JPH09291846A - Intake air amount detector and fuel injection controller for engine - Google Patents

Abstract

PROBLEM TO BE SOLVED: To accurately trace a wave form of filling efficiency in the transient time in any cylinder regardless of the timing in which a throttle valve is suddenly changed in a multiple-cylinder engine. SOLUTION: Corrector internal pressure in this time is calculated by a calculating means 44 by using corrector internal, pressure before the very small time Δt, the throttle part passing mass flow rate, atmospheric pressure, atmospheric density, and the branch pipe inside state amount before the Δt, and corrector internal density in this time is calculated by a calculating means 46 by using corrector internal density before the Δt, the throttle part passing mass flow rate, the branch pipe inside state amount before the Δt. The branch pipe inside state amount in this time is calculated by a calculating means 48 by using the density in the corrector in this time, pressure, the branch pipe inside state amount, and a basic equation of hydrodynamics, the intake valve passing mass flow rate is calculated from the branch pipe internal state amount in this time and the intake valve opening by a calculating means 49, and a value obtained by integrating this intake valve passing mass flow rate for a period of the intake stroke is calculated as the mass air amount by a calculating means 50 for each cylinder.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an engine intake air amount detection device and a fuel injection control device, and more particularly to a device for determining a mass air amount flowing into each cylinder when the engine is accelerated or decelerated.

[0002]

2. Description of the Related Art In gasoline injection, there is a mass flow system for directly measuring the intake air flow rate (intake air amount per unit time) by an air flow meter as an air amount detection system. In this mass flow method, the intake air amount per combustion cycle is obtained by dividing the measured intake air flow rate by the number of combustion cycles, and the fuel injection amount is calculated based on this value (Sankaido, Shin Electronically controlled gasoline injection ", page 9, JP-A-62-101855).

[0003]

By the way, FIG. 8 shows a general correlation between the throttle opening TVO and the airflow meter output during acceleration in the mass flow system. The overshoot of the airflow meter output includes the collector volume. The larger the intake manifold volume, the larger the amplitude of the overshoot. Therefore, the waveform of the charging efficiency ηc at the time of acceleration obtained by the general mass flow method is close to the output of the air flow meter as shown by the broken line in FIG. 9, and does not trace the accurate charging efficiency at the time of acceleration. . In addition, various correction methods have been tried to obtain the accurate filling efficiency during transient from the output of the air flow meter, but all methods are linear approximation with a simple delay system such as a first-order delay system, so that transient It is difficult to trace the exact filling efficiency.

On the other hand, a speed density method that does not use an air flow meter (a map (created in a ROM) in which the amount of air taken into the engine per one cycle is parameterized by the intake pipe pressure and the engine speed is retrieved. The amount of air injected into the engine per cycle is calculated from the throttle opening and the engine speed, and the fuel injection amount is calculated based on the amount of air. There is no problem with the mass flow method in calculating the fuel injection amount), but on the other hand, matching is required to create ROM data for obtaining the air amount, and the ROM data is determined by experiment. The calculation accuracy of the air-fuel ratio depends on the matching.

As disclosed in Japanese Patent Laid-Open No. 59-145357, a type using a reference model (dynamic characteristics of an air system which are measured and stored in advance and an air flow meter output Aa) is used.
From (n) (where n is the current value), the cylinder intake air flow rate Ac (n) is Ac (n) = d1 · Aa (n−1) −e1 · Aa (n−2) −b1 · Ac (n− 1) −c1 · Ac (n−2) (41) However, Aa (n−1): Airflow meter output before one time Aa (n−2): Airflow meter output before two times Ac (n−1) ): 1st before cylinder intake air flow rate Ac (n-2): 2nd before air flow meter output b1, c1, d1, e1: Calculated by the formula of the coefficient, and the required fuel amount is calculated from the value. The fuel injection amount is calculated from the required fuel amount and the dynamic characteristics of the fuel system.) However, since the model is not a mathematical formula that has a physical meaning, the calculation of the transient air amount is practically impossible. Have difficulty. The above formula (41) is only a linear approximation formula (empirical formula) for a certain acceleration condition and has no hydrodynamic meaning in the formula itself. Therefore, for example, the same formula is used for both rapid acceleration and gentle acceleration. It is difficult to calculate the cylinder intake air amount by using the above-mentioned coefficient b1,
Even if c1, d1, and e1 can be adapted to the one acceleration condition, an error occurs in the calculation of the cylinder intake air amount under the acceleration conditions other than the acceleration condition.

Therefore, according to the present invention, the state quantity in the collector (pressure and density) and the state quantity in the branch pipe (density, flow velocity, pressure) obtained before the minute time and the basic equation (Euler's equation) of fluid dynamics are obtained. The branch pipe state quantity of this time is obtained for each cylinder by using the boundary conditions, and the intake valve passing mass flow rate is obtained for each cylinder from the branch pipe state quantity calculated this time and the intake valve opening area.
By calculating the mass air amount flowing into the cylinder for each cylinder by integrating the intake valve passing mass flow rate over the corresponding intake valve opening period of the cylinder, the charging efficiency can be increased or decreased depending on the time when the throttle valve suddenly changes. In a multi-cylinder engine having different cylinders, the first purpose is to accurately trace the waveform of the charging efficiency during transition regardless of the time when the throttle valve suddenly changes. A second object is to maintain the air-fuel ratio flat for all cylinders before and after the transition by performing fuel injection based on the amount.

[0007] In addition, the one similar to the present invention is disclosed in
No. 74548, there is a difference in the air flow rate to each cylinder due to the intake manifold length, bending, intake inertia, etc. Therefore, an air-fuel ratio sensor is provided in the exhaust pipe for each cylinder. The intake air amount is determined for each cylinder from the detected air-fuel ratio and the fuel injection pulse signal.

However, the biggest problem at the time of acceleration is that the rise of the filling efficiency (the fall of the filling efficiency at the time of deceleration) changes depending on which cylinder is in the intake stroke when the throttle valve is rapidly opened. If the valve opening / closing timing and the intake stroke of each cylinder are not taken into consideration at the same time, the charging efficiency during a transition cannot be accurately traced for each cylinder.
Further, in Japanese Patent Laid-Open No. 3-74548, not only cannot a phenomenon (40 Hz or more) at the time of transient be detected due to the low responsiveness of the air-fuel ratio sensor, but it can be detected after the engine has been operated for a long time due to deterioration of the air-fuel ratio sensor. The error increases.

On the other hand, the present invention does not solve the fact that the charging efficiency (air-fuel ratio) varies from cylinder to cylinder even in a steady state due to the length and shape of the intake pipe in a multi-cylinder engine. Even if the shape and length of the intake pipe are exactly the same, in a multi-cylinder engine where the rise and fall of the charging efficiency differ depending on the cylinder when the throttle valve suddenly changes, regardless of when the throttle valve suddenly changes It is designed to accurately trace a transient waveform, and the problem to be solved is different from that of JP-A-3-74548.

[0010]

A first invention, as shown in FIG. 13, is a means 41 for detecting a throttle opening θt of an engine, and a means 42 for calculating a throttle opening area St from the throttle opening θt. A means 43 for calculating the mass flow rate Mt of passing through the throttle portion based on the throttle opening area St; and a collector internal pressure P before a minute time Δt.
₀ ^t , the mass flow rate Mt passing through the throttle portion, the atmospheric pressure Pa,
Means 44 for calculating air density .rho.a, branch pipe state amount before the minute time ^{_{Δt (ρ 2, i t,}} u 2, i t, P 2, i t) the current collector in the pressure P ₀ ^{t + 1} using Means 45 for storing the collector internal pressure P ₀ ^{t + 1} calculated this time for the next calculation of the collector internal pressure, the collector internal density ρ ₀ ^t before the minute time Δt, and the mass flow rate Mt passing through the throttle portion, branch pipe state amount before the minute time ^{_{Δt (ρ 2, i t,}} u 2, i t) means for calculating a collector within the density ρ ₀ ^{t + 1} the current using a 46
And a means 47 for storing the currently calculated collector internal density ρ ₀ ^{t + 1} for the next collector internal density calculation, the current collector internal density ρ ₀ ^{t + 1} and the current collector internal pressure P _0. ^{t + 1} and state quantity in branch pipe (ρ ₁ , ₁ ^t ~ before micro time Δt ~
ρ _m , _n ^t , u ₁ , ₁ ^{t to} u _m , _n ^t , P ₁ , ₁ ^{t to} P _m , _n ^t ) and basic equations of fluid dynamics (for example, first-order upwind difference method) and predetermined boundaries The condition quantity in the branch pipe (ρ ₁ , _{1
^{t + 1 ~ρ m, n t}} + 1, u 1, 1 t + 1 ~u m, n t + 1, P 1, 1 t + 1 ~
P _m , _n ^{t + 1} ) for each cylinder, a branch pipe state quantity calculated this time, and the intake valve opening area F _{n of} each cylinder
Intake valve passing mass flow M _n ^{t + 1} and means 49 for calculating the cylinder and the intake valve passing mass flow M _n ^{t + 1} mass values obtained by integrating over a period of the intake stroke of the corresponding cylinder to the air from the A means 50 for calculating the amount M _n for each cylinder and a means 51 for storing the current branch pipe state quantity for the next calculation of the collector inner state quantity and the next branch pipe inner state quantity are provided.

A second aspect of the present invention, as shown in FIG. 14, a means 61 for detecting the collector internal pressure P _0, means for calculating a collector within the density [rho ₀ with atmospheric pressure Pa and the collector internal pressure P ₀ 62, the collector internal density ρ ₀ , the collector internal pressure P _0, and the branch pipe state quantity (ρ ₁ ,
^{_{1 t ~ρ m, n t,}} u 1, 1 t ~u m, n t, P 1, 1 t ~P m, n t)
And basic equations of fluid dynamics (eg, first-order upwind difference method)
And a predetermined boundary condition, the state quantity in the branch pipe (ρ
^{_{1, 1 t + 1 ~ρ m}} , n t + 1, u 1, 1 t + 1 ~u m, n t + 1, P 1, 1
^{t + 1 to} P _m , _n ^{t + 1} ) for each cylinder, and the intake valve passage mass flow rate M _n ^{t + 1} from the cylinder 63 and the intake valve opening area F _n calculated this time. Means 4 to calculate separately
9, a means 50 for calculating a value obtained by integrating the intake valve passing mass flow rate M _n ^{t + 1} over a period of the intake stroke of the corresponding cylinder as a mass air amount M _n for each cylinder, and the present branch pipe internal state A means 64 for storing the quantity for the next calculation of the branch pipe state quantity is provided.

A third aspect of the present invention is the intake valve portion according to the first or second aspect, wherein the intake valve opening area F _n is a weight value corresponding to the intake valve opening area and the branch pipe internal state quantity calculated this time. It is the product of the passing flow velocity (u _m-1 , _n ^{t + 1} ).

As shown in FIG. 15, a fourth aspect of the present invention is a means 41 for detecting the throttle opening θt of the engine, a means 42 for calculating the throttle opening area St from the throttle opening θt, and this throttle opening area. Means 43 for calculating the mass flow rate Mt passing through the throttle portion based on St
And the collector internal pressure P ₀ ^t before the minute time Δt, the mass flow rate Mt passing through the throttle portion, the atmospheric pressure Pa, the atmospheric density ρa,
Branch pipe state amount before the minute time ^{_{Δt (ρ 2, i t,}} u 2, i t,
And P _{2, i} ^t) means 44 for calculating a collector internal pressure P _{0 t} ^{+ 1} of the current with a means 45 for storing for the collector internal pressure P _{0 t} ^{+ 1} calculated this time calculation of the next collector internal pressure, wherein a minute time Δt before the collector within the density [rho ₀ ^t throttle portion passing mass flow Mt, branch pipe state amount before the minute time Δt current collector in density with ^{_{(ρ 2, i t, u}} 2, i t) [rho ₀ and ^{t + 1} means 46 for calculating a, a means 47 for storing for a collector in density [rho ₀ ^{t + 1} calculated this time calculation of the next collector in density, the current collector within the density [rho ₀ ^{t + 1} and the current collector internal pressure P ₀ ^{t + 1} and the branch pipe state quantity before a minute time Δt (ρ ₁ , ₁ ^{t to} ρ _m , _n ^t , u ₁ , ₁
^t ~ u _m , _n ^t , P ₁ , ₁ ^t ~ P _m , _n ^t ), the basic equation of fluid dynamics (for example, first-order upwind difference method), and a predetermined boundary condition, and the current branch pipe state The quantity (ρ ₁ , ₁ ^{t + 1 to} ρ _m , _n ^{t + 1} ,
^{_{u 1, 1 t + 1 ~u}} m, n t + 1, P 1, 1 t + 1 ~P m, a means 48 for calculating a _n ^{t + 1)} in the cylinder, and the branch pipe state amount calculated this time a means 49 for calculating the intake valve passing mass flow M _n ^{t + 1} to cylinder from the intake valve opening area F _n of the cylinders, the intake stroke of the cylinder to the corresponding intake valve passing mass flow M _n ^{t + 1} to Means for calculating the value accumulated for each cylinder as the mass air amount M _n for each cylinder, and means 51 for storing the current branch pipe state quantity for the calculation of the next collector pipe state quantity and the next branch pipe state quantity. And the mass air amount M _n
The a means 71 for calculating a fuel injection amount Tp _n required for one combustion cycle by dividing the engine speed N in the cylinder at a predetermined timing, the supply of fuel for this injection amount Tp _n each cylinder at a predetermined timing And means 72 for doing so.

[0014] The fifth invention, as shown in FIG. 16, a means 61 for detecting the collector internal pressure P _0, means for calculating a collector within the density [rho ₀ with atmospheric pressure Pa and the collector internal pressure P ₀ 62, the collector internal density ρ ₀ , the collector internal pressure P _0, and the branch pipe state quantity (ρ ₁ ,
^{_{1 t ~ρ m, n t,}} u 1, 1 t ~u m, n t, P 1, 1 t ~P m, n t)
And basic equations of fluid dynamics (eg, first-order upwind difference method)
And a predetermined boundary condition, the state quantity in the branch pipe (ρ
^{_{1, 1 t + 1 ~ρ m}} , n t + 1, u 1, 1 t + 1 ~u m, n t + 1, P 1, 1
^{t + 1 to} P _m , _n ^{t + 1} ) for each cylinder, and the intake valve passage mass flow rate M _n ^{t + 1} from the cylinder 63 and the intake valve opening area F _n calculated this time. Means 4 to calculate separately
9, a means 50 for calculating a value obtained by integrating the intake valve passing mass flow rate M _n ^{t + 1} over a period of the intake stroke of the corresponding cylinder as a mass air amount M _n for each cylinder, and the present branch pipe internal state a means 64 for storing for calculation of the next branch pipe state quantity the amount, the weight given timing fuel injection amount Tp _n required for one combustion cycle by the air weight M _n is divided by engine speed n in a means 71 for calculating the cylinder, the fuel of the injection amount Tp _n is provided and means 72 for supplying each cylinder at a predetermined timing.

In a sixth aspect of the invention, in the fourth or fifth aspect of the invention, the predetermined timing for calculating the fuel injection amount required for the one combustion cycle is the timing of ending the intake stroke of each cylinder.

A seventh aspect of the present invention is the transient correction amount Ka for the fuel wall flow according to any one of the fourth to sixth aspects.
and a means for correcting the fuel injection amount Tp _n required for the combustion cycle by thos.

[0017]

[Function] Since the output of the air flow meter overshoots due to the volume in the intake manifold including the collector volume, the waveform of the charging efficiency obtained by the general mass flow method is similar to the output of the air flow meter. , It does not trace the accurate filling efficiency during acceleration. On the other hand, in the first invention, the throttle opening θt
The mass flow rate Mt passing through the throttle portion is obtained from the mass flow rate Mt passing through the throttle portion, the atmospheric state quantity (Pa, ρa),
Collector in a state of ^{_{pre-Δt (ρ 0 t, P 0}} t), the branch pipe state amount before ^{_{Δt (ρ 2, i t,}} u 2, i t, P 2, i t) from the current collector in the quantity of state (Ρ ₀ ^{t + 1} , P ₀ ^{t + 1} ) is calculated, and the current state quantity in the collector (ρ ₀ ^{t + 1} , P ₀ ^{t + 1} ) and the state quantity in the branch pipe before Δt (ρ ₁ , ₁ ^{t) _{~ρ m, n t, u 1}} , 1 t ~u m, n t,
P ₁ , ₁ ^{t to} P _m , _n ^t ), the basic equation of fluid dynamics (Euler's equation), and the boundary condition are used to determine the state quantity in the branch pipe (ρ
^{_{1, 1 t + 1 ~ρ m}} , n t + 1, u 1, 1 t + 1 ~u m, n t + 1, P 1, 1
^{t + 1 to} P _m , _n ^{t + 1} ) is calculated for each cylinder, and the branch pipe state quantity (ρ _m-1 , ₂ ^{t + 1} · u _m-1 , ₂ ^{t + 1} ) and intake valve opening calculated this time are calculated. calculated from the area F _n the intake valve passing electrolyte flow M _n ^{t + 1} to the cylinder, it flows into the cylinder by integrating over the intake valve opening period of the cylinder corresponding to the intake valve passing mass flow M _n ^{t + 1} to Since the mass air amount M _n is obtained for each cylinder, overshoot does not occur in the change of the charging efficiency at the time of transition in any cylinder, and the actual charging efficiency for each cylinder is well traced.

Moreover, in calculating the mass air amount M _n for each cylinder, only the memory for storing the state quantity in the collector before Δt and the state quantity in the branch pipe and the basic equation of the fluid dynamics are required. No matching work is required for the preparation, and the ROM can be formed only by inputting the engine intake pipe dimensions and engine specifications. The accuracy of calculating the air amount during the transition and the accuracy of controlling the air-fuel ratio during the transition can be secured without using matching.

Further, in the mass flow system, lean misfire may occur due to overshoot in the air flow meter at the time of acceleration. In the first invention, however, the lean misfire at acceleration (or rich misfire at deceleration) is eliminated. Therefore, the drivability and stability of the engine are significantly improved. It is also inexpensive because it does not require an air flow meter.

In this way, the mass air amount M _n flowing into the cylinder during the transition can be accurately obtained for each cylinder. Therefore, according to the fourth aspect of the invention, fuel injection can be performed based on this mass air amount M _n. , The supply air-fuel ratio becomes constant before and after the transition in any cylinder, which improves exhaust emission and fuel efficiency.

When performing EGR (exhaust gas recirculation), the collector internal pressure (also regarding the collector internal density) cannot be determined only by the throttle opening, so the mass air amount M _n for each cylinder is calculated by the EGR amount. although the error occurs, because the second invention is to detect the collector internal pressure P ₀ by the pressure detecting means, even when the collector internal pressure, the collector in the density can not be determined only by the throttle opening degree, mass air The amount M _n can be accurately determined for each cylinder.

In the fifth aspect of the invention, since fuel injection is performed using this mass air amount M _n , even in an engine in which the collector internal pressure and collector internal density cannot be determined only by the throttle opening, no matter which cylinder the supply air-fuel ratio is, It becomes constant before and after the transition.

[0023]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS In FIG. 1, reference numeral 1 denotes an engine body, and air is temporarily stored in an intake manifold collector 4 from an air cleaner 2 through a throttle portion 3 and then through each branch pipe 5 of an intake manifold to each cylinder. Flows into the cylinder 6. The throttle valve 5a is mechanically connected to an accelerator pedal (not shown), and adjusts the flow rate of air passing through the throttle portion 3 according to the accelerator opening.

The fuel injected from the fuel injection valve 7 provided facing the intake port of each cylinder mixes with the air flowing into the cylinder 6 to form an air-fuel mixture, and the gas after burning by ignition spark is exhausted. HC, CO, and NOx in the exhaust gas, which are discharged to the pipe 8 and are harmless by the three-way catalyst 9, are H ₂ O, CO ₂ ,
Converted to N ₂ .

Since the three-way catalyst 9 exhibits the highest conversion efficiency when the exhaust air-fuel ratio is near the stoichiometric air-fuel ratio, the supply air-fuel ratio (determined by the ratio of the air amount to the fuel amount) becomes almost the stoichiometric air-fuel ratio. It is necessary to determine the amount of fuel from the injection valve 7.

Here, since the fuel amount from the injection valve 7 is determined based on the air amount, if the air amount flowing into the cylinder 6 at the time of acceleration / deceleration can be accurately measured for each cylinder, the control accuracy of the transient air-fuel ratio is that much. Increase.

In this case, the throttle valve 3a for acceleration.
According to the mass flow method when the valve is rapidly opened, overshoot occurs in the air flow meter output due to the volume in the intake manifold including the collector volume as shown in FIG. The obtained waveform of the charging efficiency ηc at the time of acceleration is similar to the output of the air flow meter (see the broken line in FIG. 9), and the accurate charging efficiency at the time of acceleration is not traced. In addition, various correction methods have been tried to obtain the accurate filling efficiency during transient from the output of the air flow meter, but all methods are linear approximation with a simple delay system such as a first-order delay system, so that transient It is difficult to trace the exact filling efficiency.

On the other hand, the speed density method and the α-N method, which do not use an air flow meter, do not have the problems of the mass flow method, but on the other hand, the matching is required to create ROM data for obtaining the air amount. Therefore, the ROM data must be determined experimentally, and the air-fuel ratio calculation accuracy depends on matching.

To deal with this, in the present invention, the amount of mass air flowing into the cylinder at the time of transition is obtained for each cylinder in the following procedure, and fuel injection is performed based on the obtained amount of mass air.
That is, <a> The mass flow rate Mt passing through the throttle portion is calculated from the throttle opening θt, and <b> The mass flow rate Mt passing through the throttle portion, the atmospheric state quantity (atmospheric pressure Pa and atmospheric density ρa), and within the collector before a minute time Δt. From the four values of the state quantity (collector pressure P ₀ ^t and collector density ρ ₀ ^t ) and the branch pipe state quantity before the minute time Δt, the current collector state quantity (collector pressure P ₀ ^{t + 1} and collector inside pressure) Density ρ ₀ ^{t + 1} ) is calculated, and <c> State quantity in collector this time, State quantity in branch pipe before a short time Δt (density ρ ₁ , ₁ ^t ~
ρ _m , _n ^t , flow velocity u ₁ , ₁ ^{t to} u _m , _n ^t , pressure P ₁ , ₁ ^{t to} P _m , _n
^t ), basic equations of fluid dynamics, and predetermined boundary conditions are used to calculate the state quantity in the branch pipe (density ρ ₁ , ₁ ^{t + 1 to} ρ _m , _n ^{t + 1} , flow velocity u ₁ , ₁ ^{t). +1 to} u _m ,
_n ^{t + 1} , pressures P ₁ , ₁ ^{t + 1 to} P _m , _n ^{t + 1} ) are calculated for each cylinder, and <d> this branch pipe state quantity and intake valve opening area F _n
And the intake valve passing mass flow rate M _n ^{t + 1} (where n: cylinder number) is calculated for each cylinder, and <e> this mass flow rate M _n ^{t + 1} is integrated over the period of the intake stroke of the corresponding cylinder. The value is calculated for each cylinder with the mass air amount M _n .

[0030] <f> basic injection pulse width Tp _n the cylinder mass air quantity M _n from a value obtained by dividing the engine speed N
The calculated cylinder at a predetermined timing (e.g., timing of completion of the intake stroke of each cylinder), <g> fuel injection pulse width Ti _n giving the injection valve by performing various corrections on the basic injection pulse width Tp _n the calculated for each cylinder at the same timing as the calculation timing of Tp _n.

[0031] <h> for the collector in the pressure P ₀ ^{t + 1} calculated this time after calculating the Ti _n of the calculation of the next collector in the pressure, the collector in the density ρ ₀ ^{t + 1} calculated this time next collector In order to calculate the internal density, the state quantities (ρ ₁ , ₁ ^{t + 1} ~ ρ _m , _n ^{t + 1} , u ₁ , ₁ ^{t + 1} ~
u _m , _n ^{t + 1} , P ₁ , ₁ ^{t + 1 to} P _m , _n ^{t + 1} ) are stored for the next calculation of the collector state quantity and the branch pipe state quantity, respectively.

For this reason, the Ref signal from the crank angle sensor 12 (1 for four cylinders is sent to the control unit 11).
(Every 80 °, every 6 ° 120 ° in 6 cylinders), 1 ° signal, and throttle opening signal from the throttle sensor 13 are the air-fuel ratio (oxygen concentration) from the O ₂ sensor 14 provided upstream of the three-way catalyst 9. The signal and the cooling water temperature signal from the water temperature sensor 15 are input together.

The contents of this control executed by the control unit 11 will be described in detail below using the intake pipe model shown in FIG. The description of the state quantity of each part of the intake pipe model is as shown in the lower part of FIG. In FIG. 2, ρ ₀ = ρ ₁ , _n P ₀ = P ₁ , _n u ₀ = u ₁ , _n (= 0) are the boundary conditions at the joint between the collector 4 and the branch pipe 5, ρ _{m , n = ρc n P m,} n = Pc n u m, n = uc n (= 0) shall be deemed to be replaced with a boundary condition of the coupling portion of the branch pipe 5 and the cylinder 6.

The number of meshes is a number assigned to a position where each branch pipe is divided into m pieces at equal intervals, and in particular, each position of the mesh numbers 1, 2, and 3 is enlarged and shown in FIG. As shown in the figure, the position of the mesh number 1 is in the collector 4, the position of the mesh number 2 is in the connecting portion of the collector 4 and the branch pipe 5, and the position of the mesh number 3 is in the branch pipe 5.

[0035] <a-1> St = ( πD 2/4) from the throttle opening θt detected by calculating a throttle sensor 13 of the throttle opening area St · (1-cosθt / cosθmin ) ... (1) However, D : Throttle diameter θmin: Throttle opening area St is calculated by the formula of throttle opening when the throttle valve is fully closed.

<A-2> Mass flow rate Mt passing through the throttle portion
When the distance between the throttle valve 3a and the collector 4 is large, it is necessary to calculate the delay time by using a difference formula. Normally, however, since the distance between the two is short, it is not necessary to consider the delay time. _{When 0} ^t ≦ (2 / (κ + 1)) ^{κ / (κ −1)} Mt = St · (κ · ρa · Pa) ^1/2 (2) ii) P ₀ ^t > (2 / (κ + 1)) ^{When κ / (κ −1)} Mt = St · (2κ / (κ−1) · ρa · Pa {(P ₀ ^t / Pa) ^{2 / κ} − (P ₀ ^t / Pa) ^{(κ +1) / κ}) 1/2 ... (} 3) However, P ₀ ^t: P ₀ κ at time t: calculating the throttle section passing mass flow Mt by formula ratio of specific heats.

<B> Calculation of state quantity (pressure, density) in the collector The density in the collector and the pressure in the collector are cyclically obtained at every minute time Δt using the energy conservation formula and the mass conservation formula. Unless we consider the heat transfer from the collector 4,

[Equation 1] The formula is established. Based on the equations (4) and (5), the collector state quantities ρ ₀ ^{t + 1} and P ₀ ^{t + 1} after Δt time from the time t are the state quantity of the pipe flowing into the collector 4, the density ρ _in , and the pressure. P _in ,
The flow velocity u _in , the state quantity of the pipe flowing out from the collector 4 are defined as the density ρ
_out, the pressure P _out, a flow rate ^{_{u out, P 0 t + 1}} = P 0 t + Δt · [ρ in · u in · f in · {P in / ρ in + ((κ-1) / 2κ) · u _in ² } −ρ _out · u _out · f _out · {P _out / ρ _out + ((κ−1) / 2κ) · u _out ² }] / V (6) ρ ₀ ^{t + 1} = ρ ₀ ^t _{+ Δt · {ρ in · u} in · f in -ρ out · u out · f out} / V ... (7) However, P ₀ ^t: P at time ^{_{t 0 P 0 t + 1:}} P at time t + Δt ₀ ρ ₀ ^{t: ρ} at time ^t ₀ ρ ₀ t ^{+ 1:} time t + ρ in Δt ₀ f _in: coefficient f _out: factor V: can be calculated by the equation of the collector volume.

Here, when the distance between the throttle valve 3a and the collector 4 is short, ρ ₀ ^{t + 1} and P ₀ ^{t + 1} are set to

[Equation 2] Can be approximated by

It should be noted, (8), (9) branching pipe state quantities of Formula (Density [rho _{2, i} ^t, velocity u _{2, i} ^t, the pressure P _{2, i} ^t) is the number of mesh shown in FIG. 3 2 Is the state quantity at the position (described later).

<C> Calculation of state quantity in branch pipe As shown in FIG. 2, each state quantity in branch pipe is divided into 1 to m meshes according to the length of each branch pipe 5 (Δx: distance between meshes). ), The state quantity at each mesh number is calculated every Δt time (calculation cycle).

Generally, if a higher-order difference formula is used, the accuracy will be higher. However, in order to perform the calculation in the control unit 11, a logic with a short calculation time is required even if the accuracy is slightly low. First-order upwind difference method (Lelev
The calculation was performed using the method of Ier).

[Mass Conservation Formula] i) When u _i , _n ^t > 0: (ρ _i , _n ^{t + 1} −ρ _i , _n ^t ) / Δt = − {(ρu) _i , _n ^t − (ρu ) _I-1 , _n ^t } / Δx (10) ii) When u _i , _n ^t <0: (ρ _i , _n ^{t + 1} −ρ _i , _n ^t ) / Δt = − {(ρu) _{i +1} , _n ^t − (ρu) _i , _n ^t } / Δx (11) [Momentum conservation equation] i) When u _i , _n ^t > 0: {(ρ · u) _i , _n ^{t + 1} − (Ρ · u) _i , _n ^t )} / Δt = − (P _i , _n ^t −P _i−1 , _n ^t ) / Δx − {(ρ · u ² ) _i , _n ^t − (ρ · u ² ) _I-1 , _n ^t } / Δx (12) ii) When u _i , _n ^t <0: {(ρ · u) _i , _n ^{t + 1} − (ρ · u) _i , _n ^t )} / Δt = − (P _{i + 1} , _n ^t −P _i , _n ^t ) / Δx − {(ρ · u ² ) _{i + 1} , _n ^t − (ρ · u ² ) _i , _n ^t } / Δx ... (13) [Energy conservation equation] i) When u _i , _n ^t > 0: (E _i , _N ^{t + 1} −E _i , _n ^t ) / Δt = −u _i , _n ^t (E _i , _n ^t + P _i , _n ^t ) / Δx −u _i−1 , _n ^t (E _i−1 , _n ^t + P _i-1 , _n ^t ) / Δx (14) ii) When u _i , _n ^t <0: (E _i , _n ^{t + 1} −E _i , _n ^t ) / Δt = −u _{i + 1} , _N ^t (E _{i + 1} , _n ^t + P _{i + 1} , _n ^t ) / Δx −u ₁ , _n ^t (E _i , _n ^t + P _i , _n ^t ) / Δx (15) To do.

In equations (14) and (15), the energies E _i and _n ^t are E _i , _n ^t = (1/2) ρ _i , _n ^t (u ² ) _i , _n ^t + {1 / (kappa-1)} is that P _{i, n} ^t ... (16).

Furthermore, at the boundary between the mesh numbers 2 and 3, P ₁ , _n ^{t + 1} / {(ρ ₁ , _n ^{t + 1} ) ^κ } = P ₂ , _n ^{t + 1} / {(ρ ₂ , _N ^{t + 1} ) ^κ } (17a) {κ / (κ−1)} · P ₁ , _n ^{t + 1} / ρ ₁ , _n ^{t + 1} = {κ / (κ−1)} · P ₂ , _N ^{t + 1} / ρ ₂ , _n ^{t + 1} + (1/2) · (u ₂ , _n ^{t + 1} ) ² (18a) ρ ₂ , _n ^{t + 1} = ρ ₂ , _n ^t + ρ ₃ , _n ^t −ρ ₃ , _n ^{t + 1} − (Δt / Δx) · {(ρ · u) ₃ , _n ^{t + 1} + (ρ · u) ₃ , _n ^t − (ρ · u) ₂ , _n ^{t + 1−} (ρ · u) ₂ , _n ^t } (19a) where ρ ₁ , _n ^{t + 1} = ρ ₀ ^{t + 1} (20a) P ₁ , _n ^{t + 1} = P ₀ ^{t + 1} ((a) 21a) u ₁ , _n ^{t + 1} = 0 (22a) The stationary boundary condition is satisfied.

In the equations (17a) to (22a), the boundary conditions for the mesh numbers of 2 and 3 were obtained, but the equations exactly the same as these equations are inverted and used (that is, 0 → c, 1
→ m, 2 → m-1, 3 → m-2)
That is, P _m , _n ^{t + 1} / {(ρ _m , _n ^{t + 1} ) ^κ } = P _m-1 , _n ^{t + 1} / {(ρ _m-1 , _n ^{t + 1} ) ^κ } (17b) {Κ / (κ−1)} · P _m , _n ^{t + 1} / ρ _m , _n ^{t + 1} = {κ / (κ−1)} · P _m-1 , _n ^{t + 1} / ρ _m-1 , _N ^{t + 1} + (1/2) · (u _m-1 , _n ^{t + 1} ) ² (18b) ρ _m-1 , _n ^{t + 1} = ρ _m-1 , _n ^t + ρ _m-2 , _n ^t −ρ _m-2 , _n ^{t + 1} − (Δt / Δx) · {(ρ · u) _m-2 , _n ^{t + 1} + (ρ · u) _m-2 , _n ^t − (ρ · u ) _M-1 , _n ^{t + 1} − (ρ · u) _m-1 , _n ^t } (19b) where ρ _m , _n ^{t + 1} = ρc _n ^{t + 1} (20b) P _m , _n ^{t _{+1 = Pc n t + 1 ...}} (21b) um, the number of meshes in the formula of ^{_{n t + 1 = 0 ... (}} 22b) can be obtained boundary conditions of m-1 and m-2.

Equations (10) to (16) and (17)
Formula a), Formula (18a), Formula (19a), Formula (20a),
Expression (21a), Expression (22a), Expression (17b), (18
b) formula, (19b) formula, (20b) formula, (21b) formula,
By solving the equation (22b), the densities ρ ₁ , ₁ ^{t + 1 to} ρ _m , _n ^{t + 1} for each mesh number from 1 to m, and the flow velocities u ₁ , ₁ ^t for each mesh number from 1 to m. ^{_{+1 ~u m, n t + 1}} , the pressure P ₁ for each mesh number of from 1 to ^{_{m, 1 t + 1 ~P m}} ,
It is possible to obtain _n ^{t + 1} for each cylinder.

<D> Calculation of intake valve passing mass flow rate M _n ^t Among the state quantities in the branch pipe obtained by <c> above, the density ρ _m-1 , _n ^{t + 1} and the flow velocity u at the mesh number m-1 _m-1 ,
_{Using n} ^{t + 1} , M _n ^{t + 1} = F _n · ρ _m−1 , _n ^{t + 1} · u _m−1 , _n ^{t + 1} (23) where F _n : intake valve opening area (n : Cylinder No.) is used to calculate the intake valve passing mass flow rate M _n ^{t + 1} for each cylinder.

Here, the intake valve opening area F _n can be obtained by multiplying the minimum area of the gap between the intake valve head portion and the intake valve seat (this area is determined by the valve lift) by the flow coefficient. However, since it takes time to obtain the minimum area of the gap between the intake valve head portion and the intake valve seat from the valve lift, a weight value (parameter is a crank angle) corresponding to the minimum area is given in advance in the ROM, The product of this weight value obtained from the crank angle and the passage flow velocity of the intake valve portion (using this flow velocity u _m−1 , _n ^{t + 1} obtained in <c> above) is F _n
It is also possible to obtain M _n ^{t + 1} by using it instead of. F _n = 0 except for the intake stroke.

<E> Calculation of mass air amount M _n for each cylinder By integrating the intake valve passing mass flow rate M _n ^{t + 1} over the opening period of the intake valve, that is,

(Equation 3) From the equation, the mass air amount M _n (n: cylinder number) flowing from the intake valve of each cylinder in the intake stroke of each cylinder can be obtained for each cylinder.

The intake valve opening timing t _OPN and the intake valve closing timing t _{CLS in the} equation (24) are expressed in terms of the intake valve opening timing and closing timing (both are in crank angle units) using the rotational speed at that time as a time unit. It is obtained by converting to.

[0051] Tp _n = with <f> calculated mass air quantity M _n of cylinder of the basic injection pulse width _{Tp n (M n / N)} × K ... (25) However, N: rotational speed K: constant basic injection pulse width by the equation Tp _n: calculating the (n cylinder number) cylinder.

[0052] Using the <g> basic injection pulse width Tp _n different calculation cylinder of the cylinder-specific fuel injection pulse width _{Ti n, Ti n = {(} Tp n + Kathos) × Tfbya × (α + αm-1.0) + Chos _{n -Eraci n} × 2 + Ts} ... (26) However, Kathos: transient correction amount Tfbya: target air-fuel ratio equivalent weight alpha: air-fuel ratio feedback correction coefficient .alpha.m: basic air-fuel ratio learned value Chos _n, Eraci _n: cylinder correction amount (n: cylinder number) Ts: calculating a fuel injection pulse width Ti _n each cylinder by the equation of invalid pulse width (voltage correction amount).

In equation (26), Kathos, Tfb
ya, α, αm, is known both Chos _n, Eraci _n, Ts is. Kathos is a correction value for considering immediately after the start when the air-fuel ratio is greatly affected by the influence of the fuel wall flow, immediately after recovery from the fuel cut, and during acceleration / deceleration. Tfbya is a target fuel-air ratio equivalent amount, which is a value greater than 1 from the time of start-up to the completion of warm-up or at the time of full load to make the air-fuel ratio rich. α is O ₂
A value for feeding back the air-fuel ratio to near the stoichiometric air-fuel ratio based on the sensor output, αm is a learning value calculated based on the average value of this α, and Ts is for considering the operation delay until the fuel injection valve opens. Is the ineffective injection pulse width of. Ch
os _n, Eraci _n is a value introduced by considering the fast frequency response of the fuel wall flow (JP-A-3-111639
Reference).

Further, 2 in the equation (26) is a value required when the fuel injection is sequential injection (once every two engine revolutions, the exhaust stroke is the injection timing for each cylinder).

Next, a further explanation will be given according to the flowchart given to the CPU.

The flow charts of FIGS. 4 and 5 are for calculating the mass air amount M _n for each cylinder, and are executed at every minute time Δt.

First, in step 1, the crank angle sensor 12
Crank angle θc and throttle sensor 13 obtained by
The throttle opening θt, the atmospheric density ρa, and the atmospheric pressure Pa obtained by the above are read, and the throttle opening opening St is calculated from the throttle opening θt in step 2 by the above equation (1). Here, ρa and Pa are fixed values for normal lowland specifications (for example, ρa = 1.184 kg).
/ M ³ (corresponding to 25 ° C), Pa = 101325 Pascal)
In high altitude specifications, etc., it must be detected by a sensor.

In step 3, the collector internal pressure stored in the memory (initial value is atmospheric pressure Pa. P described later)
The initial values of ₁ , ₁ ^{t to} P _m , _n ^t are all Pa) P ₀ ^t and {2 /
(Κ + 1)} ^{κ / (κ + 1)} · Pa is compared, and P ₀ ^t ≦ {2
/ (Κ + 1)} ^{κ / (κ +1)} · Pa, in step 4, throttle opening area St, atmospheric density ρ
The throttle portion passing mass flow rate Mt is calculated by the above equation (3) using a and atmospheric pressure Pa. P ₀ ^t > {2 / (κ +
1)} ^{κ / (κ + 1)} · Pa, in step 5, P ₀ ^t is used in addition to St, ρa, and Pa, and the throttle portion passing mass flow rate Mt is calculated by the above equation (4).

In step 6, P ₀ ^t , Mt, ρa, Pa,
Number of meshes of the branch pipe state quantity stored in the memory those positions of ^{_{2 (ρ 2, i t,}} u 2, i t, P 2, i t) of the current by using the above equation (8) Collector pressure P ₀ ^{t + 1}
Is calculated. Similarly, in step 7, the density in the collector stored in the memory (the initial value is the atmospheric density ρa. The initial values of ρ ₁ , ₁ ^t to ρ _m , and _n ^t which will be described later are all ρa).
ρ ₀ ^t, Mt, mesh number of the branch pipe state quantity stored in the memory those positions of ^{_{2 (ρ 2, i t,}} u 2, i t)
By using the above equation (9), the current density in the collector ρ ₀
Calculate ^{t + 1} .

In step 8, these P ₀ ^{t + 1} and ρ ₀ ^{t + 1} and the branch pipe state quantities (ρ ₁ , ₁ ^{t to} ρ _m ,
^{_{n t, u 1, 1 t}} ~u m, n t, P 1, 1 t ~P m, n t), the equation (10) - (16), (17a) below, (18a)
Expression, (19a) Expression, (20a) Expression, (21a) Expression, (2
2a), (17b), (18b), (19b)
Using the equations, (20b), (21b), and (22b), the current branch pipe state quantities (ρ ₁ , ₁ ^{t + 1 to} ρ _m , _n ^{t + 1} ,
u _1, calculates ^{_{1 t + 1 ~u m, n}} t + 1, P 1, 1 t + 1 ~P m, a _n ^{t + 1)} in the cylinder. Although it is difficult to calculate the state quantity in the branch pipe in step 8, it can be easily calculated without requiring convergence calculation (calculation time is not required).

Steps 9 to 19 in FIG. 5 are portions for obtaining the mass air amount M _n for each cylinder. Here, for the sake of simplicity of the four-cylinder engine (the ignition order is # 1- # 3- # 4- # 2), as shown in FIG. 6, the intake strokes of the respective cylinders are not overlapped. . At this time, it is determined in step 9 whether or not the intake valve of the first cylinder is in the open period, and if it is in the open period, the process proceeds to step 10 and, when the number of meshes is m-1, branch from the data obtained in step 8. State quantity in pipe (density ρ _m-1 , ₁ ^{t + 1} , flow velocity u _m-1 , ₁
^{t 1} ), M ₁ ^{t + 1} = F ₁ · ρ _m-1 , ₁ ^{t + 1} · u _m-1 , ₁ ^{t + 1} (23a) where F _{1 is} the intake valve of the first cylinder The intake valve passage mass flow rate M ₁ ^{t + 1} of the No. 1 cylinder is calculated by the formula of the opening area, and this is calculated in step 11 as a variable (0
Initial setting to) Add to M ₁ .

Similarly, during the opening period of the intake valve of the third cylinder, the routine proceeds from step 12 to steps 13 and 14, where M ₃ ^{t + 1} = F ₃ · ρ _m−1 , ₃ ^{t + 1} · u _{m- 1} , ₃ ^{t + 1} (23b) However, the intake valve passage mass flow rate M ₃ ^{t + 1} of the 3rd cylinder is calculated by the formula of F ₃ : intake valve opening area of the 3rd cylinder, and this is a variable (at the time of starting) 0 is added to the initial setting) M _3, the routine proceeds to step 16 and 17 from the step 15 in the opening period of the intake valve of the fourth cylinder ^{_{M 4 t + 1 = F 4}} · ρ m-1, 4 t + 1 U _m-1 , ₄ ^{t + 1} (23c) However, the intake valve passage mass flow rate M ₄ ^{t + 1} of the fourth cylinder is calculated by the formula of F ₄ : intake valve opening area of the fourth cylinder, and this is calculated as follows. The variable (initially set to 0 at the time of starting) is added to M _4, and during the opening period of the intake valve of the second cylinder, the routine proceeds from step 15 to steps 18 and 19, and M ₂ ^{t + 1} = F ₂ · ρ _m-1 , ^{_{2 t + 1 · u m-}} 1, 2 t + 1 ... 23d) However, F _2: the equation of the intake valve opening area of the second cylinder to calculate the intake valve passing mass flow M ₂ ^{t + 1} of the second cylinder, which variable (initialized to zero at startup) to M ₂ to add.

At step 20, ρ ₀ ^{t + 1} for the next calculation.
To p ₀ ^t memory, u ₀ ^{t + 1} to u ₀ ^t memory, and P ₀ ^{t + 1} to P ₀ ^t memory. The branch pipe state quantity calculated this time ^{_{(ρ 1, 1 t + 1}} ~ρ m, n t + 1, u 1, 1 t + 1 ~u m, n t + 1, P
₁ , ₁ ^{t + 1 to} P _m , _n ^{t + 1} ), and ρ ₁ , ₁ ^{t to} ρ _m , _n ^t , u ₁ , ₁ ^t
~ U _m , _n ^t , P ₁ , ₁ ^{t to} P _m , _n ^t corresponding memory (total of m × n × 3 memories).

By _cyclically repeating the operations of FIGS. 4 and 5 every Δt, for example, in FIG. 6, the mass air amount of the first cylinder is determined by M ₁ at the end time t _CLS1 of the intake stroke of the first cylinder as shown in FIG. The mass air amount of the 3rd cylinder is M ₃ at the end time t _CLS3 of the intake stroke of the 3rd cylinder, and the mass air amount of the 4th cylinder is M ₄ at the end time t _CLS4 of the intake stroke of the 4th cylinder. M ₂ at the end time t _CLS2 of the intake stroke of the second cylinder
Thus, the mass air amount of the second cylinder can be accurately obtained.

As shown in FIG. 10, the intake strokes of the cylinders adjacent to each other in the ignition order overlap each other, so that two jobs (for example, the accumulation of M ₁ ^{t + 1} of the _first cylinder and the third work are performed at the same time). When it is necessary to perform (M ₃ ^{t + 1} integration of cylinder), TTS (Time Sharing System)
You just have to work the PU.

At step 22, it is determined whether or not one combustion cycle for all cylinders (a time equivalent to 720 ° in crank angle for a four cylinder engine) has passed. If one combustion cycle for all cylinders has passed, the routine proceeds to step 23, and the next time To prepare for the combustion cycle of M ₁ , M ₃ , M ₄ , M ₂ are cleared (0 is inserted into M _n ). As a result, one combustion cycle for all cylinders (crank angle 720 ° for a four cylinder engine)
The mass air amounts M _{1 to} M _{4 of} each cylinder are updated every time.

The flowchart of FIG. 7 is for calculating the fuel injection pulse width given to the injection valve for each cylinder, and is executed at each end timing of the intake stroke of each cylinder.

At step 31, it is determined whether the end timing of the intake stroke is for the first cylinder, and if it is for the first cylinder, the routine proceeds to step 32, where the basic injection pulse width Tp ₁ of the first cylinder is Tp ₁ = (M ₁ / N) × K (25a), and using this Tp ₁ , in step 33, Ti ₁ = {(Tp ₁ + Kathos) × Tfbya × (α + αm-1.0) + Chos ₁ − Eraci ₁ } × 2 + Ts (26a) is used to calculate the fuel injection pulse width Ti ₁ for the _first cylinder. In step 34, this Ti ₁ is transferred to the output register.

Similarly, when it comes to the end timing of the intake stroke of the third cylinder, steps 35 to 36, 37
Proceed to Tp ₃ and Ti ₃ and Tp ₃ = (M ₃ / N) × K (25b) Ti ₃ = {(Tp ₃ + Kathos) × Tfbya × (α + αm-1.0) + Chos ₃ -Eraci ₃ } × 2 + Ts According to the equation (26b), at the end timing of the intake stroke of the fourth cylinder, the routine proceeds from step 39 to steps 40 and 41, and T
p _4, Ti ₄ to _{Tp 4 = (M 4 / N} ) × K ... (25c) Ti 4 = {(Tp 4 + Kathos) × Tfbya × (α + αm-1.0) + Chos 4 -Eraci 4} × 2 + Ts ... ( 26c), when it comes to the end timing of the intake stroke of the second cylinder, the routine proceeds from step 39 to steps 43 and 44.
p ₂ and Ti ₂ are Tp ₂ = (M ₂ / N) × K (25d) Ti ₂ = {(Tp ₂ + Kathos) × Tfbya × (α + αm-1.0) + Chos ₂ −Eraci ₂ } × 2 + Ts. 26d) and the calculated Ti ₃ , Ti ₄ , and T, respectively.
i ₂ is transferred to the output register in steps 38, 42 and 45, respectively.

Inside the control unit 11, 1
At the injection timing for the # 1 cylinder (for example, the latter half of the exhaust stroke of the # 1 cylinder), Ti ₁ is injected into the injection valve of the # 1 cylinder.
When the injection signal corresponding to No. ₃ becomes the injection timing for the _third , _fourth , and _second cylinders, the respective drive signals corresponding to Ti ₃ , Ti ₄ , and Ti ₂ are supplied to the injection valves of the _third , _fourth , and _second cylinders. The fuel is output and fuel injection is performed sequentially.

The operation of the embodiment will now be described.

FIG. 8 shows a general correlation between the throttle opening TVO and the airflow meter output during acceleration in the mass flow system. The overshoot of the airflow meter output is caused by the volume in the intake manifold including the collector volume, and the larger the intake manifold volume, the larger the amplitude of the overshoot.
Therefore, the filling efficiency η obtained by the general mass flow method
The waveform of c is similar to the output of the air flow meter as shown by the broken line in FIG. 9, and does not trace the accurate charging efficiency during acceleration. In addition, various correction methods have been tried to obtain the accurate filling efficiency during transient from the output of the air flow meter, but all methods are linear approximation with a simple delay system such as a first-order delay system, so that transient It is difficult to trace the exact filling efficiency.

On the other hand, according to the present invention, the throttle portion passage mass flow rate Mt is obtained from the throttle opening θt, and the throttle portion passage mass flow rate Mt and the atmospheric state quantity (Pa, ρ).
a), state quantity in collector before Δt (ρ ₀ ^t , P ₀ ^t ), Δt
Before the branch pipe state quantity ^{_{(ρ 2, i t, u}} 2, i t, P 2, i t) collector in the state of current from seeking ^{_{(ρ 0 t + 1, P}} 0 t + 1), the current collector in the state of ^{_{(ρ 0 t + 1, P}} 0 t + 1) and Δt before the branch pipe state quantities ^{_{(ρ 1, 1 t ~ρ m}} , n t, u 1, 1 t ~u m,
_n ^t , P ₁ , ₁ ^{t to} P _m , _n ^t ), the basic equation of fluid dynamics (Euler's equation), and the boundary condition are used to calculate the state quantity in the branch pipe (ρ ₁ , ₁ ^{t + 1 to} ρ _{m). ^{, n t + 1, u 1}} , 1 t + 1 ~u m, n t + 1,
P ₁ , ₁ ^{t + 1 to} P _m , _n ^{t + 1} ) is calculated for each cylinder, and the value (ρ _m-1 , ₂ ^t when the number of meshes in the branch pipe state quantity calculated this time is m-1) ⁺¹ · u _m-1 , ₂ ^{t + 1} ) and intake valve opening area F
calculated intake valve passing electrolyte flow M _n ^{t + 1} to the individual cylinder from _n, the air mass M _n flowing into the cylinder by integrating the intake valve passing mass flow M _n ^{t + 1} for the intake valve opening period in cylinder Since it is calculated, no overshoot occurs in the change of the charging efficiency ηc in any cylinder (Fig. 9).
Solid line). It has been confirmed that the charging efficiency according to the present invention has an error of several% with respect to the actually measured charging efficiency for each cylinder.

Further, in the sequential injection of the conventional example, Tp = Qa / N × K generally at a timing of every 10 ms, where Qa: the basic injection pulse width Tp is calculated by the formula of the air flow rate obtained from the air flow meter output, Only Ti is obtained using Tp immediately before the fuel injection (see FIG. 10), and the opening / closing timing of the intake valve is not considered. Therefore, when the same acceleration is performed for the same cylinder, the same value of Qa is used regardless of whether the acceleration is performed immediately before the intake stroke or just before the intake stroke. In the invention, since the mass air amount flowing into the cylinder is individually calculated during the intake valve opening period in consideration of the opening / closing timing of the intake valve, when the throttle valve is rapidly opened immediately before entering the intake stroke for the same cylinder and when the intake stroke is increased. A mass air amount M _n having a different value from that when the throttle valve is opened a short time before entering is calculated.

In this way, the mass air amount M _n flowing into the cylinder at the time of transition can be accurately obtained in consideration of the opening / closing timing of the intake valve for each cylinder. Therefore, the fuel amount is calculated based on the mass air amount M _n. When the injection is performed, the supply air-fuel ratio becomes constant before and after the transition in any cylinder, which improves exhaust emission and fuel efficiency.

Further, in calculating the mass air amount M _n for each cylinder, only the memory for storing the state amount in the collector before Δt and the state amount in the branch pipe and the basic equation of the fluid dynamics are required. No matching work is required for the preparation, and the ROM can be formed only by inputting the engine intake pipe dimensions and engine specifications. The accuracy of calculating the air amount during the transition and the accuracy of controlling the air-fuel ratio during the transition can be secured without using matching.

In the mass flow system, lean misfire may occur due to overshoot at the output of the air flow meter at the time of acceleration. In the present invention, however, the lean misfire at acceleration (or rich misfire at deceleration) can be eliminated. Therefore, the drivability and stability of the engine are significantly improved. It is also inexpensive because it does not require an air flow meter.

The flowcharts of FIGS. 11 and 12 are the second embodiment and correspond to FIGS. 4 and 5. Explaining only the parts different from FIG. 4 and FIG. 5, in step 51, the crank angle θc, the atmospheric density ρa, and the atmospheric pressure Pa as well as the collector internal pressure P ₀ are read, and using this P ₀ , ρ ₀ = (P ₀ / Pa) ^{1 / κ} · ρa (31) The collector internal density ρ ₀ is calculated by the equation (adiabatic equation).

In steps 53 and 54, P _{0 is set} to P ₀ ^{t + 1} ,
Put ρ ₀ into ρ ₀ ^{t + 1} . The operation after this is the same as in the first embodiment.

When performing EGR (exhaust gas recirculation), the collector internal pressure (also regarding the collector internal density) cannot be determined only by the throttle opening, so the mass air amount M _n for each cylinder is calculated by the EGR amount. Although an error occurs, in the second embodiment, the collector 4 is used by the pressure sensor.
Since the internal pressure P ₀ is detected, the mass air amount M _n can be accurately determined for each cylinder even when the collector internal pressure and the collector internal density cannot be determined only by the throttle opening. According to the fuel injection amount calculated using the amount M _n , the supply air-fuel ratio becomes constant before and after the transition in any cylinder even in an engine in which the collector internal pressure and the collector internal density cannot be determined only by the throttle opening. Of.

In the lean burn system, it is necessary to suppress the fluctuation of the air-fuel ratio in the lean combustion operating range. However, by applying the present invention to the lean burn system, the fluctuation of the air-fuel ratio is suppressed so that the lean combustion operating range is expanded. Fuel consumption can be improved by just that much.

[0082]

EFFECT OF THE INVENTION Due to the volume in the intake manifold including the collector volume, the airflow meter output overshoots, so the waveform of the charging efficiency obtained by the general mass flow method approximates to the airflow meter output. However, in the first invention, the throttle part passage mass flow rate is obtained from the throttle opening, and the throttle part passage mass flow rate, the atmospheric state quantity, and the minute time before are obtained. State quantity in collector,
The state quantity in the collector this time is obtained from the state quantity in the branch pipe a minute time before, and the branch state this time is calculated using the state quantity in the collector this time, the state quantity in the branch pipe a minute time ago, the basic equation of fluid dynamics, and the boundary condition. The in-pipe state quantity is obtained for each cylinder, the intake valve passage mass flow rate is obtained for each cylinder from the branch pipe state quantity and the intake valve opening area calculated this time, and the intake valve passage mass flow rate is calculated over the intake valve opening period of the corresponding cylinder. Since the amount of mass air flowing into the cylinder is calculated for each cylinder by integrating, there is no overshoot in the change of the charging efficiency during the transition in any cylinder, and the actual charging efficiency for each cylinder is improved. Can be traced. In addition, when calculating the mass air amount, only the memory for storing the state amount in the collector and the state amount in the branch pipe a minute time ago and the basic equation of the fluid dynamics are required, so that the matching work for creating the ROM is unnecessary. Therefore, the ROM can be formed only by inputting the dimensions of the intake pipe of the engine and the specifications of the engine. Further, in the mass flow system, lean misfire may occur due to overshoot at the time of acceleration generated in the air flow meter, but in the first invention, since lean misfire at acceleration (or rich misfire at deceleration) can be eliminated, The drivability and stability of the engine are greatly improved. It is also inexpensive because it does not require an air flow meter.

In this way, the mass air amount flowing into the cylinder during the transition can be accurately obtained for each cylinder. Therefore, if fuel injection is performed based on this mass air amount according to the fourth invention, the supply air-fuel ratio Is constant before and after the transition for all cylinders, which improves exhaust emission and fuel efficiency.

When the EGR is performed, the collector internal pressure (also regarding the collector internal density) cannot be determined only by the throttle opening, so that there is a calculation error in the mass air amount M _n for each cylinder by the EGR amount. There is a second
In the invention, since the collector internal pressure P ₀ is detected by the pressure detecting means, the mass air amount M _n can be accurately determined for each cylinder even when the collector internal pressure and the collector internal density cannot be determined only by the throttle opening. be able to.

In the fifth aspect of the present invention, since fuel injection is performed using this mass air amount, even in an engine in which the collector internal pressure and collector internal density cannot be determined only by the throttle opening, the supply air-fuel ratio is transient in any cylinder. It becomes constant before and after.

[Brief description of drawings]

FIG. 1 is a control system diagram of an embodiment.

FIG. 2 is a diagram showing an intake pipe model.

FIG. 3 is an enlarged view of a joint portion between the collector and the branch pipe of FIG.

FIG. 4 is a flowchart for explaining calculation of a mass air amount M _n .

FIG. 5 is a flowchart for explaining calculation of a mass air amount M _n .

FIG. 6 is a waveform diagram showing intake valve lift in the case of a 4-cylinder engine.

7 is a waveform diagram for explaining the calculation of the fuel injection pulse width Ti _n given to the fuel injection valve.

FIG. 8 is a waveform diagram showing a general correlation between the throttle opening and the airflow meter output during acceleration in the mass flow system.

FIG. 9 is a change waveform diagram of the charging efficiency ηc during acceleration.

10 is a diagram for explaining the calculation timing of the intake valve lift and fuel injection pulse width Ti _n the case of 4-cylinder engine.

FIG. 11 is a flowchart for explaining calculation of a mass air amount M _{n according to} the second embodiment.

FIG. 12 is a flowchart for explaining calculation of a mass air amount M _{n according to} the second embodiment.

FIG. 13 is a diagram corresponding to the claims of the first invention.

FIG. 14 is a diagram corresponding to claims of the second invention.

FIG. 15 is a diagram corresponding to claims of the fourth invention.

FIG. 16 is a diagram corresponding to the claim of the fifth invention.

[Explanation of symbols]

 1 Engine Main Body 3 Throttle Part 4 Collector 5 Branch Pipe 6 Cylinder 7 Fuel Injection Valve (Fuel Supply Means) 11 Control Unit 12 Crank Angle Sensor 13 Throttle Sensor

Claims (7)

[Claims] 1. A means for detecting a throttle opening of an engine, a means for calculating a throttle opening area from the throttle opening, a means for calculating a mass flow rate of a throttle portion based on the throttle opening area, and a minute time. A means for calculating the collector internal pressure this time by using the previous collector internal pressure, the mass flow rate passing through the throttle part, atmospheric pressure, atmospheric density, and the branch pipe state quantity before a minute time, and the collector internal pressure calculated this time next time. Means for storing for calculating the collector internal pressure, means for calculating the collector internal density at this time using the collector internal density before the minute time, the mass flow rate through the throttle part, and the branch pipe state amount before the minute time, A means for storing the collector internal density calculated this time for the next collector internal density calculation, the collector internal density of the current time and the collector internal density of the current time. A means for calculating the current branch pipe state amount for each cylinder using the in-lector pressure, the branch pipe state amount before a minute time, the basic equation of fluid dynamics, and predetermined boundary conditions, and the branch pipe state amount calculated this time. A means for calculating the intake valve passing mass flow rate for each cylinder from the intake valve opening area of each cylinder, and a value obtained by integrating this intake valve passing mass flow rate over the intake stroke period of the corresponding cylinder as a mass air amount for each cylinder. An intake air amount detection device for an engine, comprising: a means for calculating and a means for storing the state quantity in the branch pipe of this time for calculating a state quantity in the collector and a state quantity in the branch pipe of the next time. 2. A means for detecting the collector internal pressure, a means for calculating the collector internal density using the collector internal pressure and the atmospheric pressure, and a state in the branch pipe before the collector internal density, the collector internal pressure and a minute time. Amount, the basic equation of fluid dynamics, and predetermined boundary conditions, and the means for calculating the current branch pipe state quantity for each cylinder, and the intake valve passage mass flow rate from the branch pipe state quantity calculated this time and the intake valve opening area. For each cylinder, a means for calculating a value obtained by integrating the intake valve passing mass flow rate over the period of the intake stroke of the corresponding cylinder as a mass air amount for each cylinder, and the current branch pipe state quantity at this time. And a means for storing the calculated state quantity in the branch pipe of the engine. 3. The intake valve opening area is a product of a weight value corresponding to the intake valve opening area and an intake valve passage velocity of the branch pipe state quantity calculated this time. Alternatively, the intake air amount detection device for the engine according to the above item 2. 4. A means for detecting a throttle opening of an engine, a means for calculating a throttle opening area from the throttle opening, a means for calculating a mass flow rate through a throttle portion based on the throttle opening area, and a minute time. A means for calculating the collector internal pressure this time by using the previous collector internal pressure, the mass flow rate passing through the throttle part, atmospheric pressure, atmospheric density, and the branch pipe state quantity before a minute time, and the collector internal pressure calculated this time next time. Means for storing for calculating the collector internal pressure, means for calculating the collector internal density at this time using the collector internal density before the minute time, the mass flow rate through the throttle part, and the branch pipe state amount before the minute time, A means for storing the collector internal density calculated this time for the next collector internal density calculation, the collector internal density of the current time and the collector internal density of the current time. A means for calculating the current branch pipe state amount for each cylinder using the in-lector pressure, the branch pipe state amount before a minute time, the basic equation of fluid dynamics, and predetermined boundary conditions, and the branch pipe state amount calculated this time. A means for calculating the intake valve passing mass flow rate for each cylinder from the intake valve opening area of each cylinder, and a value obtained by integrating this intake valve passing mass flow rate over the intake stroke period of the corresponding cylinder as a mass air amount for each cylinder. A means for calculating, a means for storing the state quantity in the branch pipe this time for calculating the state quantity in the collector and a state quantity in the branch pipe for the next time, and one combustion by dividing the mass air quantity by the engine speed. A means for calculating the fuel injection amount required for the cycle for each cylinder at a predetermined timing and a means for supplying this injection amount of fuel for each cylinder at a predetermined timing are provided. The fuel injection control apparatus for an engine according to symptoms. 5. A means for detecting the collector internal pressure, a means for calculating the collector internal density using the collector internal pressure and the atmospheric pressure, and a state in the branch pipe before the collector internal density, the collector internal pressure and a minute time. Amount, the basic equation of fluid dynamics, and predetermined boundary conditions, and the means for calculating the current branch pipe state quantity for each cylinder, and the intake valve passage mass flow rate from the branch pipe state quantity calculated this time and the intake valve opening area. For each cylinder, a means for calculating a value obtained by integrating the intake valve passing mass flow rate over the period of the intake stroke of the corresponding cylinder as a mass air amount for each cylinder, and the current branch pipe state quantity at this time. Means for storing the state quantity in the branch pipe for calculating the amount of fuel injection necessary for one combustion cycle by dividing the mass air amount by the engine speed. In means for calculating the cylinder, the fuel injection control apparatus for an engine, wherein a fuel of the injection amount is provided and means for supplying each cylinder at a predetermined timing. 6. The fuel for an engine according to claim 4, wherein the predetermined timing for calculating the fuel injection amount required for the one combustion cycle is the end timing of the intake stroke of each cylinder. Injection control device. 7. The transient correction amount relating to the fuel wall flow is used to determine the
7. The engine fuel injection control device according to claim 4, further comprising means for correcting a fuel injection amount required for a combustion cycle.