JP3036351B2 - Method of detecting rotation fluctuation of internal combustion engine - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method for detecting rotation fluctuation used for lean burn control of an internal combustion engine.

[0002]

2. Description of the Related Art In recent years, in gasoline engines for automobiles, the air-fuel ratio of an air-fuel mixture is set to a stoichiometric air-fuel ratio (14.7) in order to reduce harmful exhaust gas components such as HC and CO and improve fuel efficiency.
A much thinner lean burn engine has been proposed. In the lean burn engine, the ignition / combustion performance is improved by layered air supply that enriches the air-fuel mixture flowing near the ignition plug, swirl or tumble that enhances the turbulence of the air-fuel mixture in the combustion chamber. Further, in the lean region, nitrogen oxides (NO _x ) cannot be reduced by the three-way catalyst, and the amount of emission becomes maximum near the air-fuel ratio of 16 and decreases on the lean side, and the stable combustion limit (air-fuel ratio) On the lean side, since the torque fluctuation exceeds the allowable limit, it is necessary to control the air-fuel ratio to a narrow range near the stable combustion limit. Therefore, instead of the O ₂ sensor that uniquely detects the stoichiometric air-fuel ratio, an air-fuel ratio sensor (LAFS: linear A / F sensor) that can continuously detect the air-fuel ratio is used, and is determined based on the engine speed and the volumetric efficiency. The fuel injection amount is feedback-controlled so that the target air-fuel ratio is obtained.

[0003]

In the above-mentioned air-fuel ratio feedback control, the air-fuel ratio sensor controls the air-fuel mixture to a target air-fuel ratio. This simply controls the equivalence ratio between the fuel injection amount and the intake air amount. It does not control the combustion state of the air-fuel mixture. Further, the air-fuel ratio sensor has a characteristic that when the air-fuel ratio is largely shifted to the lean side, the detection accuracy is slightly deteriorated. For this reason, when the target air-fuel ratio is set near the stable combustion limit, abnormalities may occur in the combustion of the air-fuel mixture due to fluctuations in the outside temperature, humidity, and the like, resulting in intermittent misfires. In this case, as a matter of course, the fuel consumption is deteriorated and the harmful exhaust gas is increased, and the engine vibration and the torque fluctuation frequently occur, which gives the occupant a significant discomfort. Therefore, in the past, the target air-fuel ratio had to be set richer than the combustion stability limit in order to provide a margin for combustion fluctuations and misfires, and it was possible to pursue the reduction of NO _X emissions and the improvement of fuel efficiency to the maximum. could not.

[0004] The present invention has been made in view of the above situation,
It is an object of the present invention to provide a rotation fluctuation detection method for detecting a rotation fluctuation of an internal combustion engine caused by a combustion abnormality or the like, thereby enabling the engine to operate at a stable combustion limit or the like.

[0005]

SUMMARY OF THE INVENTION Therefore, a rotation fluctuation detecting method according to a first aspect of the present invention detects a rotation fluctuation of each cylinder in a multi-cylinder internal combustion engine from crank rotation information of the internal combustion engine. Detecting a crank angle signal output at a predetermined crank angle position of the internal combustion engine; calculating an instantaneous rotation fluctuation value in a combustion stroke of each cylinder based on the crank angle signal; By subtracting the deviation between the instantaneous rotational fluctuation value of the second cylinder, which was in the combustion stroke before the first cylinder, and its average value from the deviation between the instantaneous rotational fluctuation value of the first cylinder and its average value, Calculating a rotation variation index of the first cylinder;
Judging that the first cylinder has undergone rotation fluctuation when the detection result of the rotation fluctuation index calculation means exceeds a predetermined threshold value.

According to a second aspect of the present invention, there is provided the rotation variation detecting method according to the first aspect, wherein the rotation variation is such that the rotation variation index exceeds the threshold by a predetermined number of times within a predetermined sampling interval. Judged when exceeded.
According to a third aspect of the present invention, there is provided the rotation variation detecting method according to the first aspect, wherein the second cylinder is positioned immediately before the first cylinder in the combustion order of the internal combustion engine. Cylinder.

According to a fourth aspect of the present invention, there is provided the rotation fluctuation detecting method according to the first aspect, wherein the instantaneous value of the rotation fluctuation is determined by comparing the rotation speed of each cylinder of the internal combustion engine with the rotation speed in the first half of the combustion stroke. It is calculated based on the ratio with the rotation speed in the latter half of the combustion stroke.

[0008]

According to the rotation fluctuation detecting method of the present invention, when a combustion abnormality such as misfire occurs, the rotation fluctuation instantaneous value fluctuates, and the rotation fluctuation index indicating the change also fluctuates to detect the rotation fluctuation. At this time, the deviation between the instantaneous rotation fluctuation value of the second cylinder, which was in the combustion stroke before the first cylinder, and the average value thereof is determined from the deviation between the instantaneous rotation fluctuation value of the first cylinder and its average value. Since the rotation fluctuation index is calculated by the reduction, the variation between cylinders and between cycles is removed.

Further, according to the rotation fluctuation detecting method of claim 2,
Since it is determined that the rotation fluctuation has occurred when the rotation fluctuation index exceeds the threshold value by a predetermined number of times or more, a rotation fluctuation index that is equal to or larger than the threshold value, that is, noise caused by the influence of a road surface change or the like is removed. According to the rotation fluctuation detecting method of the third aspect, the rotation fluctuation instantaneous value of the first cylinder of the multi-cylinder internal combustion engine and the rotation fluctuation instantaneous value of the second cylinder which was in the combustion stroke immediately before the first cylinder are calculated. , The rotation fluctuation index is calculated from the deviation of the rotation fluctuation, so that the influence of the rotation fluctuation is excluded when the rotation fluctuation of a cylinder that undergoes a combustion stroke immediately after the cylinder in which the rotation fluctuation occurs is detected.

According to a fourth aspect of the present invention, there is provided a method for detecting rotation fluctuation,
From a crank angle signal output at a predetermined crank angle position that defines the first half of the combustion stroke and the second half of the combustion stroke of the internal combustion engine,
It is determined that the internal combustion engine has fluctuated based on the ratio between the rotation speed in the first half of the combustion stroke and the rotation speed in the second half of the combustion stroke. That is, when a combustion abnormality such as a combustion fluctuation occurs in a certain cylinder,
The ratio between the rotation speed of the cylinder in the first half of the combustion stroke and the rotation speed in the second half of the combustion stroke fluctuates. As a result, the rotation fluctuation instantaneous value also fluctuates and rotation fluctuation is detected.

[0011]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described below in detail with reference to the drawings. FIG. 1 is a schematic configuration diagram of an engine control system to which a rotation fluctuation detecting method according to the present invention is applied. FIG.
In FIG. 1, reference numeral 1 denotes an in-line four-cylinder gasoline engine (hereinafter simply referred to as an engine) for an automobile, and a combustion chamber, an intake system, an ignition system, and the like are designed for lean burn. An intake port 2 of the engine 1 has a fuel injection valve 3 for each cylinder.
, An air cleaner 5, an air flow sensor 6, a throttle valve 7,
An intake pipe 9 having an SC (idle speed controller) 8 and the like is connected. The exhaust port 10 is connected via an exhaust manifold 11 to an exhaust pipe 14 having an O ₂ sensor 12, a three-way catalyst 13, and a muffler (not shown). The engine 1 has a spark plug 1 in a combustion chamber 15.
6, and a crank angle sensor 18 for detecting rotation of a rotor plate 17 directly attached to a crankshaft 25 is attached. As shown in FIG. 2, two vanes 17a and 17b having an angular width of 70 ° are formed on the rotor plate 17 at intervals of 180 °.
As shown in FIG. 3, a section α of 110 ° (BTDC5 ° to ATDC105 °) including the top dead center (TDC) of each cylinder.
Followed by 70 ° (ATDC105 ° to ATDC1
75 °) is detected. In FIG. 1, reference numeral 19 denotes a throttle sensor for detecting the opening degree .theta.TH of the throttle valve 7, reference numeral 20 denotes a water temperature sensor for detecting the cooling water temperature TW, reference numeral 21 denotes an atmospheric pressure sensor for detecting the atmospheric pressure Ta, and reference numeral 22 denotes an intake air temperature T.
This is an intake air temperature sensor that detects a.

An input / output device (not shown) and a storage device (ROM, RA,
ECU (engine control unit) including M, BURAM, etc.), central processing unit (CPU), timer counter, etc.
23 are provided, and perform overall control of the engine 1. Detection information from the various sensors described above is input to the input side of the ECU 23. The ECU 23 calculates the optimum values such as the fuel injection amount and the ignition timing from the detected information,
The fuel injection valve 3 and the ignition plug 16 are driven. In the figure, 2
Reference numeral 4 denotes an ignition unit that outputs a high voltage to the ignition plug 16 according to a command from the ECU 23.

The control procedure in this embodiment will be described below with reference to the control flowcharts of FIGS. 4 to 6 and the maps and graphs of FIGS. 7 to 10. When the driver turns on the ignition key and engine 1 starts, EC
U23 is a predetermined control interval (for example, 10 ms)
Then, the fuel injection control subroutine shown in the flowchart of FIG. 4 is repeatedly executed.

When this subroutine starts, the ECU 2
In step S1, the operation information from each sensor is read into the RAM. Next, in step S2, the ECU 23 determines whether or not to perform the feedback control based on the throttle opening θTH, its time rate of change, volumetric efficiency EV, the elapsed time after starting the engine, the cooling water temperature TW, and the like.
The volumetric efficiency EV calculates an intake air amount A / N per intake stroke from the air flow rate detected by the air flow sensor 6 and the engine speed Ne, and corrects the intake air amount A / N based on the atmospheric pressure Pa, the intake air temperature Ta, and the like. It is required by If the determination is affirmative (Yes), the ECU 23
Is the volumetric efficiency EV and the engine speed Ne in step S3.
From the above, it is determined whether or not the current operating state is in a predetermined lean burn control region. Note that the lean burn control is performed in an operation region where the required torque is small, such as during idling operation or constant speed traveling.

If the determination in step S3 is Yes, the ECU 23 determines in step S4 that the volumetric efficiency E
Based on V and the engine speed Ne, the target air-fuel ratio OAF is set based on the lean air-fuel ratio map shown in FIG. next,
In step S5, the ECU 23 controls the injection amount (valve opening time TINJ) of the fuel injector 3 by a stable combustion limit control subroutine described later.

On the other hand, if the determination in step S3 is No, the ECU 23 determines in step S6 the target air-fuel ratio OAF based on the volumetric efficiency EV and the engine speed Ne based on the stoichiometric / rich air-fuel ratio map of FIG. And set
In step S7, the valve opening time TINJ of the fuel injection valve 3 for feedback control of the based on the output voltage of the O ₂ sensor 12. If the determination in step S2 is No, E
In step S8, the CU 23 sets the target air-fuel ratio OAF based on the stoichiometric / rich air-fuel ratio map. next,
After calculating the basic injection amount TINJB from the target air-fuel ratio OAF and the intake air amount A / N in step S9, the ECU 23 corrects the increase during acceleration, the increase during cold operation, and the like.
The open time of the fuel injection valve 3 is controlled in an open loop.

The above-described stable combustion limit control subroutine is repeatedly executed in the following procedure every time the crank interruption signal SGT is input. When this subroutine is started, the ECU 23 first proceeds to step S20 in FIG.
It is determined whether or not the current operating state is in a predetermined learning area. This determination is made based on, for example, whether or not the current operating state is in the learning area indicated by cross-hatching in the lean air-fuel ratio map in FIG. 7. If this determination is No, the determination in FIG. Proceeding to S36, the following steps are executed to drive and control the fuel injection valve 3. In this embodiment, the volumetric efficiency EV is between EVA and EVB,
The range where the engine speed Ne is between NeA and NeB is the above-mentioned predetermined learning area.

If the determination in step S20 is Yes, the ECU 23 determines in step S22 that the predetermined value TCDX
After subtracting 1 from the countdown timer TCD whose initial value is (256 in this embodiment), in step S23,
An instantaneous rotation fluctuation value Vmn in the n-th combustion stroke of the m-th cylinder in the ignition order is calculated by the following equation. In this embodiment, the ignition order of each cylinder is 1-3-4-2.

Vmn = (β / α−Tβ / Tα) · K Here, α (fixed value) indicates that the crank angle sensor 18 is OFF.
(110 ° including the compression top dead center), β (fixed value) is a section (70 ° following β) in which the crank angle sensor 18 is turned on, and Tα and Tβ are the crankshaft 25.
Is the time required to rotate the sections α and β, and K is a correction coefficient (> 0) set based on a map using the volumetric efficiency EV and the engine speed Ne as parameters. When the combustion is performed normally, the rotation of the crankshaft 25 is relatively slow in the first half of the combustion stroke (section α) affected by the compression stroke and the combustion is still insufficient.
In the latter half of the combustion stroke (section β) in which the combustion is completely performed, the rotation of the crankshaft 25 is relatively fast.
/ Α> Tβ / Tα, and the rotation fluctuation instantaneous value Vmn becomes a positive value. Further, when a misfire occurs, it is necessary to perform compression work of another cylinder without work due to combustion, so that the rotation of the crankshaft 25 gradually becomes slower and β / α <T
β / Tα, and the rotation fluctuation instantaneous value Vmn becomes a negative value.

When the calculation of the instantaneous rotation fluctuation value Vmn is completed, E
In step S24, the CU 23 determines in step S24 the rotational fluctuation instantaneous value Vm- of the (m-1) th cylinder in the previous combustion stroke so as to eliminate the variation between the cylinders and the variation between the cycles.
The cylinder-by-cylinder rotation fluctuation index ΔVmn is calculated from the difference from 1, n. ΔVmn = (Vmn−Emn) − (Vm−1, n−Em−1, n) where Emn is a low-pass filter average value, and is calculated using the following equation. Note that KF in the following equation is a filter coefficient, and is set to 0.95 in this embodiment.

Emn = KF · Em, n-1 + (1−KF) · Vmn After calculating the cylinder-specific rotation fluctuation index ΔVmn in step S24, the ECU 23 then determines in step S25 the cylinder-specific rotation fluctuation index ΔVmn. Is determined to be less than the threshold value ΔVx. If the determination is No, the process proceeds to step S30 in FIG. The threshold value ΔVx is set to a sufficiently low value with respect to the rotation fluctuation accompanying normal combustion, so that the cylinder-specific rotation fluctuation index ΔVmn becomes lower only when a misfire occurs.

On the other hand, if the determination in step S25 is Yes, in step S26, the misfire counter CMF is set to 1
Is added, and in step S27, the cylinder-by-cylinder variation integrated value Σ
The current cylinder-specific rotation fluctuation index ΔVmn is added to ΔVm, and the process proceeds to step S30 in FIG. The misfire counter CMF
The cylinder-by-cylinder variation integrated value ΣΔVm is reset to the initial value 0 each time the countdown timer TCD is reset to the initial value, as shown in FIG.

The ECU 23 determines in step S30 in FIG.
It is determined whether or not the countdown timer TCD has become 0, that is, whether or not the operation of the TCDX cycle has been performed within the learning region. If this determination is No, the process proceeds to step S36 described later, and the fuel injection valve 3 Drive. In the present embodiment, even if the operating state once deviates from the learning region, the misfire frequency counter CMF and the cylinder-by-cylinder variation integrated value ΣΔV
The value m and the countdown timer value TCD are stored in the RAM, and when the learning area is re-entered, the ECU 23 continuously performs integration and subtraction of each value.

When the operation of the TCDX cycle is performed in the learning area and the determination in step S30 is Yes, the ECU
In step S31, it is determined whether or not the value of the misfire counter CMF is 0. If the determination is Yes, the ECU 23 determines that the air-fuel ratio of the cylinder is still on the rich side with respect to the stable combustion limit, and determines in step S32.
Is a predetermined weight correction value KD (0.2% in this embodiment).
And the cylinder-by-cylinder feedback coefficient KmFD
To update. In this embodiment, the cylinder-by-cylinder feedback coefficient KmFD takes a value in the range of 0% to 10%,
It is stored in a non-volatile memory such as a URAM.

KmFD = KmFD-KD On the other hand, if the determination in step S31 is No, the ECU
In step S33, it is determined whether or not the value of the misfire counter CMF is equal to or greater than a threshold value CMFX (three times in this embodiment). If this determination is No, the ECU
23 judges that the air-fuel ratio of the cylinder is at the stable combustion limit, holds the cylinder-by-cylinder feedback coefficient KmFD of the fuel injection valve 3 as it is, and continues the operation at the stable combustion limit.

If the determination in step S33 is Yes, the ECU 23 determines that the air-fuel ratio of the cylinder has already exceeded the stable combustion limit and has entered the lean side. Then, in step S34, a cylinder-specific increase correction value Kam corresponding to the cylinder-specific variation integrated value ΣΔVm is set based on the map of FIG. 10, and the cylinder-specific feedback coefficient KmFD is updated by the following equation. As shown in FIG. 10, the cylinder-specific increase correction value Kam increases linearly with an increase in the cylinder-specific variation integrated value ΣΔVm.

KmFD = KmFD + Kam When the cylinder-by-cylinder feedback coefficient KmFD is updated in step S32 or step S34, in step S35, the ECU 23 resets the values of the misfire counter CMF, the cylinder-by-cylinder variation integrated value ΣΔVm, and the countdown timer TCD. When the reset of the counter, the timer, and the like in step S35 ends, the ECU 23 proceeds to step S36.
Then, it is determined from the lean air-fuel ratio map of FIG. 7 whether or not the current operation state is in a predetermined correction region. In this embodiment, the above-mentioned predetermined correction region is a range where the volumetric efficiency EV is between EVC and EVD and the engine speed Ne is between NeC and NeD. The correction area is set to be larger than the learning area, and when EVC <EVA <EVB <EVD, N
eC <NeA <NeB <NeD. In a region outside the correction region, there is no risk of misfiring, and no learning correction is required.

If the determination in step S36 is Yes, the ECU 23 determines in step S37 that the fuel correction coefficient KL
Is calculated by the following equation. If this determination is No, the ECU 23 sets the fuel correction coefficient KL to 1 in step S38. KL = 1 + KmFD After obtaining the fuel correction coefficient KL in step S37 or step S38, the ECU 23 determines in step S39 the valve opening time of the fuel injection valve 3 of the cylinder based on the target air-fuel ratio OAF and the volume efficiency EV by the following equation. TINJ is calculated, and the fuel injection valve 3 is driven in step S40. In the following equation, KT is various correction coefficients, Vol is the cylinder volume, and TD is the invalid injection time.

TINJ = KT · OAF · EV · Vol · KL + TD In the correction region, when the target air-fuel ratio OAF is too rich, it is gradually made lean every TCDX (256) cycles and approaches the stable combustion limit. Go on. Also,
When the fuel enters the lean side beyond the stable combustion limit, the air-fuel ratio of the cylinder is enriched at once, and the misfire is immediately eliminated. The above control is performed for each cylinder.
This is because the air-fuel ratio near the stable combustion limit differs for each cylinder.

Now, the description of the concrete embodiment has been completed.
Embodiments of the present invention are not limited to this embodiment. For example, in the above-described embodiment, the present invention is applied to an in-line four-cylinder engine. However, the present invention may be applied to various engines having different numbers of cylinders and arrangements thereof, such as a single-cylinder engine and a V-type six-cylinder engine, The present invention may be applied to various controls other than lean burn control, such as EGR control and ignition timing control.

As the instantaneous value of the rotation fluctuation, the average crank speed in the combustion stroke may be calculated from the ON / OFF signal of the vane type rotor plate, and this may be used. As described in the above section, an instantaneous rotation speed at a specific rotation position may be calculated using unit signal generation means for detecting a plurality of angular positions in a combustion stroke of each cylinder, and this may be used.

The initial value of the fuel injection amount in the stable combustion limit control subroutine may not be calculated from the target air-fuel ratio set based on the lean air-fuel ratio map, but may be directly set based on the fuel injection amount map. After the fuel injection amount is corrected, the target air-fuel ratio in the lean air-fuel ratio map may be learned and corrected. Further, in the above-described embodiment, when the value of the misfire counter is within the predetermined range (1 and 2), the fuel injection amount is held at the present state, but is smaller than the predetermined value (for example, 3). In such a case, the engine may be made lean uniformly. Further, the initial values of the thresholds, the countdown timer, and the like in the above control can be appropriately set, and the calculation procedure can be changed without departing from the gist of the present invention.

[0033]

According to the rotation fluctuation detecting method of the present invention, the rotation fluctuation index is calculated from the instantaneous rotation fluctuation value in the combustion stroke of each cylinder by removing the variation between cycles and the variation between cylinders. Since the rotation fluctuation is determined based on the fact that the value exceeds a predetermined threshold value, the rotation fluctuation can be determined with high accuracy.

Further, according to the rotation fluctuation detecting method of the present invention,
Since it is determined that the rotation fluctuation has occurred when the rotation fluctuation index exceeds the threshold value a predetermined number of times, the rotation fluctuation index that is equal to or larger than the threshold value, that is, noise generated due to the influence of a road surface change or the like is removed, and the accuracy of the rotation fluctuation determination is improved. According to a third aspect of the present invention, there is provided a method for detecting a rotational fluctuation of a first cylinder of a multi-cylinder internal combustion engine, and an instantaneous rotational fluctuation of a second cylinder which is in a combustion stroke immediately before the first cylinder. In order to calculate the rotation fluctuation index from the above, when performing the rotation fluctuation detection of the cylinder that reaches the combustion stroke immediately after the cylinder where the rotation fluctuation has occurred, the influence of the rotation fluctuation is excluded, and more accurate detection of the rotation fluctuation is performed. It becomes possible.

According to the rotation fluctuation detecting method of the fourth aspect, the instantaneous rotation fluctuation value is calculated based on the ratio between the rotation speed in the first half of the combustion stroke and the rotation speed in the second half of the combustion stroke in the internal combustion engine. When a combustion abnormality such as a fluctuation occurs, the fluctuation of the rotation fluctuation instantaneous value becomes large, and the rotation fluctuation can be easily determined.

[Brief description of the drawings]

FIG. 1 is a schematic configuration diagram showing an embodiment of an engine control system to which a rotation fluctuation detecting method according to the present invention is applied.

FIG. 2 is a perspective view showing a rotor plate and a crank angle sensor.

FIG. 3 is a graph showing an output signal of a crank angle sensor.

FIG. 4 is a flowchart showing a procedure of a fuel injection control subroutine.

FIG. 5 is a flowchart showing a procedure of a stable combustion limit control subroutine.

FIG. 6 is a flowchart showing a procedure of a stable combustion limit control subroutine.

FIG. 7 is a lean air-fuel ratio map using volume efficiency and engine speed as parameters.

FIG. 8 is a stoichiometric / rich air-fuel ratio map using volume efficiency and engine speed as parameters.

FIG. 9 is a graph showing a relationship between a rotation fluctuation and a fuel injection amount and the like.

FIG. 10 is a graph showing a relationship between a variation integrated value and a fuel injection increase coefficient.

[Explanation of symbols]

Reference Signs List 1 engine 3 fuel injection valve 12 O ₂ sensor 17 rotor plate 18 crank angle sensor 21 ECU 25 crankshaft

Continuation of the front page (56) References JP-A-5-195858 (JP, A) JP-A-4-311651 (JP, A) JP-A-58-5243 (JP, A) JP-A-61-11440 (JP) JP-A-4-171253 (JP, A) JP-A-4-81548 (JP, A) JP-A-4-330349 (JP, A) JP-A-4-1322862 (JP, A) JP-A-2-112646 (JP, A) JP-A-5-340293 (JP, A) JP-A-5-33703 (JP, A) JP-A-63-68750 (JP, A) A) JP-A-4-50446 (JP, A) JP-A-6-146998 (JP, A) (58) Fields investigated (Int. Cl. ⁷ , DB name) F02D 45/00 362

Claims (4)

(57) [Claims] 1. A rotation fluctuation detection method for detecting rotation fluctuation of each cylinder in a multi-cylinder internal combustion engine from crank rotation information of the multi-cylinder internal combustion engine, the crank angle being output at a predetermined crank angle position of the internal combustion engine. Detecting a signal, a step of calculating an instantaneous rotation fluctuation value in the combustion stroke of each cylinder based on the crank angle signal, and a first deviation from the instantaneous rotation fluctuation value of the first cylinder and its average value. Calculating the rotation fluctuation index of the first cylinder by reducing the deviation between the rotation fluctuation instantaneous value of the second cylinder and the average value thereof during the combustion stroke before the cylinder of the first cylinder; Determining that there is a rotation fluctuation in the first cylinder based on the rotation fluctuation index of the internal combustion engine. 2. The method according to claim 1, wherein the rotation fluctuation is determined when the rotation fluctuation index exceeds the threshold value a predetermined number of times or more within a predetermined sampling interval. . 3. The method according to claim 1, wherein the second cylinder is a cylinder located immediately before the first cylinder in a combustion order of the internal combustion engine. . 4. The engine according to claim 1, wherein the instantaneous value of the rotation fluctuation is calculated based on a ratio between a rotation speed of each cylinder of the internal combustion engine in a first half of a combustion stroke and a rotation speed in a second half of a combustion stroke. A method for detecting rotation fluctuation of an internal combustion engine.