JP2007262941A - Control device of internal combustion engine - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a control device of an internal combustion engine capable of smoothly controlling the internal combustion engine, by reducing the division number of maps used when determining a control quantity. <P>SOLUTION: This control device of the internal combustion engine 1 has a state detecting means 15 detecting a state of the internal combustion engine, and a control means 20 controlling the internal combustion engine by determining the control quantity reflected on torque by using a predetermined map, by confirming an operation state of the internal combustion engine on the basis of output of the state detecting means. The map is a learning map divided in response to an operation area of the internal combustion engine and capable of updating a condition of determining the control quantity set with every operation area by the control means. The control means 20 updates the condition by changing a learning reflection factor in response to an operation performing frequency, by performing learning with every operation area in the learning map. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

  The present invention relates to a control device for an internal combustion engine. More specifically, the present invention relates to a control device that performs optimal control according to the operating state of an internal combustion engine.

  The state of the load changes every moment depending on the operating condition of the internal combustion engine. Therefore, conventionally, various control devices for controlling the state of the internal combustion engine to an appropriate state according to the operating state have been proposed. In recent years, a control device for an internal combustion engine that performs so-called learning-type control in which the state when the internal combustion engine is actually operated is reflected in the control has been proposed. ing. In the case of a control device having such a learning function, control data (such as an ignition timing map) that has been installed in advance is sequentially updated (changed) according to actual operating conditions, so that the internal combustion engine is in operation. Efficient control can be performed according to (use environment). As a result, it is possible to improve the fuel consumption and exhaust of the internal combustion engine and improve drivability.

  Patent Document 1 discloses a control device for an internal combustion engine having the learning function. When learning a certain region on the learning map, the control device controls the air-fuel ratio by causing the learning correction values of other regions to be estimated and learned to appropriate values. More specifically, this control device includes an air-fuel ratio learning map in which the operation region is divided into a plurality of regions. When learning a certain region (A, B), the learning correction coefficient α0 of the region (A, B) is estimated and learned as a learning correction coefficient of the operation region having the same intake air flow rate as the region (A, B). . The learning correction coefficient α0 of the area (A, B) is estimated and learned as the learning correction coefficient for the operation areas belonging to the same Tp lattice as the regions (A, B) and the Tp lattices above and below the Tp lattice. Thereby, when learning the air-fuel ratio, the step of the correction level can be eliminated by interpolating the unlearned region with the value of the learning region, so that the internal combustion engine can be controlled satisfactorily.

JP-A-9-79072

  However, Patent Document 1 does not discuss how to set the learning map, that is, how to set the learning map classification is appropriate for the control of the internal combustion engine. For example, if the set division is relatively large, a torque step at the boundary may occur. On the other hand, if a large number of sections are set and subdivided, problems such as memory shortage occur.

  Accordingly, an object of the present invention is to provide a control device for an internal combustion engine that can reduce the number of map sections used when determining a control amount and can realize smooth control of the internal combustion engine.

  The object is to detect a state of the internal combustion engine and to check the operating state of the internal combustion engine based on the output of the state detector and to obtain a control amount reflected in the torque using a predetermined map. A control unit for controlling the internal combustion engine, wherein the map is divided according to an operation region of the internal combustion engine, and the control amount set for each operation region Is a learning map that can be updated by the control means, and the control means performs learning for each of the driving regions in the learning map and changes a learning reflection coefficient according to the driving execution frequency. This is achieved by an internal combustion engine controller that updates the conditions.

  According to the present invention, since the learning map divided by the operation region of the internal combustion engine is used, the number of divisions can be greatly reduced as compared with the conventional map. Therefore, the problem of insufficient memory does not occur. Furthermore, since the control means performs learning for each operation region of the learning map and updates the condition for obtaining the control amount by changing the learning reflection coefficient according to the operation execution frequency, the internal combustion engine can always be controlled appropriately.

  Further, the control means performs an interpolation process to eliminate the torque step when the torque step is larger than a predetermined value in the driving situation across the boundary between the adjacent first driving region and the second driving region. It is desirable. In this way, the internal combustion engine can be smoothly controlled even if a large torque step occurs due to the difference in learning frequency.

  Further, it is more preferable that the control means controls the internal combustion engine after correcting the learning reflection coefficient according to the appearance frequency of the operation region. In this case, it is possible to increase the weighting of the operation region with a low appearance frequency and reduce the weighting of the operation region with a high appearance frequency. Thereby, the effect by learning can be enhanced and the internal combustion engine can be accurately controlled.

  The control amount is preferably the ignition timing of the internal combustion engine, but may be an intake air amount, a fuel injection amount, an air-fuel ratio (A / F), or the like.

  According to the present invention, it is possible to provide a control device for an internal combustion engine that can reduce the number of regions of a learning map used when obtaining a control amount and can realize smooth control of the internal combustion engine.

  The best mode for carrying out the present invention will be described below with reference to the drawings. FIG. 1 is a schematic block diagram showing an internal combustion engine equipped with a control device according to the present invention. In the internal combustion engine 1, a piston 3 is disposed in a cylinder block 2 so as to be capable of reciprocating. Power is generated by reciprocating the piston 3 by burning a mixture of fuel (for example, gasoline) and air inside a combustion chamber 4 formed at the top of the piston 3. Although only one cylinder is shown in FIG. 1, the internal combustion engine 1 is preferably configured as a multi-cylinder engine, and the internal combustion engine 1 of the present embodiment is formed as a four-cylinder engine, for example.

  An intake port of each combustion chamber 4 is connected to an intake pipe 5 via an intake manifold. The same applies to the exhaust side, and the exhaust port of each combustion chamber 4 is connected to the exhaust pipe 6 via an exhaust manifold. In addition, an intake valve 16 that opens and closes an intake port and an exhaust valve 17 that opens and closes an exhaust port are disposed for each combustion chamber 4 in the cylinder head of the internal combustion engine 1. Each intake valve 16 and each exhaust valve 17 are opened and closed by a valve operating mechanism (not shown) having a variable valve timing function, for example. Further, the internal combustion engine 1 has a number of spark plugs 7 corresponding to the number of cylinders, and the spark plugs 7 are arranged on the cylinder heads so as to face the corresponding combustion chambers 4.

  An air cleaner 9, a throttle valve (in this embodiment, an electronically controlled throttle valve) 10 and a surge tank 8 are arranged in the intake pipe 5 from the upstream side. On the other hand, the exhaust pipe 6 is provided with a front-stage catalyst device 11a including a three-way catalyst and a rear-stage catalyst device 11b including a NOx storage reduction catalyst.

  The internal combustion engine 1 has a plurality of injectors 12. Each injector 12 is disposed so as to face the corresponding intake pipe 5 (inside the intake port), and injects fuel such as gasoline into each intake port. Although the internal combustion engine 1 of the present embodiment is described as a so-called port injection type gasoline engine, the present invention is not limited to this, and it goes without saying that the present invention can be applied to a so-called direct injection type internal combustion engine. . Needless to say, the present invention can be applied not only to a gasoline engine but also to a diesel engine.

  The spark plug 7, the throttle valve 10, each injector 12, the valve operating mechanism and the like are electrically connected to an ECU 20 that functions as a control device for the internal combustion engine 1. The ECU 20 includes a CPU, a ROM, a RAM, an input / output port, a storage device, etc., all not shown. As shown in FIG. 1, the ECU 20 is electrically connected to various sensors including an in-cylinder pressure sensor 15 that detects the pressure in the combustion chamber 4. The in-cylinder pressure sensor 15 is formed including a semiconductor element, a piezoelectric element, an optical fiber detection element, or the like, and is arranged in a number corresponding to the number of cylinders. Each in-cylinder pressure sensor 15 is disposed on the cylinder head so that the pressure receiving surface faces the corresponding combustion chamber 4, and is electrically connected to the ECU 20. The detection value of each in-cylinder pressure sensor 15 is sequentially given to the ECU 20 every predetermined time, and is stored and held by a predetermined amount in a predetermined storage area (buffer) of the ECU 20. The ECU 20 uses various types of maps and the like stored in the storage device, and based on the detection values of various sensors and the like, a spark plug 7, a throttle valve 10, an injector 12, Control the valve mechanism.

  In the above description, the in-cylinder pressure sensor 15 has been described as a sensor for detecting the state of the internal combustion engine. However, various sensors may be arranged to detect the state as in the case of a conventional internal combustion engine. For example, the rotational speed of the internal combustion engine can be detected by providing a crank angle sensor 14 that detects the rotational speed of the crank angle. Further, for example, an air flow meter that detects the amount of intake air, an intake pressure sensor that detects intake air pressure, an intake air temperature sensor that detects the temperature of intake air, and the like may be arranged in the intake pipe 5. Further, an exhaust pressure sensor for detecting the exhaust gas pressure is provided on the exhaust officer 6 side, and an air-fuel ratio (A / F) of the internal combustion engine is determined from the oxygen concentration and unburned gas concentration downstream of the exhaust purification catalyst 11b. A / F sensor for detecting) may be provided. Further, a water temperature sensor for detecting the temperature of the circulating cooling water and an oil temperature sensor for detecting the temperature of the circulating lubricating oil may be disposed around the crankshaft.

  Signals from the in-cylinder pressure sensor 15 and various sensors arranged at a plurality of locations as needed are supplied to the ECU 20. Therefore, the ECU 20 can accurately check the operation status of the internal combustion engine based on the detection signals from the sensors. The internal combustion engine 1 incorporates a control device that optimally adjusts the ignition timing in accordance with the operating region. This control device confirms the state detection means by the in-cylinder pressure sensor 15 and the like and the corresponding operation region of the internal combustion engine based on the output of this state detection means, specifies the ignition timing from the map, and executes the control of the internal combustion engine. It includes a control means. The control means is realized by a part of the ECU 20. Although the ECU for the control device may be provided independently, the configuration can be simplified by using the ECU 20 of the internal combustion engine 1 as in the present embodiment.

  Here, the map is configured as a learning map that can be updated. This learning map MP is designed so that the ignition timing SA can be obtained by distinguishing it according to the operating region of the internal combustion engine, for example, as shown in FIG. The ECU 20 confirms the state of the internal combustion engine from the sensor output, determines which operating region in the learning map MP belongs, and determines the ignition timing SA. Further, when the ECU 20 determines the ignition timing SA of the internal combustion engine based on the sensor output, the ECU 20 changes the learning reflection coefficient according to the operation execution frequency. Thereby, the conditions for obtaining the ignition timing SA by the learning map MP are updated according to the driving situation. Note that such a learning map may be stored in a memory such as a ROM.

  The learning map MP shown in FIG. 2 shows a case where the operation region is divided by the load factor KL and the rotational speed NE of the internal combustion engine. The load factor KL is a numerical value representing the load state, and is the relative load when the load factor when the intake air amount of the internal combustion engine 1 is equal to the exhaust amount is 100%. is there. The load factor KL is 0% when there is no load and 100% when the load is full. In particular, the load factor KL is correlated with the intake air filling rate (intake air amount) into the cylinder, and the intake air filling rate into the cylinder increases as the load factor KL increases.

  This learning map MP is extremely different from the conventional general ignition timing map, which is divided into several tens to several hundreds of regions, by defining the operation region based on the load factor KL and the rotational speed NE. There are 6 categories. Specifically, it is classified according to the load factor KL and the rotational speed NE, the first area AR1 is an idle area where the load ratio KL is “low” and the rotational speed NE is “small”, and the second area AR2 is a load factor. The transition region where KL is “medium” and the rotational speed NE is “small” is set, and the third region AR3 is set as a high load region where the load factor KL is “high” and the rotational speed NE is “small”. Similarly, the fourth area AR4 is a deceleration area where the load factor KL is “low” and the rotational speed NE is “large”, and the fifth area AR5 is a partial area where the load factor KL is “medium” and the rotational speed NE is “large”. The sixth area AR6 is set as an acceleration area where the load factor KL is “high” and the rotational speed NE is “large”.

  In each region, a function formula f is initially set as a condition defined by the load factor KL and the rotational speed NE. The general expression of the function expression f is defined as f = Pa × KL + Pb × NE, for example. The functional expression f1 = Pa1 × KL + Pb1 × NE is initially set in the idle region (first region AR1). The same applies to the other regions, and the functional equation f6 = Pa6 × KL + Pb6 × NE is set in the acceleration region (sixth region AR6). Pa and Pb in each equation are learning reflection coefficients, and the update is repeated after learning the operating state of the internal combustion engine. The learning reflection coefficients Pa and Pb are covariance matrices designed based on the output of each sensor, and this covariance matrix is a data string that corrects the ignition timing to an optimum one according to the driving situation. The covariance matrix is updated based on learning when the ECU 20 acquires the load factor KL and the rotational speed NE and determines the ignition timing SA. This covariance matrix can be updated by publicly known RLS learning using, for example, a recursive least-squares (RLS) method.

  In FIG. 2, the load factor KL is abstractly defined as “high”, “medium”, and “low”, and the rotational speed NE is abstractly defined as “large” and “small”. The range of 0 to 100 (%) is divided into three ranges. Similarly, the rotation speed NE is defined as “large” or “small” with a predetermined rotation speed (rmp) as a boundary.

  FIG. 3 is a flowchart showing a case where the ECU 20 controls the internal combustion engine while learning using the learning map MP of FIG. The ECU 20 first reads engine conditions from various sensors (S11). For example, the load factor KL is calculated from the output of the in-cylinder pressure sensor 15 and the rotational speed NE is confirmed from the output of the crank angle sensor 14. The operating state of the internal combustion engine is the first depending on whether the load factor KL belongs to “high”, “medium”, or “low” and whether the rotational speed NE belongs to “large” or “small”. -It is determined whether it is in any of the sixth operating regions (S12).

  Then, the ECU 20 sets the ignition timing SA based on the functional expression f1 = Pa1 × KL + Pb1 × NE set in the operation region (for example, the idle region of the first region AR1) determined based on the load factor KL and the rotational speed NE. Calculation (S13) and control of the internal combustion engine is executed.

  Further, at this time, the ECU 20 uses the load factor KL, the rotational speed NE, and the ignition timing SA, and by RLS learning (S14), the learning reflection coefficients Pa and Pb are updated and the function formula f is changed to one corresponding to the operating state. To do. Therefore, the function expression f of the driving region that frequently appears is learned and updated at a high frequency, whereas the driving region having a low appearance frequency has a low learning opportunity. Therefore, according to the control apparatus of the present embodiment, the ignition timing can be learned by reflecting the operation state of the internal combustion engine, so that the internal combustion engine control with high accuracy can be performed.

  By the way, when the operation region (I) having a high learning frequency and the operation region (II) having a low learning frequency are in contact with each other, the deviation of the ignition timing SA between the regions becomes gradually large, resulting in a torque step. There are concerns about the occurrence. In the control device of this embodiment, the ECU 20 suppresses the occurrence of such a torque step. This point will be described with reference to FIG.

  The function formula f in each region is initially set so that there is no deviation in the ignition timing SA between adjacent regions. However, since learning is performed according to the operating state of the internal combustion engine as described above, the function formula in the region where the appearance frequency is high is repeatedly updated. As a result, as schematically shown in FIG. 4A, it is assumed that a torque step ST occurs between the operation region (I) having a high learning frequency and the operation region (II) having a low frequency. . Such a torque step ST is eliminated by setting a gentle curve CV that the ECU 20 eliminates the step portion as shown in FIG. 4B. For example, the ECU 20 sets a point AP in the driving region (I), a point BP in the driving region (II), and an intermediate point CP of the step amount on the region boundary BL at equal distances from the region boundary BL. Then, the curve CV is designed using the values interpolated so as to eliminate the torque step ST. In this way, the control device can also suppress the occurrence of torque steps in the internal combustion engine. Note that when it is confirmed that a torque step greater than a predetermined value is generated, the ECU 20 is set to execute the torque step elimination process (smoothing process) shown in FIG. It is preferable to keep. With this setting, the smoothing process is executed only when a torque step exceeding the allowable range occurs.

  Further, when the learning map MP of FIG. 2 is used, it is necessary to consider the following points. The frequency at which each operating region appears by the internal combustion engine is almost determined. The idle region (first region AR1) frequently appears in any internal combustion engine. In contrast, the frequency of the high load region (third region AR3) varies depending on the environment in which the internal combustion engine is operated. For example, a high load region often appears in an internal combustion engine that operates at a high speed, but a high load region hardly appears in an internal combustion engine that operates only in outlying areas. Considering the situation in which each of such internal combustion engines is operated, the accuracy is improved by correcting the weighting of the information related to the operation region having a low appearance frequency to be large and the weighting of the information related to the operation region having a high appearance frequency to be small. Enhanced internal combustion engine control.

  Therefore, it is preferable to adjust the effect of learning by setting the correction coefficient K for each operation region of the learning map in accordance with the operation state of the internal combustion engine. FIG. 5 is a diagram showing a state in which a correction coefficient K for weighting the learning effect is set in each operation region of the learning map shown in FIG. FIG. 5 shows an example of setting the correction coefficient K for an internal combustion engine with few opportunities for high load operation. Here, K1 is a small value and K2 is a large value. In the operation region in which the correction coefficient K1 is set, the learning reflection coefficients Pa and Pb of the function formula f updated by learning are changed relatively small. That is, since the operation region in which the correction coefficient K1 is set has a high frequency, the learning convergence is slowed down. By doing in this way, it can prevent that the function formula f is updated too much with the information acquired much in the driving | running area | region where appearance frequency is high.

  On the other hand, in the operation region in which the correction coefficient K2 is set, the function formula f is quickly changed based on the learning described above. That is, the operation region in which the correction coefficient K2 is set is set so that the learning convergence degree becomes faster. As a result, the weight of information related to the operation region having a low appearance frequency is increased. Thus, the ignition timing of the internal combustion engine can be controlled with higher accuracy by weighting the learning using the correction coefficient K in consideration of the appearance frequency of the operation region.

  The control device for an internal combustion engine described above obtains the ignition timing of the internal combustion engine from a learning map divided into six based on the operation region, and sequentially updates the function formula (condition) of each region by learning. Therefore, ignition timing control can be performed efficiently with a small amount of memory, compared to a conventional device that performs ignition timing control using a map divided into a large number of regions. In addition, since the torque step generated between the operation regions can be suppressed, the internal combustion engine can be operated smoothly. Furthermore, by weighting learning according to the appearance frequency, ignition timing control can be performed in accordance with the situation in which the internal combustion engine is operated.

  Although the preferred embodiments of the present invention have been described in detail above, the present invention is not limited to the specific embodiments, and various modifications and changes can be made within the scope of the gist of the present invention described in the claims. It can be changed. For example, in the above embodiment, the case where the learning map is divided into six is illustrated, but the present invention is not limited to this. For example, in the above description, the rotational speed NE is large or small, but may be large, medium, or small, and the load factor may be further divided. As a result, the number of divisions increases, but the number of divisions is greatly reduced as compared with the conventional case, so that the problem of insufficient memory does not occur. In the above embodiment, an example in which the ignition timing by the plug 7 is a control amount reflected in the torque has been described. However, the present invention is not limited to this, and the control amount is an intake air amount, a fuel injection amount, an A / F value, and the like. Also good.

It is the schematic block diagram shown about the internal combustion engine provided with the control apparatus by this invention. It is the figure which showed an example of the learning map which calculates | requires the ignition timing distinguished according to the driving | operation area | region. It is the flowchart which showed the routine of ignition control according to the driving | running state of the internal combustion engine by ECU using a learning map. It is the figure shown in order to demonstrate the torque level difference which generate | occur | produces in an area | region. It is the figure which showed a mode that the correction coefficient was set to each area | region of the learning map shown in FIG.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Internal combustion engine 5 Intake pipe 6 Exhaust officer 15 In-cylinder pressure sensor (state detection means)
20 ECU (control means)
MP learning map

Claims (4)

State detecting means for detecting the state of the internal combustion engine;
An internal combustion engine comprising: control means for confirming an operating state of the internal combustion engine based on an output of the state detection means and obtaining a control amount reflected in the torque using a predetermined map to control the internal combustion engine. A control device,
The map is a learning map that is divided according to the operation region of the internal combustion engine and that can update the condition for obtaining the control amount set for each operation region by the control means,
The control unit performs learning for each of the operation regions in the learning map, and updates the condition by changing a learning reflection coefficient according to an operation execution frequency. . The control means performs an interpolation process to eliminate the torque step when the torque step is larger than a predetermined value in an operation situation across a boundary between the adjacent first operation region and the second operation region. The control device for an internal combustion engine according to claim 1. 2. The control device for an internal combustion engine according to claim 1, wherein the control unit controls the internal combustion engine after correcting the weighting of the learning reflection coefficient in accordance with the appearance frequency of the operation region. 4. The control device for an internal combustion engine according to claim 1, wherein the control amount includes an ignition timing of the internal combustion engine.

Images