating the feedback correction amount LAMBDA based on the deviation between the target air-fuel ratio TFBYA and the actual air-fuel ratio obtained by the actual air-fuel ratio detection means 5; means 7 for calculating a fuel delay learning correction coefficient KBTLRC based on the deviation or this deviation and the feedback correction 1LAM B D A; M B D
means 8 for calculating a learning correction coefficient KBLRC of the basic injection amount based on A and the deviation; and means 9 for calculating a basic fuel delay correction IKATHO3 according to the operating state;
Means 10 for calculating the fuel injection amount Ti by correcting the basic injection amount Tp based on the fuel delay correction amount KATHO6 and the learning correction coefficients KBLRC and KBTLRC is provided.

For example, the injection amount calculation means 10 sets the fuel injection amount Ti to T
i”(QAI N J XTFBYAXKBLRC1K
A T H○SXKTBLRC) Calculate with XLAMBDA.

The figure also shows the learning correction coefficient KBLRC based on the feed par correction [LAMBDA] and the target air-fuel ratio TF.
Learning correction coefficient K based on the deviation between BAYA and the actual air-fuel ratio
This shows a case where BTLRC is calculated. 1 is a means for detecting the operating state.

(Function) With this configuration, the air amount QAINJ passing through the injection valve section is calculated with high accuracy regardless of whether it is in a transient state or not, so the air amount QAINJ and the injection amount from the injection valve The air-fuel ratio can be determined uniquely, and if there is no fuel delay, accurate control to the target air-fuel ratio is possible.However, the target air-fuel ratio can only be obtained at the injection valve. By the time the engine is inhaled, a deviation from the target air-fuel ratio may occur due to fluctuations in the amount of adhering fuel in the intake system.The cause of this deviation is fuel lag, and this is due to the correction amount KATHO8 for fuel lag. will be corrected.

That is, in this invention, by first determining QAINJ, the error associated with the measurement of the air amount and the error associated with the fuel delay are determined independently. becomes easier. This improves both the accuracy of measuring the amount of air and the accuracy of correcting fuel lag, and as a result, no matter where the injector is installed, such as in the intake pipe gathering section or each boat section, and regardless of the transient state. Flat void ratio characteristics can be obtained regardless of the

On the other hand, with this configuration, there is no need to worry about control accuracy during initial settings, but if there are variations in component parts, changes over time after revision, or differences in fuel properties, the high accuracy during settings may deteriorate. Since learning is performed for such error factors, high accuracy is maintained.

This will be explained below using examples.

(Example) In Fig. 2, one fuel injection valve 24 is provided in the intake passage 22 upstream of the intake throttle valve 21 to serve all cylinders (SPIvc arrangement).
, and shows the mechanical configuration when this invention is applied to an engine that uses throttle jt-opening α (also referred to as TVO) instead of air volume as the Rinseki load signal in order to simplify the device. There is.

Therefore, in this example, the injection pulse width is controlled using α and N as basic variables (hereinafter, this will be referred to as the α-N force formula).

Therefore, an air amount sensor is not provided, but a throttle valve opening sensor 25 is provided instead.

In addition, in the passage 23 that bypasses the throttle valve 21, two idle-up solenoid valves (referred to as S■) 26.27 are provided in parallel in order to improve control at the time of starting.
2-) is also provided with a whirl control valve 28.

In addition, 8! There are no differences from the conventional device, such as the engine speed N is determined by the crank angle sensor 32 built into the distributor 31, the cooling water temperature Tu is determined by the water temperature sensor 33, and an oxygen sensor 34 is provided as a sensor for detecting the actual air-fuel ratio. , these crank angle signals (reference signal and angle signal),
The water temperature signal and the actual air-fuel ratio signal are input to the control unit 35 together with the throttle valve opening signal, and within the control unit 7, an optimum fuel injection pulse width Ti is calculated based on these signals.

Now, the characteristic part of this invention lies in the calculation content of this injection pulse width T1, which is shown in FIG.
Consisting of Between below. ) to FIG. 7. The details of these are as follows: Figure 3 shows the main routine for calculating the injection pulse width, and Figures m4 to 7 show the variables used in the main routine (fuel delay correction amount KATt).
(O8, fee 1' back correction amount L A M B D
A, target air-fuel ratio TFBYA, intake temperature correction coefficient KTA)
This is a subroutine to find. The numbers in the figure represent processing numbers. Such control can be easily performed by controlling the control unit 35 using a microcomputer. In this case, the calculation of each variable is executed in the control cycle shown in the table below.

First, the routine shown in FIG. 3 is the part that finally calculates the injection pulse width Ti using the equation (4) below.
．． Equivalent to 10 abilities.

Here, in a device where the injection valve is attached to the suction boat near the engine shilling, the amount of air flowing into the shilling is QCY.
L and the amount of air QAINJ passing through the injection valve section are approximately equal. Note that there is a slight fuel delay in the boat section. Compared to such devices, SPI devices have Q
There is a difference between Cy +- and QAINJ, and there is a difference in that there is a large supply delay until the fuel injected from the injector reaches the shilling.

This invention takes into consideration two points, the amount of air QAINJ passing through the injection valve part and the fuel delay, so that it can be used commonly in any case. However, rather than treating these as one, by calculating each of them independently (calculating QAINJ for the air amount and finding the basic fuel delay correction 1KATHO8 for the fuel delay), the consideration method can be simplified. This makes it clear whether the control error is due to a two-gas metering error or a fuel delay. This greatly improves the accuracy during setting.

Furthermore, if there are variations in component parts, changes in h over time, or differences in fuel properties after setting, these factors will reduce accuracy. Therefore, a learning function is added to deal with these factors.

Expressing this mathematically, the effective pulse width Te is calculated by the following formula (4
) (step 70). Note that the point that the invalid pulse width is Ts and the sum with Te is Ti (=Te+Ts) is the same as before (step 69.70).

Tc=(TpXKBLRC+KATHO3XKBTLR
C) XLAMBDA ... (4) However, Tp: Basic pulse width KATHOS: Basic fuel delay correction amount (transient correction fit) L A M B D A: Feedback correction amount (air-fuel ratio correction coefficient) KBLRC:, 15 injections Quantity learning correction coefficient KBTLRC
: Fuel delay learning correction coefficient. Here, Tp is used as the basic pulse width as in the conventional example, but its content is calculated using the following equation (5), unlike the conventional example.

TI)=QA I N J G XTFBYAXK-(
5) However, QA I N J G: Injection valve chamber 'A
Ing TFBYA: Target air-fuel ratio: A constant (1 ns/mg) based on the injector characteristics.

First, although there is an air amount QAINJ in the injection valve section, it is difficult to directly obtain this in this embodiment which does not have an air amount sensor, so it is obtained based on QCYL. In other words, QAINJ is QCYL and its variation dQ c Y
As mentioned above, L/dL can be approximately determined by the curve formation (3).

Qa I N J =Q CY L +c-dQ CY
L/dt...(3) Equations equivalent to this equation (3) are the following group of equations (6Δ) to (6F).

QA I N J G = Q A I N J
cXKTA −°(6A)QA INJC=Q
CYL XVCYL+DCM...(
6B) Q CY L = Q IIXK2 +Q CY L -I X (1- to 2)... (6C
) Q++=Quo XKFLAT=16D)D
CM=(QcyL QCYL-l)XKM
AN I OX T ref'-(6E)KTA=K
TAOXKTAQ CY L...(6F) However, QAINJG: Injection valve air amount/Shipping (111g
) QAINJC: Injection valve air amount / Schilling (cc) Qcyt, Air amount to Niji ring / Schilling volume (%) VCYL Niji ring volume (cc) DCM: Manifold air change amount (cc) KTA: If
i Temperature correction coefficient <mg/cc) Q u : Equilibrium air amount/
Schilling V product (%) QCYL-l: QCY at previous calculation
L K2: QCYL change rate/calculation QHo: Linearize air amount/Schilling volume (%) KFLAT Nearat air-fuel ratio coefficient (%) KMAN I
O: Cycle of manifold coefficient TreGRefm9- (μ5) KTAO: Basic intake temperature correction coefficient Log/cc) KTAQ
cyt: Load correction rate (%) for intake temperature correction.

These second groups (6^) to (6F) have become complicated due to various corrections and standardizations (converted to per cylinder, shilling volume of light, etc.), but basically they are teeth,
QAINJc is determined by the sum of a steady term (QCYLXVCYL) and a transient term (DCM).

However, since this value QAINJC is in volume units, it can change due to changes in intake air temperature, so it is converted to mass units using K T A as a correction coefficient (steps 61 to 63).
).

Also, QCYL is QII, Q with 2 as a smoothing constant.
Weighting using CYL-1 as a variable and K2 as a weight is determined by the average value (steps 54 to 57).

Next, variables such as Q Ha and K F L A T are determined from the flow path area of the intake system and the engine speed. This is because one air sensor is eliminated from the intake system to reduce costs and facilitate maintenance.

Therefore, the flow path area is determined by the following equations (6G) and (all) (steps 41 to 52).

AADNV=AAXTref/VcYL-(6G
)AA=ATVO+AI+AAC...(6+1)
However, AADNV: flow path area/(rotation speed x shilling volume) (cII+2/rpfO・CC) AA: total flow path surface B
T (0m2) ATVO: Throttle valve flow path area (CI112) AISV2 [
3 flow path surface fj? (elf12) AAC: 5V27 channel area (CTII2).

That is, in this example, the flow path surface $32ATVO based on the throttle valve opening degree TVO is used as the load signal.
If there is a passage 23 that bypasses the throttle valve 21, it is necessary to consider these areas AI and AAC, and therefore the total flow area AA is equal to the flow passage area A based on the throttle valve opening degree.
Flow area of TVO and bypass passage (AI or AAC
) (steps 41 to 49). Note that these 5V26 and 27 are two-position valves. This is to reduce costs by performing four-stage control instead of using a duty control solenoid valve.

Further, in actual control, a value AA/N (corresponding to the part AAXTref in step 52) obtained by dividing the total flow path area AA by the rotation speed N is adopted.

If AA is used as it is, it will have a region that changes suddenly as N changes, so if this is used as a parameter,
Accuracy decreases in this sudden change region. however,
In order to improve accuracy, for example, increasing the number of lattice fish in a map will increase the calculation time accordingly. Therefore, A
By employing A/N, this control problem has been solved.

Therefore, this AADNV(=AAXTref/VC
The linearized air 'It Q HO is determined using (Step 53). The flat air-fuel ratio coefficient KFLAT is obtained from a map using Qno+N as a parameter, and the crest valve flow passage area ATVO is obtained from a table using TVO as a parameter (steps 54, 42).

In addition, the basic intake temperature correction coefficient KTAO and the intake temperature load correction coefficient KTAQCYL are also calculated using the intake temperature TAIQ.
A search is made using CYL as a parameter, and the intake temperature correction coefficient KTA is determined by the product of these (steps 81 to 83 in FIG. 7).

Since the air amount QAINJ of the injection valve section has been determined by the above calculation, the next step is to determine the correction amount regarding the fuel delay. This correction amount is used in step 66.
THO3 is specifically calculated by the routine shown in FIG. This corresponds to the function of means 9 in FIG.

In this example, the amount of adhesion of intake system fuel under steady operating conditions (
Equilibrium adhesion amount) MFI (and adhesion 'ff (MP) that changes with a first-order lag with respect to this equilibrium adhesion amount. Calculated based on the deviation from the calculated value of P. Expressing this in mathematical formulas, the following equation group (7^) ~ (7
E). Note that the subscript "-1" attached to the symbols shown below means the previous calculated value.

KATHO8=VMFXGHF...(
78) VMF=(MFH-MF-l)XKMF...
(7B) MF=MF-1+VMF...(7C)KM
F=(KMFAT+KMFVMF)XKMFNXKMF
DBT...(7D) G HF ” G HF Q c Y L X G H
F F B Y A... (7E) However, K A T I-I OS: Transient correction amount (μ
S) VMF: Deposition rate (μS/injection) MFH: Equilibrium deposition amount (μS) MF: Deposition amount (μS) KMF: Volume ratio (%) KMFAT: Basic volume ratio (%) KMFVMF: Deposition speed correction factor for volume ratio (%) KMFN: Rotation correction rate of quantity ratio (%) KMFDBT:
Boost correction rate (%) of quantity ratio G HF : Correction rate (%) G HF Q c Y L : Deceleration correction rate (%) GHF
FBYA: Air-fuel ratio correction factor (%).

In other words, the adhesion speed VMF means the adhesion amount per unit period (per injection), and the deviation ( MFH-MF-+) is calculated by calculating a coefficient KMF representing the rate at which the calculated value of the adhesion amount approaches per unit cycle (step 103).

Here, the equilibrium adhesion amount M F H is the basic operating variable QA
In addition to +++J+H, it also depends on the cooling water temperature TuI, so there are a total of three parameters, and since there is one parameter too many, it is impossible to create a three-dimensional map as it is.

Therefore, in this example, this problem is solved by a combination of three-dimensional map search and linear approximation interpolation calculation.

In other words, different n (=4 or 5) preset within the range of temperature change that the cooling water temperature TII+ can actually take.
reference temperature Tu+o −Tuln(Tu+o > −
n pieces of adhesion amount data are prepared by actual measurement in order to give the adhesion amount MFHTlllln at the reference temperature Twn using QA I N J and N as parameters for each Twn).

Then, the reference temperature T u+k(k
is an integer from O to n), M F HTwk+1 is used for the adhesion amount M F HT4 at Turk+1, and T
The MFH is finally determined by interpolation calculation using If T wk + T wk ++ (step 101).

Note that high accuracy can be obtained with the method using maps and interpolation calculations, but if you want to increase the calculation speed even if the accuracy is moderate, there is also a method that uses two tables.
This is shown in the following equation (7F).

M F HTu+n=M F HQnXM F l(N
n - (7F) However, MF HQll: QA I
Coefficient MPHNn based on N J: Coefficient based on N, MFHQn is QA I N J MF H
N n is determined by table search using N as a parameter.

Note that when down>TWO, interpolation calculation cannot be performed when up Tw×T4n, so MFH=MFH
Two. Also, during fuel power/strike, MFH = FC
MFH (constant value)9 On the other hand, the adhesion amount MF calculated this time is the adhesion amount MF-1 calculated previously and the VM calculated this time.
This is the value obtained by adding F (step 104).

Next, the quantity ratio KMF may be a constant value, but in this example, the basic value KMFAT is determined by map search using AADNV and Tw as parameters, and then VM F , N ,
The amount of change in boost pressure is corrected based on the bypass value DBO3'F (step 1o2).

That is, there are three correction coefficients KMFVMF, KMFN, and KMFDBTr for the basic value KMFAT, and these have an air-fuel ratio of 7? / It is introduced so that it has strong characteristics. In other words, there is a slight lack of correction during slow acceleration, and when experiments are conducted, small deviations occur, such as a difference in rotational speed causing a false alarm, and the aim is to solve these problems individually.

In addition, the bypass value DBO8T is calculated using the following formula (7G) ~ (7])
The content is a value that gradually attenuates in synchronization with the Ref signal while integrating the amount of change in boost pressure.

(1) When setting (first time) DBO8T=DBO3T-.

+(BOO8T-BOO3TO) ...(7G) (2) Attenuation (DBOS T≧0) (from the second time onwards)
DBOST=DBO8T-I XTGEN...(71
1) (3) At the time of decay (DBO3T<0) (second time onwards) DBO
6T=DBO8T-l B○○ST is AADNV, the adhesion speed correction factor KMFVMF is VMF, the rotation correction factor KMFN is N, and the boost correction factor KM is
FDBT is obtained by table search using the absolute value of DBO8T as a parameter.

Next, the correction factor GHF is a value that takes into account differences in fuel properties, etc. (step 7, step 105). This is due to the fact that highly volatile fuel rapidly vaporizes and is inhaled into the engine due to the development of suction negative pressure during deceleration. becomes less.

For this reason, it is necessary to estimate the adhesion amount less when decelerating, and conversely, if a smaller value is assigned as the correction coefficient (GIIFQcYL), it will be easier (that is,
No correction is made during acceleration (when VMF is positive), but (
GHFQCYL=1-. 0), and during deceleration (when V MF is negative), a value of 1 or less is used. Note that the correction is also made according to @standard air-fuel ratio T F B Y A j, :: The deceleration correction factor G HFQCYL is Q Cy L, and the air-fuel ratio correction factor GHFFBYA is TFB.
It is determined by table search using YA as a parameter.

Using the thus obtained VMF and GHF, the transient correction amount K A T HOS is finally obtained (step 106).

Next, the feedback correction amount LAMBDA used in step 68.64 of FIG. 3(C), the target air-fuel ratio TFB
YA is also calculated in the conventional example, and its routines are shown in FIGS. 5 and 6, respectively. Therefore, these! Each corresponds to the configuration of means 6.3 in Figure @1.

That is, L A M B D A is the correction coefficient in the feed balance 1 control of the air-fuel ratio.6 Figure 5 shows the PI
This is an example of D control, and is based on the deviation ER between the actual air-fuel ratio (specifically, the oxygen sensor output Ip) and the target value of the air-fuel ratio (simply, the sensor output equivalent to the target value ff1T+p). LAMBDA is determined by the following equations (8^) to (8D) that add the obtained proportional component (P), integral component (1), and differential component (D) (steps 111 to 118).

LAMBDA=P1I+D...(8^)P
=KP-ER...(8th day) I=I-10, -ER...(8C)D=KD ・(
ER-ER-+) ... (8D) However, KP:
Proportional gain Kf: Integral gain KD: Differential gain.

Note that the deviation ER is given by the following equation (8E) (step 114).

ER=-rp T+ P-(lT+ 1)
...(8E) Here, the second term of Doujin (8E) is (n+
1) At times 11" (where n is the number of cylinders) R
ef (indicates the sensor output Ip when No. 1 is input. This takes into account that there is a time delay until the result of the air-fuel ratio set in the intake system is detected by the sensor 34 installed in the exhaust system. This is what I did.

Also, the target air-fuel ratio TFBYA is T 1lItQ CY
It is calculated using I-+N as a parameter (steps 91 to 95 in FIG. 6). Note that step 95 in the same figure is TFB
YA has an upper limit and a lower limit.
It has been given the function of

Next, the learning correction coefficients KBLRC and KBTLRC used in steps 65 and 67 of FIG. 3(C) are used in this example.
THO8) are determined separately, and learning correction is also performed separately and independently. In other words, the basic injection amount learning correction coefficient KB
For LRC, feedback correction fiLAMBDA
In the calculation routine, the fuel delay learning correction coefficient KBTL
RC is calculated in the transient correction amount KAT HOS calculation routine (steps 119 and 12 in Fig. 5).
0, steps 107-110 in FIG. Note that these correspond to the functions of means 8.7 in FIG.

Learning correction basically adds a control amount based on the deviation from the target value in advance to prevent deviation from occurring during the next calculation.
Then, KBTLRC is calculated based on this LAMBDA and the deviation B between the actual air-fuel ratio AFBYA and the target air-fuel ratio TFBYA (steps 119, 120, 107 to 11).
0).

In addition, by comparing the adhesion speed VMF with the reference value L1, it is determined that the steady state (VMF<L+) is different from the transient state (■MF≧L).
, +), and learning is performed only in steady state for KBL RC and only in transient state for KBTLRC (step 119,
07).

Next, to explain the operation of this embodiment, errors in injection pulse width control include those associated with air amount metering and those associated with fuel delay. Especially in SPI, high accuracy cannot be achieved unless both of these are taken into consideration. can't get it. In this example, the two air volumes of the injector part can be determined with high accuracy, so
The air-fuel ratio is accurately determined assuming there is no fuel delay. Since a deviation actually occurs due to fuel delay, this deviation is corrected by fuel delay correction. In other words, in this embodiment, in order to determine whether the error is due to air measurement or fuel delay, these are treated separately. Since it is only necessary to consider each item individually, matching becomes easy and each item can be matched with high accuracy. As a result, it is possible to maintain high air-fuel ratio control as a whole, regardless of the position of the injector or the operating conditions.

For example, the air-fuel ratio is uniquely determined by the amount of fuel from the injection valve and the amount of air passing through the area where the injection valve is located, so in order to accurately set the air-fuel ratio, first, the air that passes through the area where the injection valve is It is necessary that ff1QA+N, +.

If you try to realize this with a device that does not have an air flow sensor, it will be difficult to directly measure QA I N J, but if you know the air ff1QcyL flowing into the shilling, you can easily is required.

QA I N J =Q CY L +c-dQ CY
L/dt...(3) Therefore, the next method is to obtain QCYL, which is based on the average IfJ2'ARQ H and the previous QCYL (QCYL-
1) and by introducing 2 into the load coefficient.

QCYL: Ql +
All you have to do is decide.

Basically, in this way, using QCYL, QAI
Since NJ can be determined with high precision, the accuracy of measuring the amount of air is significantly improved compared to the case where the amount equivalent to QCY+- or QCYL is used.

For example, if QCYL is adopted in 5PIVcrli, there is no limit to the number of man-hours required as dQcyL/dt varies, making matching difficult.

In addition, in the SPI device, there is a problem of fuel delay during transient periods, which is caused by the intake system fuel adhesion fi (MFH, MFH,
) by directly specifying the transient fuel lag correction i
Since K A T HOS is calculated, matching is easier than when correction is performed indirectly based on factors that affect the adhesion amount rather than the adhesion amount itself, and it is easier to perform the control associated with various tq constants, etc. It is possible to prevent complications from occurring. Control speed is also an issue, especially during transient times, and the more complex the control, the lower the speed, so this has important implications as it relates to cost.

In this way, this embodiment simplifies the concept by separating the air amount measurement and the fuel delay, which makes matching easier and allows for accurate determination of each, resulting in This makes it possible to improve the accuracy of air-fuel ratio control. Here, improving accuracy will lead to improved drivability.
There is no need to provide a large allowance like in the conventional case, so
Improvements in fuel efficiency and exhaust emissions can also be achieved.

However, no matter how highly accurate the setting is, it is impossible to deal with error factors (changes over time, differences in fuel properties, etc.) after the setting.

Therefore, a learning function is activated for error factors after the initial setting. That is, since the error from the target air-fuel ratio TFBYA is eliminated each time learning is performed, control to the target air-fuel ratio can always be performed in the same state as initially set, and control accuracy after the initial setting can be maintained at a high level. Can be done.

Note that FIG. 8 is a block diagram of this embodiment.

In addition, in this example, the air amount sensor was abolished to reduce costs, so QAINJ is determined from QCYL, but the basic idea of separating air amount measurement and fuel delay can be applied to other devices. It can also be applied to

For example, in a conventional device (L-noetronic system) in which an air amount sensor is provided and an injection valve is attached to the intake boat, the amount of air passing through the injection valve is QCYL.
In this case, Q c y +- is calculated from the 2 air volume obtained by the air volume sensor (approximately equal to the air volume at the throttle valve part) using the following equation (9). be.

Q c Y L = (QA X K/N) Xa+Q
CYL-IX(1-a) = (9) where a is a loading coefficient.

Therefore, according to the present invention, the SPI device and the EGI device can be used in common, and the number of development steps can be reduced.

Note that in the α-N method, QCy+- is determined from the valve opening degree α and the engine speed N, but in the D-Noetronino method, it may be determined from the intake pipe pressure P[+ instead of α.

(Effects of the Invention) As explained above, in this invention, the air-fuel ratio is set accurately assuming there is no fuel delay by determining the air amount in the injection valve section, and then the air-fuel ratio is set accurately by determining the air amount in the injection valve section. Since the deviation in the air-fuel ratio is corrected by correction based on the fuel delay, it is possible to clearly distinguish whether the error in control accuracy is due to the air amount measurement error or due to the fuel delay. This makes matching easier,
It is possible to improve drivability by improving accuracy.

Furthermore, by providing a learning function, it is possible to deal with error factors after setting.

[Brief explanation of the drawing]

Fig. 1 is a conceptual diagram of this invention, Fig. rjS2 is a schematic diagram showing the mechanical groove formation of an embodiment of this invention used in SP device n, and Figs. 3 to 7 are f52 drawings. FIG. 8 is a block diagram of this embodiment. FIG. 9 is a characteristic diagram showing the change in air amount during a transient period. DESCRIPTION OF SYMBOLS 1... Operating state detection means, 2... Injection valve air amount calculation means, 3... Target air-fuel ratio pre-preparation means, 4... Basic injection amount calculation means, 5... Actual air-fuel ratio detection means , 6... Feedback correction amount calculation means, 7... Fuel delay learning correction coefficient calculation circuit, 8... Basic injection amount learning correction coefficient calculation means, 9... Basic fuel delay correction amount calculation means, 10.・
・Injection j′ [¥ calculation means, 21... intake throttle well, 22
...Intake passage, 23...Bypass passage, 24...
Fuel injection valve, 25... Throttle valve opening sensor, 26.27
...Solenoid valve for idle up, 34...Oxygen sensor (air fuel ratio sensor?), 35...Control unit)
, 130... Injection valve air amount calculation circuit, 131...
Target air-fuel ratio calculation circuit, 132... Basic pulse width calculation circuit, 133... Quantity ratio calculation circuit, 134... Equilibrium adhesion amount calculation circuit, 135... Adhesion amount calculation circuit, 136
... Adhesion speed calculation circuit, 137 ... Correction factor calculation circuit, 138 ... Transient correction amount calculation circuit, 139 ... Target sensor output calculation circuit, 140- (n+
1) Circuit for calculating the previous Ip, 141... Deviation calculation circuit, 142... Proportional and public calculation circuit, 143... Integral calculation circuit, 1.t4... Differential precalculation circuit, 145
...Feed/<・7 correction coefficient calculation circuit, 146,
147. ... learning correction coefficient calculation circuit, 148 ... invalid pulse width calculation circuit, 149 ... injection pulse width calculation circuit, 150 ... drive circuit. (1 other person)

Claims (1)

[Claims] means for calculating the amount of air passing through the fuel injection valve section according to the operating state; means for calculating the target fuel-air-fuel ratio; and means for calculating the basic fuel injection amount from these air amounts and the target fuel-air-fuel ratio. , means for calculating a feedback correction amount based on the deviation between the target air-fuel ratio and the actual air-fuel ratio obtained by the actual air-fuel ratio detection means, and learning of fuel delay based on the deviation or the deviation and the feedback correction amount. means for calculating a correction coefficient; means for calculating a learning correction coefficient for the basic injection amount based on the feedback correction amount or the feedback correction amount and the deviation; and a basic fuel delay correction amount depending on the operating state. An air-fuel ratio control device for an internal combustion engine, comprising means for calculating the basic injection amount, and means for calculating the fuel injection amount by correcting the basic injection amount based on a fuel delay correction amount and each learning correction coefficient.