JPH109038A - Air-fuel ratio detecting device of engine - Google Patents

Abstract

PROBLEM TO BE SOLVED: To detect the air-fuel ratio of the fuel mixed gas separately to each cylinder at a high accuracy, depending on the exhaust gas sensors installed to an exhaust pipe assembly. SOLUTION: This air-fuel ratio detecting device consists of a condition variable Y consisting of the exhaust gas equivalent ratio signal line at the exhaust pipe assembly, which is found from the exhaust gas sensors installed to the exhaust pipe assembly; a condition variable X consisting of the object equivalent ratio of the fuel mixed gas at each cylinder; and a transmission matrix C to relate the condition variable Y and the condition variable X; and a condition variable X' which consists of the actual equivalent ratio signal line of the fuel mixed gas of each cylinder is inferred depending on the condition variable Y and the transmission matrix C.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an air-fuel ratio detecting device for an engine, and more particularly, to a technology for detecting an air-fuel ratio of a combustion mixture for each cylinder based on a detection signal of an exhaust sensor attached to an exhaust pipe assembly. About.

[0002]

2. Description of the Related Art Conventionally, as a device for detecting an air-fuel ratio of a combustion mixture for each cylinder and controlling an air-fuel ratio (fuel injection amount), for example, Japanese Patent Application Laid-Open No. 59-101563 and Japanese Patent Application Laid-Open No.
And JP-A-5-44544. JP-A-59-101
No. 563 discloses a configuration in which the air-fuel ratio is detected for each cylinder by setting the timing of detecting the signal of the exhaust gas sensor of the exhaust collecting section according to the operating state.

Japanese Patent Application Laid-Open No. 7-83094 has a model that describes the behavior of an exhaust system. The output of a single exhaust sensor disposed in the exhaust system is used to calculate the empty space of each cylinder based on the model. It is configured to estimate the fuel ratio. More specifically, the air-fuel ratio of each of the cylinders is separated and extracted by assuming that the air-fuel ratio of the exhaust system collecting section is a weighted average in consideration of the temporal contribution of the air-fuel ratio of each cylinder.

Further, Japanese Patent Application Laid-Open No. 5-44544 has a configuration in which the air-fuel ratio of each cylinder is individually detected based on the signal of an in-cylinder pressure sensor attached to each cylinder.

[0005]

In Japanese Patent Application Laid-Open No. Sho 59-101563, the signal of the exhaust sensor is detected in consideration of the travel time of the exhaust gas from each cylinder. However, there is a problem that the air-fuel ratio of each cylinder cannot be accurately separated and detected because of the influence of the exhaust of other cylinders remaining in the exhaust pipe.

In Japanese Patent Application Laid-Open No. 7-83094, although the influence of residual exhaust of other cylinders is taken into account, only a time-series delay is taken into account. The difference in the residual exhaust characteristics and the difference in the sensitivity (the degree to which the exhaust of each cylinder is detected by the sensor based on the correlation between the flow of the exhaust and the sensor position) is not taken into consideration.

That is, in the case of Japanese Patent Application Laid-Open No. 7-83094, if the operating conditions are the same, the degree of contribution of each cylinder to the air-fuel ratio in the exhaust system collecting section is 40% for the most recently burned cylinder, The previous model is modeled as 30%, etc., and the configuration does not change the degree of contribution depending on the cylinder that burned most recently or the cylinder that burned before that.

However, in practice, there is a difference in residual exhaust characteristics and a difference in sensitivity among the cylinders as described above, so that the most recently burned cylinder is the # 1 cylinder even under the same operating conditions. The degree of contribution in each cylinder is different between the case where the most recently burned cylinder is the # 3 cylinder and the case where the most recently burned cylinder is the # 1 cylinder.
Even if it is correct to model 40% and the previous # 2 cylinder as 30%, if the most recently burned cylinder is # 3 cylinder, # 3 cylinder is 45% and the previous # 1 cylinder is May need to be modeled with different contributions, such as 25%.

Therefore, the shape of the exhaust system is complicated,
When the residual exhaust characteristics and the sensitivity are greatly different between cylinders, there is a problem that the air-fuel ratio detection accuracy for each cylinder is reduced.Moreover, since the residual exhaust characteristics and the sensitivity change depending on operating conditions, a wide range of operation is performed. There is a problem that it is difficult to stably detect the air-fuel ratio for each cylinder under the conditions.

On the other hand, in Japanese Patent Application Laid-Open No. 5-44544, although the air-fuel ratio of each cylinder can be detected with relatively high accuracy, it is necessary to attach an in-cylinder pressure sensor to each cylinder. There is a problem that the system cost is increased as compared with a case where the provided single exhaust sensor detects the air-fuel ratio for each cylinder. The present invention has been made in view of the above problems, and is based on a detection signal of an exhaust sensor attached to an exhaust pipe assembly, and based on a case where the shape of an exhaust system is complicated and residual exhaust characteristics and sensitivity greatly differ between cylinders. Even if there is, an object of the present invention is to provide an air-fuel ratio detection device that can accurately detect the air-fuel ratio of each cylinder.

[0011]

According to the present invention, there is provided an exhaust system which is attached to an exhaust pipe assembly of an engine and outputs a detection signal corresponding to an exhaust component concentration correlated with an air-fuel ratio of a combustion mixture. An apparatus for detecting an air-fuel ratio of a combustion mixture in a multi-cylinder engine for each cylinder using a sensor, wherein an exhaust air-fuel ratio in the exhaust pipe assembly is represented by a weight for each air-fuel ratio of a combustion mixture in each cylinder. A detection signal of the exhaust sensor based on an exhaust transmission model, which is based on an exhaust transmission model corresponding to a change in the degree of influence of each cylinder over time and corresponding to a difference in the degree of influence between the cylinders. From this, the air-fuel ratio of the combustion mixture in each cylinder is estimated.

That is, the exhaust air-fuel ratio in the exhaust pipe collecting section is determined by the fact that the combustion of each cylinder is performed in a time-series manner. , There is a difference in the residual exhaust characteristics and the sensitivity, and the degree of the influence also differs depending on which cylinder is affected with a time-series delay. Therefore, an exhaust gas transmission model is set up corresponding to not only the chronological delay but also the difference in residual exhaust characteristics and sensitivity between the cylinders. The air-fuel ratio of the gas is detected for each cylinder.

As the exhaust gas sensor, a sensor sensitive to the concentration of O ₂ or CO in the exhaust gas as an exhaust gas component can be used.
A so-called stoichiometric sensor whose detection signal changes off may be used, or a wide-range air-fuel ratio sensor that can detect the air-fuel ratio in a wide range. The air-fuel ratio includes parameters indicating the air-fuel ratio, such as the fuel-air ratio, the excess air ratio, and the equivalent ratio (the reciprocal of the excess air ratio), in addition to the air-fuel ratio.

According to the present invention, a plurality of types of the exhaust transmission models are provided in advance according to the rotational speed and the load of the engine, and the exhaust transmission model corresponding to the rotational speed and the load of the engine at that time is selected. The configuration was used. According to such a configuration, a change in chronological delay characteristics due to a change in engine speed and load, a change in residual exhaust characteristics,
The air-fuel ratio of the combustion air-fuel mixture can be estimated in response to a change in sensitivity or the like.

In the third aspect of the present invention, when the target air-fuel ratio of the combustion air-fuel mixture is changed independently for each cylinder in the steady operation state of the engine, the exhaust gas sensor detects the exhaust transmission model. The learning is performed based on the amount of change in the exhaust air-fuel ratio in the exhaust pipe assembly. That is, if the air-fuel ratio of the combustion mixture of a specific cylinder is changed, the air-fuel ratio of another cylinder is affected according to the residual exhaust characteristics and sensitivity characteristics of the specific cylinder, and the air-fuel ratio changes. It is possible to know the degree of influence on each cylinder from the amount of change in the exhaust air-fuel ratio before and after that, and to learn the difference in residual exhaust characteristics and sensitivity for each cylinder by detecting the degree of influence for each cylinder. become.

According to a fourth aspect of the present invention, there is provided a state variable Y consisting of an exhaust air-fuel ratio signal sequence in the exhaust pipe collecting section detected by the exhaust sensor, and a target air-fuel ratio signal sequence of a combustion mixture for each cylinder. And a transfer matrix C corresponding to the exhaust transfer model relating the state variable Y to the state variable X. The combustion variable for each cylinder is obtained from the state variable Y and the transfer matrix C. The configuration is such that the state variable X ′ composed of the actual air-fuel ratio signal sequence of the air-fuel mixture is calculated.

According to this configuration, the transfer matrix C relating the state variable X consisting of the target air-fuel ratio signal train and the state variable Y consisting of the exhaust air-fuel ratio signal train is such that the combustion air-fuel ratio of each cylinder corresponds to the exhaust air-fuel ratio. The degree of influence is set based on an element indicating each cylinder for which the exhaust air-fuel ratio is to be obtained (the transfer matrix C is a matrix including rows corresponding to the number of cylinders and columns corresponding to the number of cylinders). This makes it possible to calculate cylinder-by-cylinder air-fuel ratio estimation corresponding to differences in residual exhaust characteristics and sensitivity between cylinders based on simple equations. Further, the degree of influence of each cylinder can be easily set as the transfer matrix C, and the degree of influence of each cylinder can be read from the transfer matrix C.

According to the fifth aspect of the present invention, the transfer matrix C
Are set so that the sum of each row and each column becomes 1. That is, the sum of the rows and columns of the transfer matrix is 1
By doing so, the degree of influence of the air-fuel ratio of the combustion air-fuel ratio of each cylinder is correctly modeled as to how much the air-fuel ratio of the exhaust air-fuel ratio is formed. Has been correctly modeled as to what degree of influence is distributed to each exhaust air-fuel ratio.

[0019]

According to the first aspect of the present invention, not only the chronological delay but also the difference in residual exhaust characteristics and the difference in sensitivity among the cylinders due to the shape of the exhaust system are taken into consideration, and the exhaust pipe collecting section is considered. There is an effect that the air-fuel ratio of the combustion mixture can be accurately estimated for each cylinder based on the exhaust sensor attached to the cylinder.

According to the second aspect of the present invention, it is possible to accurately detect the air-fuel ratio for each cylinder in the entire operation range in response to a change in sensitivity due to a change in operating conditions of the engine. is there. According to the third aspect of the invention, by learning the exhaust transmission model, there is an effect that the detection accuracy of the air-fuel ratio for each cylinder can be stably maintained in response to changes over time.

According to the fourth aspect of the present invention, the air-fuel ratio detection for each cylinder in consideration of the difference in sensitivity among the cylinders due to the shape of the exhaust system and the like can be performed with a simple calculation and with high accuracy, and the transmission characteristics can be improved. Can be simply represented. According to the fifth aspect of the invention, there is an effect that the transfer matrix is normalized and the accuracy of estimating the air-fuel ratio for each cylinder can be secured.

[0022]

Embodiments of the present invention will be described below. In FIG. 1 showing a system configuration of the embodiment, an engine 1 is a four-cylinder four-cycle engine. An engine control module (hereinafter abbreviated as ECM) 2 for controlling the engine 1 includes a CPU 3, a RAM 4,
It includes a ROM 5, a timer 6, an I / O port 7, and the like.

Detection signals from various sensors are input to the ECM 2 through the I / O port 7. The various sensors include an engine 1
An air flow meter 8 for detecting an intake air amount of the engine, a throttle sensor 10 for detecting an opening degree of a throttle valve 9, a water temperature sensor 11 for detecting a cooling water temperature of the engine 1, a crank angle sensor 12 for detecting a crank angle, and an intake manifold 13. And an exhaust sensor 14 that outputs a detection signal corresponding to the exhaust component concentration is provided.

It should be noted that the exhaust sensor 14 detects the oxygen concentration in the exhaust gas in a region leaner than the stoichiometric air-fuel ratio.
In the rich region, a wide range air-fuel ratio sensor that detects the exhaust air-fuel ratio over a wide range by outputting a signal proportional to the CO concentration may be used, or it may respond to the oxygen concentration in the exhaust gas, and a low signal in the lean region. May be a stoichiometric sensor that outputs a high signal in a rich region, and any known sensor that outputs a detection signal corresponding to an exhaust component concentration correlated with the air-fuel ratio of the combustion mixture. May be used.

The ECM 2 controls the amount of fuel injected by the fuel injection valves 15 attached to the intake ports of the respective cylinders based on the detection signals from the various sensors, and faces the combustion chambers of the respective cylinders. The ignition timing by the attached spark plug 16 is controlled. In addition, the ECM2
In the air-fuel ratio control using the fuel injection amount by the fuel injection valve 15 as a control target, the actual air-fuel ratio of the combustion mixture is estimated for each cylinder based on the detection signal from the exhaust sensor 14, and the estimation result The fuel injection amount is controlled independently for each cylinder so that the actual air-fuel ratio of each cylinder matches the target air-fuel ratio.

Here, an ECM using the exhaust sensor 14 is described.
2 will be described.
In the present embodiment, since the air-fuel ratio is controlled by the fuel injection amount control based on the stoichiometric air-fuel ratio, the air-fuel ratio is set to the equivalent ratio (the ratio between the fuel-air ratio of the combustion mixture and the stoichiometric fuel-air ratio. , The reciprocal of the excess air ratio λ), but any of the air-fuel ratio, the fuel-air ratio, and the excess air ratio may be used.

As in the present embodiment, the exhaust manifold 13
When the exhaust sensor 14 is attached to the gathering portion of the cylinders, the exhaust of each cylinder cannot be completely separated and detected by the exhaust sensor 14 due to the influence of the residual exhaust in the exhaust manifold 13. Even if the detection signal of the exhaust sensor 14 is sampled at a timing corresponding to the exhaust stroke, the sampled detection signal includes other cylinders (# 2 to # 4).
If the equivalent ratio varies among the cylinders due to the influence of the exhaust of the cylinder, the equivalent ratio of the cylinder (# 1 cylinder) to be detected is erroneously detected due to the influence of the variation.

The effect of the residual exhaust between the cylinders varies depending on operating conditions (engine speed, load), and the residual exhaust characteristics vary among the cylinders and the sensitivity between the cylinders varies depending on the shape of the exhaust manifold. Because they are different
The degree of influence differs for each cylinder. For example, if the cylinder that was previously in the combustion stroke is the # 1 cylinder, the degree of influence on the current combustion cylinder is 25%, whereas the degree of influence on the previous cylinder is # 3. In the case of a cylinder, the influence degree between cylinders cannot be accurately estimated only by time-series weighting according to the combustion history, for example, the influence degree becomes 20%.

Therefore, in the present embodiment, the equivalence ratio detection for each cylinder is performed in consideration of the variation in the residual exhaust characteristics and the variation in the sensitivity among the cylinders as described below.
Note that the sensitivity indicates the degree to which the exhaust gas is detected by the exhaust sensor 14.For example, when the exhaust gas of a certain cylinder mainly passes through a portion of the exhaust sensor 14 away from the sensor element, the sensitivity becomes low, and conversely, However, when the exhaust of a certain cylinder intensively passes near the sensor element of the exhaust sensor 14, the sensitivity increases.

First, the difference in the degree of influence between the cylinders as described above is represented by a determinant as follows.
In addition, Y is a state variable composed of an exhaust equivalence ratio signal sequence yi (i = 1 to 4) for each cylinder detected based on the detection signal of the exhaust sensor 14, and the target equivalence ratio of the combustion mixture for each cylinder is Y. A state variable composed of the signal sequence xi (i = 1 to 4) is represented by X, and a transfer matrix relating the state variable Y and the state variable X is represented by C (see FIGS. 2 and 3). .

[0031]

(Equation 1)

The equivalence ratio Y (k) of the exhaust pipe collecting section is calculated as follows: Y (k) = (CX (k) + α * CX (k−1) + β * CX (k−)
2) + ...) / (1 + α + β +...) (K: discrete parameter representing time, α, β: residual exhaust ratio), and even if the equivalence ratio varies between cylinders, steady state Then, the equivalent ratio of each cylinder can be considered to be constant, and X (k) = X (k-1) = X (k-2) =
..., as shown in the above equation 1, Y = C * X
The exhaust transmission characteristics can be modeled as

Here, when the determinant is expanded, y1 = c11 * x1 + c12 * x3 + c13 * x4 + c14 * x
2 y3 = c21 * x1 + c22 * x3 + c23 * x4 + c24 * x
2 y4 = c31 * x1 + c32 * x3 + c33 * x4 + c34 * x
2 y2 = c41 * x1 + c42 * x3 + c43 * x4 + c44 * x
2 and the terms on the right-hand side of the above equation are rearranged in the order of newest combustion history according to the combustion history (ignition order: # 1 → # 3 → # 4 → # 2). * X
3 y3 = c22 * x3 + c21 * x1 + c24 * x2 + c23 * x
4 y4 = c33 * x4 + c32 * x3 + c31 * x1 + c34 * x
2 y2 = c44 * x2 + c43 * x4 + c42 * x3 + c41 * x
It becomes 1.

Here, when considering only the time-series delay due to the combustion history, the element of the transfer matrix C is c11.
= C22 = c33 = c44, c14 = c21 = c32 = c43, c13 =
c24 = c31 = c42, c12 = c23 = c34 = c41, but as described above, due to the variation of the residual exhaust characteristics and the sensitivity among the cylinders, the elements of the transfer matrix C are not necessarily related to each other. Cannot be set to.

Therefore, for example, even if the weights (c11, c22, c33, c44) are assigned to the same nearest combustion cylinder, the weights are individually set in accordance with the residual exhaust characteristics and the sensitivity of the cylinder. In this case, Y = C * X
Accordingly, it is possible to set an exhaust transmission model in consideration of not only a time-series delay but also a difference in residual exhaust characteristics and sensitivity between cylinders, and based on a detection signal of the exhaust sensor 14 according to the exhaust transmission model. The actual equivalence ratio of the combustion mixture in each cylinder can be accurately estimated.

The transfer matrix C is obtained as follows. First, the target equivalent ratio of the combustion air-fuel mixture is, for example, x1 =
Each state variable when x2 = x3 = x4 = 1 is represented by X =
X0, Y = Y0, and the target equivalent ratio is x1 = 1 + Δ
x, x2 = x3 = x4 = 1, and each state variable when the target equivalent ratio of only the first cylinder # 1 is changed by Δx is X
= X1, Y = Y1.

Here, in the steady state, the first cylinder # 1
Can be regarded as constant except for the equivalence ratio of

[0038]

(Equation 2)

And the first column of the transfer matrix C, ie, the first column
Each element multiplied by the equivalent ratio of cylinder # 1 is

[0040]

(Equation 3)

Is obtained. Similarly, x3 = 1 + Δ
Assuming that the state variables when x, x4 = 1 + Δx x2 = 1 + Δx are X3, Y3, X4, Y4, X2, and Y2, respectively,

[0042]

(Equation 4)

Thus, all the elements of the transfer matrix C are obtained. The above operations and calculations determine how much the change in the target equivalence ratio of a specific cylinder affects the exhaust equivalence ratio of each cylinder, and as a result, the residual exhaust characteristics between the cylinders and the like. A difference in sensitivity will be required. The setting of the transfer matrix C by the above operation and calculation may be performed representatively for each type of engine, or may be performed for each engine on an actual machine (on-board).

Then, the RAM or RO of the ECM2 is used.
M is the transfer matrix C (or inverse matrix C) ^-1)
The actual exhaust equivalent of each cylinder calculated from the exhaust sensor 14
Ratio based on the transfer matrix C (exhaust transfer model)
The actual equivalent ratio X '(X' = C) of the combustion mixture in the cylinder^-1* Y)
It can be estimated with high accuracy. The transfer matrix C
Varies with the engine speed and load,
For each area divided by engine speed and load,
When the engine is in a steady state,
Each time, as shown in FIG.
It is necessary to find the matrix C.

When the elements of the state variables X and Y are arranged in the order of ignition (# 1 → # 3 → # 4 → # 2) as described above,
In the transfer matrix C, large numbers are concentrated near the diagonal elements (c11, c33, c44, c22), which is advantageous in calculating the inverse matrix C- ¹ . Next, the state of the equivalent ratio detection control for each cylinder by the ECM 2 will be described with reference to the flowchart of FIG.

At S50, the rotational speed Ne of the engine 1 calculated based on the detection signal of the crank angle sensor 12 and the intake air amount Qa detected by the air flow meter 8 are read. In S51, the read engine speed Ne is read.
And a basic fuel injection amount (basic fuel injection pulse width) Tp (Tp
= K × Qa / Ne, where K is a constant).

In S52, the target exhaust equivalence ratio Y (y is referred to by referring to a map in which the target exhaust equivalence ratio is stored for each area divided according to the engine speed Ne and the basic fuel injection amount Tp.
1 = y2 = y3 = y4 = y). In S53, FIG.
As shown in (1), the transfer matrix C corresponding to the current rotational speed Ne and the load is referred to from a map in which the transfer matrix C is stored in advance for each operation region divided by the engine rotational speed Ne and the load.

In S54, the target equivalent ratio X of the combustion mixture is determined as X = C ^-1 * Y from the target exhaust equivalent ratio Y and the inverse matrix C ^-1 of the transfer matrix C corresponding to the current operating condition. Ask. In S55, the detection signal Ip from the exhaust sensor 14 is read, and in S56, the actual exhaust equivalent ratio Y 'for each cylinder is obtained. The actual exhaust equivalent ratio Y ′ is obtained by sampling the detection signal of the exhaust sensor 14 at a predetermined sampling timing that is synchronized with the exhaust stroke of each cylinder. Specifically, for example, the detection signal of the exhaust sensor 14 is sampled between 80 ° and 90 ° of the exhaust BTDC and the exhaust B
As an average value of the detection results by the exhaust sensor 14 between 80 ° and 90 ° TDC (when the detection is performed as an excess air ratio, the equivalent ratio is obtained as the reciprocal thereof), the actual exhaust equivalent ratio of each cylinder is calculated. Detection is performed in chronological order according to the ignition order (see FIG. 4).

In step S57, the actual equivalent ratio of the combustion mixture in each cylinder is determined based on the actual exhaust equivalent ratio Y 'for each cylinder and the inverse matrix C ^-1 of the transfer matrix C corresponding to the current operating conditions. X '
(X ′ = C ⁻¹ * Y ′) is calculated. When the actual equivalent ratio X ′ for each cylinder is calculated in this way, the ECM 2 controls the fuel injection amount independently for each cylinder based on the calculation result.

Next, the learning control of the transfer matrix C will be described with reference to FIG.
This will be described with reference to the flowchart of FIG. The transfer matrix C
The learning control means that the elements of the transfer matrix C are recalculated and the map data is sequentially updated in order to cope with the change over time in the residual exhaust characteristics and the sensitivity. In S60, the water temperature detected by the water temperature sensor 11 is equal to or higher than the predetermined temperature, and the exhaust sensor 14 is in an active state, that is, a fully warmed state, and the feedback control using the exhaust sensor 14 is possible. It is determined whether or not.

If it is determined in S60 that the water temperature is equal to or higher than the predetermined temperature and that the exhaust sensor 14 is in the active state, the program proceeds to S60.
In 61, the absolute value of the change amount ΔNe of the rotation speed Ne, and
The steady operation state of the engine 1 is determined based on whether or not the absolute value of the variation ΔTp of the basic fuel injection amount Tp is equal to or smaller than a predetermined value. When the vehicle is in a steady operation state, S
At 62, a learning region in the map of the transfer matrix C is determined based on the engine speed Ne and the basic fuel injection amount Tp.

At S63, the output of the exhaust sensor 14 is read in rotation synchronization. In S64, based on the output of the exhaust sensor 14 read out, the state variable Y0, that is, the actual exhaust equivalent ratio for each cylinder in a state before the cylinder-by-cylinder correction of the injection amount (target equivalent ratio) is obtained. In S65, the injection amount of the first cylinder # 1 is foot-forward controlled so as to change the target equivalent ratio of the combustion mixture of the first cylinder # 1 by Δx1.

In step S66, the exhaust gas sensor after the injection amount correction is performed.
Read 14 outputs. At S67, based on the output of the exhaust sensor 14 read at S66, the state variable Y1, that is, the actual exhaust equivalent ratio for each cylinder after the correction of the injection amount (equivalent ratio) for each cylinder is obtained. In S68,

[0054]

(Equation 5)

Thus, each element of the first column of the transfer matrix C is calculated. By repeating the same steps in S69 to S80, each element of the second, third, and fourth rows of the transfer matrix C is calculated. Then, in S81, a transfer matrix C corresponding to the current engine speed and load is set. The elements calculated in S69, S72, S76 and S80 are not directly used as the elements of the transfer matrix C, but the causality of the actual equivalent ratio X and the actual exhaust equivalent ratio Y of the combustion mixture of the engine is used. In consideration of the above, it is preferable to normalize by repeating the following equation so that the sum of each row and each example of the transfer matrix C becomes 1 respectively.

[0056]

(Equation 6)

In the above-described embodiment, a four-cylinder engine is taken as an example. However, any structure may be used as long as the exhaust of each cylinder is discharged in a time-series manner to an exhaust sensor attached to an exhaust pipe assembly. It is clear that the number of cylinders is not limited. Further, the configuration may be such that each group is divided into a plurality of groups including two or more cylinders, and the exhaust pipes are assembled for each group.

[Brief description of the drawings]

FIG. 1 is a system configuration diagram of an engine according to an embodiment.

FIG. 2 is a diagram showing a correlation between state variables in the embodiment.

FIG. 3 is a diagram showing a correlation between a transfer matrix indicating an exhaust transfer model and a state variable in the embodiment.

FIG. 4 is a view showing detection characteristics of an exhaust equivalent ratio based on an exhaust sensor in the embodiment.

FIG. 5 is a flowchart showing a state of detecting an actual equivalent ratio of a combustion mixture based on an exhaust transfer model.

FIG. 6 is a diagram showing a map according to operating conditions of a transfer matrix.

FIG. 7 is a flowchart showing a state of learning control of a transfer matrix (an exhaust transfer model).

[Explanation of symbols]

 1 Engine 2 ECM 3 CPU 4 RAM 5 ROM 6 Timer 7 I / O Port 8 Air Flow Meter 9 Throttle Valve 10 Throttle Sensor 11 Water Temperature Sensor 12 Crank Angle Sensor 13 Intake Manifold 14 Exhaust Sensor 15 Fuel Injection Valve 16 Spark Plug

Claims (5)

[Claims] 1. An air-fuel ratio of a combustion mixture of a multi-cylinder engine using an exhaust sensor attached to an exhaust pipe assembly of the engine and outputting a detection signal corresponding to an exhaust-gas concentration correlated with the air-fuel ratio of the combustion mixture. Is an exhaust transmission model in which the exhaust air-fuel ratio in the exhaust pipe assembly is expressed by weighting the air-fuel ratio of the combustion mixture in each cylinder, and the time series of each cylinder Estimating an air-fuel ratio of a combustion mixture of each cylinder from a detection signal of the exhaust sensor based on an exhaust transmission model set corresponding to a change in the degree of influence and corresponding to a difference in the degree of influence between the cylinders. An air-fuel ratio detection device for an engine. 2. A plurality of exhaust transmission models are provided in advance according to the engine speed and load, and an exhaust transmission model corresponding to the engine speed and load at that time is selected and used. The engine air-fuel ratio detection device according to claim 1. 3. When the target air-fuel ratio of the combustion air-fuel mixture is changed independently for each cylinder in the steady-state operating state of the engine, the exhaust gas is detected by the exhaust sensor. 3. The engine air-fuel ratio detecting device according to claim 1, wherein learning is performed based on a change amount of the air-fuel ratio. 4. A state variable Y consisting of an exhaust air / fuel ratio signal train detected by the exhaust sensor at the exhaust pipe collecting section, and a state variable X consisting of a target air / fuel ratio signal train of a combustion mixture for each cylinder. , A transmission matrix C corresponding to the exhaust gas transmission model relating the state variable Y and the state variable X,
4. A state variable X 'comprising an actual air-fuel ratio signal sequence of a combustion mixture for each cylinder is calculated from the state variable Y and the transfer matrix C. Engine air-fuel ratio detector. 5. An air-fuel ratio detecting device for an engine according to claim 4, wherein the sum of each row and each column of said transfer matrix C is set to be 1.