JP3538874B2 - Internal combustion engine stability control device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a stability control apparatus for an internal combustion engine, and more particularly to a technique for improving control accuracy of stabilization control based on detection of engine speed fluctuation.

[0002]

2. Description of the Related Art Conventionally, as a stability control technique for an internal combustion engine, there is one disclosed in, for example, JP-A-51-104130 and JP-A-60-128947. These devices detect the fluctuation of the rotation of the engine, monitor the stability of the engine based on the rotation fluctuation value, and control the air-fuel ratio.

[0003]

However, such a conventional stability control technique based on engine speed fluctuation has the following problems. That is, there is an individual difference between the sensor for detecting the rotation of the engine and the engine itself. For this reason, even if the rotation fluctuations are actually at the same level, there is a possibility that the level of the rotation detection signal may vary greatly due to individual differences between the sensors and the engine itself. There is a problem that the fluctuations in the signal level may cause erroneous detection of the engine stability and unnecessarily increase or decrease the air-fuel ratio, thereby deteriorating the exhaust performance or operability. Was.

The present invention has been made in view of such a conventional problem. By learning and correcting a target level of stability, an individual difference between a sensor and an engine itself is absorbed, and the stability of the engine is improved. It is a feature of the present invention to provide a stability control device for an internal combustion engine that enhances control accuracy of control.

[0005]

Therefore, in the internal combustion engine stability control apparatus according to the present invention, as shown in FIG. 1, the engine speed is detected based on the detected value of the engine speed detector and the engine speed detector. Rotation fluctuation calculation means for calculating a rotation fluctuation value of the engine, and a correction amount of a control amount related to combustion stabilization of an engine control factor such that the rotation fluctuation value calculated by the rotation fluctuation calculation means approaches a stability target value. Correction amount calculating means, a control amount calculating means for calculating a control amount of the engine control factor based on the correction amount calculated by the correction amount calculating means, and a control amount calculated by the control amount calculating means. Control means for controlling combustion of the engine under lean conditions, learning value calculation means for calculating a learning value of the rotation fluctuation value calculated by the rotation fluctuation calculation means under operating conditions which are not lean conditions , and learning value calculation means. Write the calculated learning value Rewritable learning value storage means, learning value update means for rewriting the learning value stored in the learning value storage means with the learning value calculated by the learning value calculation means, and learning value storage means stored in the learning value storage means. And a stability target value setting means for setting the stability target value based on the learning value.

The stability target value setting means calculates the stability target value by correcting the learning value stored in the learning value storage means with different correction values according to the gear position. Good to do. The learning value calculation means may be configured to calculate a learning value during a stoichiometric air-fuel ratio operation.

The learning value calculation means compares the rotation fluctuation value calculated by the rotation fluctuation calculation means with a predetermined value, and the comparison result of the comparison means turns the rotation value from the predetermined value. When the fluctuation value is large, it is preferable to include a determination unit that determines that the engine is abnormal and prohibits the calculation of the learning value.

[0008]

According to this structure, the operating condition is not a lean condition.
By learning the rotational fluctuation value in the case , it is possible to absorb variations in individual differences between sensors and the engine itself, and to set a stability target value based on the learning value, so that appropriate stability according to each internal combustion engine A target value can be set, and erroneous detection of stability can be prevented.

The lower the gear position is, the greater the increase / decrease of the rotation speed and the speed of change with respect to the rotation fluctuation on the wheel side. The learning value of the fluctuation is corrected by making the correction value different, or the learning of the rotation fluctuation is performed during the stoichiometric air-fuel ratio operation in which the engine stability is good, or the calculated rotation fluctuation value is set in advance. By comparing with a predetermined value, when the rotation fluctuation value is larger than the predetermined value, it is determined that the engine is abnormal and the calculation of the learning value is prohibited, so that the calculation accuracy and reliability of the learning value can be improved. Become.

[0010]

Embodiments of the present invention will be described below with reference to the drawings. FIG. 2 is a diagram showing a system configuration according to one embodiment of the present invention. In FIG. 2, an engine body 1 includes an intake passage 2
And the exhaust passage 3 are connected. The intake passage 2 is provided with an air flow meter 4 for detecting an intake air flow rate, a throttle valve 5, and a fuel injection valve 6. The exhaust passage 3 is provided with an O ₂ sensor 7 for detecting the oxygen concentration in the exhaust gas. 8 is a throttle sensor for detecting the opening of the throttle valve 5, 9 is a crank angle sensor as engine speed detecting means for detecting the engine speed, 10 is a water temperature sensor, and 11 is a gear position sensor for detecting the gear position of the transmission. It is.

The control unit 20 has a built-in microcomputer, and the engine rotation fluctuation detected based on the engine rotation detection signal from the crank angle sensor 9 is set based on a learning value of the rotation fluctuation described later. The stability control described later is executed so as to approach the degree target value. The control unit 20 includes a memory as a storage unit that freely stores the learning value data calculated during the stability control, and this memory corresponds to the learning value storage unit. And the control unit 20
As shown in the flowcharts to be described later, each function of the rotation fluctuation calculating means, the correction amount calculating means, the control amount calculating means, the control means, the learning value calculating means, the learning value updating means and the stability target value setting means is implemented by software. It is prepared for.

Next, the stability control operation of this embodiment will be described with reference to FIGS. FIG. 3 shows an engine control factor in the stability control of the present embodiment (the fuel amount in the present embodiment).
This is a routine for calculating a control amount related to the combustion stabilization, that is, a fuel-air ratio correction coefficient Dml, and is a function of the control amount calculating means. In this embodiment, this routine detects the engine speed based on a reference signal (REF signal) for each cylinder output from the crank angle sensor, and is executed at the reference signal cycle. When the engine speed is detected based on the angle signal from the crank angle sensor, it is executed in a time-synchronized manner.

First, in step 1 (indicated by S1 in the drawing and the same applies hereinafter), the cycle of a reference signal (REF signal) for each cylinder from the crank angle sensor 9 is read. Next, in step 2, the fluctuation of the engine speed is calculated based on the cycle of the reference signal read in step 1. This calculation method will be described later with reference to the flowchart of FIG.

In step 3, it is determined whether or not to perform feedback (F / B) control based on rotation fluctuation detection.
Also, it is determined whether or not to learn a rotation fluctuation value for setting a slice level SL as a stability target value. These determination methods are shown in the flowchart of FIG. 5 described later. In step 4, it is determined whether or not feedback control based on rotation fluctuation detection has been determined in step 3.
Otherwise, skip step 5 and proceed to step 6.

In step 5, feedback (F /
B) The correction rate, that is, the correction amount of the calculated fuel-air ratio correction coefficient Dml (stabilized fuel-air ratio correction coefficient Lldml) is calculated. Details will be described later with reference to a flowchart of FIG. In step 6, the lean flag Flean is set (Flean =
1) is determined, and if standing (Flean = 1), proceed to Step 7; if sleeping (Flean = 0), proceed to Step 8.

In step 7, the target fuel-air ratio correction coefficient Tdm
Calculate l by the following formula. Tdml = Mdml × Lldml Here, Mdml is obtained as a map fuel-air ratio as shown in FIG. Lldml is a stabilized fuel-air ratio correction coefficient as a correction amount of the control amount calculated in step 5 described above, and is a feedback correction rate obtained by detecting rotation fluctuation.

In step 8, the target air-fuel ratio correction coefficient Tdml is set to 1 and the stoichiometric air-fuel ratio is set by the determination in step 6 that the lean operation is prohibited. From the next step 9, a damper operation for switching the fuel-air ratio is performed. The purpose of this is to prevent a sudden change in torque and to optimize driving performance by gently switching the air-fuel ratio.

First, in step 9, the latest fuel-air ratio correction coefficient Dml currently stored and held is compared with the target fuel-air ratio correction coefficient Tdml set in step 7 or 8.
If Dml ≧ Tdml (YES), it is determined that the air-fuel ratio is richer than the target, and the routine proceeds to step 12, where Dml stored in the memory is replaced with the previous value Dml _n-1 , and the air-fuel ratio change in the preset lean direction is performed. A new Dml is obtained by subtracting the speed value Ddmll. Further, in step 13, the fuel-air ratio correction coefficient Dml is updated while limiting Dml so that Dml does not become less than Tdml, and the flow ends.

If Dml ≧ Tdml is not satisfied (NO),
It is determined that it is leaner than the target, the process proceeds to step 10, and the new Dml is obtained by replacing the Dml stored in the memory with the previous value Dml _n-1 and adding a preset air-fuel ratio change speed value Ddmlr in the rich direction. , And in step 11, D
While limiting Dml so that ml does not exceed Tdml,
The fuel-air ratio correction coefficient Dml is updated, and the flow ends.

The fuel-air ratio correction coefficient Dml thus obtained is used for calculating the fuel injection amount shown in the flowchart of FIG. Next, the rotation fluctuation calculation routine executed in step 2 of the flowchart of FIG. 3 will be described with reference to the flowchart of FIG. This routine is the function of the rotation fluctuation calculation means.

First, in step 21, an initial process, for example, a rotational speed conversion process for measuring an output interval (period) of a cylinder-by-cylinder reference signal every 180 ° for each cylinder and calculating an engine speed from the measured value is performed. Do. At step 22, the rotation signal 0.
Performs filter processing to remove the fifth-order component. Step 23
Then, a filtering process for removing a component of the rotation signal that changes according to the vehicle speed is performed. The above processing is performed in rotation synchronization.

Next, at step 24, band-pass filtering is performed, and the result is stored as rotation fluctuation calculation data Lljes, thereby ending this flow. Here, in the band-pass filter processing, for example, a filter represented by a transfer function represented by the following equation (1) may be converted into software of an ECU (engine control unit) and used.

[0023]

(Equation 1)

Next, the respective operations of the feedback (F / B) control judgment and the rotation fluctuation value learning judgment by the rotation fluctuation detection performed in step 3 in the above-mentioned flowchart of FIG. 3 will be described with reference to the flowchart of FIG. First, in step 31, it is determined whether or not the operating condition of the engine is a lean condition by using a lean condition determination executed in a background job of FIG. Execute rotation fluctuation value learning. The learning of the rotation fluctuation value is shown in a flowchart of FIG. In this learning flow, it is also determined whether or not the lean flag Flean is to be cleared (Flean = 0) based on the stability determination result. And during this learning, step 3
At 6, the F / B control is prohibited.

On the other hand, if the condition is a lean condition, the routine proceeds to step 32, where a lean flag Flean is set (Flean = 1), and in step 33, it is determined whether the air-fuel ratio is being switched. This determination is based on the fuel-air ratio correction coefficient D obtained in steps 9 to 13 in FIG.
If ml is equal to the target fuel-air ratio correction coefficient Tdml, it is determined that switching is not being performed, and if not, it is determined that switching is being performed. If it is determined that the switching is in progress, the F /
Jump to B control prohibition. If the switching is not being performed, the process proceeds to step 34, where the F / B control is performed, and the process ends.

Next, the F / B correction rate calculation routine of FIG. 6 will be described. This routine is the function of the correction amount calculating means. First, in step 41, calculation of a slice level SL as a stability target value is executed. This slice level calculation method will be described later with reference to a flowchart shown in FIG.

In step 42, it is determined whether or not the current operation area has been learned based on a learned flag FSL set in a learning flow of FIG. Where FS
L = 1 indicates that learning has been completed, and FSL = 0 indicates non-learning. Therefore, in this step, when FSL = 1, it is determined that learning has been completed, and the routine proceeds to step 43, where FSL =
If it is 0, it is determined that the learning has not been completed, and the present flow ends.

In step 43, the rotation fluctuation calculation data Lljes obtained by the rotation fluctuation calculation routine of FIG. 4 is sampled, and the number of samples is counted. In step 44, the number of samples L is set. This sample number L is, for example, as shown in FIG.
Is set based on the engine speed Ne by a table search as described above. In step 45, it is checked whether or not the number of samples is L, and if it is, in step 46, the total of the sample data is divided by the number of samples L to obtain
An average value S (= total value of sample data / number of samples L) of the sampled rotation fluctuation data is obtained.

Next, at step 47, the average value S calculated at step 46 is compared with the slice level SL calculated at step 41, and when S> SL, the routine proceeds to step 48, where the rotational fluctuation is assumed to be large and the stability is determined. The correction amount Dlldml of the stabilized fuel-air ratio correction coefficient Lldml is set in the direction of increasing the fuel density so as to approach the target value.
Then, contrary to step 48, the correction amount Dlldml is set in a direction to make the fuel thinner so as to approach the stability target value. Further, when S = SL, the correction amount Dlldml is left as it is because the stability is substantially the target stability.

Then, in step 50, the stabilized fuel-air ratio correction coefficient Lldml used in FIG. 3 is updated by the following equation. Lldml = Lldmln _-1 + Dlldml Here, Dlldml has a value of +/-, and is set by the table shown in FIG. 12 based on the difference between the average value S of the rotational fluctuation values and the slice level SL. Although omitted in the flow, Lldml is not less than 1.0 and a predetermined maximum value LL
The range is limited below DMMX #, and the flow ends.

Next, the fuel injection amount is calculated and output using the fuel-air ratio correction coefficient Dml calculated in the flow of FIG. 3 using the stabilized fuel-air ratio correction coefficient Lldml obtained as described above. FIG. 7 shows and explains the routine. This routine is the function of the control means. This routine is executed every 10 ms. First, in step 61, the target fuel-air ratio Tfbya is calculated by the following equation using the fuel-air ratio correction coefficient Dml.

Tfbya = Dml + Ktw + Kas where Ktw is a water temperature increase correction coefficient, and Kas is a post-start increase correction coefficient. Next, at step 62, the detection signal of the air flow meter 4 is A / D converted and linearized to obtain the intake air flow rate Q. In step 63, an average value Avtp for a predetermined time is obtained from a basic injection amount Tp (= K × Q / Ne, where K is a proportional constant) calculated based on the engine speed Ne and the intake air flow rate Q.

In step 64, the actual fuel injection amount Ti is obtained by the following equation. Ti = Avtp × Tfbya × Ktr × (α + αm) + Ts Here, Ktr is a transient correction coefficient, α is an air-fuel ratio feedback correction coefficient, αm is an air-fuel ratio learning correction coefficient, and Ts is an invalid pulse width. Next, in step 65, a fuel cut determination is made, and in step 66, it is determined whether or not a fuel cut condition is satisfied. If NO, the fuel injection amount Ti is stored in an output register in step 67, so that the injection timing is determined. Prepare for injection at If YES, Ts is set in step 68.
Is stored in the output register. Except for the calculation of Tfbya, the operation is the same as the conventional calculation processing of the fuel injection amount.

Next, the rotation fluctuation value learning routine will be described with reference to the flowchart of FIG. In step 71, it is determined whether or not a learning condition is satisfied. The learning condition is, for example, a predetermined operation region in a steady state. Here, if it is not a learning condition, this flow is ended, and if it is a learning condition, the process proceeds to step 72. In step 72, the current operating conditions are read.

In step 73, a weighted average value Sfild of the rotation fluctuation values up to now including the rotation fluctuation value Lljes calculated in the flow of FIG. 4 is calculated. In step 74, it is determined whether or not the weighted average value Sfild of the rotation fluctuation exceeds a predetermined value SLS set in advance. Here, when Sfild> SLS, the fluctuation of the engine is excessive, and it is determined that the engine is abnormal, and the routine proceeds to step 75, where the lean flag Flean is set to 0 and learning of steps 76 to 78 described later is not performed. . That is, when it is determined that the engine is abnormal, the lean operation (lean combustion operation) is prohibited, and the learning of the rotation fluctuation value is prohibited.
As a result, the stability of the engine is restored, and the reliability of the learning value is ensured. On the other hand, if Sfild ≦ SLS, the routine proceeds to step 76.

In step 76, the weighted average value Sfild
Is divided by a gear position correction coefficient SLG corresponding to the current gear position detected by the gear position sensor 11 and normalized as a learning value independent of the gear position. Ask for. here,
The gear position correction coefficient SLG is retrieved from a map shown in FIG. The gear position correction coefficient SLG is such that the lower the speed, the higher or lower the rotation speed of the engine with respect to the wheel rotation.
Since the change speed is large and has a large effect on the detection of rotation fluctuation, as shown in FIG. This step 76 corresponds to the function of the learning value calculation means.

Next, at step 77, the learned flag F
In step 78 of setting SL = 1 and storing that the learning has been performed, the learning value GSL and the learned flag FSL are stored in a predetermined area of the learning map instead of the currently written value. The learning map is obtained by dividing the learning operation region into a plurality of regions based on, for example, the basic injection amount Tp serving as a load on the engine and the engine speed Ne, and calculating the learning value in the region corresponding to the read operating condition. GSL
And the learned flag FSL are stored, and this flow ends. This step 78 corresponds to the function of the learning value updating means.

Next, a routine for calculating the slice level, which is the stability target value, will be described with reference to the flowchart of FIG. This routine corresponds to the function of the stability target value setting means. In step 81, the current operation state is read.
In step 82, the state of the learned flag FSL stored in the operation area read in step 81 is searched from the learning map.

In step 83, it is determined whether the learned flag FSL retrieved in step 82 is 1 or 0, and FSL =
In the case of 1, it is determined that learning has been completed, and the routine proceeds to step 84, where F
When SL = 0, it is determined that learning has not been performed, and the present flow ends.
In step 84, the gear position at that time is detected, and the corresponding gear position correction coefficient SLG is searched from the map of FIG.

In step 85, the learning value GSL of the current operation area read in step 81 is read from the learning map. In step 86, the learning value GSL is multiplied by the searched gear position correction coefficient SLG to correct the learning value, and the slice level SL as the stability target value is calculated. this is,
This is because, even if the stability of the engine is the same, the fluctuation range of the engine rotation changes depending on the gear position at that time.By correcting this gear position, it is possible to set an appropriate slice level suitable for the operating state at that time. it can.

FIG. 10 shows a background job (BGJ).
The map fuel-air ratio Mdml used in the flow of FIG. 3 is calculated. First, at step 91, a lean condition is determined. This is to set a region in which lean combustion is desired from the requirements of fuel efficiency, exhaust, and drivability, and determine whether the current operating condition is within the range. However, a detailed description is omitted here.

Next, if it is determined in step 92 that the condition is lean, a lean fuel-air ratio map MDMLL is generated in step 94, otherwise a stoichiometric air-fuel ratio map MDMLS not including the lean region is generated in step 93. A map selected by using the selected engine speed Ne and the basic injection amount Tp is referred to, and the result is defined as Mdml. The target fuel-air ratio correction coefficient Tdml, the fuel-air ratio correction coefficient Dml,
This is to prepare for calculation of the target fuel-air ratio Tfbya and the fuel injection amount Ti.

As described above, by learning the stability target value during the stoichiometric air-fuel ratio operation in which the engine is stable, individual differences in the crank angle sensor 9 for detecting the rotation of the engine, the engine body 1, and the like can be reduced. This prevents erroneous detection of engine stability resulting from unnecessary increase or decrease in fuel density, eliminating adverse effects on drivability and exhaust performance, and improving stability control accuracy. Enhanced. When calculating the learning value, the correction based on the gear position is introduced to remove the influence of the gear position on the rotation fluctuation of the engine, and the learning is prohibited when the engine is abnormal. Therefore, the calculation accuracy and reliability of the learning value can be improved, and an appropriate stability target value can be set.

In this embodiment, the air-fuel ratio (fuel amount) is shown as an engine control factor. However, the same control can be applied to the exhaust gas recirculation (EGR) amount.

[0045]

According to the present invention as described above, according to the present invention, Li
By learning the target value of the stability under operating conditions other than the run condition , it is possible to absorb individual differences between the engine rotation detecting means and the engine itself, thereby preventing erroneous detection of the engine stability.
Therefore, it is possible to prevent adverse effects on the operability and the exhaust performance due to individual differences of parts, and it is possible to improve the control accuracy of stability.

When calculating the learning value, correction based on the gear position is introduced to remove the influence of the gear position on the rotation fluctuation of the engine, and learning is prohibited when the engine is abnormal. If this is the case, the calculation accuracy and reliability of the learning value can be improved, and a more appropriate stability target value can be set.

[Brief description of the drawings]

FIG. 1 is a block diagram illustrating a configuration of the present invention.

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

FIG. 3 is a flowchart for calculating a fuel-air ratio correction coefficient according to the first embodiment;

FIG. 4 is a flowchart of rotation fluctuation calculation in FIG. 3;

FIG. 5 is a flowchart of feedback control determination and rotation fluctuation learning determination in FIG. 3;

FIG. 6 is a flowchart for calculating a feedback correction rate in FIG. 3;

FIG. 7 is a flowchart for calculating a fuel injection amount using the fuel-air ratio correction coefficient calculated in FIG. 3;

FIG. 8 is a flowchart of learning a rotation fluctuation value.

FIG. 9 is a flowchart for calculating a stability target value.

FIG. 10 is a flowchart of a background job.

FIG. 11 is a diagram showing a table for setting the number of samples for calculating a rotation fluctuation value.

FIG. 12 is a characteristic diagram of a correction amount of a stabilized fuel-air ratio correction coefficient.

FIG. 13 is a diagram showing a relationship between a gear position and a gear position correction coefficient.

[Explanation of symbols]

1 Engine body 4 Air flow meter 6 Fuel injection valve 8 Throttle sensor 9 Crank angle sensor 11 Gear position sensor 20 control unit

──────────────────────────────────────────────────続 き Continued on the front page (58) Field surveyed (Int. Cl. ⁷ , DB name) F02D 41/14 310 F02D 45/00 330 F02D 45/00 340 F02D 45/00 362

Claims (4)

(57) [Claims] An engine rotational speed detecting means, a rotational fluctuation calculating means for calculating an engine rotational fluctuation value based on a detection value of the engine rotational speed detecting means, and a rotational fluctuation value calculated by the rotational fluctuation calculating means. Correction amount calculating means for calculating a correction amount of a control amount related to combustion stabilization of the engine control factor so as to approach the stability target value, and a correction amount of the engine control factor based on the correction amount calculated by the correction amount calculating means. A control amount calculating means for calculating a control amount, and a control amount based on the control amount calculated by the control amount calculating means.
And control means for controlling the combustion of the engine in a lean condition, the lean
A learning value calculation means for calculating a learning value of the rotational variation value calculated by the rotational movement calculation means with operating conditions not condition, a rewritable learning value storage means for storing a learned value calculated by the learning value calculation section Learning value updating means for rewriting the learning value stored in the learning value storage means with the learning value calculated by the learning value calculating means; and the stability target based on the learning value stored in the learning value storage means. A stability control device for an internal combustion engine, comprising stability target value setting means for setting a value. 2. The stability target value setting means calculates a stability target value by correcting a learning value stored in the learning value storage means by changing a correction value according to a gear position. The stability control device for an internal combustion engine according to claim 1. 3. The internal combustion engine stability control apparatus according to claim 1, wherein said learning value calculation means calculates a learning value during a stoichiometric air-fuel ratio operation. 4. The learning value calculating means according to claim 1, wherein said learning value calculating means compares the rotation fluctuation value calculated by said rotation fluctuation calculating means with a predetermined value set in advance. 2. The internal combustion engine stability control device according to claim 1, further comprising: a determination unit configured to determine that the engine is abnormal when the fluctuation value is large, and prohibit calculation of a learning value.