JPH0523815Y2 - - Google Patents

Description

[Detailed explanation of the idea] (Industrial application field) This invention relates to an air amount detection device for an engine.

(Prior Art) Since a fuel injection engine is configured to inject and supply a fuel amount commensurate with the amount of air taken into the engine, it is important to accurately detect the amount of intake air. For this reason, there are detection devices that directly detect the amount of air using a flow rate sensor such as a hot-wire type, and devices that indirectly detect the amount of air from the intake pipe pressure and engine rotation speed measured by a pressure sensor.
Additionally, a system has been proposed in which a throttle valve opening sensor is installed in addition to the pressure sensor, and the amount of air is detected from the throttle valve opening and intake pipe pressure.
(See Publication No. 4981, etc.)

(Problem that the invention attempts to solve) However, in the detection device using the flow rate sensor or pressure sensor, the detected value is greatly affected by intake pulsation, and the fuel injection amount calculated based on this detected value fluctuates. As a result, engine torque fluctuations occur. Further, in the flap type flow rate sensor, the configuration for suppressing pulsation limits accurate response during transient times, and there is also the problem of high cost.

On the other hand, these detection devices detect the amount of air at the sensor installation location, so if the sensor is installed far away from the cylinder, the amount of air detected by the sensor and the air actually flowing into the cylinder will differ during transient periods. The amount does not match by the amount of response delay. Furthermore, even if the cylinder air amount can be accurately detected by taking this response delay into consideration, if the fuel injection valve is located far from the cylinder,
The amount of air in the cylinder and the amount of air in the injection valve section do not necessarily match. As a result, if injection control is performed based on an air amount that is not accurately detected, the air-fuel ratio will deviate from the target air-fuel ratio and become rich or lean during transient times such as acceleration and deceleration.

For this reason, the steady flow rate Q _H of the throttle valve section is determined from the throttle valve opening degree α and the engine rotational speed N, and by delay-correcting this with a delay coefficient K ₂ , the cylinder flow rate is determined.
Q _CYL was calculated, and a predetermined addition amount ΔCM was added to the product of this cylinder flow rate Q _CYL and cylinder volume V, thereby calculating the air amount Q _AINJ in the injection valve section. (Unexamined Japanese Patent Publication No. 1983-
(See No. 32323).

In the device of this prior application, as shown in FIG.
Calculate the linearized flow rate (steady flow rate per unit cylinder volume of the throttle valve section) Q _HO from (N・V)), and correct this using the correction factor KFLAT to calculate the steady flow rate of the throttle valve section per unit cylinder volume. flow rate
Set Q _H (=Q _HO × KFLAT) (steps 41 to 45).

For this Q _H , Q _CYL = Q _H × K ₂ + Q _CYL-1 × (1-K ₂ ) However, K ₂ ; Delay coefficient Q _CYL-1 ; Performing first-order delay correction using the previous Q _CYL cylinder flow rate
Q _CYL and the additional air amount ΔCM are calculated as follows: ΔCM=KMANI×(Q _CYL −Q _CYL-1 ) However, KMANI is determined by the coefficient, and the air amount Q _AINJ in the injection valve section is set as Q _AINJ = Q _CYL ×V+ΔCM. Yes (steps 46-49).

According to the device of this prior application, the influence of intake pulsation is eliminated by using the throttle valve opening degree α and the engine rotational speed N, and by introducing the delay coefficient K ₂ and the air addition amount ΔCM, from the throttle valve position to the cylinder. The response delay of the intake air during a transient period is considered to be the deviation corresponding to the pressure change in the intake pipe between the position of the injection valve during the transient period and the position near the cylinder. The amount of air flowing can now be calculated accurately.

However, it has been found that when the correction factor KFLAT is used in the process of determining the air addition amount ΔCM, large fluctuations occur in ΔCM.

In order to find the correction factor KFLAT from the parameters N and Q _HO , a three-dimensional map of KFLAT must be searched, but if 2 bytes are given to KFLAT data, the calculation time and program will be longer. It is preferable to give byte data. In this case, in the operating range near the throttle valve fully open,
A change in 1 bit in 1 byte data of KFLAT is reflected as a change in Q _CYL or Q _CYL-1 at the same rate, and this causes a large fluctuation in ΔCM.

The purpose of this invention is to provide an air amount detection device that can prevent fluctuations in ΔCM by preventing the correction factor KFLAT from entering the process of determining the additional air amount ΔCM.

(Means for Solving the Problems) As shown in FIG. =Aα) means 3
, a means 2 for detecting the engine rotational speed N, a means 4 for dividing the flow passage area A by the rotational speed N, and a linearization flow rate (per unit cylinder volume of the throttle valve section) based on this division value A/N. Steady flow rate) Q _HO
means 5 for calculating the delay coefficient K ₂ regarding the air flow from the throttle valve section to the cylinder using both _the detected values (N, A) _; means 7 for calculating the cylinder air amount Q _CYLO per unit cylinder volume by delay correction at , means 8 for calculating a predetermined air addition amount ΔCM from the amount of change in this cylinder air amount Q _CYLO , and the aforementioned division value. The above cylinder air amount is calculated using any one of A/N, linearization flow rate Q _HO , cylinder air amount Q _CYLO and rotational speed N.
Q Means of calculating correction factor KFLAT for _CYLO 9
and this correction factor KFLAT, the cylinder air amount
Means 10 for calculating the air amount Q _AINJ of the injection valve section using Q _CYLO , cylinder volume V, and air addition amount ΔCM
Equipped with.

(Function) When calculating the air addition amount ΔCM using the cylinder air amount related to the correction factor KFLAT, the correction factor
In order to shorten the calculation time and program length of KFLAT, when 1 byte data is given to the correction factor KFLAT, the correction factor is
A 1-bit change in 1-byte data of KFLAT changes the cylinder air amount at the same rate.
This causes a large variation in the air addition amount ΔCM, but in this invention, if the air addition amount ΔCM is determined using the cylinder air amount Q _CYLO , which is not affected by the correction factor KFLAT, even if one byte data is given to the correction factor KFLAT, , fluctuations in the air addition amount ΔCM in the operating range near the fully open throttle valve are prevented.

(Embodiment) FIG. 2 shows the mechanical configuration of a first embodiment in which the detection device of the present invention is applied to a fuel injection engine. however,
This example is the so-called single point injection method (SPI method),
One (or a plurality of) fuel injection valves 17 are provided in the intake passage 16 upstream of the intake throttle valve 15 .

In the figure, 19 is a sensor that detects the opening degree α of the throttle valve 15 (throttle valve opening sensor), and 20 is a sensor that detects the engine rotation speed N (for example, a crank angle sensor), and both detection signals are based on operating variables. These signals are input to the control unit 23 as basic values, and the unit 23 detects the air amount Q _AINJ in the injection valve section based on these signals, and performs fuel injection control using this Q _AINJ .

In addition, other operating variables include engine cooling water temperature T _W ,
There is an intake air temperature T _A , which is detected by a water temperature sensor 21 and an intake air temperature sensor (not shown), respectively. Further, the actual air-fuel ratio required for feedback control is detected by the air-fuel ratio sensor 22. Note that the passage 24 that bypasses the throttle valve 15 is provided with an electromagnetic valve (idle control valve) 25 that makes the flow area A _BY of the passage 24 variable.

FIG. 3 shows a calculation routine for the injection pulse width Ti that is executed within the CPU when the control unit 23 is configured with a microcomputer.
This figure is executed at predetermined intervals. The numbers shown are step numbers. Here, the calculation formula for Ti used in the L-dietronic method is as shown in equation (1) below, which calculates the basic pulse width Tp from various correction coefficients (the sum of these coefficients is COEF) and the actual air-fuel ratio. The air-fuel ratio is corrected using the obtained air-fuel ratio feedback correction coefficient LAMBDA (step 40). However, Ts
is the invalid pulse width.

Ti=Tp×COEF×LAMBDA+Ts...(1) Since Tp corresponds to the amount of fuel supplied from the injection valve 17, in order to obtain a constant air-fuel ratio, the difference between Tp and the amount of air Q _AINJ in the injection valve section is Therefore, Tp should be Tp=Ka×Q _AINJ ×KTA (2) (Step 39). However, Ka is a coefficient based on the characteristics of the injection valve 17, and KTA is an intake temperature correction coefficient. The units of these parameters are, for example, Tp (msec), Ka (msec/g), Q _AINJ (cc),
KTA (g/cc). As a result, how accurately Q _AINJ is calculated determines the air amount detection accuracy.

First, from the throttle valve opening α, the flow path area of the throttle valve part is
Calculate Aα (step 31). This can be easily obtained, for example, by searching a two-dimensional table (Aα table) containing the channel area characteristics shown in FIG. However, when the bypass passage 19 is opened, an error occurs by the passage area A _BY of this passage 19, so in this case, the total passage area A
(=Aα+A _BY ) must be adopted. Here, since the flow path area A (Aα) is uniquely determined by α, the influence of intake pulsation accompanying intermittent intake is eliminated.

Next, a value A/(N·V) is calculated by dividing this A by the engine rotational speed N and the cylinder volume V (step 32). The only reason to divide by N is A.
Depending on the change in N, there will be an operating range where A suddenly changes,
This is because the resolution decreases in this region.
The reason for dividing by V is to obtain a value per unit cylinder volume, that is, to obtain a value that is not related to the magnitude of V.

The linearization flow rate _QHO is calculated based on this division value A/(N·V) (step 33). Here,
Since the linearization flow rate Q _HO is a flow rate in a steady state, unlike the flow rate in a transient state, it can be accurately determined in advance by matching. As shown in FIG. 5, the obtained results show that almost the same flow rate characteristics are obtained regardless of the size of the cylinder volume. Therefore, the fifth
Q _HO is determined by searching a two-dimensional table containing the flow characteristics shown in the figure. however,
Considering the resolution, Q _HO must be calculated by linear interpolation, and therefore requires 2 bytes (8 bits).
use.

Next, the amount of air flowing into the cylinder Q _CYLO is calculated using the delay coefficient K ₂ (K ₂ <1) using the following equation (3) (step 35).

Q _CYLO = Q _HO ×K ₂ +Q _CYLO-1 × (1−K ₂ ) ……(3) Equation (3) takes into account transient response. For example, for a step change in the throttle valve opening α,
Q _HO shows the same change, but the amount of air cannot flow stepwise into cylinders that are farther away than the location where the throttle valve 15 is installed, and Q _CYLO responds with a first-order lag. Here, the coefficient K ₂ is a value that determines how Q _CYLO follows Q _HO , and is determined in advance by matching. That is, (N・V)
and A to calculate the delay coefficient _K2 (step
34). This is obtained by searching for a three-dimensional map containing the characteristics shown in FIG.

Note that since the control routine shown in FIG. 3 is executed at predetermined intervals (for example, 10 msec), the current value is determined using the value each time. That is, in the above equation (3), the subscript "-1" appended to _QCYLO means the previous value, and the current calculated value is sequentially determined using the previous calculated value. Thereby, even during a transient period, the cylinder air amount can be accurately calculated even though the air amount in the throttle valve section is being determined.

As a result, a deviation occurs between Q _HO and Q _CYLO during a transient period, but this deviation is factored into the delay coefficient K ₂ , and even during a transient period, the amount of air to the cylinder is
Q _CYLO is required accurately.

Finally, calculate the additional air amount ΔCM from the amount of change in Q _CYLO (Q _CYLO − _{Q CYLO-1} ) (step 36), and add this additional amount ΔCM to Q _CYLO × KFLAT × V to Find the air amount Q _AINJ (step 38). That is, ΔCM=(Q _CYLO −Q _CYLO-1 )×KMANI……(4) _QAINJ =Q _CYLO ×KFLAT×V+ΔCM……(6). Here, in the SPI method, the installation position of the injection valve 17 is farther away than the cylinder, so
Q _AINJ and Q _CYLO ×KFLAT × V do not necessarily match, and during a transient period, a deviation occurs by a predetermined amount (ΔCM) corresponding to the pressure change in the intake pipe pressure. Therefore, ΔCM in equation (6) is introduced as a value that takes into account the deviation between the two during transient times, and has the meaning of the amount of air that compensates for the inside of the intake pipe as the pressure changes in the intake pipe.

In addition, without KFLAT in equation (6), the actual air-fuel ratio will deviate slightly from the target air-fuel ratio due to the difference in cylinder volume (cylinder volume) V, and even if V is the same, the operating condition will change. This difference causes a slight deviation from the target air-fuel ratio, so this value is introduced to eliminate these differences. Different values are used depending on the size of the cylinder volume V, and it is calculated based on N and Q _HO . be done. For example, it can be obtained by searching for a three-dimensional map containing the characteristics shown in FIG.

For example, Q _AINJ (=Q _CYLO
×V+ΔCM) when performing fuel injection,
If the actual air-fuel ratio deviates slightly to the lean side from the target air-fuel ratio, give a value larger than 1 to KFLAT to return the air-fuel ratio to the target air-fuel ratio, and vice versa. By returning to the target air-fuel ratio with a value of KFLAT smaller than 1,
Minor deviations from the target air-fuel ratio are corrected,
Regardless of the size of the cylinder volume V, flat air-fuel ratio characteristics can be obtained in all operating ranges.

On the other hand, equation (4) is an equation that gives ΔCM, but unlike the device of the prior application, Q _CYLO and Q _CYLO-1 in equation (4) are KFLAT
Since it is a value that does not involve , according to equation (4),
ΔCM can be determined without using KFLAT.

Note that the coefficient KMANI in equation (4) is the value obtained by dividing the manifold coefficient KMANI ₀ by N, and as a result,
Errors caused when the control period and the rotation period of the engine are significantly different are prevented. In addition, when executing the control shown in Figure 3 in synchronization with engine rotation, KMANI=
KMANI ₀ is fine. The manifold coefficient KMANI ₀ is a constant determined by the intake volume from the throttle valve 15 to the intake port and the calculation cycle when the injection valve 17 is installed immediately upstream or downstream of the throttle valve 15. This is to ensure that the instantaneous air flow rate and Q _AINJ are in the same phase as a result.

Calculating Tp as a fuel amount using Q _AINJ obtained in this way means that it is possible to supply fuel according to the amount of air at the fuel injection position, which allows for a nearly constant air-fuel ratio. This allows the air-fuel mixture to be supplied to the cylinder.

Here, the operation of this example will be explained.

FIG. 8 is a diagram using the device of the earlier application as mentioned above.
The major difference between the calculation procedure of Q _AINJ and the present invention shown in Fig. 3 is that KFLAT is involved in the calculation of ΔCM. In other words, in the device of the prior application, using the linearization flow rate Q _HO as the basic value, the steady flow rate Q _H is determined by Q _H = Q _HO × KFLAT (7), and based on this Q _H , the cylinder flow rate Q _CYL , ΔCM, Q _AINJ (Steps 45-49) Note that these equations (8) to (10) are the previous equation
Corresponds to (3), (4), and (6), respectively.

Q _CYL = Q _H ×K ₂ +Q _CYL-1 × (1-K ₂ ) …(8) ΔCM = (Q _CYL −Q _CYL-1 ) ×KMANI …(9) Q _AINJ = Q _CYL ×V+ΔCM … …(10) However, with such a procedure,
It has been confirmed through experiments that fluctuations occur in ΔCM. When this is shown in FIG. 9A, the bit error occurring in KFLAT is magnified, causing large irregularities in ΔCM. This is due to the expression problem of how many bytes of hexadecimal number (or how many bits of binary number) each parameter should be expressed, and the expression has a large effect on detection accuracy.

For example, the parameter Aα, A/(N・V),
Regarding calculation formulas and two-dimensional table searches for Q _HO , Q _CYLO , ΔCM, and Q _AINJ , even if they are assigned as 2 bytes, the calculation time and program length will not become so long, and therefore 8-bit It can be fully functional with good responsiveness using a microcomputer. Note that these parameters are specifically assigned as 2 bytes because 1 byte does not provide sufficient resolution during idling or deceleration.

However, the parameters K ₂ and 3 of KFLAT
Searching for a dimensional map becomes a big problem because using 2 bytes increases calculation time and program length. Therefore, these parameters (K ₂ ,
KFLAT) is assigned in 1 byte.
However, for _K2 , there is no problem since the performance is sufficient even if it is 1 byte, but if KFLAT is assigned as 1 byte, a problem will occur near the fully open throttle valve. In other words, in this operating range, K ₂ is required to be almost 1.0, so in the previous equation (9), the change in 1 bit in 1 byte is calculated as the change in Q _CYL and Q _CYL-1 at the same ratio. ,
This results in large irregularities shown in FIG. 9A in ΔCM. As a result, Ti fluctuates through Tp, causing torque fluctuation.

Therefore, if KFLAT is assigned as 2 bytes, the assignment ratio to data with a 1-bit change will be reduced, so ΔCM can be changed smoothly. However, as mentioned above, assigning KFLAT to 2 bytes will reduce the computation time and program. This is a problem due to the length. For example, if 2 bytes are used, the ΔCM calculated when operating conditions suddenly change will not match the actual amount of air filling the intake pipe, resulting in a delay in response.
The deviation of KFLAT from the target value causes an undesirable phenomenon in which unevenness occurs in ΔCM.

On the other hand, according to this embodiment, although KFLAT is assigned as 1 byte to ensure responsiveness, values (Q _CYLO , Q _CYLO-1 ) that are not corrected by KFLAT are used. ΔCM is calculated (steps 34-36). In other words, ΔCM is calculated using a value that does not involve KFLAT.

Therefore, the response waveform during acceleration according to this example is shown in FIG. 9B under the same acceleration conditions as FIG. 9A for comparison. From the same figure, even if KFLAT
Even if a bit error occurs in ΔCM
changes smoothly without any unevenness caused by this error. As a result, changes in Ti also become smoother, thereby achieving good responsiveness while preventing torque fluctuations.

Next, FIG. 10 is a block diagram of a second embodiment of this invention. In this case, Q _CYLO is used only for calculating ΔCM, and Q _AINJ is calculated using this ΔCM, linearization flow rate Q _HO , and correction factor KFLAT. In other words, means 51 for determining the steady flow rate Q _H (=Q _HO × KFLAT) by correcting Q _HO by KFLAT.
Then, using this Q _H and delay coefficient K ₂ , the cylinder flow rate is
A means 52 for calculating Q _CYL (=Q _H ×K ₂ +Q _CYL-1 × (1-K ₂ )), and a means 52 for calculating Q _CYL , ΔCM and cylinder volume V.
The injection valve air amount calculation means 10 is constituted by the means 53 for calculating Q _AINJ (=Q _CYL ×V+ΔCM) by adding the product. In this example as well, KFLAT is not used in the calculation of ΔCM, and the same effect as in the first embodiment is achieved.

However, in this embodiment, the overall calculation time increases slightly due to the increase in the number of components.
Since it also has the flow rates (Q _H , Q _CYL ) corrected by KFLAT, it is easy to apply it to other controls that require these signals (such as ignition timing, swirl control valve, or lean burn control). .

In addition, in the first embodiment, Q _HO is used as a parameter.
I am calculating KFLAT, but instead of this Q _HO ,
The parameters before and after that are A/(N・V) and
Q There is no functional problem using _CYLO . Also, SPI
Needless to say, this invention can be applied not only to the injection method but also to an injection method in which a fuel injection valve is provided for each cylinder.

(Effects of the invention) As described above, according to the invention, an additional air amount is introduced, and this air addition amount changes the intake pipe pressure during a transient period between the cylinder air amount and the air amount of the injection valve part. While compensating for the deviation corresponding to Since the air addition amount is determined in such a manner that the air addition amount is determined without any difference, even if one byte data is given to the correction factor, it is possible to prevent the air addition amount from fluctuating in the operating range near the fully open throttle valve.

[Brief explanation of the drawing]

Figure 1 is a block configuration diagram of the present invention, Figure 2 is a mechanical configuration diagram showing the first embodiment of the invention, Figure 3 is a flowchart showing the calculation contents of this embodiment, and Figures 4 to 7. is a characteristic diagram showing the contents of each table used in this calculation. FIG. 8 is a waveform diagram explaining the operation of the earlier application, FIGS. 9A and 9B are waveform diagrams explaining the operation of the earlier application and the above embodiment, respectively, and FIG. 10 is a second embodiment of this invention. FIG. DESCRIPTION OF SYMBOLS 1... Throttle valve opening detection means, 2... Engine rotational speed detection means, 3... Channel area calculation means, 4... Division means, 5... Linearization flow rate calculation means, 6...
...Delay coefficient calculating means, 7... Cylinder air amount calculating means, 8... Addition amount calculating means, 9... Correction factor calculating means, 10... Injection valve air amount calculating means, 15...
... Throttle valve, 17... Fuel injection valve, 19... Throttle valve opening sensor, 20... Crank angle sensor, 23...
...Control unit, 51... Steady flow rate correction means, 52... Cylinder flow rate calculation means, 53...
means of addition.

Claims (1)

[Scope of utility model registration request] means for detecting a throttle valve opening; means for calculating a flow passage area of the throttle valve section from the throttle valve opening; means for detecting an engine rotation speed; and means for dividing the flow passage area by the rotation speed. , means for calculating a linearization flow rate based on this division value, means for calculating a delay coefficient regarding the air flow from the throttle valve section to the cylinder using the above-mentioned detected values, and delay correction for the linearization flow rate using the delay coefficient. means for calculating the amount of cylinder air per unit cylinder volume by calculating the amount of cylinder air per unit cylinder volume; means for calculating a predetermined additional amount of air from the amount of change in the amount of cylinder air;
means for calculating a correction factor for the cylinder air amount using either one of the linearization flow rate or the cylinder air amount and the rotational speed; An air amount detection device for an engine, comprising: means for calculating the air amount of the engine.