JPH06100149B2 - Control device for internal combustion engine - Google Patents

Description

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a control device for an internal combustion engine, and more particularly to a control device for coping with a change in volatility of fuel properties.

(Prior Art) A device has been proposed that prevents exhaust emission and drivability from deteriorating even when the fuel supplied to the engine changes from standard fuel to heavy fuel due to refueling. 56-32451).

This example is applied to those that control the fuel to the engine so that the target air-fuel ratio can be obtained. The air-fuel ratio is targeted for the exhaust concentration of exhaust three components (HC, CO, NOx) This is because the fuel ratio has a great influence. For example, from the characteristics of the conversion rate of the three-way catalyst with respect to the air-fuel ratio shown in FIG. 21, if it is attempted that all the conversion rates of the three components of the exhaust gas are 80% or more, the predetermined width W in the figure It is necessary to keep the actual air-fuel ratio within the range. Therefore, for standard fuel, the specified width W
The air-fuel ratio (which is almost the theoretical air-fuel ratio) A at the center of is defined as the target air-fuel ratio.

In order to realize this in the fuel injection engine of the L-Getronic system shown in FIG. 20, the injection pulse width Tp for obtaining the basic air-fuel ratio given to the fuel injection valve 7 provided in the intake port 2
(= K × Qa / N, where K is a constant, Qa is the intake air amount detected by the air amount sensor 5 upstream of the throttle valve 4, and N is the engine speed detected by the crank angle sensor 6) Is corrected by the feedback correction amount α (calculated based on the output of the sensor 9 whose output suddenly changes at the stoichiometric air-fuel ratio) to eliminate the deviation between the basic air-fuel ratio and the theoretical air-fuel ratio. In the figure, 3 is an intake pipe, 8 is an exhaust pipe, 11 is a water temperature sensor, 12 is a spark plug, and 13 is a control unit.

In this case, if a fuel (heavy fuel) that is less volatile than the standard fuel is used, the actual air-fuel ratio becomes leaner than the target air-fuel ratio when accelerating, which may cause drivability problems or exhaust emission. Make bad. Since the volatility in the intake pipe is determined by the distillation property of the supplied fuel, as the content of the heavy fraction increases, the amount of the fuel that flows in the liquid state on the intake pipe wall surface (wall Flow)
This is due to the fact that the ratio of is larger than the amount of fuel flowing in the intake pipe in a gaseous state.

In other words, during steady operation, the balance between the amount of fuel flowing into the intake pipe that is the wall flow and the amount that is sucked into the cylinder from the wall flow is in a balanced state (equilibrium state). The amount of wall flow itself due to the difference in properties does not affect the air-fuel ratio required for the engine. However, during the transient operation, the amount of fuel to be sucked into the cylinder is taken as an increase in the wall flow amount until the wall flow amount settles in the equilibrium state after the operation change. Here, the amount of fuel taken away as an increase in the wall flow amount is larger in the heavy fuel than in the standard fuel, and thus the heavy fuel has a large lean air-fuel ratio during the transition.

Therefore, by setting the target air-fuel ratio to the air-fuel ratio B, which is slightly richer than A in Fig. 21, by setting the target air-fuel ratio, it is possible to make a transition even in an area where heavy fuel is likely to be used (even in North America). It achieves both drivability and exhaust emission.

(Problems to be Solved by the Invention) By the way, in such a device, the target air-fuel ratio is set to the position of B in FIG. 20 by assuming the fuel (heavy fuel) that will be used in advance. Therefore, if a fuel that is different from the expected value (for example, a light fuel that is more volatile than the standard fuel, such as gasoline mixed with alcohol) is used, the actual air-fuel ratio will suddenly become rich during the transition and exhaust emission Becomes worse. The reason is that if the target air-fuel ratio is set for the light fuel, the air-fuel ratio C should be slightly leaner than A in Fig. 21, but in the heavy fuel specification where B is the target air-fuel ratio. This is because there will be a gap between BCs. Moreover, considering that the deviation between ACs is sufficient when the light fuel is used for the standard fuel specification with A as the target air-fuel ratio, the degree of deterioration is rather large in the specification with B as the target air-fuel ratio. Has become.

On the other hand, recently, a sensor (wide-range air-fuel ratio sensor) having a linear characteristic for a wide range of air-fuel ratios, not limited to the theoretical air-fuel ratio, has been developed. Alternatively, the lean air-fuel ratio can be accurately detected. Therefore, if this sensor is used to detect the exhaust air-fuel ratio at the time of transition and this detected value is used as a feedback control signal, it is possible to deal with the difference in the fuel property (volatility). Is required to calculate a correction amount (correction amount related to fuel property) with considerable frequency, and the calculation time of the fuel injection amount and the ignition timing is increased, so that the fuel injection amount and the like cannot be given with good response. . Further, if a microcomputer having a high calculation speed is used to compensate for the responsiveness, the cost will increase.

The present invention has been made in view of such a conventional problem, and introduces a correction amount related to volatility, and calculates the correction amount when fuel is supplied and according to the fuel property. An object of the present invention is to provide a control device in which a changing factor is limited to within an allowable range.

(Means for Solving Problems) In the present invention, as shown in FIG. 1, means 21 for calculating a control amount (air-fuel ratio or ignition timing) involved in combustion based on a detected value of engine operating conditions, Means for detecting the remaining amount of fuel 22
A means 23 for determining whether or not refueling is performed based on the detected amount, a means 24 for starting correction of the control amount when it is determined that refueling is performed, and a factor that changes depending on the difference in volatility. (For example, the specific gravity of the fuel, the occurrence of engine stall, the period when the transient air-fuel ratio fluctuates beyond the allowable range of the target air-fuel ratio, the difference between the supplied fuel amount and the combusted fuel amount during the transition) The detected value exceeds the allowable range. Means 26 for determining whether or not, means 27 for correcting the control amount so as to be within the permissible range when the permissible range is exceeded, and a correction amount when the permissible range is within the permissible range (for example, KF for the air-fuel ratio, Ignition timing K
The means 28 for continuously using A) until the next refueling determination is provided. Note that 25 is a means for detecting a factor that changes depending on the difference in volatility.

(Function) Whether the standard fuel changes to heavy fuel or light fuel due to the difference in commercially available fuel, this is judged, and the correction amount for the combustion related control amount is calculated according to the changed heavy fuel or light fuel. It Since it is not necessary to assume the fuel that will be used here, it is possible to avoid a situation in which exhaust emission and drivability become poor due to the use of unscheduled fuel.

In addition, the correction amount is calculated only when refueling is determined and when the factor is not within the allowable range, and when it is within the allowable range, the correction amount at that time continues until the next refueling determination. used. In other words, after the correction amount is within the allowable range, it is only necessary to read out the correction amount, so the calculation time of the fuel injection amount and ignition timing is relatively shortened by that amount, which makes the responsiveness even during transient correction. Fully enhanced.

(Embodiment) FIG. 2 is a first embodiment of the present invention and is a system diagram applied to a fuel injection engine in which a throttle valve opening and an engine speed are used as basic values of operating variables. In the figure, the fuel supply system comprises a fuel tank 33, a fuel supply passage 34, a fuel pump 35, a damper 36, a pressure regulator 37 and a fuel return passage 38.

Here, the fuel tank 33 is provided with a sensor for measuring the weight and volume of the fuel in the tank. Regarding the fuel weight, for example, four piezoelectric load cells 45 that are distorted in proportion to the weight and generate a voltage according to the distorted amount are prepared. As shown in FIGS. 3 and 4, If the fuel tank 33 has a floating structure by interposing between the flange 33A of the fuel tank 33 and the vehicle body 46, the total weight is supported by the four load cells 45, and therefore the total output from the load cells 45 is , The total weight including the tank weight is obtained. Therefore, the fuel weight (FJ) is measured by subtracting the tank weight from the total weight. Note that 33B is a rib.

Further, a fuel gauge (for example, capacitance type) 39 for detecting the remaining amount of fuel, that is, the fuel volume (FG2) is provided in the tank 33.
It is provided inside. The signal from the gauge 39 is also used to determine whether refueling has been performed.

On the other hand, the control system includes various sensors, a control unit 40 to which these signals are input, and a fuel injection valve 7 and an ignition plug 12 from which control signals from the control unit 40 are output. Specifically, a sensor 32 for detecting the intake throttle valve opening, a sensor (crank angle sensor) 6 for detecting a reference position and a unit angle of the engine crank angle, and a sensor 11 for detecting a cooling water temperature Tw of the engine are provided in each part of the engine. To be done. Here, the engine speed N is calculated from the unit angle signal of the crank angle, and the cylinder discrimination is made from both the unit angle signal and the reference position signal.

Further, a sensor 31 for detecting the air-fuel ratio of exhaust gas is attached to the exhaust pipe 8. The sensor 31 is a sensor (wide range air-fuel ratio sensor) having linear characteristics on both the rich side and the lean side of the theoretical air-fuel ratio. This detection signal is used as a signal for calculating the air-fuel ratio feedback correction coefficient α.

In the control unit 40, the fuel injection valve 7 and spark plug
Air-fuel ratio control and ignition timing control are performed with 12 as the control target.
Here, for the air-fuel ratio control, the fuel injection pulse width Ti applied to the fuel injection valve 7 is calculated by the basic equation Ti = Tp × COEF × α + Ts. In the equation, Tp is the pulse width corresponding to the injection amount for obtaining the basic air-fuel ratio, COEF is the sum of various correction factors (for example, the water temperature increase correction factor based on the cooling water temperature Tw), and α is the basic air-fuel ratio and the target. The air-fuel ratio feedback correction coefficient Ts, which has a meaning as a correction value for the deviation of the air-fuel ratio (for example, the theoretical air-fuel ratio), is an invalid pulse width based on the battery voltage V _B , and is conventionally known in the L-Getronic system. By the way.

By the way, since the above formula is for standard fuel, when heavy or light fuel with different volatility from standard fuel is used,
For example, during a transition, the air-fuel ratio greatly deviates from the stoichiometric air-fuel ratio and becomes lean or rich, and the startability becomes poor during a cold start. Therefore, in this example, a correction coefficient (KF) relating to volatility is introduced, and the injection amount (Ti) is corrected and calculated by this KF. That is, as shown in FIG. 5, Ti × KF is newly set as Ti during the transition and during the cold when Tw is lower than the predetermined value (Tw ₀ ) (steps 53 to 55).

The method of directly correcting Ti here is that the method of calculating Tp based on the throttle valve opening and the number of revolutions provides high calculation accuracy. This is because there is no problem. The correction format is not limited to the method of directly correcting Ti. For example, in the L- Jetronic system to basic value and the intake air quantity Qa and the engine speed N, rather than correcting the entire Ti by KF, acceleration increase correction as _{COEF = 1 + Kt + K A} cc × KF It is conceivable to correct this using the coefficient K _A cc as a basic value. However, in the same equation, Kt
Is the sum of the various correction coefficients other than K _A cc.

Similarly, because the difference in fuel property changes the required ignition advance value, a correction coefficient (KA) related to volatility is also introduced for the ignition advance value, and the ignition advance angle to be output to the ignition device at this KA. The value (ADV) is corrected and calculated. That is, as shown in FIG. 5, ADV + KA is replaced with ADV (step 53).
~ 55). The ADV for the standard fuel is a numerical value that represents the crank angle before the top dead center of the compression, and is an operating condition (for example, Tp and N when the intake throttle valve is open, N when the throttle valve is fully closed).
Needless to say, it is calculated according to (only).

Next, it is judged whether the volatility has changed from the standard fuel to the heavy fuel or the light fuel. In this example, it is based on the specific gravity of the fuel. This is because when the fuel is changed to the heavy fuel, the fuel specific gravity is correspondingly increased, and conversely, when the fuel is the light fuel, the fuel specific gravity is correspondingly reduced.

FIG. 6 is a routine for calculating the correction coefficients (KF and KA) relating to the air-fuel ratio and the ignition advance value in accordance with this determination and the determination result, which is started when the key switch or the door switch is turned on. The door switch is a switch that outputs an ON signal when the door is opened.

First, read the detection value (FG2) of the fuel gauge 41,
Difference from the previously read detection value (FG1) ΔFG (= FG2-FG
By comparing 1) with the reference value (FG0), ΔFG ≧ FG0
If so, it is determined that the fuel has been supplied (step 62). The reason for refueling is that the volatility of the fuel used cannot change unless refueled, and if the volatility changes, it depends on refueling. Also, it is sufficient to determine whether or not the volatility has changed, even if the fuel has been refueled, without having to determine it at a high frequency during operation. If the fuel has not been refueled, FG2 is stored in the memory as FG1 for the next determination (steps 62 and 71).

On the other hand, FG2 is the fuel volume, and this FG2 and the load cell 45
Fuel specific gravity in the tank from the fuel weight (FJ) obtained from
FR2 (= FJ / FG2) is calculated (steps 61, 63, 65). Since the FJ varies depending on the fuel temperature, the fuel temperature is corrected in step 64 so that the fuel weight becomes the fuel weight with respect to the reference fuel temperature. This improves the accuracy of fuel weight measurement.

Next, the difference ΔFR between FR2 and specific gravity (FR1) for standard fuel
It is checked whether (= FR2-FR1) is larger than a predetermined value (FR0), and if ΔFR ≧ FR0, the specific gravity is large, so it is determined that the fuel is heavy fuel. On the other hand, ΔFR
<FR0, proceed to step 69 and calculate the absolute value of the difference (| ΔFR
|) And FR0 are compared, and if | ΔFR | ≧ FR0, it is determined to be a light fuel.

Then, depending on the determination result, each correction coefficient for heavy fuel (KF _H and KA _H ) or each correction coefficient for light fuel (KF _L)
And KA _L ) are obtained by referring to the table (steps 66, 67, 6)
6,69,70). For example, in the cold state, as the heavy fraction increases, the amount of fuel lost as the wall flow increases, so for heavy fuel, the increase in the starting amount is increased as much as the wall flow increases. Also, each KF _H and KA _H is determined so that the ignition timing is advanced by the amount of the combustion delay. On the contrary, for the light fuel, it is necessary to reduce the increase in the starting amount by the amount of excess supply, and retard the amount by the amount of early combustion.
To determine the KF _L and KA _L. An example of KF with Tw as a parameter is shown in FIG.

Finally, each coefficient obtained by referring to the table (for air-fuel ratio
KF _H and KF _L , KA _H and KA _L regarding ignition advance values)
It is set to F and KA, and this is stored in the memory (nonvolatile memory) (steps 67, 68, 70, 68).

Then, these KF and KA are read from the memory in step 52 of FIG. 5, and the TiF for the standard fuel is read using this value.
And ADV is corrected.

Therefore, according to this example, whether the standard fuel changes to heavy fuel or light fuel, this is judged based on the fuel specific gravity, and the correction coefficient (KF and KA) according to the heavy fuel or light fuel is used. The controlled variables (Ti and ADV) for standard fuel are corrected (steps 66 to 68,52,55, 66,69,70,68,52,5
Five). It is not necessary to assume the fuel that will be used here, and since it can be applied to all fuels as far as the difference in volatility, it is possible to use unscheduled fuels in regions where commercial fuels are unknown. Even if this happens, a situation in which exhaust emission and drivability are poor is avoided. That is, the startability at the cold start, the drivability from cold to warm-up, and the startability at the hot restart are improved.

The calculation of KF is limited to the case where refueling is determined (step 62), and since the calculation is completed only once, the value (constant value) stored in the memory is read and used after the calculation. To be In this state, KF need only be read out from the memory, so the calculation time of the injection amount is relatively shortened by that amount, which allows sufficient responsiveness. Therefore, it is possible to maintain high control accuracy even during a transition.

The factors that detect the change in volatility include changes in the transient air-fuel ratio, backfire, complete explosion time at cold start, and engine stall, which will be described later, and the transient air-fuel ratio exceeds the allowable fluctuation range. There is a difference between the amount of fuel supplied and the amount of fuel burned during a certain period. All of these factors are matters to be detected only after burning, but there is an advantage that the specific gravity of fuel can be detected without burning.

Further, some of these factors are detected only during a transient operation.According to these matters, it may be possible to experience a problem in drivability before the transient operation. Since it is good, these problems do not occur.

Next, FIG. 8 and FIG. 9 show the second embodiment of the present invention.
Corresponds to the figure. However, in this example, the calculation of the correction coefficient for the ignition advance value is omitted. This also applies to the third to fifth embodiments described later.

Now, the difference from the first embodiment is the method of judging that the volatility has changed, and in this example, when the engine stalls, it is judged that the fuel is mainly heavy fuel (step 93). This is because most of the causes of engine stall are associated with lean misfires, so engine stall tends to occur with heavy fuel, which is largely depleted as wall flow. For example, the engine stall is determined by the engine speed N falling below a predetermined value and the exhaust air-fuel ratio being leaned above a predetermined value for a predetermined time (for example, 1 second or more). It should be noted that rich misfires are not taken into consideration. This is because the recent air-fuel ratio control tends to make the air-fuel ratio leaner for the purpose of improving fuel efficiency, and this case can be ignored because rich misfire cannot occur.

Then, when it is determined that the fuel is heavy fuel due to the engine stall, the cooling water temperature Tw is read, and KF is determined by referring to the table having the characteristics shown in FIG. 7 as in the first embodiment. Yes (steps 93,95-97,100).

However, stalling often occurs due to a driver's driving mistake, so it is necessary to deal with this case. Therefore, a flag (FLAGS) indicating whether or not the engine stalls for the first time and a flag (FLAGK) for counting the number of times the engine stall occurs are introduced. That is, in the case of the first engine stall (FLAGS = 0), the routine is ended by setting FLAGS after calculating the correction coefficient (steps 94 and 99). It should be noted that raising the flag means changing the value of the flag from "0" to "1", and lowering the flag means reversing the value. It will also be used in the following description.

If an engine stall occurs next time regardless of the correction,
This time FLAGS is standing, so proceed to step 101, FL
Until AGK becomes "2", the routine is finished by simply increasing the value of FAGK (step 102). Therefore, the occurrence of engine stall that occurs between the two times after correction is considered to be due to a driver's mistake, and FLAGS and FLAGK are lowered for the third time to prepare for the next engine stall occurrence determination (101,10).
3).

On the other hand, if the engine stall does not occur in the correction, has the flag (FLAGE) that counts the number of times the engine stall has not occurred to reach the allowable value KE (for example, 5 times for the number of times of operation including start) in order to check the situation for a while? If it reaches KE, a flag (FLAGOK) is set to indicate that the correction of the fuel oil has been completed (steps 104 to 106). With this, step 93 and thereafter are not executed until the next refueling determination (step 92).

In this example, the refueling determination routine is configured differently as shown in FIG. 8, and a flag (FLGC) indicating that refueling has been performed is set in this routine, and only when this flag is set is shown in FIG. The correction routine runs (steps 82, 83, 91). In step 83, another flag (FLAGOK
And FLAGK) is the initialization.

The second example also has the same effects as the first example in that the determination that the volatility has changed is not limited to the transient state.

Next, FIG. 10 and FIG. 11 are the third embodiment, and FIGS. 13 and 14 are the fourth embodiment. As in the second embodiment, the refueling determination routine has a different structure. That is, the symbols shown in FIG. 8 are used as they are in FIGS. 10 and 13.

Now, the first and second embodiments employ factors that are not affected by steady state or transient, but in the third and fourth examples, the volatility change is determined based on factors that cause a difference during the transient. In this sense, the determination is limited to the transient time, which is different from the first and second embodiments (step 123 in FIG. 11 and step 153 in FIG. 14). That is, in the third and fourth examples, the proportion of the wall flow changes due to the difference in volatility, and this change causes a difference in the deviation from the target air-fuel ratio during transition and the amount of fuel sucked into the cylinder. It pays attention to the coming point.

First, in the third example, it is determined that the volatility has changed based on the period (time) during which the transient air-fuel ratio exceeds the allowable fluctuation range. This is because when the standard fuel is used, the fluctuation range of the air-fuel ratio during acceleration is the allowable fluctuation range (lean side limit value AFL, rich side limit value AFR) set for the target air-fuel ratio as shown in Fig. 12. ), The heavy fuel and the light fuel become leaner or richer than the allowable fluctuation range as shown in the figure, so the time (TL and TR) during leaning or enriching is volatile, This is because there will be a correlation with drivability problems such as stumble and hesitation. It should be noted that measuring the time for leaning or enriching means indirectly detecting a defect in drivability.

Specifically, in FIG. 11, until the deviation ΔA / F (= exhaust air-fuel ratio-target air-fuel ratio) between the exhaust air-fuel ratio and the target air-fuel ratio exceeds the lean side limit value (AFL) and crosses the AFL again. Is counted up, and the count value (TL) is compared with the allowable value (LL) to determine that the fuel is heavy fuel if TL ≧ LL (steps 126 to 128, 131). Similarly, the counter value (TR) counted up while the absolute value of ΔA / F exceeds the limit value (AFR) on the rich side is compared with the allowable value (RR), and if TR ≧ RR It is determined that the fuel is light fuel (steps 127, 129, 130, 131, 133). Although AFL and AFR are obtained by referring to a table having Tw as a parameter (steps 124 and 125), it is desirable to narrow AFL.

After that, the correction coefficient (KF _H or HF _L ) for heavy or light fuel is read according to the determination result, and this is stored in the nonvolatile memory as KF (132, 134, 135). Also, ΔA
When / F falls within the limit value, FLAGOK is set and the vehicle is ready for the next refueling determination (steps 136 and 122).

According to this example, there is an advantage that the memory capacity required for determining the change in volatility can be reduced. For example, as shown in FIG. 18 described later, ΔA / F> AFL or | Δ
If A / F |> AFR is judged as the heavy or light fuel, ΔA / F over the entire transient time
However, the memory capacity increases accordingly, but in this example, it is possible to deal with the memory capacity in the period corresponding to TL and TR in the transient time.

Next, in the fourth embodiment, the difference in fuel property is determined based on the difference between the supplied fuel amount and the burned fuel amount. Here, the combustion fuel amount is the amount of fuel sucked into the cylinder (cylinder intake fuel amount).

A part of the fuel injected from the fuel injection valve 7 rides on the air flow or jumps directly into the cylinder, but the rest adheres to the intake port, intake valve or valve guide, and then the cooling water temperature, intake temperature or intake It is vaporized by the vaporization rate determined by a function such as negative pressure and is sucked into the cylinder. At this time, if the fuel has good fuel distilling properties (volatility), as shown in FIG. 15, more fuel is sucked into the cylinder even at the same temperature.

Therefore, if the fuel difference ΔF (= FC−FIN) between the fuel supply amount (FC) and the cylinder intake fuel amount (FIN) for each cylinder is calculated,
ΔF means the amount of fuel that has not been sucked into the cylinder out of the supplied fuel amount, and ΔF is small for light fuel and ΔF is large for heavy fuel. As a result, the characteristics of ΔF with respect to volatility are shown in Fig. 16, and in the same figure, if ΔF <DFL by comparing ΔF with the lower limit (DFL) or upper limit (DFU) of the allowable range, then light fuel If it has changed, and if ΔF> DFU, it can be determined that the fuel has become heavy fuel.

However, it is difficult to directly obtain the FIN. However, since the air-fuel ratio in the cylinder can be detected as the exhaust air-fuel ratio (AF), the cylinder intake air amount (AI
If N) is known, FIN can be easily obtained by AIN / AF. There is boost pressure in the AIN equivalent.

Such control operation is performed by the steps of the routine shown in FIG.
It is completed in 154-158,161,165.

However, in the figure, instead of ΔF, an average value (DF) of a predetermined number (for example, 10) of combustion cycles of ΔF is adopted. That is, ΔF is integrated until the counter value (CNTS) that counts the number of cycles reaches 10,
When it becomes 10, the DF which is also the integrated value is divided by 10
(Steps 159,169,170,159,160). The use of the average value takes into consideration the stability of control.

Next, FIGS. 17 to 19 show a fifth embodiment, which considers the fuel remaining in the fuel supply system when refueling is judged. It goes without saying that FIG. 17 corresponds to FIG. 10 and FIG. 13, and FIG. 18 corresponds to FIG. 11 and FIG.

That is, at the time of starting immediately after refueling, there is a predetermined amount of fuel (for example, about 150 cc) remaining in the fuel supply system from the fuel tank 33 to the injection valve 7 shown in FIG. Can be switched to. Therefore, it is sufficient to use the correction coefficient for the fuel before replenishment while the residual fuel is consumed, instead of immediately calculating the correction coefficient immediately after determining the refueling.

Therefore, a residual fuel consumption determination routine shown in FIG. 19 is newly added to check whether or not the residual fuel has been consumed in this routine. For example, fuel is injected after starting and the elapsed time is counted, and the count value (TM
When RRF becomes equal to or more than a predetermined value F (for example, 5 seconds), a flag (FLAGF) indicating the completion of consumption is set (step 2
12-214). Since the fuel injection from the injection valve is performed in synchronization with the engine rotation, the predetermined rotation may be performed instead of the predetermined time.

Then, in the correction routine shown in FIG.
Check F, step 192 while flag is down
Do not allow them to enter later (step 191).

As a result, the correction coefficient KF for the residual fuel is used as it is until the residual fuel before refueling is exhausted from the upstream side of the injection valve, that is, until the newly refueled fuel is supplied to the engine. The optimum correction coefficient can be given to the fuel to be used.

Each of the routines of the above-described embodiment is a control unit 40.
This is a routine that is executed in the CPU when is configured by a microcomputer. The fuel injection device may be either a multi-point injection method or a single-point injection method.

(Effects of the Invention) As described above, in the present invention, it is determined whether or not the detected value of the factor that changes according to the difference in volatility exceeds the allowable range, and if it exceeds the allowable range, it falls within the allowable range. As described above, the control amount related to combustion is corrected, so that it is possible to apply to all fuels as far as the difference in volatility is concerned, and exhaust emission and drivability are poor due to the use of unplanned fuel. Is avoided.

The correction amount is calculated only when refueling is determined, and the correction amount when it falls within the allowable range is used continuously until the next refueling determination. And thus transient drivability and exhaust emissions can be optimized.

[Brief description of drawings]

FIG. 1 is a configuration diagram of the present invention, FIG. 2 is a system diagram of a fuel supply system and a control system of a first embodiment of the present invention, and FIGS. 3 and 4 are front views of a fuel tank of this embodiment. Plan views, FIGS. 5 and 6 are flow charts showing the calculation contents of this embodiment, FIG. 7 is a diagram showing the characteristics of the correction coefficient KF of this embodiment, FIGS. 8 and 9
The drawing is a flow chart showing the contents of calculation in the second embodiment of the present invention. 10 and 11 are flow charts showing the contents of calculation of the third embodiment of the present invention, FIG. 12 is a waveform chart showing changes in the air-fuel ratio during transition,
FIG. 13 and FIG. 14 are flow charts showing calculation contents of the fourth embodiment,
FIG. 15 is a characteristic diagram showing the distillation rate with respect to the distillation temperature, and FIG. 16 is a characteristic diagram of the fuel difference ΔF in this embodiment. FIGS. 17 to 19 are flow charts showing the calculation contents of the fifth embodiment of the present invention, FIG. 20 is a system diagram of a conventional example, and FIG. 21 is a characteristic diagram showing the conversion rate of a three-way catalyst. 6 ... Crank angle sensor, 7 ... Fuel injection valve, 11 ... Water temperature sensor, 12 ... Spark plug, 21 ... Combustion-related control amount calculation means, 22 ... Remaining amount detection means, 23 ... Refueling determination means ,twenty four
...... Compensation start means, 25 …… Change factor detection means, 26 …… Decision means, 27 …… Correction means, 28 …… Correction amount continuous use means,
31 …… Wide range air-fuel ratio sensor, 32 …… Throttle valve opening sensor,
33 …… Fuel tank, 39 …… Fuel gauge, 40 …… Control unit, 45 …… Piezoelectric load cell.

Claims (1)

[Claims] 1. A means for calculating a control amount involved in combustion based on a detected value of an engine operating condition, a means for detecting a remaining amount of fuel, and a means for judging whether or not refueling is performed based on the detected value. , Means for starting correction of the control amount when it is determined that fuel has been refilled, means for determining whether or not a detected value of a factor that changes according to a difference in volatility exceeds an allowable range, and an allowable range When the value exceeds the allowable range, a means for correcting the control amount so as to be within the allowable range and a means for continuously using the correction amount when the allowable range falls within the allowable range until the next refueling determination are provided. A control device for an internal combustion engine.