JP3550821B2 - Engine air-fuel ratio control device - Google Patents

Description

[0001]
TECHNICAL FIELD OF THE INVENTION
The present invention relates to an air-fuel ratio control device for an engine.
[0002]
[Prior art]
Air-fuel ratio sensors are provided upstream and downstream of the exhaust gas purification catalyst.The air-fuel ratio feedback control is performed by the upstream air-fuel ratio sensor, and the control constant used for the air-fuel ratio feedback control is determined based on the output of the downstream air-fuel ratio sensor. So-called double O ₂ There is a sensor system device (see JP-A-62-60941, JP-A-63-97851, and JP-A-3-217636).
[0003]
In this device, the upstream O ₂ By comparing the sensor output with the slice level, the air-fuel ratio feedback correction amount α is generated as follows.
[0004]
(1) Immediately after inversion from rich to lean, in order to return the air-fuel ratio to the rich side in a stepwise manner, the air-fuel ratio feedback correction amount α calculated last time is increased by a proportional amount PL, and the inversion from lean to rich is performed. Immediately afterward, in order to gradually return the air-fuel ratio to the lean side, α is reduced by a proportional amount PR from the previous value.
[0005]
(2) In this case, when the air-fuel ratio is the same lean as the previous time, the air-fuel ratio is gradually returned to the rich side, so that α is increased by the integral IL (IL <PL) from the previous value. At some point, α is reduced from the previous value by an integral amount IR to gradually return the air-fuel ratio to the lean side.
[0006]
The α thus formed has a periodic waveform as shown in FIG.
[0007]
On the other hand, during the air-fuel ratio feedback control, the downstream O ₂ In order to perform the correction control of the control constant (for example, the proportional component PL) of the air-fuel ratio feedback control based on the sensor output, the downstream O ₂ By comparing the sensor output with the slice level, a correction value LP is created as follows.
[0008]
When lean, the air-fuel ratio is shifted to the rich side as a whole, so that the correction value LP is increased by a constant value ΔLP. When rich, the air-fuel ratio is shifted to the lean side, so the correction value LP is constant. It is increased by ΔLP (see FIG. 19).
[0009]
The correction value LP thus created is added to the proportional component PL immediately after the inversion from rich to lean, and is subtracted from the proportional component PR immediately after the inversion from lean to rich.
[0010]
[Problems to be solved by the invention]
By the way, downstream O ₂ The movement of the air-fuel ratio at the sensor is ₂ Since there is a response delay with respect to the air-fuel ratio at the sensor, ₂ In a part of the period (time or crank angle) in which the sensor output is determined to be rich or lean, an erroneous determination occurs due to a response delay. Furthermore, the actual downstream O ₂ As shown in FIG. 20, the characteristics of the sensor output are different between the response from rich to lean and the response from lean to rich. If the period difference between the two is relatively large, the downstream O ₂ An erroneous determination is made that the sensor output as a whole is rich or lean. If the control constant of the air-fuel ratio feedback control is corrected based on the result of the erroneous determination, the air-fuel ratio feedback control will perform an unnecessary rich shift or a lean shift, and good driving performance and exhaust purification performance cannot be obtained. Can occur.
[0011]
Here, the specifications of the engine, the configuration of the exhaust system including the catalyst, the fuel injection valve and the O ₂ If the characteristics of each part related to the correction control of this control constant, such as the characteristics of the sensor and the filter characteristics of the input / output unit of the control circuit, are clear, the above-mentioned unnecessary air-fuel ratio shift amount is determined in advance by experiments or the like. At the time of actual control, it is possible to make a correction so as to cancel this. ₂ The sensor deteriorates and its characteristics change as the operating time of the engine increases, and even if the exhaust gas flow rate changes due to the operating conditions of the engine, the downstream O ₂ Since the response of the sensor output changes, it is actually difficult to prevent an erroneous determination over a long period of time or in response to a wide range of operating conditions of the engine using only the correction amount obtained in advance through experiments or the like.
[0012]
To avoid such misjudgments, the downstream O ₂ It is conceivable that the response characteristic of the sensor is actually measured, and the result of the correction control of the control constant is further corrected based on the measurement result. As an example, by superimposing a response measurement signal such as a rectangular wave on the basic signal of the fuel injection amount of the engine, the downstream O ₂ Japanese Patent Application Laid-Open No. Hei 5-187296 discloses a technique for measuring a response characteristic from a response waveform of a sensor.
[0013]
However, this has the following two problems.
[0014]
1) When measuring the response characteristics, at least the downstream O ₂ Correction control of the control constant by the sensor needs to be stopped.
[0015]
2) When a rectangular wave is input, it is necessary to fix the center air-fuel ratio to a predetermined value (preferably a stoichiometric air-fuel ratio). In the case of a step input, the air-fuel ratio at the start point and the end point must also be fixed at a predetermined value. If these settings are not made with high accuracy, an error will occur in the response measurement result.
[0016]
Due to these two problems, a signal for response measurement is input, and the response waveform is observed. ₂ It is actually difficult to change and adjust the sensor output and the control result based on the modification of the control constant.
[0017]
Accordingly, the present invention provides a response period from the rich-to-lean output of the downstream air-fuel ratio sensor during the correction control of the control constant of the air-fuel ratio feedback control in order to know how rich or lean the misjudgment was made. And the response time from lean to rich is measured, and a correction according to this response time difference is added to the correction value of the control constant of the air-fuel ratio feedback control, so that the output from the downstream air-fuel ratio sensor changes from rich to lean. The purpose of the present invention is to obtain good driving performance and exhaust gas purification performance even when there is a large response period difference between the response from the lean and the rich.
[0018]
[Means for Solving the Problems]
In the first invention, as shown in FIG. 21, the air-fuel ratio sensors 31 and 32 on the upstream and downstream sides of the catalyst and the basic control constants of the air-fuel ratio feedback control (for example, proportional components PL and PR, integral component IL, Means 33 for calculating IR, a slice level SLF to be compared with the upstream air-fuel ratio sensor, etc.) One of the conditions is that there is an operation condition in a predetermined correction control region which is in the control region for performing the air-fuel ratio feedback control and is narrower than the control region for performing the air-fuel ratio feedback control. Means 34 for determining whether the correction control condition of the control constant is satisfied or not, and calculating a correction value based on the output of the downstream air-fuel ratio sensor 32 when the correction control condition is satisfied based on the determination result. Means 35, means 36 for calculating the first control constant by correcting the basic control constant with the corrected value, means 38 for storing the learning value, and based on the corrected value when the corrected control condition is satisfied. Means 39 for updating the learning value stored in the storage means 38; and a difference between a response from rich to lean and a response from lean to rich from the output of the downstream air-fuel ratio sensor 32 when the correction control condition is satisfied. (A time difference or a crank angle difference), a means 41 for calculating a correction value for the learning value according to the response difference, a means 42 for storing the correction value, and a storage means 42 A value obtained by correcting the learning value stored in the storage means 38 with the stored correction value is set as a correction value, and the basic control constant is corrected with the set correction value to set a second control constant. Means 43 for calculating, based on the output of the upstream air-fuel ratio sensor 31 using the first control constant when the modified control condition is satisfied, and using the second control constant when the modified control condition is not satisfied. Means 44 for performing feedback control of the air-fuel ratio.
[0019]
In the second invention, as shown in FIG. 22, the air-fuel ratio sensors 31 and 32 on the upstream and downstream sides of the catalyst and the basic control constants of the air-fuel ratio feedback control (for example, proportional components PL and PR, integral component IL, IR, a slice level SLF to be compared with the output of the upstream air-fuel ratio sensor, etc.), and the output of the downstream air-fuel ratio sensor 32 is delayed. 51, means 52 for calculating a correction value based on the delayed and limited signal when the air-fuel ratio feedback control condition is satisfied, and the basic control constant Means 53 for calculating a control constant by correcting the air-fuel ratio, means 54 for performing feedback control of the air-fuel ratio based on the output of the upstream air-fuel ratio sensor 31 using the control constant, Means 56 for calculating the difference (time difference or crank angle difference) between the response from rich to lean and the response from lean to rich based on the delayed and limited signal when the feedback control condition is satisfied. Means 57 for calculating a correction value HD for each of the basic values DR0 and DL0 of the upper limit and the lower limit, and calculating the basic values DR0 and DL0 of the upper and lower limits by the correction value. Without changing the width between the two base values DR0, DL0 Means 58 for correcting and calculating the upper limit DR and the lower limit DL is provided.
[0020]
In a third aspect based on the second aspect, the learning value is updated and stored based on the correction value when the learning condition is satisfied, and the learned value is stored immediately after the air-fuel ratio feedback control condition is satisfied. It was used as the initial value of the correction value.
[0021]
In the fourth invention, as shown in FIG. 23, the air-fuel ratio sensors 31 and 32 on the upstream and downstream sides of the catalyst and the basic control constants of the air-fuel ratio feedback control (for example, proportional components PL, PR, integral IL, Means 33 for calculating IR, a slice level SLF to be compared with the output of the upstream air-fuel ratio sensor, and the like; This control corrects a control constant used for the air-fuel ratio feedback control based on the output of the downstream air-fuel ratio sensor 32. Means 61 for calculating a basic control constant of the correction control (a step toward the increasing side and a step toward the decreasing side when calculating the correction value, a slice level SLR to be compared with the output of the downstream air-fuel ratio sensor, etc.); Means 62 for calculating a difference (time difference or crank angle difference) between a rich-to-lean response and a lean-to-rich response from the output of the downstream air-fuel ratio sensor 32 when the air-fuel ratio feedback control condition is satisfied; According to the correction control Basic Means 63 for calculating a correction value for the control constant, and correcting the basic control constant of the correction control with the correction value Of the correction control Means 64 for calculating a control constant; Correction control Using control constants Repair Means 65 for calculating a positive value when said air-fuel ratio feedback control condition is satisfied; Means 53 for calculating the control constant of the air-fuel ratio feedback control by correcting the basic control constant with this correction value, and controlling the air based on the output of the upstream air-fuel ratio sensor 31 Means 54 for performing feedback control of the fuel ratio; Was provided.
[0022]
In a fifth aspect based on the fourth aspect, the learning value is updated and stored based on the correction value when the learning condition is satisfied, and this learning value is stored immediately after the air-fuel ratio feedback control condition is satisfied. It was used as the initial value of the correction value.
[0023]
According to a sixth aspect of the present invention, as shown in FIG. 24, the air-fuel ratio sensors 31 and 32 on the upstream and downstream sides of the catalyst and the basic control constants of the air-fuel ratio feedback control (for example, proportional PL, PR, integral IL, Means 33 for calculating IR, a slice level SLF to be compared with the upstream air-fuel ratio sensor, etc.) One of the conditions is that there is an operation condition in a predetermined correction control region which is in the control region for performing the air-fuel ratio feedback control and is narrower than the control region for performing the air-fuel ratio feedback control. Means 34 for determining whether the correction control condition is satisfied or not, This control corrects the control constant used for air-fuel ratio feedback control based on the output of the downstream air-fuel ratio sensor. Means 61 for calculating a basic control constant of the correction control (a step toward the increasing side and a step toward the decreasing side when calculating the correction value, a slice level SLR to be compared with the output of the downstream air-fuel ratio sensor, etc.); Said From the judgment result Means 62 for calculating the difference (time difference or crank angle difference) between the response from rich to lean and the response from lean to rich from the output of the downstream air-fuel ratio sensor 32 when the modified control condition is satisfied; Of the correction control Basic Means 63 for calculating a correction value for the control constant, and correcting the basic control constant of the correction control with the correction value Of the correction control Means 64 for calculating a control constant; Correction control Using control constants Repair Means 72 for calculating a positive value when said correction control condition is satisfied; A means 36 for calculating a first control constant by correcting the basic control constant of the air-fuel ratio feedback control with the corrected value, a means 38 for storing a learning value, and adding the corrected value when the corrected control condition is satisfied. Means 39 for updating the learning value stored in the storage means 38 based on the learning value stored in the storage means 38 as a correction value, and performing the air-fuel ratio feedback control with the set correction value. Means 71 for calculating a second control constant by correcting the basic control constant of the above, using the first control constant when the corrected control condition is satisfied, and using the second control constant when the corrected control condition is not satisfied. Means 44 for performing feedback control of the air-fuel ratio based on the output of the upstream air-fuel ratio sensor 31 using Was provided.
[0024]
In a seventh aspect based on the sixth aspect, when the learning condition is satisfied, the learning value is updated based on the correction value and stored, and the learning value is stored in the correction value immediately after the correction control condition is satisfied. Was used as the initial value of.
[0025]
In an eighth aspect based on any one of the first, fourth to seventh aspects, a difference between a period in which the downstream air-fuel ratio sensor output is lean and a period in which the downstream air-fuel ratio sensor is rich is defined as the response difference.
[0026]
In a ninth aspect based on any one of the first, fourth to seventh aspects, the slice level for comparison with the downstream air-fuel ratio sensor output has a lower limit SLRL and an upper limit SLRR, The time required for the downstream air-fuel ratio sensor output to pass from the upper limit SLRR to the lower limit SLRL and the time required for the downstream air-fuel ratio sensor output to pass from the lower limit SLRL to the upper limit SLRR The difference between the two is defined as the response difference.
[0027]
In a tenth aspect based on any one of the first, fourth to seventh aspects, a period from when the output of the downstream air-fuel ratio sensor becomes lean to when it reaches a lower limit peak, and when the downstream air-fuel ratio sensor The difference from the period from when the output becomes rich to when it reaches the upper limit peak is defined as the response difference.
[0028]
In an eleventh aspect based on the second or third aspect, a difference between a period in which the delayed and limited signal is lean and a period in which the signal is rich is defined as the response difference.
[0029]
In a twelfth aspect based on the second or third aspect, the slice level for comparison with the delayed and limited signal has a lower limit SLRL and an upper limit SLRR, and the delayed and limited signal is The difference between the time required for the signal to pass from the upper limit SLRR to the lower limit SLRL and the time required for the delayed, restricted signal to pass from the lower limit SLRL to the upper limit SLRR is defined as Response difference.
[0030]
According to a thirteenth aspect, in the second or third aspect, a period from when the delayed and limited signal reaches a lean to a lower limit peak and when the delayed and limited signal becomes rich. The difference from the period up to the upper limit peak is defined as the response difference.
[0031]
In a fourteenth aspect based on any one of the first, sixth and seventh aspects, the correction control condition is a predetermined operating range.
[0032]
[Action]
Since the learning value obtained during the correction control is affected by the imbalance in the response characteristics of the downstream air-fuel ratio sensor, when the correction control condition is not satisfied, the learning value is used as it is as the correction value for calculating the control constant. However, in the first invention, when the modified control condition is satisfied, the unbalance of the response characteristics of the downstream air-fuel ratio sensor is measured, and a correction value based on the unbalanced air-fuel ratio is determined. In the meantime, when the modified control condition is not satisfied, a value obtained by correcting the learning value with the corrected value is set as a modified value. Therefore, in an operation range where the modified control condition is not satisfied, the response characteristic of the downstream air-fuel ratio sensor is reduced. The air-fuel ratio can be made closer to the stoichiometric air-fuel ratio by eliminating the effect of imbalance.
[0033]
In the second and fourth inventions, since the correction value is reflected in the correction control itself, it is possible to eliminate the influence of the imbalance in the response characteristic of the downstream air-fuel ratio sensor output while performing the correction control.
[0034]
In the third and fifth aspects, the control accuracy of the air-fuel ratio feedback control is improved immediately after the air-fuel ratio feedback control condition is satisfied by using the learning value as the initial value of the correction value immediately after the air-fuel ratio feedback control condition is satisfied. I do.
[0035]
In the sixth invention, the correction value is reflected in the correction control itself, so that it is possible to eliminate the influence of the unbalance of the response characteristics of the output of the downstream air-fuel ratio sensor while performing the correction control. Since the corrected control result obtained is used as the learning value, it is possible to eliminate the influence of the unbalance of the response characteristics of the downstream air-fuel ratio sensor even in an operation range outside the correction control condition without correcting the learning value. .
[0036]
In the seventh invention, the control accuracy of the air-fuel ratio feedback control is improved immediately after the correction control condition is satisfied by using the learning value as the initial value of the correction value immediately after the correction control condition is satisfied.
[0037]
In the ninth and twelfth aspects, the speed of the response from rich to lean and the speed of the response from lean to rich are measured, and the response characteristics appear more clearly, so that the accuracy of the correction value is improved.
[0038]
In the tenth and thirteenth aspects, for example, the control side that has detected the inversion from rich to lean switches from the leaning process (by decreasing the correction value) to the enrichment process (by increasing the correction value). Since the period until the control appears as a result (that is, the air-fuel ratio at the downstream position of the catalyst starts to be enriched) is measured, the response characteristic of the output of the downstream air-fuel ratio sensor appears best, thereby further improving the accuracy of the correction value. I do.
[0039]
Since the influence of the imbalance of the response characteristics of the downstream air-fuel ratio sensor can be eliminated in the operating range where the modified control condition is deviated, according to the fourteenth aspect, the smaller the predetermined operating range, the lower the downstream air-fuel ratio. The operating range in which the influence of the imbalance in the response characteristics of the sensor can be eliminated is expanded.
[0040]
BEST MODE FOR CARRYING OUT THE INVENTION
In FIG. 1, reference numeral 1 denotes an engine main body. A fuel injection valve 7 is provided in an intake passage 8 at a position downstream of an intake throttle valve 5. An injection signal from a control unit (abbreviated as C / U in the figure) 2 is provided. Thus, fuel is injected and supplied into the intake air so that a predetermined air-fuel ratio is obtained according to the operating conditions.
[0041]
The Ref signal (reference position signal) and Pos signal (1 ° signal) from the crank angle sensor 4, the intake air amount signal from the air flow meter 6, the engine cooling water temperature signal from the water temperature sensor 11, and the like are input to the control unit 2. Based on these, the basic injection pulse width Tp is calculated, and feedback control of the air-fuel ratio is performed based on the air-fuel ratio (oxygen concentration) signal from the oxygen sensor 3 installed on the exhaust passage 9 on the upstream side of the three-way catalyst 10. Then, the proportional component used for the air-fuel ratio feedback control is corrected by the air-fuel ratio (oxygen concentration) signal from the oxygen sensor 13 installed on the downstream side of the three-way catalyst 10.
[0042]
Here, the air-fuel ratio feedback control is a control in which the exhaust air-fuel ratio periodically fluctuates around the stoichiometric air-fuel ratio. At this time, the three-way catalyst 10 provided in the exhaust passage 9 has the maximum conversion efficiency. Reduction of NOx in exhaust gas and oxidation of HC and CO are performed.
[0043]
The flowchart of FIG. ₂ This is a routine for calculating the air-fuel ratio feedback correction coefficient α based on the sensor output OSR1, and is executed at rotation synchronization (Ref signal synchronization) or every predetermined time (for example, every 4 ms).
[0044]
In S10, it is determined whether the air-fuel ratio feedback control condition is satisfied. (1) When the cooling water temperature Tw is equal to or lower than a predetermined value, (2) upstream O ₂ When the sensor is inactive, {circle around (3)} when the load is high, etc., are the cases where the air-fuel ratio feedback control condition is not satisfied. In this case, the process proceeds to S25, where 1 is set to α (α is clamped), and FIG. The flow of 2 is ended.
[0045]
If none of the above (1) to (3) is satisfied (when the air-fuel ratio feedback control condition is satisfied), the process proceeds to S11, and the upstream O ₂ The sensor output OSR1 is A / D converted and captured, and in S12, OSR1 is compared with the slice level SLF. If OSR1> SLF, it is determined that it is on the rich side, and "1" is set in the flag AFF1 in S13. If OSR1≤SLF, it is determined that it is on the lean side, and "0" is inserted in the flag AFF1 in S14. . AFF1 = 0 indicates a lean side, and AFF1 = 1 indicates a rich side.
[0046]
In S15, the value of the flag AFF0 is read. The flag AFF0 is a flag indicating whether the air-fuel ratio was on the rich side or the lean side last time. AFF0 = 0 indicates that it was on the lean side last time, and AFF0 = 1 indicates that it was on the rich side last time.
[0047]
In S16, the two flags AFF0 and AFF1 are compared. If the values are not equal, it is determined that OSR1 is immediately after the inversion from rich to lean or conversely, immediately after the inversion from lean to rich, and in S17, the proportionality is determined. The minutes PL and PR are calculated. The calculation of the proportional components PL and PR will be described with reference to the flowchart of FIG.
[0048]
In S30 of FIG. 3, the proportional basic values PL0 and PR0 are read. PL0 and PR0 are upstream O ₂ The basic value when the sensor 3 and the catalyst 10 are matched in a state where there is no deterioration. The basic value is assigned by a map using the rotation speed Ne and the basic injection pulse width Tp corresponding to the engine load as parameters as shown in FIGS. Search this map and ask.
[0049]
In S31, it is determined whether or not the proportional component correction control condition (to be described collectively in S40 of FIG. 7). When the proportional component correction control condition is satisfied, the correction value LP is read in S32 and S33, and the correction value LP is set to the basic value PL0. The proportional values PL and PR are calculated by setting the value obtained by subtracting the correction value LP from the basic value PR0 to the proportional value PL, and calculating the proportional components PL and PR, and then return to S18 in FIG.
[0050]
On the other hand, when the proportional component correction control condition is not satisfied, the process proceeds from S31 to S34 and S35, where the learning value LPA is read and is set as LP. The correction values HLP in S34 and S35 will be described later.
[0051]
Returning to FIG. 2, in S18, the value of the flag AFF1 is checked. If AFF1 = 0, it is determined that it is immediately after the inversion from rich to lean, and in step S19, the air-fuel ratio feedback correction coefficient α (initial value is 0) is increased by a proportional amount PL, and if AFF1 = 1, lean Is determined to be immediately after the reversal from to rich, and α is reduced by a proportional amount PR in S20.
[0052]
On the other hand, if the values of the two flags AFF0 and AFF1 are equal, it is determined that it is not immediately after inversion, and the process proceeds to S21 to check the value of the flag AFF1. If AFF1 = 0, it is determined that the current time is lean in both the previous and current times, and in step S22, α is increased by the integral amount IL. If AFF1 = 1, it is determined that the current time is rich in both the previous and current times. Is reduced by the integral IR.
[0053]
In S24, the value of AFF1 is moved to AFF0, and the flow of FIG. 2 ends.
[0054]
The flowchart of FIG. 7 is for calculating the above-mentioned correction value LP, and is executed independently of FIGS. 2 and 3 in synchronization with the Ref signal.
[0055]
In S40, it is checked whether or not the condition is the proportional correction control condition. (4) Downstream O ₂ When the sensor is inactive, (5) the operating condition is not in the predetermined proportional correction control range, or (6) when the change in the operating condition is larger than the predetermined range, the proportional correction control condition is not satisfied. . As shown in FIG. 6, the proportional component correction control area A in (5) has noise (upstream O) related to the calculation of the learning value LPA. ₂ It is sufficient to set an area where the deviation from the stoichiometric air-fuel ratio due to factors other than deterioration of the sensor and the catalyst is small.
[0056]
When the proportional component correction control condition is satisfied, the process proceeds to S41 and thereafter, but S41 to S54 are almost the same as S11 to S16 and S18 to S24 in FIG.
[0057]
In S41, the downstream O ₂ The sensor output OSR2 is read, and this OSR2 is compared with the slice level SLR in S42. If OSR2> SLR, "1" is set to the flag AFR1 in S43, and if OSR2≤SLR, "0" is set to the flag AFR1 in S44. AFR1 = 0 indicates a lean side, and AFR1 = 1 indicates a rich side.
[0058]
In S45, the value of the flag AFR0 is read. AFF0 = 0 indicates that it was on the lean side last time, and AFF0 = 1 indicates that it was on the rich side last time.
[0059]
In S46, the two flags AFR0 and AFR1 are compared. If the two values are equal, it is determined that it is not immediately after the inversion, and the process proceeds to S51 to check the value of the flag AFR1. If AFR1 = 0, it is determined that the current value is lean in both the previous time and the current time. In S52, the correction value LP (initial value is 0) is increased by a predetermined value DLP. It is determined that there is, and in S53, LP is reduced by a predetermined value DLP.
[0060]
On the other hand, when the values of the two flags AFR0 and AFR1 are not equal, it is determined that the OSR1 has just been inverted, and the process proceeds to S47 to check the value of the flag AFR1. If AFR1 = 0, it is determined that it is immediately after the inversion from rich to lean, and the process proceeds to S48 and S49. If AFR1 = 1, it is determined that it is immediately after the inversion from lean to rich, and the process proceeds to S50.
[0061]
Here, since there is no portion corresponding to S19 and S20 in FIG. 2 in FIG. 7, the correction value LP has a waveform that repeats a gradual increase and a gradual decrease as shown in FIG.
[0062]
On the other hand, S48, S49 and S50 not shown in FIG. 2 are added to FIG.
[0063]
In FIG. 7, the flow to S50 is at the timing of t2 and t4 in FIG. 10 (immediately after the inversion from lean to rich). When LP is shifted to LPR at this timing, the latest maximum value of LP is stored in LPR. enter.
[0064]
Also, the flow to S48 and S49 is at the timings t1 and t3 in FIG. 10 (immediately after the inversion from rich to lean), and the LP (that is, the minimum value of LP) and the average value LPM of the LPR at this timing are obtained. , A learning value LPA from the average value LPM,
LPA ← LPM × n + LPA × (1-n) (1)
Where n: weighted average coefficient (n <1)
It is calculated by the following equation. The learning value LPA is stored in the backup RAM.
[0065]
In S54, after the value of AFR1 is shifted to AFR0 for the next control, the calculation of the proportional components PL and PR is terminated.
[0066]
The flowchart of FIG. 11 is for calculating the fuel injection pulse width Ti, and is executed, for example, at a period of 4 ms.
[0067]
In step S1, the intake air amount Qa and the rotational speed Ne are read, and the basic injection pulse width Tp is determined from these by the equation Tp = K · Qa / Ne (where K is a constant). The fuel injection pulse width Ti is calculated by correcting with the air-fuel ratio feedback correction coefficient α and the invalid pulse width Ts, and this is set in a register in S4. Prepare for injection.
[0068]
In this way, the upstream O ₂ Basic values PL0 and PR0 obtained by matching without deterioration of the sensor and catalyst and the upstream O ₂ When calculating the proportional components PL and PR from the correction value LP for considering the deterioration state of the sensor and the catalyst (S33 in FIG. 3), if the proportional component correction control is being performed, the catalyst downstream position is set as the correction value LP. The value calculated according to the air-fuel ratio is used as it is (S32 in FIG. 3, FIG. 7), and when the proportional component correction control is not performed, the learning value LPA is used (S35 in FIG. 3).
[0069]
It should be noted that the fact that the value of the appropriate proportion differs depending on the operating conditions is stored in a map (ROM) in which the basic values PL0 and PR0 are divided according to the operating conditions (that is, the basic values are matched for each region). Since the correction value LP and the learning value LPA are not calculated for each area, only the values LP and LPA that are updated during operation are limited to two. Done.
[0070]
By the way, the learning value LPA calculated during the proportional correction control includes the downstream O ₂ Since the response characteristic of the sensor is affected by the imbalance, if the LPA is directly used as the correction value LP for the calculation of the proportional component, it will deviate from the stoichiometric air-fuel ratio. In order to prevent this deviation, the present invention newly introduces a correction value HLP for the learning value LPA. Specifically, S34 and S35 in FIG. 3 are changed, and the flow in FIG. 8 is newly added.
[0071]
The flowchart in FIG. 8 is for calculating the correction value HLP, and is executed at regular time intervals Δt independently of FIGS. 2 and 7. The same parts as those in FIG. 7 are denoted by the same reference numerals, and parts different from those in FIG. 7 will be mainly described.
[0072]
The calculation of the correction value HLP is performed on condition that the proportional correction control is being performed (S60). The calculation of the HLP is performed on condition that the proportional component correction control is being performed. ₂ This is because the influence of the imbalance of the response characteristics of the sensor is generated as a deviation from the stoichiometric air-fuel ratio without being corrected.
[0073]
If the proportional amount correction control is being performed, the process proceeds from S46 and S47 to S62 when the inversion from rich to lean is performed, and the counter T1 is reset to 0. During the lean operation, the process proceeds from S46 and S51 to S68 and the counter T1 is Δt. By adding each time, the time during which the air-fuel ratio is lean is measured by the counter T1 (see FIG. 10).
[0074]
Similarly, at the time of inversion from lean to rich, the process proceeds from S46 and S47 to S63 to reset the counter T2 to 0, and while rich, proceeds from S46 and S51 to S69 and adds the counter T2 by Δt. , T2, the time during which the air-fuel ratio is rich is measured (see FIG. 10).
[0075]
At the time of inversion, the process proceeds from S46 to S61 to calculate the difference ΔT between T1 and T2, and adds the difference ΔT to the integrated value SUMT (initial value is 0) in S64 and S65, and also sets the counter C (initial value is 0). Count up. From SUMT, C to S66, S67,
TUB = SUMT / C (2)
The average value TUB of the response difference during the proportional correction control is calculated according to the following formula, and the table of FIG. 9 is searched from this TUB to obtain the correction value HLP, which is stored in the backup RAM.
[0076]
Downstream O ₂ The relationship between the response difference of the sensor and the resulting air-fuel ratio shift amount (a deviation from the stoichiometric air-fuel ratio) and the correction value HLP for eliminating the air-fuel ratio shift can be obtained in advance by experiments or the like. Therefore, a table for setting the correction value HLP corresponding to the response difference can be created in advance, and the correction value HLP corresponding to the (average value of) the current response difference TUB is searched from this table.
[0077]
When the response difference is 0, the correction value HLP is not always 0. When the response difference is equal to the response difference when matching the control constants (for example, PL0 and PR0) of the air-fuel ratio feedback control, the correction value HLP becomes 0, but when a deviation from the response difference occurs, the correction value HLP becomes zero. In order to eliminate the air-fuel ratio shift due to the deviation, the correction value HLP of a value other than 0 is calculated according to the flow of FIG.
[0078]
In the flowchart of FIG. 8, the details are omitted. Actually, when the calculation condition of the correction value HLP is not satisfied last time and this time is satisfied, the processing for resetting both T1 and T2 and the elapsed time from the satisfaction of the calculation condition of HLP are short, and either T1 or T2 is not satisfied. When the value is 0 (that is, when the measurement is not performed), processing such as not performing the processing of S61 and S64 to S67 is required.
[0079]
The correction value HLP thus obtained is read together with the learning value LPA in S34 and S35 when the proportional component correction control condition is not satisfied in FIG. 3, and a value obtained by adding the correction value HLP to the learning value LPA is the correction value LP. Set as
[0080]
Thus, in the first embodiment, the downstream O ₂ During the proportional component correction control in which the effect of unbalance in the response characteristics of the sensor is output without correction, the downstream O ₂ The unbalance of the response characteristics of the sensor is measured, and a correction value HLP based on the unbalance is measured. When the operating range is not in the proportional correction control, the value obtained by correcting the learning value LPA with the correction value HLP is corrected to the correction value LP. In the operating range, the downstream O ₂ The air-fuel ratio is made closer to the stoichiometric air-fuel ratio by eliminating the effect of the imbalance in the response characteristics of the sensor.
[0081]
In the first embodiment, the correction value HLP is calculated in the entire proportional component correction control area A shown in FIG. 6, but the correction value HLP is calculated only in a narrower area in the control area A. You can make it work.
[0082]
In the first embodiment, as shown in FIG. 12A, a lean time T1 and a rich time T2 are measured, and the time difference ΔT is used as a response difference. Is not limited to this. Note that the waveform of OSR2 merely represents a tendency, and is different from an actual waveform.
[0083]
As shown in FIG. 12B, a lower limit SLRL and an upper limit SLRR are provided in a slice level for comparison with OSR2, and a time T1 required for OSR2 to pass from SLRR to SLRL, and vice versa, OSR2 is switched from SLRL to SLRL. The time T2 required to pass through the SLRR is measured, and the time difference between the two is taken as the response difference (second embodiment). Also, as shown in FIG. The time T1 required to reach the maximum value and the time T2 required until the OSR2 reaches the upper limit peak after the richness is measured, and the time difference between the two is used as the response difference (third embodiment).
[0084]
Here, the method shown in FIG. 12B measures the speed of the response from rich to lean and the speed of the response from lean to rich, and the response characteristics are clearer than the method of FIG. However, since a process of comparing two slice levels SLRL, SLRR and OSR2 is newly added, the process is slightly more complicated than the method of FIG. In the method of FIG. 12C, for example, the control side that has detected the reversal from rich to lean changes from the leaning process (by LP decrease) to the enrichment process (by LP increase), and the control is performed as a result. (Ie, the air-fuel ratio at the downstream position of the catalyst starts to be enriched). ₂ Although this is the measurement method in which the response characteristics of the sensor appear most frequently among the three methods, processing for detecting the peak position of OSR2 is required, and the processing is the most complicated among the three methods. As described above, if response characteristics are measured by different methods, different response differences can be obtained, so that the HLP table is completely different depending on which method is used.
[0085]
13, 15, and 16 are flowcharts of the fourth embodiment. FIG. 13 is FIG. 3, FIG. 15 is a portion for calculating the correction value LP in FIG. 7, and FIG. FIG. 8 corresponds to the part for calculating the value LPA and FIG.
[0086]
As shown in FIG. 18, the method of the fourth embodiment is basically the same as FIG. 12C, and measures the time from the inversion until the OSR2 reaches a peak.
[0087]
However, the point that time measurement is started is not the timing when OSR2 is inverted, but the timing when the signs (positive and negative) of the delay counter CAFR are inverted from FIG. 12C. Note that the delay counter CAFR is a signal (a waveform itself is different from OSR2) obtained by temporally delaying OSR2 as described later.
Hereinafter, differences from the first embodiment will be mainly described.
[0088]
First, in S70 of FIG. 13, the basic values PL0 and PR0 for the proportion and the learning value LPA are read.
[0089]
Here, in the first embodiment, PL0 and PR0 are map values according to the operating conditions, and the learning value LPA is one value in the entire operating range. However, PL0 and PR0 in the fourth embodiment are 1 in the entire operating range. The one value, the learning value LPA, is a map value (having a learning value for each learning region) according to the driving conditions as shown in FIG.
[0090]
In addition, since FIG. 13 does not include S31 in FIG. 3, the flow basically proceeds from S70 to S32 and S33 in the entire operation range where the air-fuel ratio feedback control is performed, and executes the correction control of the proportional components PL and PR. . If the correction control of the proportional components PL and PR is performed in the entire operation range in which the air-fuel ratio feedback control is performed, the deviation of the stoichiometric air-fuel ratio based on the difference in the operation conditions is also reduced to the upstream side O. ₂ The deviation of the stoichiometric air-fuel ratio based on the deterioration state of the sensor and the catalyst is also corrected by the correction value LP to obtain an appropriate proportional component (the air-fuel ratio converges on the stoichiometric air-fuel ratio). The values PL0 and PR0 may be constants independent of operating conditions. Of course, as in the first embodiment, PL0 and PR0 may be set as map values.
[0091]
Further, in FIG. 13, S71 and S72 are added, and the first time immediately after the air-fuel ratio feedback control condition is satisfied (or immediately after the operating condition shifts from one learning region to another learning region different from this in FIG. 14). At the first time, the learning value LPA is shifted to the correction value LP. The method of using the learning value LPA here is different from that of the first embodiment, by using the learning value LPA as the initial value of the correction value LP, immediately after the air-fuel ratio feedback control condition is satisfied or immediately after shifting to another learning region. This contributes to improving the control accuracy of the air-fuel ratio feedback control.
[0092]
Referring to the flowchart of FIG. 15, this is executed at predetermined time intervals Δt, unlike in FIG.
[0093]
S42 and S83 to S88 are portions for performing a process of delaying the signal of OSR2. When OSR2> SLR is rich, the delay counter CAFR is increased by a predetermined time Δt in S83, and in S84 and S85, the CAFR is set to the upper limit value (positive value). Value) If DR is exceeded, limit to DR. Similarly, when lean than OSR2 ≦ SLR, the delay counter CAFR is decreased by a predetermined time Δt in S86, and when CAFR becomes smaller than the lower limit (negative value) DL in S87 and S88, it is limited to DL. As a result, as shown in FIG. 18, the delay counter CAFR is obtained as a signal delayed from OSR2 (the waveform itself is different from OSR2).
[0094]
The CAFR thus obtained is used in place of OSR2. In S89, CAFR is compared with 0. If CAFR is smaller than 0, the correction value LP is increased by DLP in S52, and CAFR becomes 0. In the above case, LP is calculated by decreasing LP by DLP in S53. The skipped S81 and S82 will be described later.
[0095]
In the flowchart of FIG. 16, S90, S91, S92, S93, S48, S49, and S50 (the rest will be described later) are parts related to the calculation of the learning value LPA (the same parts as in FIG. ing).
[0096]
In S90, it is determined whether the air-fuel ratio feedback control condition is satisfied. If the air-fuel ratio feedback control condition is satisfied, the process proceeds to S91, and it is determined whether the learning condition is satisfied.
[0097]
When the learning condition is not satisfied (for example, when the change in the operating condition is larger than a predetermined range), the flow of FIG. 16 is terminated without calculating the learning value LPA, and when the learning condition is satisfied, the process finally proceeds to S49. Then, the learning value LPA is calculated. In the proportional correction control of the first embodiment, it is important to calculate the learning value LPA during the proportional correction control, whereas in the fourth embodiment, it is important to perform the proportional correction control itself. Since the calculation of the learning value LPA is secondary, a condition for ensuring the calculation accuracy of the learning value during the proportional component correction control (specifically, whether the operating condition change is larger than a predetermined range or not). Or not). That is, the correction control is performed as much as possible, and when the learning condition is satisfied, the learning value LPA is calculated.
[0098]
The learning value LPA in S49 of FIG. 16 is obtained by retrieving a learning value corresponding to the learning region to which the current driving condition belongs from the LPA map and updating the retrieved value as the latest learning value. It will be written in the same learning area of the map.
[0099]
In this way, even when the proportional component correction control is performed, the correction value LP (and the learning value LPA) is added to the downstream O ₂ In the fourth embodiment, a correction value HD for the upper limit value and the lower limit value of the delay counter is introduced because the imbalance of the response characteristic of the sensor is added. ₂ The ratio between the upper limit of the delay counter and the lower limit of the delay counter is changed by the correction value HD so that the unbalance of the response characteristics of the sensor output is eliminated. More specifically, in FIG. 16, portions other than the portion related to the learning value LPA and S81 and S82 in FIG. 15 are newly added. However, in FIG. 16, the same portions as those in FIG. 8 are denoted by the same reference numerals.
[0100]
In FIG. 16, when the learning condition is not satisfied in S91, the correction value HD is not calculated.
[0101]
When the learning condition is satisfied, the process proceeds from S92 and S93 to S62 at the time of inversion from rich to lean, and resets the counter T1 to 0. During lean and only when OSR2 is decreasing, the process proceeds from S92, S95, S96 to S68. And increment the counter T1 by Δt. Similarly, at the time of inversion from lean to rich, the process proceeds from S92, S93 to S63, resets the counter T2 to 0, and proceeds from S92, S95, S97 to S69 while rich and while OSR2 is increasing. Increments the counter T2 by Δt.
[0102]
At the time of inversion, the process proceeds from S92 to S61 to calculate the difference ΔT between T1 and T2, and adds the difference ΔT to the integrated value SUMT in S64 and S65, and counts up the counter C. From these two values, in S66, the average value TUB of the response difference when the learning condition is satisfied is calculated according to the above equation (2), and in S94, the table in FIG. This is stored in the backup RAM.
[0103]
The correction value HD stored in this manner is stored in S8 of FIG. 1 , S8 2 The value obtained by adding the correction value HD to the basic value DR0 of the upper limit value of the delay counter is used as the upper limit value DR of the delay counter, and the basic value DL0 of the lower limit value of the delay counter. To HD Added The value is obtained as the delay counter lower limit value DL. Since the absolute values of DR0 and DL0 have the same magnitude, if the magnitude of the correction value HD added to DR0 and DL0 is changed, the ratio of the magnitude (absolute value) of DR and DL changes.
[0104]
As described above, in the fourth embodiment, the correction value HD is reflected on the delay counter lower limit value DL and the upper limit value DR used in the proportional correction control (that is, the correction value HD is reflected on the proportional correction control itself). Therefore, the downstream O ₂ The influence of the unbalance of the response characteristics of the sensor can be eliminated by the correction value HD.
[0105]
This means that the correction value HD is reflected in other control constants of the proportional correction control (DLP, which is the size of the step toward the increasing side and the size of the step toward the decreasing side in calculating the LP, the slice level SLR, etc.). The same applies to the case of Note that DL and DR are also control constants for the proportional correction control.
[0106]
In the first embodiment, only when the proportional component correction control is not performed (other than the predetermined operation range A), the downstream O ₂ As described above, the influence of unbalance of the response characteristics of the sensor can be eliminated by the correction value HLP. This is because the correction value HLP is controlled by other control constants of the air-fuel ratio feedback control (integral IL, IR, slice level SLF). And so on).
[0107]
Further, a predetermined region for performing the control for correcting the control constant of the air-fuel ratio feedback control is set as in the first embodiment, and the correction value HD is reflected in the control for correcting the control constant itself as in the second embodiment, and the predetermined region is controlled. The learning value obtained during the correction control in (1) can be directly reflected (no correction) in the air-fuel ratio feedback control in a region where the correction control is not performed.
[0108]
【The invention's effect】
Since the learning value obtained during the correction control is affected by the imbalance in the response characteristics of the downstream air-fuel ratio sensor, when the correction control condition is not satisfied, the learning value is used as it is as the correction value for calculating the control constant. However, in the first invention, when the modified control condition is satisfied, the unbalance of the response characteristics of the downstream air-fuel ratio sensor is measured, and a correction value based on the unbalanced air-fuel ratio is determined. In the meantime, when the modified control condition is not satisfied, a value obtained by correcting the learning value with the corrected value is set as a modified value. Therefore, in an operation range where the modified control condition is not satisfied, the response characteristic of the downstream air-fuel ratio sensor is reduced. The air-fuel ratio can be made closer to the stoichiometric air-fuel ratio by eliminating the effect of imbalance.
[0109]
In the second and fourth inventions, since the correction value is reflected in the correction control itself, it is possible to eliminate the influence of the imbalance in the response characteristic of the downstream air-fuel ratio sensor output while performing the correction control.
[0110]
In the third and fifth aspects, the control accuracy of the air-fuel ratio feedback control is improved immediately after the air-fuel ratio feedback control condition is satisfied by using the learning value as the initial value of the correction value immediately after the air-fuel ratio feedback control condition is satisfied. I do.
[0111]
According to the sixth aspect of the present invention, it is possible to eliminate the influence of the imbalance of the response characteristics of the output of the downstream air-fuel ratio sensor while performing the correction control, and to perform the correction in the operating range outside the correction control condition without correcting the learning value. Also, it is possible to eliminate the influence of the unbalance of the response characteristics of the downstream air-fuel ratio sensor.
[0112]
In the seventh invention, the control accuracy of the air-fuel ratio feedback control is improved immediately after the correction control condition is satisfied by using the learning value as the initial value of the correction value immediately after the correction control condition is satisfied.
[0113]
In the ninth and twelfth aspects, the speed of the response from rich to lean and the speed of the response from lean to rich are measured, and the response characteristics appear more clearly, so that the accuracy of the correction value is improved.
[0114]
In the tenth and thirteenth aspects, for example, the control side that has detected the inversion from rich to lean switches from the leaning process (by decreasing the correction value) to the enrichment process (by increasing the correction value). Since the period until the control appears as a result (that is, the air-fuel ratio at the downstream position of the catalyst starts to be enriched) is measured, the response characteristic of the output of the downstream air-fuel ratio sensor appears best, thereby further improving the accuracy of the correction value. I do.
[0115]
Since the influence of the imbalance of the response characteristics of the downstream air-fuel ratio sensor can be eliminated in the operating range where the modified control condition is deviated, according to the fourteenth aspect, the smaller the predetermined operating range, the lower the downstream air-fuel ratio. The operating range in which the influence of the imbalance in the response characteristics of the sensor can be eliminated is expanded.
[Brief description of the drawings]
FIG. 1 is a control system diagram of a first embodiment.
FIG. 2 is a flowchart for explaining calculation of an air-fuel ratio feedback correction coefficient α.
FIG. 3 is a flowchart for explaining calculation of proportional components PL and PR.
FIG. 4 is a diagram showing a map of a basic value PL0 for a proportional component.
FIG. 5 is a diagram showing a map of a basic value PR0 for a proportional component.
FIG. 6 is a region diagram of a predetermined operation region A.
FIG. 7 is a flowchart for calculating a correction value LP.
FIG. 8 is a flowchart for explaining the calculation of a correction value HLP.
FIG. 9 is a diagram illustrating a table of correction values HLP.
FIG. 10 is a waveform chart for explaining the operation of the first embodiment.
FIG. 11 is a waveform chart for explaining calculation of a fuel injection pulse width Ti.
FIG. 12 is a waveform chart for explaining a difference in a method for obtaining a response difference according to three embodiments.
FIG. 13 is a flowchart for explaining the calculation of proportional components PL and PR according to the fourth embodiment.
FIG. 14 is a diagram showing a map of a learning value LPA.
FIG. 15 is a flowchart illustrating a calculation of a correction value LP according to the fourth embodiment.
FIG. 16 is a flowchart illustrating a calculation of a correction value HD according to a fourth embodiment.
FIG. 17 is a diagram illustrating a table of correction values HD.
FIG. 18 is a waveform chart for explaining the operation of the fourth embodiment.
FIG. 19 shows a conventional O ₂ It is each waveform diagram of sensor output OSR1, OSR2, air-fuel ratio feedback correction coefficient (alpha), correction value PL, PR.
FIG. 20: Downstream side O ₂ It is a response waveform diagram of a sensor output.
FIG. 21 is a diagram corresponding to the claims of the first invention.
FIG. 22 is a diagram corresponding to the claims of the second invention.
FIG. 23 is a diagram corresponding to a claim of the fourth invention.
FIG. 24 is a diagram corresponding to the claim of the sixth invention.
[Explanation of symbols]
2 Control unit
3 Upstream O ₂ Sensor (upstream air-fuel ratio sensor)
4 Crank angle sensor
6 Air flow meter
7 Fuel injection valve
9 Exhaust passage
10 Three-way catalyst
13 Downstream O ₂ Sensor (downstream air-fuel ratio sensor)

Claims (14)

Air-fuel ratio sensors upstream and downstream of the catalyst,
Means for calculating a basic control constant of the air-fuel ratio feedback control;
Correction control condition of the control constant, which includes one of the conditions that an operating condition is in a predetermined correction control region that is narrower than the control region that performs the air-fuel ratio feedback control in the control region that performs the air-fuel ratio feedback control. Means for determining whether or not is established,
Means for calculating a correction value based on the output of the downstream air-fuel ratio sensor when the correction control condition is satisfied based on the determination result;
Means for correcting the basic control constant with the corrected value to calculate a first control constant;
Means for storing a learning value; means for updating a learning value stored in the storage means based on the correction value when the correction control condition is satisfied;
Means for calculating a difference between a response from rich to lean and a response from lean to rich from the output of the downstream air-fuel ratio sensor when the correction control condition is satisfied;
Means for calculating a correction value for the learning value according to the response difference;
Means for storing the correction value;
A value obtained by correcting the learning value stored in the storage means with the correction value stored in the storage means is set as a correction value, and the basic control constant is corrected with the set correction value to obtain a second correction value. Means for calculating a control constant;
When the correction control condition is satisfied, the first control constant is used. When the correction control condition is not satisfied, the air-fuel ratio feedback control based on the output of the upstream air-fuel ratio sensor is performed using the second control constant. Means for controlling the air-fuel ratio of an engine. Air-fuel ratio sensors upstream and downstream of the catalyst,
Means for calculating a basic control constant of the air-fuel ratio feedback control;
Means for delaying the output of the downstream air-fuel ratio sensor, limiting the delayed value between a positive upper limit and a negative lower limit,
Means for calculating a correction value based on the delay and the limited signal when the air-fuel ratio feedback control condition is satisfied;
Means for correcting the basic control constant with the corrected value to calculate a control constant,
Means for performing feedback control of the air-fuel ratio based on the output of the upstream air-fuel ratio sensor using the control constant,
Means for calculating a difference between a response from rich to lean and a response from lean to rich from the delayed, limited signal when the air-fuel ratio feedback control condition is satisfied,
Means for calculating a correction value for each basic value of the upper limit value and the lower limit value according to the response difference;
Means for calculating the upper limit value and the lower limit value by correcting each of the basic values of the upper limit value and the lower limit value with the correction value without changing the width between the two basic values. Air-fuel ratio control device. When the learning condition is satisfied, the learning value is updated based on the correction value and stored, and this learning value is used as an initial value of the correction value immediately after the air-fuel ratio feedback control condition is satisfied. The air-fuel ratio control device for an engine according to claim 2, wherein: Air-fuel ratio sensors upstream and downstream of the catalyst,
Means for calculating a basic control constant of the air-fuel ratio feedback control;
Means for calculating a basic control constant of correction control, which is control for correcting a control constant used for air-fuel ratio feedback control based on the output of the downstream air-fuel ratio sensor ,
Means for calculating the difference between the response from rich to lean and the response from lean to rich from the output of the downstream air-fuel ratio sensor when the air-fuel ratio feedback control condition is satisfied,
Means for calculating a correction value for a basic control constant of the correction control according to the response difference;
Means for correcting the basic control constant of the correction control with the correction value and calculating a control constant of the correction control ;
Means for calculating the Fix positive using the control constant of the correction control at the time of establishment of the air-fuel ratio feedback control condition,
Means for calculating the control constant of the air-fuel ratio feedback control by correcting the basic control constant with the corrected value;
An air-fuel ratio control device for an engine, comprising: means for performing feedback control of the air-fuel ratio based on the output of the upstream air-fuel ratio sensor using the control constant of the air-fuel ratio feedback control . When the learning condition is satisfied, the learning value is updated based on the correction value and stored, and this learning value is used as an initial value of the correction value immediately after the air-fuel ratio feedback control condition is satisfied. The air-fuel ratio control device for an engine according to claim 4, characterized in that: Air-fuel ratio sensors upstream and downstream of the catalyst,
Means for calculating a basic control constant of the air-fuel ratio feedback control;
When the modified control condition is satisfied, which includes one of the conditions that the operating condition is in a predetermined modified control region which is in the control region for performing the air-fuel ratio feedback control and is smaller than the control region for performing the air-fuel ratio feedback control. Means for determining whether or not
Means for calculating a basic control constant of correction control, which is control for correcting a control constant used for air-fuel ratio feedback control based on the output of the downstream air-fuel ratio sensor ,
Means for calculating a difference between a response from rich to lean and a response from lean to rich from the output of the downstream air-fuel ratio sensor when the correction control condition is satisfied from the determination result ;
Means for calculating a correction value for a basic control constant of the correction control according to the response difference;
Means for correcting the basic control constant of the correction control with the correction value and calculating a control constant of the correction control ;
Means for calculating the Fix positive using the control constant of the correction control at the time of establishment of the corrected control conditions,
Means for correcting a basic control constant of the air-fuel ratio feedback control with the corrected value to calculate a first control constant;
Means for storing a learning value;
Means for updating a learning value stored in the storage means based on the correction value when the correction control condition is satisfied;
Means for setting a learning value stored in the storage means as a correction value, calculating a second control constant by correcting a basic control constant of the air-fuel ratio feedback control with the set correction value,
When the correction control condition is satisfied, the first control constant is used. When the correction control condition is not satisfied, the air-fuel ratio feedback control based on the output of the upstream air-fuel ratio sensor is performed using the second control constant. Means for controlling the air-fuel ratio of an engine. When the learning condition is satisfied, the learning value is updated based on the correction value and stored, and the learning value is used as an initial value of the correction value immediately after the correction control condition is satisfied. The air-fuel ratio control device for an engine according to claim 6. 8. The engine according to claim 1, wherein a difference between a period in which the downstream air-fuel ratio sensor output is lean and a period in which the downstream air-fuel ratio sensor is rich is defined as the response difference. 9. Fuel ratio control device. A slice level for comparison with the downstream air-fuel ratio sensor output has a lower limit and an upper limit, and a period required for the downstream air-fuel ratio sensor output to pass from the upper limit to the lower limit and the downstream. 8. The response difference according to claim 1, wherein a difference from a period required for an output of the side air-fuel ratio sensor to pass from the lower limit to the upper limit is set as the response difference. Engine air-fuel ratio control device. The difference between the period from when the downstream air-fuel ratio sensor output becomes lean to when it reaches the lower limit peak and the period when the downstream air-fuel ratio sensor output becomes rich until it reaches the upper limit peak is defined as the response difference. The air-fuel ratio control device for an engine according to any one of claims 1, 4 to 7, wherein: 4. The air-fuel ratio control device for an engine according to claim 2, wherein a difference between a period in which the delayed and limited signal is lean and a period in which the signal is rich is set as the response difference. 5. The delay, the slice level for comparing with the signal subjected to the limit processing has a lower limit and an upper limit, and the delay, the period required for the signal subjected to the limit processing to pass from the upper limit to the lower limit. 4. The air-fuel ratio of the engine according to claim 2, wherein a difference from a period required for the delayed and limited signal to pass from the lower limit to the upper limit is set as the response difference. 5. Control device. The difference between the period from when the delayed, limited signal becomes lean to when it reaches the lower limit peak and the time when the delay and the limited signal becomes rich until it reaches the upper limit peak is the response difference. The air-fuel ratio control device for an engine according to claim 2 or 3, wherein: The air-fuel ratio control device for an engine according to any one of claims 1, 6, and 7, wherein the correction control condition is a predetermined operation range.

Images