JPH03246350A - Combustion condition detector for internal combustion engine - Google Patents

Abstract

PURPOSE:To improve the detection precision of a combustion condition, by seeking figural mean effective pressure from pipe inside pressure obtained by correcting a pipe inside pressure detection signal on the basis of the quantity of bias, and from pipe inside pressure, at a non-sampling period, predicted from the corrected pipe inside pressure and a combustion chamber capacity. CONSTITUTION:At the sampling period of a predetermined crank angle range, a pipe inside pressure detection signal is conducted with sampling by means of a sampling means C, and at the same time a combustion chamber capacity at every crank angle is set by means of a setting means D. And the bias quantity of the pipe inside pressure detection signal is set by means of a setting means E on the basis of the combustion chamber capacity and the pipe inside pressure detection signal, and the pipe inside pressure detection signal is corrected by means of a correcting means F on the basis of the bias quantity, and pipe inside pressure is sought. Meanwhile, at a non-sampling period, pipe inside pressure is predicted by means of a predicting means G on the basis of the combustion chamber capacity and the pipe inside pressure after correction, and figural mean effective pressure is operated by means of an operating means H on the basis of the predicted pipe inside pressure and the pipe inside pressure after correction.

Description

[Detailed description of the invention] <Industrial application field> The present invention relates to a combustion state detection device for an internal combustion engine.

<Conventional technology> In order to understand combustion conditions such as engine misfire, cylinder pressure (
There is a method of detecting the combustion chamber pressure), and the actual combustion state is determined by determining the maximum in-cylinder pressure and indicated mean effective pressure Pi (hereinafter abbreviated as P) in a specific operating state. Conventional examples of combustion state detection devices include the following (see "Internal Combustion Engine JVOL, 1, page 19, September 1962, published by Sankaido Co., Ltd.").

In other words, the in-cylinder pressure is measured by the in-cylinder pressure sensor at 2 times per cylinder.
The period of rotation (720 degrees of crank angle in a 4-cycle engine) is detected, and P is calculated from the detected cylinder pressure.

<Problems to be Solved by the Invention> By the way, most of the engines installed in vehicles are usually multi-cylinder engines with four or more cylinders, and when the above-mentioned conventional example is applied to these multi-cylinder engines, the following problems occur.

In other words, if the in-cylinder pressure for each cylinder is detected for a period of two revolutions and Pi is calculated for each cylinder, even if a microcomputer is used, the detected in-cylinder pressure will be A/
A high-speed A/D converter for D conversion is required, and a high-speed CPU is also required for calculating Pl, resulting in high costs.

In addition, in order to eliminate the above-mentioned problems, we have developed
There is one shown in Japanese Patent No. 5548.

This is basically a predetermined crank angle range (e.g. 1
Pseudo P1 is calculated by the following equation from the cylinder pressure P detected within 80 degrees) and the minute cylinder volume change amount Δ■ for each crank angle.

A method is disclosed in which P is calculated using the above calculation formula in a section where the cylinder pressure is equal to or higher than a predetermined value.

Further, Japanese Patent Application Laid-Open No. 62-238434 discloses a system in which pseudo P is calculated from the above-mentioned calculation formula in the range from the ignition timing to the period in which the cylinder pressure is equal to or higher than a predetermined value.

However, in these methods, the pseudo P□ is calculated within a predetermined crank angle range, so when the combustion condition deteriorates and the combustion period becomes longer as shown by the chain line in FIG. Since the correlation between pseudo P and P2 during the 720° period deviates widely as shown in Fig. 23, the detection accuracy of the combustion state deteriorates, and the control accuracy of ignition timing, air-fuel ratio, exhaust gas recirculation amount, etc. There is a problem that makes it worse.

Furthermore, in JP-A-62-203036, the actual in-cylinder pressure is detected near the compression top dead center, and the operating conditions (intake air flow rate, throttle valve opening, rotational speed, etc.) are detected in other sections. ), search the cylinder pressure from the map, and use these cylinder pressures to calculate P,
A method for calculating is disclosed.

However, since the combustion state constantly changes, it is difficult to predict the actual in-cylinder pressure using a method that searches the cylinder pressure from preset map data, and the detection accuracy of P, that is, the detection accuracy of the combustion state deteriorates. There is a problem with this.

The present invention has been made in view of the above circumstances, and an object of the present invention is to provide a combustion state detection device for an internal combustion engine that can improve the detection accuracy of the combustion state.

<Means for Solving the Problems> For this reason, the present invention, as shown in FIG. angle detection means B; sampling means C for sampling the cylinder pressure detection signal in a sampling period of a predetermined crank angle range;
The combustion chamber volume setting means sets the combustion chamber volume for each crank angle, and the bias amount of the cylinder pressure detection signal is determined based on the combustion chamber volume during the set sampling period and the sampled cylinder pressure detection signal. A bias amount setting means E to set, a cylinder pressure correction means F to correct the sampled cylinder pressure detection signal to obtain the cylinder pressure based on the set bias amount, and a cylinder pressure correction means F to calculate the cylinder pressure based on the set bias amount; a cylinder pressure prediction means G that predicts the cylinder pressure based on the combustion chamber volume and the corrected cylinder pressure; and an indicated average based on the corrected cylinder pressure and the predicted cylinder pressure. Effective pressure calculating means H for calculating effective pressure is provided.

<Operation> In this way, during the predetermined sampling period, the sampled cylinder pressure signal is corrected by the bias amount to obtain the cylinder pressure, and during the non-sampling period, the corrected cylinder pressure and combustion chamber volume are calculated. Predict the cylinder pressure from . Then, the indicated mean effective pressure is determined from the corrected cylinder pressure and the predicted cylinder pressure.

<Example> An example of the present invention will be described below based on FIGS. 2 to 17.

In FIG. 2, cylinder pressure sensors 1 to 6 are provided for each cylinder (six cylinders in this embodiment) as cylinder pressure detection means for detecting the cylinder pressure in the engine combustion chamber. 6 converts the cylinder pressure into a charge signal using a piezoelectric element and outputs the signal to charge amplifiers 7 to 12. The charge amplifiers 7 to 12 convert the charge signals into voltage signals and output the voltage signals to the multiplexer 13.

The multiplexer 13 outputs detection signals from the in-cylinder pressure sensors 1 to 6 of the selected cylinder based on a switching signal, which will be described later, to the I10 interface 16 of the control device 15 via the low-pass filter 14. The low-pass filter 14 removes high-frequency components such as knocking vibrations and ignition noise that are unnecessary for detecting cylinder pressure and cause erroneous detection, and passes only low-frequency components below a predetermined frequency.

The control device 15 includes a CPU 17. ROM18. RAM
19. An A/D converter 20 is provided, and a CPU 17
is executed according to the program written in ROM18.
10 reads necessary external data from the interface 16, and while exchanging data with the RAM 19, calculates the processing values necessary for calculating parameters related to the combustion state, and as necessary. The processed data is output to the I10 interface 16. The I10 interface 16 includes the low pass filter 14. Crank angle sensor 21 as crank angle detection means. As the signal from the air flow meter 22 is input, I1
A switching signal is output from the 0 interface 16 to the multiplexer 13 in accordance with a command from the CPU 17.

The A/D converter 20 operates according to instructions from the CPU 17.
10 The external signal input to the interface 16 is
D-convert. Further, the ROM 1B stores programs for the CPU 17, and the RAM 19 stores data used for calculations etc. in the form of a map or the like.

The crank angle sensor 21 outputs a reference signal at a predetermined crank angle position before compression top dead center of each cylinder at every predetermined crank angle (120 degrees in crank angle for a 6-cylinder engine), and also outputs a reference signal at a predetermined crank angle position before compression top dead center of each cylinder. A unit signal is output every 1°). Therefore, the engine rotational speed can be detected based on the input cycle of the reference signal and the number of counts. Also,
Air flow meter 22 outputs a signal corresponding to the intake air flow rate.

Here, the CPU 17 constitutes a sampling means, a combustion chamber volume setting means, a bias amount setting means, an in-cylinder pressure correction means, an in-cylinder pressure prediction means, and an effective pressure calculation means.

Next, the operation will be explained according to the flowcharts shown in FIGS. 3 to 14.

First, the sampling routine for in-cylinder pressure data will be explained according to the flowchart of FIG. This routine is executed at a predetermined crank angle position (in this embodiment, 60 degrees before compression top dead center) determined by the reference signal and unit signal.

In Sl, the cylinder pressure detected by the cylinder pressure sensors 1 to 6 is set at a predetermined crank angle (in this example, the crank angle is 4).
The A/D converter 20 performs A/D conversion and reads the data.

In S2, the read cylinder pressure PE is stored in the RAM 19 in association with the cylinder number i and the sampling order (corresponding to the crank angle) j.

Here, the reading order j based on the detected values is 30 to 60.

In S3, the A/D conversion end timing (6 after compression top dead center)
0''), and if YES, proceed to S4, and if NO, return to S1.

In S4, a calculation flag, which will be described later, is set to 1.

In S5, it is determined whether it is the next A/D conversion timing,
When the answer is YES, the process advances to S6, and when the answer is No, the process returns to S5.

In S6, the cylinder number i is incremented by 1.

In S7, it is determined whether the cylinder number i has become 6, and Y
When the answer is ES, the process advances to S8, and when the answer is No, the routine is ended.

In S8, the cylinder number i is set to zero and the routine is ended.

In this way, the cylinder pressure P during a crank angle period of 120' from 60 degrees before compression top dead center to 60 degrees after compression top dead center
E and j are detected for each cylinder.

Next, the calculation routine will be explained according to the flowchart shown in FIG.

In Sll, it is determined whether the calculation flag is 1 or not, and Y
When the answer is ES, the process advances to 512, and when the answer is NO, the routine is ended.

In S12, the polytropic coefficient PN in the compression stroke is searched from the map. This polytropic coefficient PN is generally set to about 1.3.

In S13, the bias amount X is calculated according to the flowchart of FIG. 5, which will be described later.

In S14, the cylinder pressure PE, , is corrected by the bias amount X according to the flowchart of FIG. 6, which will be described later, and the corrected cylinder pressure P! Store j in RAM 19 (j
=30-60).

515, the in-cylinder pressure Pij is predicted by calculating the in-cylinder pressure Pij in the range of 0 to 29 (06 to 116" in terms of crank angle with reference to the start of the compression stroke) according to the flowchart in FIG. 7 described later. do.

In S16, the polytropic coefficient PNE in the expansion stroke is
is calculated according to the flowchart of FIG. 8, which will be described later.

In S17, when j is in the range of 61 to 90 (244° to 360° in crank angle as above), the cylinder pressure p tJ
is predicted by calculating according to the flowchart of FIG. 9, which will be described later.

31B, P 2 and + are calculated by calculating according to the flowchart of FIG. 10, which will be described later.

In step 319, the calculation flag is set to zero, and then the routine ends.

Next, the routine for the bias amount X will be explained according to the flowchart of FIG.

In S21, 2 before the ignition timing (which can be assumed to be an adiabatic change)
The in-cylinder pressures PE, , l, PEff1J of the same cylinder detected at the two crank angle positions lx, y are read from the RAM 19.

In S22, the timing of the X and y (crank angle)
Search the combustion chamber volumes VX and VY at from the map.

In S23, a coefficient A is calculated from the searched combustion chamber welds VX, VY and the polytropic coefficient PN using the following equation.

A=(VX/VY)PM In S24, the bias amount X is calculated using the following equation.

X= (PEi, Errors are likely to occur, especially in the case of a cylinder pressure sensor that is attached to the ignition plug seat, the output value is likely to fluctuate due to changes in ambient temperature, etc. Therefore, the combustion at the 2nd crank angle position Chamber volumes vx, vy and polytropic coefficient PN
The coefficient A corresponding to the true in-cylinder pressure change during an adiabatic change from
At the same time, the bias amount X (see Figure 15) is calculated from this coefficient A and the cylinder pressure detected by the cylinder pressure sensors 1 to 6, and the true cylinder pressure It is designed to detect internal pressure. here,
Since the bias amount X changes every cycle due to startup, warm-up, load, etc., the bias amount X is always calculated.

Note that when using the combustion chamber volumes at two preset crank angles, the routine shown in the flowchart of FIG. 11 may be used instead of the routine shown in the flowchart of FIG. 5.

That is, in 325, coefficient A is searched from the map, and S2
In step 6, the bias amount X is calculated using the previous equation.

Next, the cylinder pressure correction routine will be explained according to the flowchart of FIG.

In S31, the initial value of j is set to 30.

In 332, it is determined whether j is less than 61. If YES, the process advances to S33, and if NO, the routine is ended.

At step 333, a bias amount X is added to the detected cylinder pressure PE, , to calculate a corrected cylinder pressure P1, and this cylinder pressure Pij is stored in RAM19.

In S34, j is incremented by 1 and the process returns to S32.

In this way, the cylinder pressure P□j will change from j=30 to j=
It is corrected sequentially up to 60.

Next, the seventh routine predicts the cylinder pressure during the compression stroke.
The explanation will be given according to the flowchart shown in the figure.

In step 541, the initial value of j is set to 30.

In step 342, it is determined whether j has become zero. If YES, the routine is terminated, and if NO, the routine proceeds to S43.

In S43, the combustion chamber volume Vj is adjusted in accordance with the crank angle.
+Vj-1 is searched from a map stored corresponding to a preset crank angle.

At 344, a coefficient C is calculated from the combustion chamber volumes V, , V, and the polytropic coefficient PN using the following equation.

C= (V j / V J-1) In PNS45, j
From the cylinder pressure Pi4 corrected when =30 or the previously predicted cylinder pressure Plj and the coefficient C, the cylinder pressure P at a rank angle position that is smaller by a predetermined crank angle position is calculated!
+j-1> is predicted by calculating by the following equation.

P5 (J-, , = pth×C At 546, j is decremented by 1, and the process returns to 342.

In this way, the cylinder pressure Pi(j-1+, j=30
to 0 is predicted and stored in the RAM 19.

Incidentally, instead of the routine shown in the flowchart of FIG. 7 above, as shown in the flowchart of FIG. 12, a coefficient C preset for j, that is, corresponding to the crank angle, is searched from the map in S47. Good too.

Next, a calculation routine for the polytropic coefficient PNE in the expansion stroke will be explained according to the flowchart of FIG.

351, the combustion chamber volume v6° at 60° after compression top dead center (240° in crank angle based on the start of the compression stroke) and expansion bottom dead center (360° in crank angle)
Combustion chamber volume V, at . and from the map based on the crank angle.

In S52, the searched combustion chamber volume V60. v.

Based on , the coefficient is calculated using the following formula.

D=log(■, ./v6°) In S53, the coefficient E is calculated from the detected cylinder pressure P4bo at 60@ after the compression top dead center and the constant cylinder pressure Pt□ using the following equation.

E = log (P Nho/ P FIX) where,
Constant cylinder pressure at 360° crank angle position P F
The reason for choosing IX is that since the exhaust valve is open at this position, no problem will occur assuming that the cylinder pressure at this time is constant regardless of the operating state.

In S54, a polytropic coefficient PNE (=E/D) is calculated.

As shown in the flowchart of FIG. 13, the coefficient is searched from the map in S55, and the PL
The coefficient E may be searched from a map set according to bO.

Next, the prediction routine for the cylinder pressure during the expansion stroke is performed in the ninth step.
The explanation will be given according to the flowchart shown in the figure.

In 361, the initial value of j is set to 60.

In S62, it is determined whether or not j is less than 90. If YES, the process proceeds to S63, and if NO, the routine is ended.

In S63, the combustion chamber volume V,
, Vj. Search for 1 from the map.

In step 364, the coefficient F is calculated from the combustion chamber volume VJ+Vj*1 and the polytropic coefficient PNE using the following equation.

F = (V j / V J-1) "In ES65, j
From the cylinder pressure Pij corrected at -60 or the previously predicted cylinder pressure Pij and the coefficient F, the cylinder pressure P at a rank angle position that is larger by a predetermined crank angle is calculated.
The prediction is made by calculating i(j+I) using the following equation.

P□j. +1=P ijX F In S66, j is incremented by 1 and the process returns to S62.

In this way, the cylinder pressure P,. . 1) is predicted from j=60 to 90 and stored in the RAM 19.

Incidentally, as shown in the flowchart of FIG. 14, the coefficient F may be retrieved from the map in S67 to predict the cylinder pressure Pj.

Next, the calculation routine of Pl will be explained according to the flowchart of FIG. 1θ.

In S71, the initial value of the total work amount is set to zero, and S
In step 72, the initial value of j is set to zero.

In S73, it is determined whether j is less than 91. If YES, the process advances to S74, and if NO, the process advances to S79.

In S74, the in-cylinder pressure sampling period (in this example, 4
The amount of change in combustion chamber volume Δ■j at 100 °C is searched from the map in association with j.

In S75, the corrected or predicted cylinder pressure Pij is read from the RAM 19.

In step 376, the partial work amount l is determined by multiplying the retrieved combustion chamber volume Δvj and the read cylinder pressure Pij.

S7? Now, the calculated partial work amount l is added to the total work amount k set in the previous routine to obtain a new total work amount k.

In step 37B, j is incremented by 1 and the process returns to S73. In this way, the total amount of work k in the crank angle range from j=0 to j=90, that is, from θ to 360° (compression stroke and expansion stroke) is calculated.

In S79, P4''i is calculated and predicted for each cylinder by the following equation based on the total amount of work k in the crank angle range.

7"i=to/V3 (2) is the total combustion chamber volume at the crank angle of 0 to 360 degrees.

As explained above, the cylinder pressures from the cylinder pressure sensors 1 to 6 are read in the crank angle range (120° to 240°) centered on compression top dead center, and the crank angle is The in-cylinder pressure is predicted in a range of 120° before and after the range, and based on these in-cylinder pressures, Pi″i
Since it is calculated, the following effects are obtained.

That is, since the detection values of the cylinder pressure sensors 1 to 6 of a predetermined cylinder are read in a range of 120° centered on the compression top dead center, each cylinder (6 The cylinder pressure can be detected in a timely manner. Here, in this embodiment, the sampling period is set to 120 degrees because it is a six-cylinder engine, but the sampling period can be expressed as 720''/number of cylinders.

In addition, since the in-cylinder pressure detected by the in-cylinder pressure sensors 1 to 6 is corrected by the bias amount, the true in-cylinder pressure in the combustion chamber can be detected with high accuracy, thereby increasing the detection accuracy of Pi"i to the highest level. As shown in Figure 17, it can be greatly improved.
Based on the corrected cylinder pressure and the predicted cylinder pressure, Pvi is calculated in the 360° crank angle range that is the compression stroke and expansion stroke, so it is possible to reduce the combustion period due to deterioration of the combustion condition. Even if it becomes long (see the chain line in Fig. 22), P, "i can be detected with high accuracy. By the way, 1
Total P of the cycle! As shown in Fig. 16, is the sum of Pi''i in the compression stroke and expansion stroke and PH-i in the intake stroke and exhaust stroke, but P
4-i is determined by operating conditions 9 engine characteristics, etc., and the combustion state corresponds to Pll, so P
, "i" are determined in this embodiment.

Next, an example in which this device is applied to misfire determination is shown in Figs.
This will be explained based on FIG. 21. Incidentally, since the configuration is the same as that shown in FIG. 2, the explanation will be omitted.

FIG. 18 shows a first embodiment of misfire determination.

That is, in S81, the P value calculated in S79 is read out.

In S82, it is determined whether the read P and i are less than a predetermined value, and if YES, the process proceeds to S82 (N).

If so, the process advances to S84.

In S83, the misfire flag is set to 1 and stored in the RAM 19 to indicate that a misfire has occurred.

In S84, the misfire flag is set to 0 to indicate that no misfire has occurred.
It is stored in the RAM 19 as .

In this way, a misfire is determined for each cylinder.

When a misfire is determined in this way, the P
4"i can be predicted with high accuracy in accordance with the actual Pi, so the misfire determination accuracy can be greatly improved.

Next, a second embodiment of misfire determination will be described according to the flowchart of FIG. 19.

In S91, the detected engine rotational speed is read.

In S92, the basic injection t T P (corresponding to the engine load) is
=KQ/N; K is a constant, Q is the intake air flow rate, and N is the engine rotation speed).

In S93, the Pll calculated in S79 is read.

In S94, the average learning number C is searched from the map according to the detected engine rotational speed, basic injection amount T, and cylinder number. The average learning number C is stored in the RAM 19 in correspondence with the engine rotation speed and the basic injection amount TP for each cylinder.

In S95, the average value AVEP of P and i is searched from the map according to the engine rotation speed, basic injection amount Tr, and cylinder number. AVEPi'' is searched based on the engine rotation speed and basic injection amount Tp for each cylinder. They are stored in the RAM 19 in correspondence.

In S96, it is determined whether or not the retrieved average number of learning times C is less than a predetermined value. If YES, the process advances to S97; if NO, the process advances to 399.

397, the average learning number C and 3 retrieved in S94
M is calculated by multiplying 95 by the retrieved AVEP.

At step 39B, AVEPi'' is calculated based on the PH"i read at step 393, the M, and the average number of learning times C using the following simple average formula.

A V E P i” = (M 0P =.i)
/ (C+ 1 ) Meanwhile, in S99, AVEP,'' is calculated using the following weighted average formula.

AVEPi”=AVEP+”+ (Pt”i AVE
Pi'')/Y Y is a predetermined value exceeding 1.

In 5100, A calculated in 398 or S99
VEPi'' is stored in a map in correspondence with the engine rotational speed, basic injection ITP, and cylinder number.

In S101, the average learning number C is incremented by 1 and stored in the map.

In 5102, A calculated in 398 or S99
The ratio q (=
P, "i/AVEP,").

In 3103, it is determined whether the calculated ratio q is less than or equal to a predetermined value (for example, zero), and if YES, the process advances to S104, and if NO, the process advances to S105.

In 5104, a misfire flag = 1 indicates that a misfire has occurred.
It is stored in the RAM 19 as .

In 5105, a misfire flag = indicates that no misfire has occurred.
It is stored in the RAM 19 as 0.

In this way, a misfire is determined for each cylinder.

Next, a third example of misfire determination will be described according to the flowchart of FIG. 20. In the following embodiments, the same elements as the steps in FIG. 19 are given the same step numbers as in FIG. 19, and the description thereof will be omitted.

That is, in 5110, A searched in S95
The ratio between VEPso and Pl"i read in S93 (=P+"i/AVEP+") is performed.

In 5111, it is determined whether the calculated T is less than or equal to a predetermined value, and if YES, that is, the Pc amount is AVEP! When it is smaller than +, it is 3 without passing through 396-5101.
The process advances to step 102, and when the answer is No, the process advances to step S96.

In this way, when the operating state where P and i are small (poor combustion state) continues, the AVEP value is not updated, so the misfire determination accuracy can be improved.

Next, a fourth example of misfire determination will be described according to the flowchart of FIG. 21.

5121te is AVEPi” (7) Maximum value M A
OMAXPi, which searches P+ from a map based on the engine rotational speed and the basic injection amount T, is set in the same map for all cylinders in correspondence with the engine rotational speed and the basic injection amount Tp.

In 5122, MAXP is searched for AVEPi'' calculated in 398 or S99 or 399.

512 if YES.
Go to step 3, and if No, go to step 5124.

In step 5123, the calculated AVEP, + is stored in the map of the RAM 19 as MAXP.

In 5124, A calculated in 398 or S99
VEP, + and P read in S93, “i” and MAXP,
Based on this, the ratio S is calculated by the following equation.

S=MAXP=XP, "i/AVEP," 5125, determines whether the calculated ratio S is less than a predetermined value,
When , it is determined that a misfire has occurred and the process proceeds to 5104 with N.
When it is O, the process proceeds to S105.

In this way, the same MAXP is set for all cylinders, and the ratio S is calculated for each cylinder using this MAXP, so the output values of cylinder pressure sensors 1 to 6 will vary from cylinder to cylinder. Also, misfires can be determined with high precision for each cylinder.

<Effects of the Invention> As explained above, the present invention predicts the cylinder pressure in other crank angle ranges based on the cylinder pressure detected in a predetermined crank angle range, and predicts the cylinder pressure based on these cylinder pressures. Since P is calculated, P! can be detected with high precision, and thus combustion conditions such as misfire can be detected with high precision.

[Brief explanation of drawings]

Fig. 1 is a diagram corresponding to the claims of the present invention, Fig. 2 is a configuration diagram showing an embodiment of the present invention, Figs. 3 to 14 are flowcharts of the same, and Figs. A diagram for explaining, FIG. 18 is a flowchart of the apparatus of the present invention,
FIG. 21 is a flowchart of the fourth embodiment, which is applicable to misfire determination, and FIGS. 22 and 23 are diagrams for explaining the drawbacks of the conventional technology. 1 to 6...Cylinder pressure sensor -5...Control device
17...CPU 19...RAM 21.
・・Crank angle sensor

Claims (1)

[Claims] In-cylinder pressure detection means for detecting in-cylinder pressure in the engine combustion chamber;
a crank angle detection means for detecting a crank angle of the engine; a sampling means for sampling the in-cylinder pressure detection signal during a sampling period of a predetermined crank angle range; and a combustion chamber volume setting means for setting a combustion chamber volume for each crank angle. , a bias amount setting means for setting a bias amount of the in-cylinder pressure detection signal based on the combustion chamber volume in a set sampling period and the sampled in-cylinder pressure detection signal, and based on the set bias amount, In-cylinder pressure correction means corrects the sampled in-cylinder pressure detection signal to obtain the in-cylinder pressure, and in the non-sampling period, the in-cylinder pressure is calculated based on the combustion chamber volume in the non-sampling period and the corrected in-cylinder pressure. An internal combustion engine characterized by comprising: a cylinder pressure prediction means for predicting the cylinder pressure; and an effective pressure calculation means for calculating the indicated mean effective pressure based on the corrected cylinder pressure and the predicted cylinder pressure. Engine combustion state detection device.