JP5091013B2 - Fuel injection device for internal combustion engine - Google Patents

Description

The present invention obtains a fuel injection amount based on the number of revolutions of the crank of the internal combustion engine and the throttle opening, and based on an output signal output from an O ₂ sensor provided in the exhaust system, a predetermined air-fuel ratio is obtained. The present invention relates to a fuel injection device for an internal combustion engine that calculates a feedback correction amount so as to approach to obtain a final fuel correction amount and performs learning control based on the feedback correction amount.

  The fuel injection amount for obtaining the ideal air-fuel ratio of the internal combustion engine is roughly set based on the number of revolutions of the crank and the intake air amount, but the intake air amount may be substituted by the throttle opening. When the throttle opening is large, the actual intake amount and the throttle opening have a fairly high correlation, but when the throttle opening is small, the correlation with the actual air amount is low due to the effects of air viscosity and bypass passages. Thus, there may be an error with respect to the ideal air-fuel ratio in the fuel injection amount obtained based on the throttle opening.

  In a region where the throttle opening is small, a method of estimating and calculating the intake amount based on a negative pressure sensor can be considered. In Patent Document 1, a negative pressure sensor is provided in the intake system, and when there is a change in the environment or aging of the vehicle, the fuel injection amount is controlled based on the output of the sensor. However, the negative pressure sensor is more expensive than the throttle opening sensor, and providing this increases the cost.

Therefore, in Patent Document 2, based on an output signal output from an O ₂ sensor provided in the exhaust system, a feedback correction amount is calculated so as to approach a previously obtained air-fuel ratio, and the fuel injection amount is corrected. It has been proposed to determine the final fuel injection amount. In Patent Document 2, the learning correction amount based on the feedback correction amount is extended to control the fuel injection amount and the ignition timing.

  Further, in Patent Document 3, learning control is performed in which an engine operating state is divided into a plurality of regions and a feedback correction amount for compensating for excess or deficiency of the fuel supply amount by the feedback control means is calculated. As described above, by setting individual control parameters divided into a plurality of regions, it is possible to perform appropriate control according to the regions.

JP-A-5-59997 Japanese Patent Laid-Open No. 9-4554 Japanese Patent Publication No. 5-26935

In the control method described in Patent Document 3, control according to a plurality of regions is possible, but it is desirable to update the control parameters as appropriate based on O ₂ feedback values and the like in order to cope with environmental changes and the like. .

  However, the plurality of regions divided according to the operating state of the engine are not necessarily executed at the same frequency, and when a general vehicle is traveling in the city, the traveling frequency with a small throttle opening is considerable. The control parameter is frequently updated in the region corresponding to this state, but the control parameter in the region corresponding to the state where the throttle opening is large is not updated so much. Therefore, for example, immediately after entering a highway from the city and moving to a driving state with a large throttle opening, the corresponding control parameter is not an appropriate value, and an ideal air-fuel ratio cannot be obtained immediately. Become.

  Furthermore, when a general vehicle is traveling in the city, the traveling frequency in a state where the throttle opening is small is considerably high. Therefore, it is desirable to update the correction coefficient in such a region with particularly high accuracy.

  The present invention has been made in consideration of such problems, and divides the operating state of the engine into a plurality of regions and updates the control parameters recorded for each region to appropriate values as quickly as possible. Further, it is an object of the present invention to provide a fuel injection device for an internal combustion engine that can update a correction coefficient in a frequently used region with high accuracy.

  The fuel injection device for an internal combustion engine according to the present invention has the following features.

1st characteristic; With reference to the basic fuel-injection map based on the throttle opening degree of the throttle valve provided in the intake system of the internal combustion engine, and the rotation speed of the internal combustion engine, based on the throttle opening degree and the rotation speed A basic injection amount calculating means for obtaining a basic fuel injection amount;
Feedback correction coefficient calculating means for obtaining a feedback correction coefficient for performing feedback control of fuel injection at a predetermined cycle based on an output signal output from an O ₂ sensor provided in the exhaust system;
A final fuel injection amount calculating means for multiplying the basic fuel injection amount by the feedback correction coefficient and determining a final fuel injection amount so as to approach a target air-fuel ratio;
Feedback correction learning coefficient storage means for storing the feedback correction coefficient as a learning correction coefficient in a nonvolatile storage unit at a predetermined timing;
A fuel injection device for an internal combustion engine having
The feedback correction learning coefficient storage means stores the calculated learning correction coefficient for each area of the basic injection map that is divided into a plurality of areas in advance according to the driving situation,
The plurality of regions are characterized in that the region width in the axial direction of the throttle opening is reduced as the region on the side where the throttle opening is smaller compared to the same rotational speed in the basic fuel injection map. To do.

  Thus, by setting the region width to be smaller as the throttle opening is smaller, the region of the normal use region is limited, and the correction coefficient of the region can be updated with high accuracy.

  2nd characteristic; The said area | region where the throttle opening is the highest is set to one area | region which is not related to the said rotation speed, and the side where the said throttle opening is the lowest corresponds to the said rotation speed in the said several area | region. It is characterized by being set in a plurality of areas.

  On the side where the throttle opening is the highest, there is a difference in the frequency of use based on the engine speed, so it is reasonable to make it an integrated area. On the other hand, on the side where the throttle opening is the lowest, a plurality of regions corresponding to the number of revolutions can be set to enable more accurate control. In a region where the throttle opening is medium, an appropriate region classification may be performed according to the design conditions.

  Third feature: A representative region is provided in one region set on the side with the highest throttle opening, and the learning correction coefficient obtained in the representative region is reflected in the entire one region. Features.

  By obtaining the learning correction coefficient based on this representative area, the learning correction coefficient can be applied simply and accurately for the entire area.

4th characteristic; The said area | region is from the idle area | region corresponding to idle rotation, the high throttle area | region where the said throttle opening is the highest, and the medium and low throttle area | regions other than the said idle area | region and the said high throttle area | region. Become
The learning correction coefficient in the middle / low throttle area is obtained from the learning correction coefficient in the idle area and the learning correction coefficient in the high throttle area.

  Thus, when three regions are used, the number of regions is sufficiently small, the storage capacity for storing parameters such as learning correction coefficients is reduced, and the control procedure is simplified.

Fifth feature: With reference to a basic fuel injection map based on the throttle opening of a throttle valve provided in the intake system of the internal combustion engine and the rotational speed of the internal combustion engine, based on the throttle opening and the rotational speed A basic injection amount calculating means for obtaining a basic fuel injection amount;
Feedback correction coefficient calculating means for obtaining a feedback correction coefficient for performing feedback control of fuel injection at a predetermined cycle based on an output signal output from an O ₂ sensor provided in the exhaust system;
A final fuel injection amount calculating means for multiplying the basic fuel injection amount by the feedback correction coefficient and determining a final fuel injection amount so as to approach a target air-fuel ratio;
Feedback correction learning coefficient storage means for storing the feedback correction coefficient as a learning correction coefficient in a nonvolatile storage unit at a predetermined timing;
The fuel injection device for an internal combustion engine, wherein the final fuel injection amount calculation means calculates the final fuel injection amount using the learning correction coefficient in an operation situation in which feedback control based on an output signal from the O ₂ sensor is not performed. ,
The basic change rate of the learning correction coefficient for each of the plurality of areas or several adjacent areas is held in a predetermined recording unit, and the current area change with respect to the reference value of the learning correction coefficient in the current area An estimation correction means for correcting the learning correction coefficient for at least one other region based on the current region change amount and the corresponding basic change rate when the amount is equal to or greater than a predetermined amount;
It is characterized by having.

  In this way, when the engine operating state is divided into a plurality of regions and the correction coefficient is updated for one region, the other regions are estimated and corrected in accordance with the variation tendency, so that the other regions The correction coefficient can be updated to an appropriate value as quickly as possible.

  Sixth feature; the estimated correction means is characterized in that a predetermined limit is provided for a correction amount for another region. The estimation correction is only an estimation, and it is desirable to prevent excessive correction. Therefore, by providing a predetermined restriction, it is possible to prevent the correction coefficient for the region to be estimated and corrected from being excessively corrected.

  7th characteristic; The said restriction | limiting is made into the said current area change amount, It is characterized by the above-mentioned. As described above, if the restriction is the current region change amount, it is appropriate as a value, and the processing becomes simple.

  Eighth feature: The plurality of regions include independent idle regions corresponding to idle rotation. The idle area is an area where the frequency of use is high and the disturbance is large, and it can be made an independent area and can cope with it finely.

  Ninth feature: When the current region change amount is equal to or greater than a predetermined amount during idle operation of the vehicle, all regions other than the idle region are determined based on the change amount of the learning correction coefficient and the basic change rate. The correction coefficient is estimated and corrected. Since the disturbance is large in the idle region, it is appropriate to correct all other regions accurately based on the basic change rate.

  According to the fuel injection device for an internal combustion engine according to the present invention, the operating state of the engine can be divided into a plurality of regions, and the control parameters recorded for each region can be updated to appropriate values as quickly as possible. Furthermore, it is possible to update the correction coefficient in the frequently used area with high accuracy.

  Embodiments of a fuel injection device for an internal combustion engine according to the present invention will be described below. The fuel injection device for an internal combustion engine according to the present embodiment shown in FIGS. 1 to 12 is applied to a vehicle such as a motorcycle or a four-wheeled vehicle.

As shown in FIG. 1, a fuel injection device 10 for an engine (internal combustion engine) according to the present embodiment is applied to an engine system 11 of a motorcycle. The engine system 11 includes an engine 12, a controller 14, a fuel injection 16, a spark plug 18, a throttle valve 20 provided in an intake pipe, a catalyst 22 and an O ₂ sensor 24 provided in an exhaust pipe. And an accelerator sensor 26 and a rotation speed sensor 30 for detecting the rotation speed of the crank 28.

The controller 14 reads the value of the accelerator sensor 26 and opens / closes the throttle valve 20 via the actuator 32. The O ₂ sensor 24 detects the O ₂ concentration in the exhaust pipe and supplies it to the controller 14. The controller 14 performs predetermined calculations based on signals from the accelerator sensor 26, the O ₂ sensor 24, the rotation speed sensor 30, and the like, determines the fuel injection amount, ignition timing, and the like, and controls the fuel injection 16 and the spark plug 18. Do. In order to facilitate understanding, it is assumed that the operation amount of the accelerator sensor 26 and the opening of the throttle valve 20 (hereinafter referred to as throttle opening Th) are proportional, and the throttle opening is detected by detecting the signal of the accelerator sensor 26. Th can be specified. The means for detecting the throttle opening Th is not limited to the accelerator sensor 26, and for example, an opening sensor may be provided in the throttle valve 20.

  The controller 14 can realize different controls by software processing, and hereinafter, first to third control methods will be described. The first control method is a control method based on the first map 40, the second control method is a control method based on the second map 80 (see FIG. 10), and the third control method is the third control method. This is a control method based on the map 90 (see FIG. 11). First, the first control method will be described.

As shown in FIG. 2, in the first control method, the controller 14 refers to the first map 40 and determines the rotation speed Ne obtained from the rotation speed sensor 30 and the throttle opening degree Th obtained from the accelerator sensor 26. Based on the basic fuel injection amount calculation unit (basic injection amount calculation means) 42 for determining the basic fuel injection amount based on this, and the O ₂ concentration (output signal) output from the O ₂ sensor 24, the ideal air-fuel ratio (target air-fuel ratio) ) With respect to the basic fuel injection amount, an O ₂ feedback section (feedback correction coefficient calculation means) 44 that performs feedback control by correcting the feedback correction coefficient so as to approach O and correcting unit 48 for correcting, based on the obtained correction amount _second feedback unit 44, the fuel injection time corresponding to the final fuel injection amount obtained determined That the fuel injection time calculating section (final fuel injection amount calculating means) having 50 and. The controller 14 is also connected to a water temperature sensor 51 that detects the water temperature of the engine 12.

The fuel injection time calculation unit 50 opens the fuel injection 16 at a predetermined timing based on the obtained fuel injection time and performs fuel injection. As a result, the air-fuel ratio changes, and the influence thereof is detected by the O ₂ sensor 24 in the exhaust pipe, and a feedback system is formed.

The O ₂ feedback unit 44 includes a rich / lean determination unit 52 that determines the rich / lean degree of exhaust based on the O ₂ concentration, and parameters for correcting the feedback correction coefficient and the basic fuel injection amount based on the determination result. And a parameter calculation unit (feedback correction learning coefficient storage means) 54 to be obtained. The O ₂ feedback unit 44 does not perform feedback control at the start of operation and during warm-up.

  The parameter calculation unit 54 records predetermined parameters in the correction amount recording unit 46 at a predetermined cycle, and reads these parameters when the system is activated (when the ignition key is turned on). The correction amount recording unit 46 is a non-volatile recording unit such as an EEPROM or a flash memory.

  1 and 2 show an example of a single cylinder, it goes without saying that a multi-cylinder engine can be similarly configured. In the case of a multi-cylinder engine, for example, the controller 14 and the accelerator sensor 26 can be shared.

  As shown in FIG. 3, the first map 40 is two-dimensional data that is referred to based on the rotational speed Ne and the throttle opening degree Th, and is recorded in a predetermined recording unit. FIG. 3 schematically shows the contents of the first map 40 for easy understanding, and the recording format for the recording unit is not limited.

In the first map 40, the basic fuel injection amount corresponding to the rotational speed Ne and the throttle opening degree Th is recorded. This basic fuel injection amount is set based on calculations, simulations, experiments, or the like in advance so as to obtain an ideal air-fuel ratio. In other words, by performing fuel injection based on this basic fuel injection amount, combustion close to the ideal air-fuel ratio can be realized. However, depending on the operating state of the engine 12, combustion deviates from the ideal air-fuel ratio. In order to eliminate such an error, feedback by the O ₂ sensor 24 is performed. In the first map 40, the data is distributed in a right upward manner. This is because the throttle opening Th increases when the rotational speed Ne is large.

A map 60 shown in FIG. 4 shows an error from the ideal air-fuel ratio in an experimental operation with the basic fuel injection amount based on the first map 40 when the feedback control by the O ₂ sensor 24 is not performed (that is, in the open loop). The degree of is schematically represented. This map 60 indicates that the darker the net pattern, the larger the error, and the thinner the net pattern, the smaller the error. The map 60 is used in feedback control estimation learning. Specifically, by using the data of the map 60 as the initial value of the deviation of the estimation learning, it is possible to suppress a change in drivability due to a large deviation (change amount) at the time of estimation.

  As can be understood from FIG. 4, the error with respect to the ideal air-fuel ratio can be divided into a certain range. Specifically, the error is smaller as the throttle opening Th is larger, and the error is smaller as the throttle opening Th is smaller. Tend to be big. Further, the larger the rotation speed Ne, the larger the error, and the smaller the rotation speed Ne, the smaller the error tends to be. Further, the error tends to be large in the idle rotation region.

  In view of such a tendency, it is reasonable to collectively control the same tendency region. In other words, it is expected that the same tendency results will be obtained to the extent that there is substantially no problem if the same tendency is controlled, and the number of parameters is suppressed by the number of regions. The first map 40 in FIG. 3 is divided into a plurality of regions in advance according to such deviation from the ideal air-fuel ratio and the operating conditions (that is, the rotational speed Ne and the throttle opening degree Th).

  As shown in FIG. 3, for example, about 10 sections are preferable. For example, approximately vertical lines 62 and 64 that partition the rotational speed Ne and a straight line (or a curve) that rises to the right that sections the throttle opening Th. ) 66 and 68 are divided into 9 regions, and the independent idle regions are combined into 10 regions. The straight lines 66 and 68 are lines corresponding to the upwardly rising shape of the entire first map 40. These lines 62 and 64 and straight lines 66 and 68 are set so that the errors shown in the map 60 in FIG.

  The first region 70a is an idle region, and the second region 70b, the third region 70c, and the fourth region 70d are regions where the rotational speed Ne is the lowest, and are set in order from a lower throttle opening Th to a higher one. . The fifth region 70e, the sixth region 70f, and the seventh region 70g are regions where the rotational speed Ne is medium, and are set in order from the lower throttle opening Th to the higher one. The eighth region 70h, the ninth region 70i, and the tenth region 70j are regions where the rotational speed Ne is the highest, and are set in order from the lower throttle opening Th to the higher one.

  As can be seen from FIG. 3, when the rotation speed Ne is compared at the same position, the region width H is set smaller as the throttle opening degree Th is smaller. For example, at a location where the engine speed Ne is a small constant value Ne1, the region width H4 of the upper fourth region 70d, the region width H3 of the middle third region 70c, and the region width H2 of the lower second region 70b are decreased in this order. It is set to be.

  As a result, the area of the normal use area where the throttle opening degree Th is small (that is, the first area 70a, the second area 70b, the fifth area 70e, the eighth area 70h, etc.) is limited to the area and is the correction coefficient for the area. It is possible to update KBUK and the like with high accuracy and high frequency.

  At first glance, it can be considered that the first area 70a may be included in the second area 70b. However, the idle area is an area that is frequently used and has a large disturbance. Can do.

For the first region 70a to the tenth region 70j, a variable KO2 [x] and a variable KBUK [x] (correction coefficient) for performing O ₂ feedback are provided. Here, the argument x is an identifier of 10 areas and takes a value of 1 to 10. Hereinafter, as necessary, the variable KO2 [x] and the variable KBUK [x] are simply expressed as KO2 and KBUK.

KO2 is a variable that is primarily used for each predetermined control period when O ₂ feedback is performed. Basically, O ₂ feedback is performed based on this KO ₂ to approach the ideal air-fuel ratio. KBUK is an auxiliary variable of KO2, and supplements KO2 with an appropriate period so as to return KO2 to the initial value 1.

  In other words, KO2 corresponds to a state transition process on control, or to cope with a considerably high cycle fluctuation, and KBUK tends to have a medium-long-term trend of a medium cycle, a low cycle, or a non-cycle. It is for correspondence. KBUK is recorded in the correction amount recording unit 46 at a predetermined cycle, and the value is retained even after the system is stopped (when the ignition key is turned off). The value is read when the system is started, and so-called learning control is performed.

Specifically, KO2 as a feedback correction amount is obtained based on the rich / lean degree of exhaust gas determined by the rich / lean determination unit 52 from the O ₂ concentration read from the O ₂ sensor 24. Here, the integrated correction coefficient KT is obtained by the following equation (1). KO2 takes a value in the range of +0.5 to +2.0, for example, and KO2 and KBUK are the reference value 1 in the default state. KTW is an engine water temperature coefficient, KBUR is a time-dependent change auxiliary coefficient, and KALT is an environmental auxiliary coefficient. In order to simplify the description, KTW = KBUR = KALT = 1 is assumed below.
KT ← KO2 × KTW × KBUK × KBUR × KALT (1)

Thereafter, assuming that the basic fuel injection amount read from the first map 40 is T0, the corrected fuel injection amount T1 is obtained by the equation (2) and supplied to the fuel injection time calculation unit 50. That is, the final fuel injection amount is determined so as to approach the ideal air-fuel ratio by multiplying the basic fuel injection amount T0 by KT as a feedback correction coefficient.
T1 ← T0 × KT (2)

KO2 acts to decrease the fuel injection amount from the current value when in the range of +0.5 to +1.0, and increases the fuel injection amount from the current value when in the range of +1.0 to +2.0. It acts and becomes a feedback system. KO2 becomes 1 when combustion is performed at the ideal air-fuel ratio. The O ₂ feedback control is, for example, a feedback system in which KO2 takes positive and negative values, addition / subtraction with respect to the basic fuel injection amount, and multiplication by a predetermined gain or correction coefficient (for example, PID coefficient) is output. Also good.

  By the way, although KO2 can cope with a considerably high cycle fluctuation, it also superimposes a low cycle and aperiodic fluctuation component, so that the component is recorded in the correction amount recording unit 46 after moving to KBUK. It is desirable to keep it. Therefore, the correction coefficient reflecting step for reflecting the value of KO2 to KBUK at a predetermined rate is performed.

That is, while maintaining the value of KO2 × KBUK, KO2 is brought close to the reference value 1 at a predetermined rate, and the change from the reference value 1 is complemented with KBUK. For example, KO2 _AVE = 1.2 (subscript AVE indicates an average value (for example, moving average for a predetermined period)) and KBUK = 1.0 (KO2 _AVE × KBUK = 1.2). Sometimes, KO2 = 1.1 and KBUK = 1.0909 (KO2 _AVE × KBUK≈1.2). By repeating such conversion processing at a predetermined cycle, KO2 or its average value approaches 1, and only a considerably high frequency component is included, and some high-cycle, medium-cycle, low-cycle, or non-cycle components are KBUK. Will be included. Since KBUK complements KO2 _AVE which is an average value of KO2, an error (noise etc.) component is removed.

FIG. 5 shows changes in KO2 _AVE and KBUK. KBUK is converted at a predetermined period, or converted when KO2 _AVE is equal to or higher than a predetermined value or lower than a predetermined value. In FIG. 5, KO2 _AVE and KBUK change smoothly, but microscopically change stepwise.

In this manner, KBUK is complemented and converted so that KO2 _AVE approaches 1.0, and is recorded in the correction amount recording unit 46. Therefore, when the system is restarted after the system is stopped, KBUK is recorded before the system is stopped. Even if KO2 is initialized to 1, the feedback system immediately returns to the state before the system stop, and the combustion of the engine 12 becomes close to the ideal air-fuel ratio. An effect is obtained.

  Such KO2 and KBUK are provided for each of the first region 70a to the tenth region 70j, and appropriate feedback control is performed according to the characteristics of each region.

By the way, KO2 and KBUK are provided and processed and updated for each of the first area 70a to the tenth area 70j, but these areas 70a to 70j are not necessarily executed at the same frequency. When traveling, the traveling frequency tends to be high when the throttle opening degree Th and the engine speed Ne are small. In the actual running state, for example, in the first area 70a, the second area 70b, the third area 70c, the sixth area 70f, and the seventh area 70g, the corresponding KO2 and KBUK (that is, KO2 [1], KO2 [2], KO2 [3], KO2 [6], KO2 [7], KBUK [1], KBUK [2], KBUK [3], KBUK [6], KBUK [7]) are frequently updated. On the other hand, if there is no special means, in the fourth area 70d, the fifth area 70e, the eighth area 70h, the ninth area 70i, and the tenth area 70j, the corresponding KO2 and K BUK (that is, KO2 [4], KO2 [ 5], KO2 [8], KO2 [9], KO2 [10], KBUK [4], KBUK [5], KBUK [8], KBUK [9], KBUK [10]) are not updated much.

  In such a region, the fuel injection apparatus 10 according to the present embodiment performs estimation learning control described below so that the parameters can be updated quickly.

  In the estimation learning control, a basic change rate RATE [x] of the basic fuel injection amount in each of the regions 70a to 70j is obtained in advance and recorded in a predetermined recording unit.

  As shown in FIG. 6, the basic change rate RATE [x] is provided corresponding to the first region 70a to the tenth region 70j, and indicates the relative change degree of KBUK between the regions 70a to 70j. Yes. The basic change rate RATE [x] is obtained by experiment, for example, and the amount of change in KBUK in a state where the amount of air passing through the throttle changes by a predetermined amount is measured as basic data, and the ratio between each region is maintained. The values of RATE [2] to RATE [10] in other areas are adjusted and recorded so that RATE [1] = 1 in one area 70a. The reason why the RATE [1] of the first area 70a is adjusted and set to 1 as the reference value is that it is convenient for performing the estimated learning control in the idle area as will be described later.

  Further, the average value of RATE [4], RATE [7], and RATE [10] in the upper stage is obtained as the upper average RATE1, and the average value of RATE [3], RATE [6], and RATE [9] in the middle stage is determined as the middle average. Obtained as RATE2, the average value of RATE [2], RATE [5], and RATE [8] at the lower stage is obtained as lower stage average RATE3 and recorded.

  The timing for obtaining the basic rate of change RATE [x] and RATE1, RATE2, and RATE3 may be at the time of initial driving of the vehicle (a time corresponding to so-called running-in) or at an adjustment stage before being handed to the user (for example, bench It may be performed in advance (on a tester), or an experimentally determined value may be recorded in advance. The processing for obtaining the basic rate of change RATE [x] and RATE1, RATE2, and RATE3 is not necessarily performed once in the initial stage, but is repeated at an appropriate cycle (for example, every 10,000 km or at the time of vehicle inspection, etc.). Also good.

  The estimation learning control is divided into idle estimation learning control and travel estimation learning control.

Idle estimation learning control is performed when the vehicle is in an idle state. As shown in FIG. 7, each KBUK is initially 1.000. Assume that from this state, KBUK of the first area 70a, which is an idle area, decreases by a threshold value of 0.100 and changes to 0.900. At this time, the values of all the KBUK [2] to KBUK [10] in the second area 70b to the tenth area 70j are changed to the current area change amount V (here, -0.100) of KBUK [1] and the basic change rate RATE. [2] to RATE [10] Based on correction. For example, for KBUK [4], KBUK _NEW [4] ← KBUK _OLD [4] + V × RATE [4] = 1.000 + (− 0.100) × 0. It is updated as 080 = 0.992. The subscripts “OLD” and “NEW” indicate before and after the update. The other areas can be obtained by the same calculation.

  Since the disturbance is large in the idle region, it is appropriate to accurately correct all other regions based on the basic change rate RATE as described above.

  Next, travel estimation learning control will be described. The estimation learning control during traveling is performed during traveling.

  As shown in FIG. 8, it is assumed that KBUK [6] in the sixth region 70f is decreased by a threshold value 0.100 from the state of 0.988 and changed to 0.888 while the vehicle is traveling. At this time, correction is performed on the upper, middle, and lower KBUKs based on the current region change amount V (here, -0.100) and the upper average RATE1, the middle average RATE2, and the lower average RATE3, or the correction amount is limited. . The correction amount limitation prevents the amount of change due to correction from exceeding the current region change amount V.

As shown in FIG. 9, as an example, for the upper KBUK [4], using the upper average RATE1, based on the original value _{0.992, KBUK NEW [4] ←} KBUK OLD [4] + V × RATE1 /RATE[6]=0.922+(−0.100)×0.067/0.12=0.936. In this case, the amount that is changed by the correction is 0.992−0.936 = 0.056, which is smaller than the current region change amount V, and thus there is no limitation. The other areas can be obtained by the same calculation. For KBUK [4], RATE [4] may be used instead of the upper average RATE1. By using the upper average RATE1, the processing for the upper KBKB [4], KBUK [7], and KBUK [10] is simplified.

Next, the middle KBUK [3] is basically corrected in the same manner as the above-mentioned KBUK [4], but the RATE2 / RATE [6] with respect to the current region change amount V which is the basis in the calculation process. ], When RATE2 / RATE [6] ≧ 1, the amount that changes due to the correction is equal to or greater than the current region change amount V. Therefore, in such a case, the current region change amount V is limited. In this case, since RATE2 / RATE [6] = 0.12 / 0.12 ≧ 1, the current region change amount V is limited, and KBUK _NEW [3] ← KBUK _OLD [3] + V = 0.984− Obtained as 0.100 = 0.484. The same applies to the middle KBUK [9].

The case of the lower stage is the same as that of the middle stage. As an example, for KBUK [2], RATE3 / RATE [6] = 0.44 / 0.12 ≧ 1, so that the restriction is made by the current region change amount V, and KBUK _NEW [2] ← KBUK _OLD [2] + V = 0.968-0.100 = 0.868 The same applies to KBUK [5] and KBUK [8].

  In this way, since there is little disturbance during running in the running estimation learning control, simple calculation is sufficient, and the upper average RATE1, the middle average RATE2, the lower average RATE3 are used, and the restriction process by the current region change amount V is simplified. Is planned.

  The estimation correction is only an estimation, and it is desirable to prevent excessive correction. Therefore, by providing a restriction based on the current region change amount V, it is possible to prevent KBUK from being excessively corrected. The limit value is appropriate as the current region change amount V, and the processing becomes simple.

  As described above, according to the fuel injection device 10 for an engine according to the present embodiment, when the engine operating state is divided into 10 regions and KBUK is updated by a predetermined amount for one region, By performing estimation correction for KBUKs in other areas in accordance with the fluctuation tendency, the KBUKs in other areas can be updated to appropriate values as quickly as possible. Therefore, for example, immediately after entering a highway after driving in a state where the throttle opening degree Th is small, the corresponding control parameter KBUK is almost at an appropriate value. Therefore, an ideal air-fuel ratio can be obtained immediately.

  Next, a second control method performed by the controller 14 will be described. In the second control method and the third control method, the same constituent parts and the same process parts in the first control method are denoted by the same reference numerals and the same names, and detailed description thereof is omitted. In the second control method, the controller 14 refers to the second map 80.

  As shown in FIG. 10, the second map 80 is set based on the engine speed Ne and the throttle opening degree Th similarly to the first map 40, and the outer frame is set in the same region. ing. The areas 82a (idle areas), 82b, 82c, 82e, 82f, 82h and 82i in the second map 80 are set to be the same as the areas 70a, 70b, 70c, 70e, 70f, 70h and 70i in the first map 40. The area 82d in the second map 80 has a large throttle opening Th in the first map 40, and is set to an area where the upper three areas 70d, 70g and 70j are integrated on the graph. The regions 82g and 82j are omitted in the second map 80 so that the correspondence with each region of the first map 40 is easy.

  In the second map 80, in the regions 82a, 82b, 82c, 82e, 82f, 82h and 82i, the same control as the regions 70a, 70b, 70c, 70e, 70f, 70h and 70i in the first map 40 is performed.

The upper region 82d is provided with a calculation representative area 84 slightly above the center, and the parameters set in the representative area 84 are reflected in the entire area 82d. The representative area 84 is a simple rectangle with a moderate width and a slightly wider width than the areas 82e and 82f, and is easy to set. It is preferable that the representative area 84 is set to an area including the intermediate rotation speed Nec in the use range of the engine rotation speed Ne (from the idle rotation speed Ne1 to the red zone rotation speed Ne2). Furthermore, the width of the engine speed Ne direction in the representative area 84 is suitably about 4000 rpm to 8000 rpm, and the height in the direction of the throttle opening Th is 10 ° to 40 ° (under the condition of O ₂ feedback). The degree is appropriate.

  The representative area 84 is an area substantially equal to the area 70g, and the parameter corresponding to KBUK [7] corresponding to the area 70g is reflected in the area 82d.

  The representative area 84 is a place where the throttle opening Th is not easily affected by aging deterioration and the change in the intake air is stable. By learning the air density change in such an area, the representative area 84 is stabilized. Highly accurate control. The parameter in the representative area 84 acts as a deterioration / environment correction coefficient for the first time (a coefficient for correcting a long-term trend change such as a change over time in an internal combustion engine or a short-term trend change such as a change in atmospheric pressure). Is reflected in the entire area 82d.

  According to the experiment result of the present inventor, in the portion where the throttle opening degree Th is high as in the region 82d, the left and right portions (that is, the portion where the engine rotation Ne is low and the portion where the engine rotation Ne is high) are higher than the frequency where the central portion is frequently used. Is not used frequently. Therefore, while the parameter is frequently updated in the central portion, the parameter does not change so much in the left and right portions, and there is a possibility that convergence to an appropriate value may be delayed, but as in the second map 80 By integrating the portions where the throttle opening degree Th is high in one region 82d, the parameters can be updated collectively for the region 82d and set to an appropriate value promptly. Further, since the region 82d is one region, even when the engine speed Ne changes from low to high in the region 82d, it is controlled with the same parameters and there is little useless state change.

  Furthermore, since the parameters in the representative area 84 in the central part where the frequency of use is high are reflected in the entire area 82d, the parameters can be updated with high frequency and learning control can be performed. Furthermore, since the number of areas is small compared to the first map 40, the control procedure is simplified, the control load is reduced, and the parameter storage capacity is reduced.

  Note that the number of regions may be further increased depending on design conditions such as allowance for CPU calculation load. At this time, the region 82d where the throttle opening degree Th is large may be left as it is, and the regions 82c, 82f and 82i where the throttle opening degree Th is medium may be divided into two upper limits as indicated by a virtual line 86. In this case, it is preferable that the straight line 66 is slightly lowered so that the region in the axial direction of the throttle opening Th is smaller in the region where the throttle opening is smaller.

  As described above, according to the second control method, in the second map 80, the side with the highest throttle opening degree Th is set to one region 82d irrespective of the engine speed Ne, and the throttle opening degree Th The lowest side is set in a plurality of regions 82a, 82b, 82e and 82h corresponding to the engine speed Ne.

  On the side where the throttle opening Th is the highest, there is a difference in the frequency of use based on the engine speed Ne, so it is reasonable to make it an integrated one region. On the other hand, on the side where the throttle opening Th is the lowest, a plurality of regions corresponding to the engine speed Ne are set, and control can be performed with higher accuracy. In a place where the throttle opening degree Th is medium, an appropriate area division according to the design condition may be performed, and the map 80 is divided into three areas 82c, 82f and 82i.

  A representative area 84 is provided in one area 82d set on the side where the throttle opening Th is the highest, and the learning correction coefficient obtained in the representative area 84 is reflected in one whole area, thereby the one area 82d. The learning correction coefficient can be applied simply and accurately for the entire region.

  Next, a third control method performed by the controller 14 will be described. In the third control method, the controller 14 refers to the third map 90.

  As shown in FIG. 11, the third map 90 is set based on the engine speed Ne and the throttle opening degree Th similarly to the first map 40, and the outer frame is set in the same region. ing.

  The area 92a (idle area) related to idle in the third map 90 is set to be the same as the area 70a in the first map 40. A region 92b in which the throttle opening degree Th is medium / low in the third map 90 (medium / low throttle region) 92b is a region in which the regions 70b, 70c, 70e, 70f, 70h, and 70i in the first map 40 are integrated. In addition, a region where the throttle opening degree Th is high (high throttle region) 92c is a region where regions 70d, 70g and 70i are integrated. Thus, in the third map 90, only three regions 92a to 92c are provided. The area 92c corresponds to the area 82d in the second map 80, and the same representative area 94 as the representative area 84 is provided. In the third map 90, the same control as that in the region 70a in the first map 40 is performed in the region 92a, and the same control as that in the region 82d in the second map 80 is performed in the region 92c.

In the third control method, parameters KAD [1], KAD [2] and KAD [3] (hereinafter also referred to as KAD) corresponding to KBUK are obtained for the regions 92a, 92b and 92c. KAD has a reference value of 1.0. Then, similarly to KBUK, KAD is updated as appropriate, the integrated correction coefficient KT is obtained by equation (3), the corrected fuel injection amount T1 is obtained by equation (2), and is supplied to the fuel injection time calculation unit 50. .
KT ← KO2 × KTW × KAD × KBUR × KALT (3)

KAD [1] in the region 92a related to the idle state is obtained by subtracting the KO2 average value at that time from the target KO2 in the idle state, similarly to KBUK [1] in the region 70a. KAD [3] in the region 92c where the throttle opening degree Th is high is obtained based on the representative region 94, and is obtained as substantially the same value as KBUK [7] in the region 70g. This KAD [3] is obtained by the following equation (4) based on KAD [1] and KAD [3] for KAD [2] in the region 92b where the throttle opening degree Th is medium and low.
KAD [2] ← ((KAD [1] −1.0) × Kc + 1.0) × KAD [3] (4)

  Here, the parameter Kc is a value obtained by referring to the table 100 shown in FIG. The table 100 is recorded by dividing the region 92b into 7 levels for the throttle opening Th and 5 levels (for example, every 1000 rpm) for the engine speed Ne. The parameter Kc is blank in the upper left part according to the shape of the area 92b, but of course, data may be recorded in this part.

  In the equation (4), first, a change amount with respect to the reference value 1.0 is obtained for KAD [1], weighted by multiplying the parameter Kc, and after adding the reference value 1.0, KAD [1] 3]. Therefore, KAD [3] is corrected by adjusting the influence of KAD [1] according to the category. The equation for obtaining KAD [2] is not limited to equation (4), and other appropriate means based on KAD [1] and KAD [3] may be used.

  Further, the amount of change in KAD [3] is learned as a result of changes due to atmospheric pressure, engine deterioration, or the like, and the amount of change is reflected in KAD [1], thereby absorbing deviations from these effects. It is also possible to respond as described above.

  As described above, according to the third control method, the third map 90 includes the region 92a in the idle region corresponding to the idle rotation, the region 92c in the high throttle region on the side with the highest throttle opening Th, KAD [2] that is a learning correction coefficient of the region 92b is obtained from KAD [1] of the region 92a and KAD [3] of the region 92c. It is done. As described above, when the three regions 92a to 92c are used, the number of regions is sufficiently small, and the storage capacity for storing parameters such as learning correction coefficients becomes small. In addition, since the number of parameters is reduced, the calculation load on the CPU is reduced and the processing procedure is simplified.

  In the present embodiment, the electronic throttle valve system in which the throttle valve is opened and closed by the actuator has been described. However, the present invention can also be applied to a mechanical throttle valve system in which the throttle valve is opened and closed by a wire. Further, an oil temperature sensor that detects the temperature of the engine lubricating oil may be used instead of the water temperature sensor.

  The fuel injection device for an internal combustion engine according to the present invention is not limited to the above-described embodiment, and it is needless to say that various configurations can be adopted without departing from the gist of the present invention.

It is a schematic diagram of an engine system It is a block block diagram of a controller. It is a figure which shows the several area | region of the 1st map where the basic fuel injection quantity was recorded. It is a figure which represents typically the grade of the difference | error with an ideal air fuel ratio when not performing feedback. It is a time chart showing transition of variables KO2 and KBUK. It is a table | surface showing the basic change rate RATE for every area | region. It is a table | surface of the state in which KBUK changed in the idle area | region. It is a table | surface which shows the update of KBUK by idle time estimation learning control. It is a table | surface which shows the update of KBUK by driving | running | working estimation learning control. It is a figure which shows the several area | region of the 2nd map where the basic fuel injection quantity was recorded. It is a figure which shows the several area | region of the 3rd map where the basic fuel injection quantity was recorded. It is a table in which the parameter Kc is recorded.

Explanation of symbols

10 ... Fuel injection system 11 ... engine system 12 ... engine 14 ... controller 16 ... Fuel Injection 18 ... spark plug 20 ... throttle valve 24 ... O ₂ sensor 26 ... accelerator sensor 28 ... crank 30 ... rotational speed sensor 40, 60, 80, 90 ... Map 42 ... Basic fuel injection amount calculation unit 44 ... Feedback unit 46 ... Correction amount recording unit 48 ... Correction unit 50 ... Fuel injection time calculation units 70a-70j, 82a-82f, 82h, 82i, 92a-92c ... Area Ne ... rotation speed Th ... throttle opening

Claims (3)

Referring to the basic fuel injection map based on engine speed and (Ne) of the throttle opening (Th) and the engine throttle valve disposed in an intake system of an internal combustion engine (12) (20) (12), the throttle Basic injection amount calculating means (42) for determining a basic fuel injection amount based on the opening degree (Th) and the rotational speed (Ne) ;
Feedback correction coefficient calculation means (44) for obtaining a feedback correction coefficient for performing feedback control of fuel injection at a predetermined cycle based on an output signal output from an O ₂ sensor (24) provided in the exhaust system;
A final fuel injection amount calculating means (50) for multiplying the basic fuel injection amount by the feedback correction coefficient and determining a final fuel injection amount so as to approach a target air-fuel ratio;
Feedback correction learning coefficient storage means (54) for storing the feedback correction coefficient as a learning correction coefficient in the nonvolatile storage unit (46) at a predetermined timing ;
The feedback correction learning coefficient storage means (54) stores the calculated learning correction coefficient for each area of the basic fuel injection map that is divided into a plurality of areas in advance according to the driving situation,
A plurality of said areas compares the rotational speed in the basic fuel injection map (Ne) is in the same place, the more area on the side throttle opening (Th) is small, the axial direction of the throttle opening (Th) In a fuel injection device for an internal combustion engine with a reduced area width ,
In the plurality of regions, the side with the highest throttle opening (Th) is set to one region (82d or 92c) irrespective of the rotational speed (Ne), and the throttle opening (Th) is the lowest. The side is set in a plurality of regions according to the rotation speed (Ne),
A representative area (84 or 94) is provided in one area (82d or 92c) set on the side with the highest throttle opening (Th), and the learning required in the representative area (84 or 94) is provided. Reflecting the correction coefficient over the entire area (82d or 92c),
The representative area (84 or 94) is an area including an intermediate rotation speed (Nec) in a use area of the rotation speed (Ne) of the internal combustion engine (12), and the rotation speed (Ne) is low. When the fuel injection system for an internal combustion engine, wherein an area der Rukoto without the high region. The fuel injection device for an internal combustion engine according to claim 1,
In the representative region (84 or 94), the range of the rotational speed (Ne) is 4000 rpm to 8000 rpm, and the range of the throttle opening (Th) is 10 ° to 40 °. Engine fuel injection device. The fuel injection device for an internal combustion engine according to claim 1 or 2 ,
The plurality of regions include an idle region (92a) corresponding to idle rotation, a high throttle region (92c) on the side with the highest throttle opening (Th) , the idle region (92a), and the high throttle region ( 92c) is comprised of three medium and low throttle areas (92b) ,
The fuel for the internal combustion engine, wherein the learning correction coefficient in the middle / low throttle area (92b) is obtained from the learning correction coefficient in the idle area (92a) and the learning correction coefficient in the high throttle area (92c). Injection device.

Images