Control device for internal combustion engines

Abstract

To provide an internal combustion engine control device capable of accurately learning a degree of clogging of a throttle valve in a wide opening region while reducing a computational load and a memory capacity.SOLUTION: An internal combustion engine control device calculates a clogging ratio KTHC using a first flow rate function (KTH when a throttle valve is new and KTH with maximum clogging) when a degree of clogging of a throttle valve 5 is in a reference state and a second flow rate function (estimated KTH) estimated on the basis of an intake air amount GAIR. The internal combustion engine control device is configured to: acquire sample points obtained by combining the second flow rate function with throttle valve openings TH for respective predetermined periods (step 1); calculate a learning point by averaging a plurality of sample points for each predetermined opening region (Fig.8); calculate coefficients a to c in an approximation function (formula (8)) of second flow rate function characteristics (estimated KTH characteristics) through the least-square method on the basis of the plurality of learning points (step 41); and calculate the clogging ratio KTHC on the basis of the second flow rate function characteristics approximated with the approximation function using the coefficients a to c and the first flow rate function.SELECTED DRAWING: Figure 2

Description

Technical Field

[0001] The present invention relates to a control device for an internal combustion engine, and particularly to a control device that controls the internal combustion engine according to the degree of clogging of a throttle valve.

Background Art

[0002] As a conventional control device for this type of internal combustion engine, for example, the one disclosed in Patent Document 1 is known. In this control device, as a parameter representing the degree of clogging of deposits in the opening of the throttle valve (hereinafter referred to as "the degree of clogging of the throttle valve"), the clogging rate of the throttle valve is calculated and learned during the operation of the internal combustion engine, and is used for control of the intake air amount and the like. Therefore, accurately learning the clogging rate is important for improving fuel consumption and improving energy efficiency.

[0003] The method for calculating and learning the clogging rate of the throttle valve in Patent Document 1 is as follows. First, during the idle operation of the internal combustion engine, every predetermined time, an estimated KTH (estimated actual flow function) is calculated based on the intake air amount detected by an air flow meter and the like, and is stored as a sample point together with the throttle valve opening. When the operation of the internal combustion engine ends, for each predetermined opening region of the throttle valve, a representative point 1 representing the estimated KTH is calculated by applying the least squares method to a plurality of stored sample points. Next, a representative point 2 for each opening region is calculated by weighted averaging of these representative points 1 and the representative points obtained in the previous operation cycle. Then, an approximate function representing the relationship between the estimated KTH and the throttle valve opening is obtained by applying the least squares method to the calculated plurality of representative points 2.

[0004] On the other hand, as other flow functions that serve as a basis for calculating the throttle valve clogging rate, the new KTH, which is the flow function when no deposits have accumulated on the throttle valve, and the maximum clogging KTH, which is the flow function when the throttle valve has accumulated the maximum amount of deposits, are pre-set over the full throttle valve opening range. Then, the clogging rate is newly calculated and learned from the relationship between these new KTH and maximum clogging KTH and the estimated KTH calculated in the current operating cycle. The clogging rate learned in this way is used for controlling the internal combustion engine, for example, by setting the target throttle valve opening. [Prior art documents] [Patent Documents]

[0005] [Patent Document 1] Patent No. 6768031 [Overview of the project] [Problems that the invention aims to solve]

[0006] In the conventional control device described above, all estimated KTH values ​​calculated at predetermined intervals during the operation of the internal combustion engine are stored as sample points. At the end of the operation of the internal combustion engine, the least squares method is applied to the numerous stored sample points to calculate representative point 1. This may result in an enormous amount of memory capacity required to store the numerous sample points. Furthermore, it is necessary to calculate representative point 1 using the least squares method for each of the multiple throttle valve opening ranges, and to derive an approximate function representing the estimated KTH across the entire opening range using the least squares method, which increases the computational load.

[0007] The present invention was made to solve the above-mentioned problems, and aims to provide a control device for an internal combustion engine that can learn the degree of clogging of the throttle valve accurately over a wide opening range while reducing the computational load and memory capacity required for that learning, and can improve fuel efficiency and energy efficiency by controlling the internal combustion engine using the learned degree of clogging. [Means for solving the problem]

[0008] To achieve this objective, the control device for an internal combustion engine according to claim 1 includes a clogging parameter calculation means that calculates a clogging parameter (clogging rate KTHC) representing the degree of clogging of the throttle valve 5, which is provided in the intake passage 11 of the internal combustion engine 3, using a first flow function (KTH when new, KTH when fully clogged in the embodiment (hereinafter the same applies in this section)) when the degree of clogging of the throttle valve 5 is in a reference state and a second flow function (estimated KTH) estimated based on the intake air amount GAIR detected by the airflow meter 22. The system includes an ECU2 (Step 8 in Figure 2, Figure 12) and a control means (ECU2, Figure 17) that controls the internal combustion engine 3 using the calculated clogging parameters. The clogging parameter calculation means calculates a second flow function at predetermined intervals during the operation of the internal combustion engine 3 and acquires a sample point acquisition means (ECU2, Step 1 in Figure 2) that acquires a sample point which is a combination of the second flow function and the opening degree of the throttle valve 5 (throttle valve opening degree TH). For each of a predetermined number of opening degree regions of the throttle valve 5, it acquires a plurality of sample points belonging to the opening degree region. Each time a sample point is acquired, By averaging , sample points A learning point calculation means (ECU2, step 20 in Figure 5, Figure 8) calculates average values ​​(throttle valve opening average value THave(n+1), KTH average value KTHave(n+1)) and calculates learning points (THave, KTHave) based on these average values, and multiple opening ranges Each against calculation Released studiesThe system includes a coefficient calculation means (ECU2, step 7 in Figure 2, step 41 in Figure 10) that calculates coefficients a to c of a predetermined polynomial approximation function (equation (8)) for approximating the relationship between the second flow function and the opening degree of the throttle valve 5, based on the learning point, using the least squares method, and is characterized by calculating a blockage parameter (equation (2)) based on the second flow function characteristic approximated by the approximation function using the calculated coefficients a to c and the first flow function.

[0009] In this control device, a clogging parameter representing the degree of clogging of the throttle valve is calculated using a first flow function (the flow function when the degree of clogging of the throttle valve is in a standard state) and a second flow function (a flow function estimated based on the amount of intake air detected by the airflow meter), and the internal combustion engine is controlled according to the clogging parameter.

[0010] The blockage parameter is calculated as follows. First, the sample point acquisition means calculates the second flow function at predetermined intervals during the operation of the internal combustion engine and acquires sample points which are combinations of the second flow function and the throttle valve opening. Next, the learning point calculation means calculates the multiple sample points belonging to each predetermined opening range of the throttle valve. Each time a sample point is acquired, By averaging , sample points Calculate the average value (Step 20 in Figure 5, Figure 8) Based on this average value, the learning score is calculated. Next, the coefficient calculation means calculates multiple opening regions Each against calculation Released studies Based on the input, coefficients are calculated using the least squares method. These coefficients define a predetermined polynomial approximation function for approximating the second flow function characteristic, which represents the relationship between the second flow function and the throttle valve opening. Then, the blockage parameter is calculated based on the second flow function characteristic approximated by the approximation function using the calculated coefficients, and the first flow function.

[0011] As described above, according to the invention of claim 1, a plurality of sample points are obtained for each of a predetermined plurality of opening ranges of the throttle valve. Each time a sample point is acquired,By averaging , sample points The average value is calculated, and the learning points are calculated based on this average value. This reduces the computational load and memory capacity compared to the conventional method, which calculates the representative point 1 of the estimated KTH for each opening degree region using the least squares method. In addition, the coefficients of the approximation function of the polynomial that approximates the second flow function characteristics are set for multiple opening degree regions of the throttle valve. Each against calculation Released studies Based on the learning points, the clogging parameter is calculated using the least squares method, and then calculated based on the approximated second flow function characteristics and the first flow function. This allows for accurate learning of the throttle valve clogging rate over a wide range of opening angles.

[0013] In this configuration, the learning score is calculated by sequentially averaging multiple sample scores each time a sample score is obtained. This sequential averaging is performed by calculating the current average xave(n+1) using the following equation (1), where n is the number of sample scores up to the previous step, xave(n) is the average up to the previous step, and x(n+1) is the current sample score. xave(n+1) =(n xave(n)+x(n+1)) / (n+1) ··· (1)

[0014] Thus, in the case of sequential averaging, the data required to calculate the mean value xave(n+1) is only the current sample point x(n+1), the previous sample point count n, and the mean value xave(n). Regardless of the number of sample points, it is only necessary to sequentially store the sample point count n and the mean value xave(n). This significantly reduces the memory capacity compared to the conventional method of storing all the numerous sample points and averaging them all at once.

[0015] Claim 2 The invention relating to this claim is 1In the control device for an internal combustion engine described above, the learning point calculation means calculates learning points in the normal operating mode of the internal combustion engine 3 (step 5 in Figure 2), the coefficient calculation means calculates coefficients of an approximate function based on the calculated learning points in the mode immediately after stopping (step 7 in Figure 2), the clogging parameter calculation means calculates clogging parameters based on an approximate function using the calculated coefficients in the next initial operating mode (step 8 in Figure 2), and the control means controls the internal combustion engine 3 using the clogging parameters in the next normal operating mode.

[0016] In this configuration, the calculation of learning points, the calculation of coefficients for the approximation function, the calculation of clogging parameters, and the control of the internal combustion engine are performed in the normal operation mode, the mode immediately after stopping, the next initial operation mode, and the normal operation mode, respectively. As a result, the processing load can be reduced by distributing the processing for these calculations and control in each operation mode. Furthermore, the data that needs to be passed between the operation modes of the internal combustion engine is basically only multiple learning points for each throttle valve opening range between the normal operation mode and the mode immediately after stopping, the coefficients for the approximation function between the mode immediately after stopping and the next initial operation mode, and the clogging parameters between the initial operation mode and the normal operation mode. Therefore, the storage and transfer of this data can be easily performed with a very small memory capacity.

[0017] Claim 3 The invention relating to this claim is 1 or 2 In the control device for an internal combustion engine described above, the learning point calculation means calculates a learning point for each operating cycle of the internal combustion engine 3, and the coefficient calculation means calculates the coefficient of the approximation function using the learning point calculated in the previous operating cycle as the learning point for the throttle valve opening region where the number of sample points acquired in the current operating cycle is less than a predetermined value (step 40 in Figure 10, Figure 11).

[0018] In this configuration, the calculation of the learning points is performed for each operating cycle of the internal combustion engine. Therefore, when only idling operation is performed or the operation time is short, etc., in a certain opening range of the throttle valve, almost no sample points may be obtained. In such a case, when calculating the learning points of that opening range from a small number of sample points, the accuracy of the learning points is low, and the accuracy of the approximation function calculated based on a plurality of learning points including that learning point may decrease. Or, when calculating the approximation function based only on a plurality of learning points in other opening ranges, assuming that there are no learning points in that opening range, the accuracy of the approximation is low, and there is also a risk that the accuracy of the approximation function will decrease. On the other hand, since the throttle valve clogging is caused by the deposition of deposits, the degree of clogging usually does not change rapidly.

[0019] From such a perspective, according to the present invention, when there is an opening range of the throttle valve in which the number of sample points obtained in the current operating cycle is less than a predetermined value, since a sufficient number of sample points are not obtained, as the learning points of that opening range, the learning points calculated in the previous operating cycle are used to calculate the coefficients of the approximation function. Thereby, the accuracy of the approximation by the approximation function based on a plurality of learning points and the learning accuracy of the clogging parameter based thereon can be maintained well.

[0020] Claim 4 The invention according to 3 In the control device for an internal combustion engine described in claim

[0021] Depending on the operating conditions of the internal combustion engine, the low-opening or high-opening regions of the throttle valve may be hardly used. For example, as the throttle valve is used and clogging progresses, the throttle valve opening is controlled to gradually shift to the high-opening side, resulting in zero sample points in the low-opening region. In this case, when using the learned points obtained in other opening regions to approximate the entire throttle valve opening range with an approximation function, the accuracy of the approximation in other opening regions may decrease due to being dragged down by the low-opening region where there are no learned points.

[0022] From this perspective, according to the present invention, when there is a low-opening or high-opening region of the throttle valve that is not used during the operation of the internal combustion engine, the low-opening or high-opening region is set as a learning prohibited region, and learning of the clogging parameter is prohibited. As a result, by using the learning points obtained in the opening region other than the learning prohibited region (hereinafter referred to as the "learning permitted region") and performing approximation with an approximation function targeting only the learning permitted region, the learning accuracy of the clogging parameter can be maintained well. Furthermore, the clogging parameter in the learning prohibited region can be set without any problems by using the same value as the clogging parameter calculated in the opening region adjacent to the learning prohibited region.

[0023] Claim 5 The invention relating to this claim is 1 to 4 In the control device for an internal combustion engine described in any of the above, the clogging parameter calculation means is characterized by correcting the downward-sloping portion of the second flow function characteristic approximated by the approximation function to become upward-sloping when the downward-sloping portion has a downward-sloping portion (step 57 in Figure 12, Figure 14).

[0024] The second flow function characteristic represents the relationship between the second flow function and the throttle valve opening, and therefore inherently has an upward sloping characteristic (increasing as the throttle valve opening increases). Furthermore, for example, when determining the target throttle valve opening based on the second flow function characteristic, if the second flow function characteristic slopes upward to the left (estimated KTH decreases as the TH opening increases), control hunting can occur, such as the existence of multiple solutions for the target throttle valve opening.

[0025] From this perspective, according to the present invention, when the second flow function characteristic approximated by the approximation function has a downward sloping portion, this downward sloping portion is corrected to become an upward sloping portion. As a result, the second flow function characteristic becomes an appropriate upward sloping portion, and the control hunting described above can be avoided.

[0026] Claim 6 The invention relating to this claim is 1 to 5 In the control device for an internal combustion engine described in any of the above, the sample point acquisition means is characterized by acquiring the calculated second flow function as a sample point after correcting it with respect to a predetermined reference rotational speed #NEKTC (step 11 in Figure 5, Figure 6).

[0027] The second flow function changes according to the rotational speed of the internal combustion engine, and has the characteristic of becoming larger as the rotational speed increases. Taking this characteristic into consideration, according to the present invention, the calculated second flow function is corrected based on a predetermined reference rotational speed and then acquired as a sample point. This correction allows the second flow function calculated under different internal combustion engine rotational speed conditions to be unified and converted to the value at the reference rotational speed, thereby effectively compensating for the variation in the second flow function due to rotational speed.

[0028] Claim 7 The invention relating to this claim is 1 to 6The control device for an internal combustion engine described in any of the above further comprises a peeling determination means (ECU2, step 60 in Figure 12) for determining whether or not peeling of deposits from the throttle valve 5 has occurred, and the clogging parameter calculation means is characterized in that, when it is determined that peeling has occurred, it uniformly corrects the clogging parameter to the decrease side regardless of the opening degree of the throttle valve 5 (step 61 in Figure 12, Figure 16).

[0029] When deposits detach from the throttle valve, the second flow function increases sharply, and the degree of throttle valve clogging decreases sharply at any opening. Taking this characteristic into consideration, according to the present invention, when it is determined that detachment has occurred from the throttle valve, the clogging parameter is uniformly corrected to the decrease side regardless of the throttle valve opening. This makes it possible to appropriately correct the clogging parameter in response to the detachment of deposits from the throttle valve. [Brief explanation of the drawing]

[0030] [Figure 1] This diagram schematically shows an internal combustion engine and control device to which the present invention is applied. [Figure 2] This flowchart shows the process for calculating the clogging rate of the throttle valve. [Figure 3] This figure shows the relationship between estimated KTH, KTH when new, KTH at maximum blockage, and blockage rate. [Figure 4] This diagram illustrates the relationship between the throttle valve opening range, sample points, and learning points. [Figure 5] This is a flowchart showing the process for calculating learning points. [Figure 6] This diagram illustrates the method for correcting the rotational speed for a sample point (estimated KTH). [Figure 7] This diagram illustrates a method for limiting the sample points (estimated KTH) to an acceptable range when the sample points are noisy points with large errors. [Figure 8] This is a flowchart showing the sequential averaging process of sample points. [Figure 9]This diagram illustrates how to set the learning restriction region on the low-opening side of the throttle valve. [Figure 10] This flowchart shows the process for calculating the coefficients of the approximation function. [Figure 11] This diagram illustrates how to supplement learning points in the throttle valve opening range where sample points are not available. [Figure 12] This flowchart shows the process for calculating the coefficients of the approximation function. [Figure 13] This diagram illustrates how to set the jamming rate during the initial learning phase. [Figure 14] This diagram illustrates a method for correcting the upward trend in estimated KTH characteristics. [Figure 15] This diagram illustrates how to set the jamming rate when a learning prohibition region is set on the low-opening side. [Figure 16] This diagram illustrates how to set the clogging rate when peeling occurs in the throttle valve. [Figure 17] This flowchart shows the process for calculating the target throttle valve opening. [Figure 18] This diagram illustrates a method for setting the target throttle valve opening degree according to the target flow rate function. [Modes for carrying out the invention]

[0031] Preferred embodiments of the present invention will now be described in detail with reference to the drawings. The internal combustion engine (hereinafter referred to as "engine") 3 shown in Figure 1 is mounted as a power source in a vehicle (not shown), for example, and has a plurality of cylinders (not shown). The engine 3 is provided with a fuel injection valve 4 for injecting fuel, an intake passage 11 through which air (fresh air) flows, an exhaust passage 12 through which exhaust gas flows, and an EGR passage 13 through which a portion of the exhaust gas is recirculated to the intake passage 11 as EGR gas.

[0032] The intake passage 11 is connected to each cylinder of the engine 3 via multiple branches of the intake manifold. The exhaust passage 12 is connected to each cylinder of the engine 3 via multiple branches of the exhaust manifold. The EGR passage 13 bypasses the cylinders of the engine 3 and is connected to the intake passage 11 and the exhaust passage 12.

[0033] An EGR valve 14 is provided in the EGR passage 13. The EGR valve 14 is connected to an EGR motor 15, which is, for example, a DC motor. The opening degree of the EGR valve 14 is controlled by adjusting the duty cycle of the drive current supplied to the EGR motor 15 using an ECU (electronic control unit) 2, which will be described later, and thereby the flow rate of the EGR gas is controlled.

[0034] A rotatable throttle valve 5 is provided within the intake passage 11. The throttle valve 5 is connected to the TH motor 7 as an actuator via a drive mechanism 8. The drive mechanism 8 is a combination of multiple gears (not shown). The TH motor 7 is, for example, a DC motor. The opening degree of the throttle valve 5 is controlled by adjusting the duty cycle of the drive current supplied to the TH motor 7 in the ECU 2, thereby controlling the amount of intake air drawn into the cylinders of the engine 3.

[0035] The intake passage 11 is provided with a stopper 10 to restrict the rotation of the throttle valve 5 toward the closed position. The opening degree at which the throttle valve 5 is in contact with the stopper 10 is the fully closed position of the throttle valve 5. When the TH motor 7 is not driven, the throttle valve 5 is positioned at an opening degree slightly open from the fully closed position.

[0036] A throttle valve 5 is equipped with a throttle valve opening sensor 21, and an airflow meter 22 is provided upstream of the throttle valve 5 in the intake passage 11. The throttle valve opening sensor 21 detects the opening degree (throttle valve opening) TH of the throttle valve 5 and outputs the detection signal to the ECU 2. The airflow meter 22 detects the flow rate of air flowing through the intake passage 11 as the intake air volume GAIR and outputs the detection signal to the ECU 2. Based on the input intake air volume GAIR, the ECU 2 calculates the average intake air volume GAIRAVE, which is the average value of the intake air volume GAIR.

[0037] Furthermore, a first pressure sensor 23 is provided upstream of the throttle valve 5, and a second pressure sensor 24 and an intake air temperature sensor 25 are provided downstream of the throttle valve 5. The first and second pressure sensors 23 and 24 detect the pressure in the intake passage 11 upstream and downstream of the throttle valve 5 as the throttle pre-pressure PAAC and intake manifold pressure PBA, respectively. The intake air temperature sensor 25 detects the temperature (intake air temperature) TA in the intake passage 11 downstream of the throttle valve 5. These detection signals are input to the ECU 2.

[0038] Furthermore, the ECU2 receives a detection signal from the rotation speed sensor 26 representing the rotation speed of the engine 3 (hereinafter referred to as "engine rotation speed") NE, a detection signal from the EGR valve opening sensor 27 representing the opening degree of the EGR valve 14, and a detection signal from the accelerator opening sensor 28 representing the opening degree (accelerator opening degree) of the vehicle's accelerator pedal (not shown).

[0039] The ECU2 is a microcomputer consisting of a CPU, RAM, ROM, EEPROM, and input / output interfaces (none of which are shown), and is configured to operate while the engine 3 is running with the IG (ignition) switch 31 in the ON position, and immediately after the engine 3 stops with the IG switch 31 turned OFF. The RAM is provided with a ring buffer for storing sample points.

[0040] The ECU2, in response to detection signals from the various sensors 21 to 28 mentioned above, performs engine control, including control of fuel injection by the fuel injector 4 and control of intake air volume by the throttle valve 5, according to a control program stored in ROM. In this embodiment, in particular, the ECU2 calculates (learns) the clogging rate KTHC of the throttle valve 5 as a clogging parameter representing the degree of clogging of deposits at the opening of the throttle valve 5 for each operating cycle of the engine 3, and performs engine control such as setting the target throttle valve opening degree according to the calculated clogging rate KTHC. In this embodiment, the ECU2 comprises clogging parameter calculation means, control means, sample point acquisition means, learning point calculation means, coefficient calculation means, and peeling determination means.

[0041] Figure 2 shows the calculation process for the clogging rate KTHC of the throttle valve 5. As shown in the figure, this process is performed as a series of cycles, from the normal operation of the engine 3 (normal operation mode), to the period immediately after the IG switch 31 is turned OFF and the engine 3 stops (immediate stop mode), and to the initial period after the IG switch 31 is turned ON for the next operating cycle of the engine 3 (initial operation mode).

[0042] In normal operation mode, steps 1 to 4 (S1 to S4) are repeated at predetermined intervals. First, in step 1, the estimated KTH is calculated. This estimated KTH is calculated by estimating the actual flow function at the throttle valve 5 as the second flow function, and is calculated by applying the detected average intake air volume GAIRAVE, throttle pre-pressure PAAC, intake manifold pressure PBA, and intake air temperature TA to a known nozzle formula. The calculated estimated KTH is combined with the throttle valve opening TH at that time and stored as a sample point (TH, estimated KTH).

[0043] In this embodiment, in addition to the estimated KTH, the new KTH and the maximum clogging KTH are used as the first flow function reference when calculating the clogging rate KTHC of the throttle valve 5. The new KTH is the flow function under the condition that no deposits have accumulated at all in the opening of the throttle valve 5, and the maximum clogging KTH is the flow function under the condition that the maximum amount of deposits has accumulated in the opening of the throttle valve 5. The new KTH and the maximum clogging KTH are pre-mapped (not shown) according to the throttle valve opening TH and engine speed NE through experiments, etc., and stored in ROM.

[0044] The clogging rate KTHC of the throttle valve 5 is calculated using these three flow functions (KTH) by the following equation (2). KTHC=(Estimated KTH - KTH when new) / (KTH at maximum clog - KTH when new) ...(2) Furthermore, the relationship between these four parameters is shown in Figure 3.

[0045] Returning to Figure 2, in Step 2, following Step 1 above, a determination is made as to whether to allow learning of the clogging rate KTHC. As will be described later, this learning of the clogging rate KTHC (hereinafter referred to as "clogging learning" as appropriate) is performed by calculating one learning point that is a representative point of the multiple sample points for each region of the throttle valve opening degree TH (hereinafter referred to as "TH region" as appropriate) based on the multiple sample points obtained in Step 1 (see Figure 4), then calculating the coefficients of the approximation function that approximates these multiple learning points, and calculating the clogging rate KTHC based on the estimated KTH characteristics approximated by this approximation function and the above equation (2).

[0046] In Step 2, if all of the following learning conditions A to E are met, block learning is permitted and the learning permission flag F_KTHCCND is set to "1". A. The pressure ratio across the throttle valve 5 (= intake manifold pressure PBA / throttle valve front pressure PAAC) must be less than or equal to a predetermined value. B. The throttle valve opening TH is below a specified opening. C. The airflow meter 22 is activated. D. The engine speed NE must be below the specified speed. E. The change in estimated KTH (the absolute difference between the current and previous estimated KTH values) remains below a predetermined value for a predetermined period of time.

[0047] Next, in step 3, it is determined whether the learning permission flag F_KTHCCND is "1". If the answer is NO, meaning learning is not permitted, the learning points are not recalculated, and the learning points are held at the previous learning points (step 4).

[0048] On the other hand, if the answer to Step 3 is YES, meaning that learning is permitted, the process proceeds to Step 5 to calculate the learning points. Figure 5 shows this calculation process. In this process, first in Step 11, an NE (revolutions per minute) correction is applied to the sample points (estimated KTH) obtained in Step 1. This NE correction takes into account the characteristic that the estimated KTH becomes larger as the engine speed NE increases, and is performed as follows, for example (see Figure 6).

[0049] Figure 6 shows an example where the current engine speed (NE) is 1500 rpm, and the estimated KTH at this current NE is calculated in step 1 (point a in Figure (a)). Next, using the estimated KTH calculated for the current NE, the KTH when the NE was new, and the KTH when it was at its maximum blockage, the blockage rate KTHC is calculated using formula (2) above (point b (e.g., 0.2)). Next, as shown in Figure (b), while maintaining the current clogging rate KTHC at NE (point c), the KTH of a new machine and the KTH of a machine with maximum clogging at a predetermined reference rotation speed #NEKTHC are used to convert them to an estimated KTH at the reference rotation speed #NEKTHC using the following equation (3) (point d). Estimated KTH = (1-KTHC) x KTH when new + KTHC x KTH at maximum clog ...(3)

[0050] The reference rotational speed #NEKTHC is a predetermined engine rotational speed that serves as the basis for NE correction, and equation (3) is equation (2) expressed for estimated KTH. Through the above NE correction, estimated KTH obtained at different engine rotational speeds NE can be unified and converted to the estimated KTH at the reference rotational speed NEKTHC while maintaining the clogging rate KTHC, thereby effectively compensating for the variation in estimated KTH due to engine rotational speed NE.

[0051] Returning to Figure 5, step 12, following step 11 above, determines whether the initial learning of the clogging rate KTHC has been completed. If the answer is NO, for example, if this corresponds to the first operating cycle after replacing the ECU2, and this clogging learning corresponds to the initial learning, proceed directly to step 14, which will be described later.

[0052] On the other hand, if the answer to step 12 is YES, and this blockage learning corresponds to the period after the initial learning, then in step 13, limit processing is performed on the estimated KTH. This limit processing is performed to restrict the estimated KTH corrected in step 11 to an acceptable range when it is a noisy point with a large error, and is done, for example, as shown in Figure 7. First, the upper and lower limits of KTH are set by adding and subtracting a predetermined tolerance (for example, 8% of the estimated KTH) from the estimated KTH characteristics obtained in the previous operating cycle (D / C). Next, if the estimated KTH falls outside the acceptable range defined by the upper and lower KTH limits, this estimated KTH is considered a noise point with a large error, and the estimated KTH is limited by setting it to the upper or lower KTH limit on the side that falls outside the limit.

[0053] As described above, if the estimated KTH is a noisy point with a large error, the estimated KTH is limited by the upper or lower KTH limit, thereby suppressing the effect of variability in the estimated KTH and improving the accuracy of the learning points and the learning accuracy of the congestion rate KTHC.

[0054] Returning to Figure 5, if the answer to step 12 is NO and this blockage learning corresponds to the initial learning, or if step 13 is followed, step 14 performs a delay processing of the sample points. In this case, the sample points consist of a combination of the throttle valve opening TH and the estimated KTH corrected and / or limited in steps 11 and 13.

[0055] The purpose of this delay processing is as follows: For example, if the conditions for clogging learning are no longer met as the throttle valve opening TH increases from a stable state (hereinafter referred to as "when the learning conditions are no longer met"), the estimated KTH changes with a delay relative to the throttle valve opening TH immediately before this point, and the relationship between the two deviates from the relationship in the stable state. As a result, the learning accuracy of the clogging rate KTHC calculated based on this relationship may decrease. Considering these points, when calculating the learning points when the learning conditions are no longer met, the sample points obtained immediately before are excluded, and this is why delay processing is performed.

[0056] This delay process is performed, for example, using a ring buffer in the RAM of ECU2. Specifically, in step 14, the sample points obtained this time (TH, estimated KTH) are stored in buffer number 1 of the ring buffer (number of buffers = N), and at the same time, the sample points stored in buffer number n are shifted to buffer number n+1. By repeating this process, the buffer up to number N stores sample points obtained a predetermined time (execution interval of this process × (N-1)) before the current time. Therefore, by reading the sample points from buffer number N, they can be delayed by a predetermined time, and when the learning condition is abandoned, sample points obtained within the predetermined time immediately preceding that can be reliably excluded.

[0057] Returning to Figure 5, step 15, following step 14, determines whether or not "detachment" of the step valve 5 (detachment of deposits from the opening of the throttle valve 5) has occurred. This detachment determination is made, for example, when the difference between the current value and the previous value of the estimated KTH calculated in step 1 of Figure 2 is greater than a predetermined value for determination, and the estimated KTH has decreased sharply, it is determined that detachment has occurred.

[0058] If the answer to step 15 is YES and it is determined that peeling has occurred, the reliability of the estimated KTH calculated up to that point is presumed to be poor, and in step 16, the stored sample points are cleared.

[0059] On the other hand, if the answer to step 15 is NO, step 17 determines whether or not blockage learning is prohibited due to the detection of abnormalities in sensors such as the throttle valve opening sensor 21 or the EGR valve opening sensor 27. If the answer is YES, the system proceeds to step 16 to clear the sample points, as the accuracy of the estimated KTH calculated up to that point may be low due to the low detection accuracy of the sensors.

[0060] On the other hand, if the answer to step 17 is NO, that is, if peeling has not occurred and learning has not been prohibited due to abnormal detection of sensors such as the throttle valve opening sensor 21, then in step 18, it is determined whether the learning permission flag F_KTHCCND is "1". If the answer is NO and clogging learning is not permitted, the process proceeds to step 19, where the learning points are not calculated and the sample points are held at the previous sample points.

[0061] On the other hand, if the answer to step 18 is YES and blockage learning is permitted, in step 20, the learning score is calculated by sequentially averaging multiple sample points for each opening range of the throttle valve 5.

[0062] The reason for calculating the average of multiple sample points using sequential averaging is to reduce memory usage, as follows. First, when there are multiple data values ​​(x(1) ... x(n), x(n+1)), their average value xave(n+1) is calculated by the following formula. xave(n+1) =(x(1)+ … +x(n)+x(n+1)) / (n+1) ···(4) =(n·(x(1)+ … +x(n)) / n+x(n+1)) / (n+1) =(n xave(n)+x(n+1)) / (n+1) ··· (5)

[0063] Equation (4) is a general calculation formula that calculates the mean xave(n+1) all at once using all data values ​​(x(1) to x(n+1)). In this case, it is necessary to store all data values ​​in order to calculate the mean xave(n+1). Therefore, when used to calculate the average of sample points, the required memory capacity may become enormous when the number of sample points is very large.

[0064] In contrast, equation (5) above sequentially calculates the current average value xave(n+1) from the previous average value xave(n) and the current data value x(n+1). In this case, it is not necessary to store all the data values ​​in order to calculate the average value xave(n+1); the previous average value xave(n), the number of data points n, and the current data value x(n+1) are sufficient. Therefore, by using such sequential calculation for calculating the average of sample points, the memory capacity can be significantly reduced regardless of the number of sample points.

[0065] Figure 8 shows the sequential averaging process of sample points performed in step 20 above. In this process, first in step 21, the throttle valve opening TH(n+1) and KTH(n+1), which are the sample points for the current test, are read from the RAM ring buffer. Next, the TH region to which the read throttle valve opening TH(n+1) belongs is identified (step 22). Then, the previous average throttle valve opening THave(n), estimated KTH average KTHave(n), and the number of sample points n, which are stored for the identified TH region, are read (step 23).

[0066] Next, using the above parameters, the average throttle valve opening value THave(n+1) and the average KTH value KTHave(n+1) are calculated, respectively, by the following equations (6) and (7) based on equation (5) (steps 24 and 25). THave(n+1) = (n·THave(n)+TH(n+1)) / (n+1) ··· (6) KTHave(n+1) = (n·KTHave(n)+KTH(n+1)) / (n+1) ··· (7)

[0067] Then, by repeating the above process while the learning conditions are met, one learning point (THave, KTHave), which is a combination of THave and KTHave values, is ultimately calculated for each TH region (see Figure 4).

[0068] Returning to Figure 2, after the learning points are calculated in the normal operation mode as described above, in the mode immediately after the ignition switch 31 is turned off, in step 6, a learning prohibition area is first set. The reason for this is as follows: For example, as the throttle valve 5 is used and clogging progresses, the throttle valve 5 is controlled to a higher opening to compensate for the resulting shortage of intake air, which may result in the frequency of use in the low TH region becoming 0. In this case, the number of sample points in the low TH region becomes 0, which not only reduces the learning accuracy of the learning points in the low TH region but may also adversely affect the learning accuracy of the clogging rate KTHC in other TH regions. Therefore, learning in the low TH region is prohibited.

[0069] The setting of this learning prohibition region is performed, for example, as shown in Figure 9. First, the throttle valve opening TH corresponding to the intersection of a line representing a predetermined KTH lower limit value KTHmin, which corresponds to the usage limit on the low flow rate side of the flow function, and a line representing the estimated KTH characteristics based on multiple learning points obtained in the previous normal operating mode, is determined as the learning lower limit opening THLMT. Then, the TH region to which this learning lower limit opening THLMT belongs, along with the TH region on the lower opening side, is set as the learning prohibition region, while the TH region on the higher opening side than the learning prohibition region is set as the learning permission region.

[0070] When such domain settings are made, the approximation of estimated KTH characteristics using an approximation function based on learning points, and the learning of the congestion rate KTHC based on the estimated KTH characteristics, are prohibited in the learning prohibited domain and are only performed in the learning permitted domain. In this case, the congestion rate KTHC in each TH domain of the learning prohibited domain is set to the same value as the congestion rate KTHC in the learning permitted domain adjacent to the learning prohibited domain, as shown in Figure 15.

[0071] Returning to Figure 2, in step 7 following step 6 in the immediate post-stop mode, the coefficients of the approximation function are calculated. This approximation function approximates the estimated KTH characteristic, which represents the relationship between the throttle valve opening TH and the estimated KTH, and in this embodiment, the following quadratic polynomial (8) is used. Estimated KTH = a·TH2 + b·TH + c ··· (8)

[0072] Figure 10 shows the process for calculating these coefficients. As will be described later, in this process, when predetermined conditions are met, coefficients a to c of the approximation function are calculated (identified) based on multiple learning points calculated for each TH region, and the approximation function is updated. When the learning conditions are not met, the coefficients a to c are not newly calculated, but are held or initialized as a general rule.

[0073] In this process, first, in step 31, it is determined whether the coefficients a to c of the approximation function (hereinafter referred to as "coefficients") have already been calculated in the current immediate-stop mode. If the answer is YES, no further calculation of the coefficients is performed, and this process is terminated. In other words, the calculation of the coefficients is performed only once in the immediate-stop mode.

[0074] If the answer to step 31 is NO, the process proceeds to step 32 to determine whether the throttle valve 5 has been cleaned during a service inspection or the like. In this determination, for example, the maximum value KTHCMAX is determined from the multiple clogging rates KTHC for each TH region that have been obtained so far, and if the determined maximum value KTHCMAX is smaller than a predetermined value, it is determined that cleaning has been performed. If the answer to this is YES, the process proceeds to step 33, assuming that the actual flow function has become like new due to the cleaning of the throttle valve 5, and the coefficients of the approximation function are initialized to the equivalent values ​​of when the throttle valve 5 is new.

[0075] If the answer to step 32 is NO, the process proceeds to step 34 to determine whether or not an abnormality has been detected in the sensors, such as the throttle valve opening sensor 21. If the answer is YES, the accuracy of the sample points and learning points obtained in the previous normal operating mode may be low due to the low detection accuracy of the sensors. Therefore, in step 35, the coefficients are not recalculated, but are held at the previous values ​​obtained in the previous operating cycle.

[0076] If the answer to step 34 above is NO, it is determined whether the number of sample points obtained in the current operating cycle is insufficient and below a predetermined value in any TH region (step 36), and whether or not peeling of the throttle valve 5 has occurred (step 37). If the number of sample points is insufficient in any TH region and peeling has not occurred, the learning points for any TH region may have low accuracy, so the process proceeds to step 35, and the coefficients are held at the previous values ​​without recalculating them.

[0077] On the other hand, in cases other than those described above, that is, when there are enough sample points in at least some TH regions, or when delamination has occurred, the process proceeds to step 38 to determine whether the initial learning of the clogging rate KTHC has been completed. If the answer is NO, and this clogging learning corresponds to the initial learning, there is a risk that a sufficient number of reliable learning points have not been secured, so the process proceeds to step 39, and the coefficient is initialized to an appropriate predetermined value without recalculating it.

[0078] If the answer to step 38 is NO, and this blockage learning corresponds to the period after the initial learning, proceed to step 40 to adjust the number of learning points required to calculate the coefficient. Specifically, as shown in the TH region n of Figure 11(a), there are cases where the number of sample points in the normal operating mode is 0, making it impossible to calculate the learning points. In this case, as shown in Figure 11(b), read the blockage rate KTHC at the throttle valve opening TH at the center of that TH region from the blockage rate table learned in the previous operating cycle, and use equation (3) above to calculate the estimated KTH from this blockage rate KTHC. This calculates and supplements the learning points in that TH region, thereby securing the number of learning points required to calculate the coefficient.

[0079] Returning to Figure 10, in step 41 following step 40, the coefficients a to c of the approximation function of equation (7) that approximates the estimated KTH characteristics are calculated. Specifically, as shown in Figure 11(c), the coefficients a to c of the approximation function are calculated (identified) by the least squares method based on multiple learning points for each TH region. This updates the approximation function. Note that if a learning prohibition region is set in step 6 of Figure 2, the calculation of coefficients a to c is performed based only on the learning points within the learning permission region, excluding the learning prohibition region.

[0080] Returning to Figure 2, after calculating the coefficients of the approximation function in the immediate post-stop mode as described above, in the initial operating mode of the next operating cycle when the ignition switch 31 is turned on, the clogging rate KTHC is calculated in step 8.

[0081] Figure 12 shows the calculation process. In this process, first in step 51, it is determined whether or not the throttle valve 5 was cleaned in the previous operating cycle. If the answer is YES, it is assumed that the degree of clogging of the throttle valve 5 has been restored to like-new condition by cleaning, and the process proceeds to step 52, where the clogging rate KTHC for the entire TH region is set to 0, and the process ends.

[0082] If the answer to step 51 is NO, proceed to step 53 to determine whether the clogging rate KTHC has been calculated in the current initial operating mode. If the answer is NO, proceed to step 54 to determine whether the initial learning of the clogging rate KTHC has been completed. If the answer is NO and this clogging learning corresponds to the initial learning, there is a risk that sufficient reliable learning points have not been secured, so proceed to step 55, set the clogging rate KTHC of the entire TH region to the maximum value, and terminate this process.

[0083] Specifically, as shown in Figure 13, the clogging rate KTHC is calculated from multiple learning points for each TH region obtained in the mode immediately after stopping the previous operating cycle. Next, the maximum value among the multiple calculated clogging rates KTHC is set as the uniform clogging rate KTHC for all TH regions. In this way, by estimating a larger clogging rate KTHC only during the initial learning, the opening of the throttle valve 5 is controlled to a larger side to compensate for this, thereby controlling the intake air volume to an increase, which is safer from the perspective of exhaust gas characteristics and other factors.

[0084] Returning to Figure 12, if the answer to step 54 is YES and this blockage learning corresponds to the period after the initial learning, proceed to step 56 and calculate the estimated KTH by applying the coefficients a to c calculated in the mode immediately after the previous operating cycle's stop to the approximation function of equation (8).

[0085] Next, in step 57, an upward sloping correction is applied to the estimated KTH characteristics approximated by the approximation function. This upward sloping correction takes into account that, due to the nature of the flow function, the estimated KTH characteristics inherently have an upward sloping characteristic, and that if they have a downward sloping characteristic, control hunting may occur, for example, when determining the target throttle valve opening based on the estimated KTH characteristics, such as the existence of multiple solutions.

[0086] Specifically, if the estimated KTH characteristic has a downward sloping portion as shown by the dotted line in Figure 14, this downward sloping portion is corrected to become an upward sloping portion with minimal slope, using the high-opening side as the reference, as shown by the solid line. This makes the estimated KTH characteristic an appropriate upward sloping portion and avoids the control hunting described above.

[0087] Next, in step 58, the congestion rate KTHC in the learning prohibited region is calculated. This process is performed when a learning prohibited region is set in step 6 of Figure 2, and approximation of the estimated KTH characteristics using an approximation function based on learning points, and calculation of the congestion rate KTHC are prohibited in the learning prohibited region. Specifically, as shown in Figure 15, the congestion rate KTHC in each TH region of the learning prohibited region is set to the same value as the congestion rate KTHC obtained in the TH region closest to the learning prohibited region among the learning permitted regions (black circle in the same figure). This allows the congestion rate KTHC to be set without problems in the learning prohibited region where no sample points have been obtained.

[0088] Next, in step 59, the previously calculated clogging rate KTHC is limited to a range of 0 to 1, and this process is terminated.

[0089] On the other hand, if the answer to step 53 is YES and the clogging rate KTHC has already been calculated, the process proceeds to step 60 to determine whether or not peeling of the throttle valve 5 occurred in the previous operating cycle. If the answer is YES, it is estimated that the clogging rate KTHC has decreased sharply across the entire TH region due to the occurrence of peeling. Therefore, the process proceeds to step 61, where, as shown in Figure 16, the clogging rate KTHC for each TH region is uniformly subtracted by a predetermined amount ΔKTHC. This avoids incorrect learning of the clogging rate KTHC due to peeling and allows for a good setting of the clogging rate KTHC.

[0090] If the answer to step 60 above is NO, that is, if the clogging rate KTHC has already been calculated and no peeling of the throttle valve 5 has occurred, no further calculation of the clogging rate KTHC will be performed, and this process will be terminated. In other words, the calculation of the clogging rate KTHC is performed only once in the initial operating mode.

[0091] Next, the process for calculating the target throttle valve opening will be explained with reference to Figure 17. This process uses the clogging rate KTHC calculated in the initial operating mode of the current operating cycle and is performed at predetermined intervals in the subsequent normal operating mode.

[0092] In this process, first, in step 71, the target intake air volume GAIRCMD is calculated. This calculation is performed, for example, by searching a predetermined map (not shown) according to the detected engine speed NE and the required torque TRQ. The required torque TRQ is calculated by searching a predetermined map (not shown) according to the engine speed NE and the detected accelerator opening AP.

[0093] Next, in step 72, the target flow rate function KTHCMD is calculated by searching a predetermined map (not shown) according to the target intake air volume GAIRCMD and the clogging rate KTHC. Finally, in step 73, as shown in Figure 18, the target throttle valve opening degree THCMD is calculated from the target flow rate function KTHCMD and the estimated KTH characteristics, and this process is completed. This makes it possible to appropriately set the target throttle valve opening degree THCMD while reflecting the clogging rate KTHC of the throttle valve 5.

[0094] As described above, according to this embodiment, for each of a predetermined number of TH regions of the throttle valve 5, a learning point is calculated by averaging multiple sample points that combine the throttle valve opening TH and the estimated KTH. This reduces the computational load and memory capacity compared to the conventional method in which a representative point of the estimated KTH is calculated for each TH region using the least squares method. Furthermore, the coefficients a to c of the approximation function of the polynomial that approximates the estimated KTH characteristics are calculated using the least squares method based on multiple learning points calculated for each of the multiple TH regions, and the clogging rate KTHC is calculated based on the estimated KTH characteristics approximated by the approximation function, the KTH when new, and the KTH when maximum clogging occurs. This allows for accurate learning of the clogging rate KTHC of the throttle valve 5 over a wide TH region.

[0095] Furthermore, as shown in Figure 8, the learning score is calculated by sequentially averaging the sample scores each time a sample score is acquired. Therefore, regardless of the number of sample scores, it is only necessary to sequentially store the number of sample scores and the average sample score up to the previous calculation. This significantly reduces the memory capacity compared to the conventional method of storing all the numerous sample scores and averaging them all at once.

[0096] Furthermore, as shown in Figure 2, the calculation of learning points, the calculation of coefficients a to c of the approximation function, the calculation of the clogging rate KTHC, and the control of engine 3 are performed in the normal operation mode, the mode immediately after stopping, the next initial operation mode, and the normal operation mode of engine 3, respectively. This reduces the processing load by distributing the processing for these calculations and control in each operation mode. In addition, the data that needs to be passed between the operation modes of engine 3 is basically only multiple learning points for each TH region between the normal operation mode and the mode immediately after stopping, the coefficients a to c of the approximation function between the mode immediately after stopping and the next initial operation mode, and the clogging rate KTHC between the initial operation mode and the normal operation mode. Therefore, the storage and transfer of this data can be easily performed with a very small memory capacity.

[0097] Furthermore, as shown in Figure 11, when there are TH regions where the number of sample points acquired in the current operating cycle is 0, the coefficients a to c of the approximation function are calculated using the learning points calculated in the previous operating cycle. This allows for good maintenance of the accuracy of the approximation using an approximation function based on multiple learning points, and the learning accuracy of the clogging rate KTHC based on that.

[0098] Furthermore, as shown in Figures 9 and 15, when there is a low TH region that is not being used during the operation of engine 3, this low TH region is set as a learning prohibition region, and learning of the clogging rate KTHC is prohibited. This allows the accuracy of the clogging rate KTHC to be maintained well by using the learning points obtained in the learning permission region and performing approximation with an approximation function only on the learning permission region. In addition, the clogging rate KTHC in the learning prohibition region can be set without any problems by using the same value as the clogging rate KTHC calculated in the TH region adjacent to the learning prohibition region.

[0099] Furthermore, as shown in Figure 14, when the estimated KTH characteristic approximated by the approximation function has a downward sloping portion, this downward sloping portion is corrected to become an upward sloping portion. This ensures that the estimated KTH characteristic has an appropriate upward sloping portion and avoids control hunting when setting the target throttle valve opening according to the estimated KTH characteristic.

[0100] Furthermore, as shown in Figure 6, the calculated estimated KTH is corrected based on a predetermined reference rotational speed #NEKTHC and then acquired as a sample point. This correction allows estimated KTH calculated under different engine rotational speed NE conditions to be unified and converted to the value at the reference rotational speed #NEKTHC, thereby effectively compensating for the variation in estimated KTH due to engine rotational speed NE.

[0101] Furthermore, as shown in Figure 16, when it is determined that deposits have detached from the throttle valve 5, the clogging rate KTHC is uniformly corrected to decrease, regardless of the throttle valve opening TH. This allows the clogging rate KTHC to be appropriately corrected in response to the detachment of deposits from the throttle valve 5.

[0102] It should be noted that the present invention is not limited to the embodiments described and can be implemented in various ways. For example, in the embodiment shown in Figure 11, when the number of sample points acquired in the current operating cycle is 0, the learning points calculated in the previous operating cycle are used as the learning points for the TH region. However, this replenishment of learning points may also be performed when the current number of sample points is very small, less than a predetermined value close to 0 (for example, 3).

[0103] Furthermore, in this embodiment, the learning prohibition region that prohibits clogging learning is set in the low TH side region. However, if a situation occurs in the high TH side region where the usage frequency of the throttle valve 5 becomes 0, the high TH side region may be set as the learning prohibition region. In this case, learning points obtained in the learning permission region other than the high TH side region can be used to perform clogging learning accurately, targeting only the learning permission region, and the clogging rate KTHC in the high TH side region will be set to the same value as the clogging rate KTHC of the adjacent TH region.

[0104] Furthermore, in this embodiment, the clogging rate KTHC defined by equation (2) is used as a clogging parameter representing the degree of clogging of the throttle valve 5. However, other suitable parameters, such as the rate or amount of deposits at the opening of the throttle valve 5, may be used as long as they represent the degree of clogging of the throttle valve 5.

[0105] Furthermore, in this embodiment, an example is given in which the target throttle valve opening degree THCMD is set as the control of the engine 3 using the clogging rate KTHC of the throttle valve 5. However, the invention is not limited to this, and the clogging rate KTHC may be used to control the intake air amount or fuel injection amount, or to estimate the intake air amount. In addition, the details of the configuration can be appropriately changed within the scope of the spirit of the present invention. [Explanation of Symbols]

[0106] 2 ECU (means for calculating blockage parameters, control means, sample point acquisition means, learning point calculation means, coefficient calculation means, peeling determination means) 3. Engine (internal combustion engine) 5. Throttle valve 11 Intake passage 22 Airflow Meter GAIR Intake Air Volume KTHC throttle valve clogging rate (clogging parameter) TH Throttle valve opening (opening degree of the throttle valve) a, b, c: coefficients of the approximation function #NEKTHC Standard rotation speed

Claims

1. A clogging parameter calculation means calculates a clogging parameter representing the degree of clogging of a throttle valve provided in the intake passage of an internal combustion engine, using a first flow function when the degree of clogging of the throttle valve is in a standard state and a second flow function estimated based on the amount of intake air detected by an airflow meter. The system includes a control means for controlling the internal combustion engine using the calculated blockage parameter, The aforementioned blockage parameter calculation means is During the operation of the internal combustion engine, a sample point acquisition means calculates the second flow function at predetermined intervals and acquires a sample point which is a combination of the second flow function and the opening degree of the throttle valve. A learning point calculation means calculates the average value of the sample points by successively averaging the multiple sample points belonging to each predetermined opening degree region of the throttle valve each time a sample point is acquired, and calculates a learning point based on the average value. The system includes a coefficient calculation means that calculates, by least squares method, the coefficients of a predetermined polynomial approximation function for approximating the relationship between the second flow function and the opening of the throttle valve, based on the learning points calculated for each of the plurality of opening ranges, A control device for an internal combustion engine, characterized in that it calculates the clogging parameter based on the second flow function characteristics approximated by the approximation function using the calculated coefficient and the first flow function.

2. The control device for an internal combustion engine according to claim 1, characterized in that the learning point calculation means performs the calculation of the learning points in the normal operating mode of the internal combustion engine, the coefficient calculation means performs the calculation of the coefficient of the approximation function based on the plurality of learning points in the mode immediately after the internal combustion engine stops, the clogging parameter calculation means performs the calculation of the clogging parameter based on the approximation function using the coefficient in the next initial operating mode of the internal combustion engine, and the control means performs control of the internal combustion engine using the clogging parameter in the next normal operating mode of the internal combustion engine.

3. The learning point calculation means calculates the learning point for each operating cycle of the internal combustion engine, The control device for an internal combustion engine according to claim 1 or 2, characterized in that, when there is an opening range of the throttle valve in which the number of sample points acquired in the current operating cycle is less than a predetermined value, the coefficient calculation means calculates the coefficient of the approximation function using the learning points calculated in the previous operating cycle as the learning points for that opening range.

4. The control device for an internal combustion engine according to claim 3, characterized in that, when there is a low-opening region or a high-opening region of the throttle valve that is not used during the operation of the internal combustion engine, the clogging parameter calculation means sets the low-opening region or the high-opening region to a learning prohibition region in which learning of the clogging parameter is prohibited, and sets the clogging parameter in the learning prohibition region to the same value as the clogging parameter calculated in the opening region adjacent to the learning prohibition region.

5. The control device for an internal combustion engine according to any one of claims 1 to 4, characterized in that the blockage parameter calculation means corrects the second flow function characteristic approximated by the approximation function to have a downward sloping portion so that it becomes upward sloping when the downward sloping portion has a downward sloping portion.

6. The control device for an internal combustion engine according to any one of claims 1 to 5, characterized in that the sample point acquisition means acquires the calculated second flow function as a sample point after correcting it with respect to a predetermined reference rotational speed.

7. The system further includes a peeling determination means for determining whether or not peeling of deposits from the throttle valve has occurred, The control device for an internal combustion engine according to any one of claims 1 to 6, characterized in that the clogging parameter calculation means uniformly corrects the clogging parameter to a decrease regardless of the opening degree of the throttle valve when it is determined that the peeling has occurred.

Images