JPH0749054A - Fuel injection controller - Google Patents

Abstract

PURPOSE:To maintain combustion stability and improve detection system by standardizing the value of a parameter for showing the combustion state in its value combustion state to be considered a standard, in the case of judgement of deterioration in a combustion state by a parameter showing the combustion state of an engine. CONSTITUTION:Rotational angle speed in extremely small rotation is measured by an angle signal of a cam shaft enclosed in a distributer 16, and a signal for detecting rotation of a crank shaft, in a control unit 15. A combustion stability index is calculated from the rotational angle speed. It is judged whether engine part (parts) has (have) some trouble or not, and when there is no trouble it is judged whether it is during lean operation or not, if combustion stability is accurately evaluated by the acceleration time or starting time of the engine. A learning value for judging deterioration of combustion stability is renewed when it is not during the lean operation. Weight is added to difference between the combustion stability index and a conventional learning value, and this value is added to the conventional learning value so as to reneas it. Dispersion of the combustion condition of the engine is detected/corrected by this repetition, and thereby a required combustion state can be obtained.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a fuel injection device for an internal combustion engine, and more particularly to a fuel injection device for an internal combustion engine that corrects a combustion state to a desired state.

[0002]

2. Description of the Related Art In the prior art, for example, JP-A-58-2177 is used.
As described in Japanese Patent No. 32, the engine rotation angular velocity is detected, and when the variation in the angular velocity is large, ignition and fuel control parameters are corrected to improve combustion.

[0003]

In the above-mentioned prior art, it is premised that the detection accuracy of the rotation information of the sensor for detecting the rotation is sufficiently higher than the accuracy required for control. There was no consideration of how to deal with variations in detection accuracy.

[0004]

The above-mentioned object is achieved by learning a rotation information detection error due to variations in individual rotation detection sensors and correcting a parameter representing combustion stability according to the value.

[0005]

The deterioration of the combustion state is determined by the parameter indicating the combustion state of the engine, using the value of the parameter indicating the combustion state in the reference combustion state as the reference value.

[0006]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS A fuel injection control device for an internal combustion engine according to the present invention will be described in detail below with reference to the illustrated embodiments.

FIG. 2 shows an example of an engine system applied to the present invention. In the figure, the air to be taken in by the engine is taken in from an inlet portion 2 of an air cleaner 1, and a throttle valve for controlling the intake flow rate is housed. Throttle valve body 5
Pass through and enter collector 6. Then, the intake air is distributed to each intake pipe 8 connected to each cylinder of the engine 7 and guided into the cylinder.

On the other hand, fuel such as gasoline is sucked and pressurized by a fuel pump 10 from a fuel tank 9, and then a fuel damper 11, a fuel filter 12, a fuel injection valve (injector) 13, and a fuel pressure regulator 14 are connected to a pipe. Is supplied to the fuel system being operated. Then, this fuel is regulated to a constant pressure by the above fuel pressure regulator 14,
Fuel is injected into the intake pipe 8 from a fuel injection valve 13 provided in the intake pipe 8 of each cylinder.

A signal representing the intake air flow rate is output from the air flow meter 3 and input to the control unit 15.

Further, a throttle sensor 18 for detecting the opening degree of the throttle valve 5 is attached to the throttle valve body 5, and its output is also input to the control unit 15. Next, 16 is a distro (distributor)
A crank angle sensor is built in this distort, and a reference angle signal REF representing the rotational position of the crankshaft and an angle signal POS for detecting the rotational speed (rotation speed) are output. These signals are also output from the control unit. 15 is input.

Reference numeral 20 denotes an A / F sensor provided in the exhaust pipe, which detects whether the desired air-fuel ratio is rich or thin in order to detect the actual operating air-fuel ratio. Is also input to the control unit 15.

Main parts of the control unit 15 are MPU 15A, ROM 15B, RAM 1 as shown in FIG.
It consists of 5C and I / O LSI 15D, takes in signals from various sensors that detect the operating state of the engine as input, executes predetermined arithmetic processing in MPU 15A, and outputs various control signals calculated as the arithmetic result. Then
A predetermined control signal is supplied to the above-mentioned fuel injection valve 13 and ignition coil 17 to perform fuel supply amount control and ignition timing control.

In such an engine, when the air-fuel ratio of the intake air-fuel mixture is set leaner than the stoichiometric air-fuel ratio, the characteristics shown in FIG. 5 are obtained. When the air-fuel ratio is made lean while keeping the torque and the engine speed constant, the pumping loss is reduced because the intake air amount is increased to maintain the torque, and the fuel consumption rate is improved because the specific heat ratio is improved.
On the other hand, the NOx emission concentration decreases without increasing the combustion temperature as the air-fuel ratio becomes lean. Further, the combustion stability which can be quantitatively grasped by the torque fluctuation is gradually deteriorated to a certain lean air-fuel ratio because the ignitability of the air-fuel mixture is deteriorated due to the lean air-fuel ratio, and suddenly deteriorates when it exceeds that. As described above, the combustion stability and NOx emission concentration in the lean region largely depend on the air-fuel ratio.

During operation at a lean air-fuel ratio, the crankshaft rotation is sucked and compressed by an angle signal of the camshaft built in the distort 16 or a signal for directly detecting the rotation of the crankshaft, for example, rotation detection in the ring gear portion. ， Measure in a sufficiently short time for the stroke of explosion and exhaust, and measure the rotational angular velocity in minute rotations. The rotational angular velocity in each phase fluctuates as shown in FIG. 8 depending on the stroke of the engine. The fluctuation of the rotational angular velocity is analyzed mainly because the explosive force in the explosion stroke of each cylinder is the main source. By doing so, it is possible to know the combustion state of the engine.

However, the above rotation detection sensor outputs rotation information having an error with respect to the true rotation of the crankshaft due to variations between the rotation detection sensor itself and the individuals existing in the rotation transmission path. Therefore, the combustion stability index P calculated from the rotation information has a different relationship with the combustion stability depending on the individual as shown in FIG. Here, the combustion stability index P
The magnitude of the influence of the rotation information error on P is always constant regardless of the combustion stability, and therefore depends only on that individual. As a result, the combustion stability of each individual and the combustion stability index P
The relationship with is a parallel movement relationship as shown in FIG.
The inclination does not vary depending on the individual.

Therefore, the stable operation state in which the combustion stability is considered to be substantially constant is set as the reference position, and the combustion stability index P at that point is used as the reference for the combustion deterioration determination. That is, as shown in FIG. 6, the combustion stability index P at the reference position is stored as a learning value D for the combustion deterioration determination, and the combustion deterioration determination is performed by comparing the combustion stability index P with the learning value D + slice level S. By. As a result, the variation in the relationship between the combustion stability of each individual and the combustion stability index P can be corrected, and the true deterioration of combustion can be accurately determined. According to the result of the determination, for example, in the case of combustion deterioration due to lean burn, the correction is performed in the rich direction when the combustion stability is poor, and in the lean direction when the combustion stability is good. This allows
A desired combustion state can be obtained.

FIG. 1 shows an embodiment in which the above processing is represented by a schematic processing flow. In this embodiment, step 10
1, 102 calculates the combustion stability index P from the rotational angular velocity.

Next, if there is a failure in an engine component, combustion stability cannot be ensured.
The process ends without performing the following determination process. Next, under the condition that the combustion stability cannot be accurately evaluated, for example, during acceleration or engine start, in step 104, the processing is ended without performing the following processing. The judgment material for this judgment may be engine speed, engine water temperature, vehicle speed, engine load, starter motor operation signal, throttle valve opening, gear position, and the like.

Then, in step 105, it is determined whether or not the combustion stability is determined, and in the present embodiment, the lean operation is being performed. If lean operation is in progress, combustion stability evaluation separately shown in step 108 is performed. When the lean operation is not performed, the routine proceeds to step 106, and in order to obtain the learning value D for determining the deterioration of the combustion stability, it is first determined whether or not the learning condition is satisfied. The learning value D is learned in an operating region where the engine operating conditions are stable and a constant and stable combustion stability can always be obtained, so the operating conditions that satisfy the conditions are determined. Specifically, the judgment is made by the judgment material as shown in step 104, but the judgment condition is different from that in step 104. For example, there are specific operating conditions that show a constant combustion stability, such as when the engine is in a no-load operating state such as idling, when the engine speed and the load are within certain conditions.
Alternatively, in the fuel cut state, combustion is not occurring in the engine, so the combustion stability is 0. By learning the combustion stability index P at this time, it is possible to accurately absorb variations due to individuals. In addition, it is advisable to add a time judgment condition for condition stabilization to the judgment.

Next, in step 107, the learning value D is updated. In this embodiment, the combustion stability index P at that point is
Is multiplied by a weight W and added to the conventionally learned value D to update.
By repeating this operation, the learning value D becomes 1
It becomes equal to the combustion stability index P under the operating condition determined in 06, and the learning convergence ends. By changing the weight W according to the magnitude of the difference between the combustion stability index P and the learning value D, the weight W is quickly converged and is less likely to diverge.

Since the learning convergence in step 107 serves as a criterion for determining the deterioration of combustion during lean operation, prohibiting the lean operation until the convergence of learning prevents combustion deterioration. It is valid. In particular,
The means for counting the number of times learning is established in step 107 and prohibiting lean operation until a predetermined number of times, or prohibiting lean operation until the difference between the combustion stability index P and the learned value D is within a predetermined value is provided. Conceivable.

Further, by storing the learning value D in a non-volatile memory, the result of convergence once obtained can be utilized thereafter, and the frequency of lean operation prohibition can be reduced.

The above processing is an example in which the combustion stability index is one parameter, but in the case of a multi-cylinder engine, the combustion stability is calculated for each cylinder and the above flow is performed for each cylinder. More precise control is possible.

Next, step 108 of FIG. 1, the combustion stability evaluation correction routine will be described in detail with reference to FIG. In step 110, first, a learning value D, which is a reference for comparison, is searched for by the rotation speed and load in the corresponding operating state. In this embodiment, the combustion stability learning value D shown in FIG.
It is learned as D ₁₁ and D ₁₂ for each operating region as shown in FIG. This is because when the combustion stability differs depending on the operating region in the standard operating state, it is necessary to distinguish and learn this, so the operating region is subdivided according to the engine speed and load, and the learning value is independently set for each region. This is an example in which D is provided. Here, the rotation speed and load are used to define the engine operating range in order to deal with an example in which combustion stability can be accurately reproduced by classification of this parameter. If there are parameters that are effective in identifying combustion stability, such as
The learning value may be searched using the value.

Next, at step 111, the slice level S used for evaluation of the combustion stability under the relevant operating condition.
To search. This is a control for coping with the difference in the margin to the allowable upper limit of the combustion stability because the reference combustion stability differs in each operating state. The D and S retrieved in steps 110 and 111, respectively, are compared with the combustion stability index P in steps 112 and 114. At step 112, it is judged if the combustion stability index P is larger than the sum of D and S, that is, if the combustion stability is worse than the allowable value, and if so, it is leaner than the desired air-fuel ratio, so at step 113. Enrich process. On the contrary, in step 114, the combustion stability is better than the allowable value,
In FIG. 5, it is determined when the rich side limit of the lean air-fuel ratio operation range is exceeded, and if so, the air-fuel ratio is rich, so the leaning process is performed in step 115. Here, the criterion in step 114 is not the sum of D and S in step 112, but it is necessary to obtain the combustion stability when the air-fuel ratio does not exceed the allowable upper limit of the NOx concentration. A value obtained by subtracting the value Z serves as a criterion.

By repeating this process, the air-fuel ratio can be introduced into the lean air-fuel ratio operation range shown in FIG.

Here, in the embodiment shown in FIG. 4, the learning value D
Is required to be sufficiently learned under the operating conditions for evaluating the combustion stability. Therefore, a method of estimating the learning value D so that the combustion stability can be evaluated even in a region where learning has not progressed sufficiently will be described with reference to FIG. 7. FIG. 7 is an example showing each learning value assigned to the operating region and the distribution of combustion stability in each operating state. Here, D ₂₂ and D ₃₂ are almost equal combustion stability. Therefore, when a reliable learning value is obtained in either of the two regions, the combustion stability can be evaluated under two operating conditions by using the same learning value for the other learning value. Further, when the relative difference between the learned values in each operating region is known in advance, when the learning progresses sufficiently under one operating condition, the learned value in each operating region can be estimated according to the value. The specific processing flow is shown in FIG. First, the process is started when learning is established in any of the learning areas. This is step 130. Next, in the learning at that time, it is determined in step 131 whether the learning has progressed sufficiently. Specifically, the number of times learning is established is a predetermined value or more, the difference between the combustion stability index P and the learning value D is within a predetermined value, and the like. When learning has not progressed sufficiently, other learning regions cannot be estimated, and the process ends. If learning is in progress, the process proceeds to step 132. Here, an area other than the area where learning is established is selected, and the next step 133
Then, the learning progress state in the area is determined. When the learning is progressing sufficiently, it is not necessary to estimate the learning value of the area, and therefore, step 134 and subsequent steps are not executed. If the learning progress is insufficient, in step 134, step 13
The relative difference between the learning value and the region where learning is established at 0 is searched. In Step 135, Step 1
Based on the relative difference obtained in step 34 and the learning value obtained in step 130, the learning value in the area where the learning value is estimated is calculated and written.
Next, in step 136, it is checked whether or not the above processing has been performed on all learning regions. If not completed, the processing of step 132 and subsequent steps is repeated. By this processing, a reliable learning value can be obtained even in a region where learning has not progressed sufficiently, and the combustion stability can be determined in a wide operating region.

FIG. 9 shows an embodiment of a combustion stability evaluation and correction routine when the learning value D is learned at the time of fuel cut. In this case, the learning value is one of D _FCUT because the operating condition for learning is one condition during fuel cut. Since D _FCUT is a combustion stability index when the combustion stability is 0, it represents an offset amount of the combustion stability index P of each individual. Therefore, from the combustion stability index P to D _FCUT
By subtracting, the offset amount for each individual is absorbed,
The value can be used to determine the combustion stability. Therefore, in step 120, D _FCUT is subtracted from P and the value is set to P _REAL . Next, in step 121, it is compared with the slice level S1 which is the upper limit of combustion stability. When _PREAL exceeds this, the air-fuel ratio is made rich so that the combustion stability is within the allowable value. In step 123, the air-fuel ratio is made lean so that the combustion stability is within a predetermined value as compared with the slice level S2 of the combustion stability lower limit. The basic principle of the series of processes of this embodiment is the same as the principle of the embodiment shown in FIG. 4, and by repeating this process, it is possible to bring the combustion stability within a desired range. Further, when the learning condition of the learning value D is limited to one place such as at the time of idling, the same process as this process is performed.

Further, in the above-mentioned embodiment, the method of learning the combustion stability index for each individual is used. However, regarding the method of correcting the input from the sensor for the calculation of the combustion stability, the combustion stability is determined by the engine rotation. An example of evaluating will be described below.

The rotation of the engine fluctuates in synchronization with the stroke of each cylinder as shown in FIG. 8, but the fluctuation of the rotational angular velocity is mainly caused by the explosive force in the explosive stroke of each cylinder. As mentioned above, it is possible to know the combustion state of the engine by analyzing the fluctuation. Therefore, the rotational angular velocity is measured in a sufficiently short time for the stroke of the engine, and the combustion stability index is calculated. Specifically, a sensor with markings at the angular intervals you want to measure is attached to the crankshaft or camshaft that represents the rotation of the engine, and the displacement of the rotating shaft is detected by the output of the detection unit of the sensor that detects the passage of the mark. know. The rotational angular velocity is obtained by measuring the time between two or more markings. Here, since it is impossible to apply the marking position without error, the measurement result of the rotational angular velocity includes an error,
The value of the error depends on the individual. Furthermore, there are errors that occur irregularly due to backlash and the like that exist in the rotating system.

FIG. 11 shows an example of the engine speed measured in such a measuring system. The horizontal axis is time,
TR _i-2 , TR _i-1 , and TR _i will be described in detail in the explanation of FIG. 12, but they are the average values of the engine speed correction values calculated by the angular displacement at each time, and therefore occur irregularly. The error to make is almost removed. Here, since the timing for calculating each average value is a short time, the change in angular acceleration during that time is limited to a certain range. Therefore, T
The inclination between R _i-2 and TR _i-1 , that is, the angular acceleration is TR _i-1 , T _i.
It is also held to some extent in R _i . Explaining with reference to the figure, there is an expected value TI _i of TR _i on the extension of the slopes of TR _i-2 and TR _i-1 , and TR _i is TI if there is no error.
It is within the range of the broken line centered on _i . Here, the slopes of the broken lines are the maximum and minimum values of the change in angular acceleration that can occur between TR _i-1 and TR _i . Therefore, the measured value TR
_{When i} is outside the range of the broken line as shown in the figure, it can be said that the measurement of TR _i includes an error due to individual difference. Since the magnitude of the error can be grasped by the magnitude of the deviation from the broken line range, the correction coefficient can be learned.

FIG. 1 shows an embodiment of a processing flow for concretely executing the above method for absorbing individual variations in the rotation measurement system.
2 shows. First, in step 140, the position i of the crank angle displacement to be corrected is recognized. Next, Step 14
In step 1, the required time T _i between markings in the process is measured. In step 142, the measured T _i is multiplied by the learning value KCO that absorbs the individual variation to obtain TR _i . Before learning, KCO is 1. Next, in step 143, it is determined whether or not the conditions are such that the subsequent learning process can be performed. In order to learn, it is necessary for the operating state of the engine to be stable, and specifically, there are things such as that the engine is not being started and that large acceleration / deceleration is not being performed. If this condition is satisfied, the process proceeds to step 144, an average value TR _i of TR _i. In the present embodiment, a weighted average that uses less memory is used to calculate the average value. By this processing, irregular errors can be almost eliminated. In step 145, it is determined whether the averaging process in step 144 has obtained a sufficient population and has a reliable average value. When the average value becomes a reliable value, the process proceeds to step 146, and the average values TR _i-2 and T of the required times of one and two times before.
The expected value TI _i of TR _i is calculated using R _i-1 . In the present embodiment, the prediction is obtained by linear interpolation, but the number of average values to be used, the order of complement, and the method may be appropriately selected according to the required accuracy. The average value TR _{i of the} predicted value TI _i and the actual measurement value
The correction amount ΔKCO of the learning value KCO is obtained based on the difference of In the present embodiment, ΔKCO is obtained by searching the table in the figure using the ratio TR _i / TI _i of two values as a parameter. When the two values are the same or the difference is small, in other words, when TR _i is within the broken line in FIG. 11, the KCO used at that time is considered to be correct.
Does not need to be corrected, and ΔKCO becomes 0. Also, 2
If the difference between the two values is large, the value of KCO is not appropriate, so the table is set to search for ΔKCO such that TR _i approaches TI _i . △ KC obtained in this way
The O is used to modify the KCO in step 148 and the process ends. By performing this operation, for example, each time T _i is measured, it is possible to learn the KCO that has absorbed individual variations.

In the above description, an example in which the parameter of combustion stability is calculated on the basis of the rotational angular velocity is shown. However, other engine parameters such as combustion pressure in the cylinder, vibration of the cylinder block, and ignition discharge state change. The same effect can be obtained even if it is performed based on the above.

Further, in the above description, the air-fuel ratio is operated by lean operation, but a method of operating the EGR amount, the intake air amount, and the ignition timing may be considered.

When a means for quantitatively detecting the exhaust air-fuel ratio is provided, the output of the means for detecting the exhaust air-fuel ratio is corrected by the air-fuel ratio in the desired combustion state obtained by the present invention, and the means is provided. It is also effective to absorb the individual variation of.

[0036]

According to the present invention, it is possible to detect and correct the variation in the combustion state of the engine and set the combustion state to a desired state.

[Brief description of drawings]

FIG. 1 is a flowchart showing an embodiment of the present invention.

FIG. 2 is a diagram showing an embodiment of the present invention.

FIG. 3 is a diagram showing an embodiment of the present invention.

FIG. 4 is a flowchart showing an embodiment of the present invention.

FIG. 5 is a diagram showing a relationship between an air-fuel ratio and engine performance.

FIG. 6 is a diagram showing characteristics of a combustion stability index.

FIG. 7 shows an example of the relationship between the operating region and combustion stability.

FIG. 8 is a diagram showing the behavior of the rotational angular velocity of the engine.

FIG. 9 is a flowchart showing an embodiment of the present invention.

FIG. 10 is a flowchart showing an embodiment of the present invention.

FIG. 11 is an example of an experimental result according to an embodiment of the present invention.

FIG. 12 is a flowchart showing an embodiment of the present invention.

[Explanation of symbols]

3 ... Air flow sensor, 5 ... Throttle valve, 7 ... Engine, 13
… Fuel injection valve.

 ─────────────────────────────────────────────────── ─── Continuation of the front page (72) Inventor Hiroshi Kimura 7-1, 1-1 Omika-cho, Hitachi-shi, Ibaraki Hitachi Ltd. Hitachi Research Laboratory (72) Inventor Kimio Hoshi 2520, Takada, Katsuta-shi, Ibaraki Address Automobile Technology Research Association

Claims (5)

[Claims] 1. In a fuel injection control device for quantitatively detecting a combustion state of an engine and performing a correction for improving combustion, a combustion state serving as a reference for determining deterioration of the combustion state by a parameter representing the combustion state of the engine. The fuel injection control device is characterized in that the value of a parameter representing the combustion state in step S1 is used as a reference value. 2. The fuel injection control device according to claim 1, wherein a plurality of memories for storing the values of parameters representing the combustion state in the reference combustion state are provided according to the operating state of the internal combustion engine. . 3. The fuel injection control device according to claim 1, wherein the reference combustion state is a fuel cut state. 4. The fuel injection control device according to claim 1, wherein the reference combustion state is a no-load operation state. 5. A fuel injection control device for quantitatively detecting a combustion state of an engine based on rotation information of the engine to perform correction for improving combustion, and a sensor for detecting rotation when the engine is in a predetermined operating state. A fuel injection control device characterized by calibrating the information accuracy of.