JPH09112309A - Engine controller - Google Patents

Abstract

PROBLEM TO BE SOLVED: To conduct canister purging sufficiently without generating disorder in the feedback control of air-fuel ratio at the time of starting canister purging. SOLUTION: In an engine CE, an average feedback correction value is calculated by a control unit CU, and a trap amount is estimated based on the average feedback correction value. Purging air amount is calculated based on differential pressure between the front and rear of purging control valve and drive duty ratio, and vapor fuel discharge amount from a canister 25 is calculated based on estimated trap amount and purging air amount. Vapor fuel inflow amount into a combustion chamber 4 is calculated based on the vapor fuel discharge amount, and the vapor fuel inflow amount is subtracted from required fuel injection amount to set actual fuel injection amount. At the time of starting canister purging, canister purging amount is increased gradually toward a target value. It is thus possible to prevent the feedback control of air-fuel ratio from being disordered.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention feedback-controls a fuel injection amount so that an air-fuel ratio becomes a target air-fuel ratio in a predetermined operation region, while purging evaporated fuel trapped in a canister into an intake system. The present invention relates to the engine control device.

[0002]

2. Description of the Related Art Generally, in a fuel injection engine for an automobile, basically, a basic fuel injection amount corresponding to an intake air amount detected by an air flow sensor is set so that the air fuel ratio becomes a target air fuel ratio. The fuel is injected from the fuel injection valve into the intake system or the combustion chamber. However, the accuracy of the flow rate control of the fuel injection valve is limited, and a part of the fuel injected from the fuel injection valve does not enter the combustion chamber immediately because it adheres to the wall of the intake passage. Further, the injection characteristics of the fuel injection valve may change over time. Therefore, it is difficult to exactly match the air-fuel ratio with the target air-fuel ratio by simply injecting the fuel with the basic fuel injection amount corresponding to the intake air amount.

Therefore, normally, in a fuel injection engine, in a predetermined operating region (feedback region), the linear O ₂ sensor (or λO ₂ sensor) detects the O ₂ concentration in the exhaust gas to detect the O ₂ concentration. The air-fuel ratio is obtained from the deviation of the air-fuel ratio from the target air-fuel ratio (air-fuel ratio deviation), and a feedback correction value that acts in a direction to eliminate the air-fuel ratio deviation is set, and the basic fuel injection amount is set by the feedback correction value. The air-fuel ratio feedback control is performed such that the air-fuel ratio is corrected to follow the target air-fuel ratio. Here, when the feedback correction value is fixed to the neutral value, the feedback control of the air-fuel ratio is not performed and the open loop control is performed. The linear O ₂ sensor can detect the O ₂ concentration linearly,
The O ₂ sensor basically only detects whether the excess air ratio λ is greater than 1.

On the other hand, when the air in the fuel tank of a vehicle is directly discharged into the atmosphere, it causes air pollution and causes a loss of fuel. Therefore, the vehicle usually captures the evaporated fuel contained in the air discharged from the fuel tank. A canister for collecting (adsorbing) and an evaporated fuel purging means for appropriately purging the evaporated fuel collected by the canister into the intake system are provided.
Such evaporative fuel purging means is usually provided with a purge passage connecting the canister and the intake system, and a purge control valve for opening and closing the purge passage. When the purge control valve is opened, the evaporation inside the canister is evaporated. Fuel is purged into the intake system.

Therefore, when the intake system is purged with the evaporated fuel, the evaporated fuel in the canister is supplied to the intake system. Therefore, when the canister purge is performed during the open loop control, the air-fuel ratio is changed from the target air-fuel ratio. It will shift. Therefore, in an engine in which feedback control of the air-fuel ratio is performed in the feedback region, normally, the evaporated fuel is purged into the intake system only in the feedback region.

When the fuel vapor is purged into the intake system during the air-fuel ratio feedback control, it is possible to treat the fuel vapor supplied to the intake system by the purge as a disturbance. In this case, for example, if the evaporated fuel is purged when the air-fuel ratio matches the target air-fuel ratio, the air-fuel ratio is biased to the rich side with this, so the feedback correction value changes to the lean side, The fuel injection amount is reduced by the feedback correction value and the air-fuel ratio is returned to the target air-fuel ratio.

However, if the evaporated fuel is simply treated as a disturbance in this way, the disturbance becomes large and the feedback control is likely to be disturbed. In particular, it takes a considerable amount of time for the air-fuel ratio to return to the target air-fuel ratio during a transition such as immediately after the start of purging, and the feedback control is likely to be disturbed. Therefore, when purging evaporative fuel to the intake system during feedback control of the air-fuel ratio, the amount of evaporative fuel that is released (sucked) into the combustion chamber is obtained by some method, and the amount of evaporative fuel is calculated from the required fuel injection amount. Is preferably subtracted from to set the actual fuel injection amount, and the purge is excluded from the disturbance. For example, the required fuel injection amount is obtained based on the basic injection amount corresponding to the intake air amount and the feedback correction value, and the amount of evaporated fuel released to the intake system (combustion chamber) from the required fuel injection amount It is preferable to set the actual fuel injection amount by, for example, subtracting "the amount of evaporated fuel into the intake system".

As described above, when the actual fuel injection amount is set by subtracting the amount of evaporated fuel discharged to the intake system (the amount of evaporated fuel to the intake system), the amount of evaporated fuel to this intake system is accurately obtained. Although it is necessary, a practical sensor that can directly detect such evaporated fuel amount with high accuracy is not found at this time. Therefore, an engine is proposed that indirectly estimates the amount of evaporated fuel to the intake system and subtracts the estimated amount of evaporated fuel from the required fuel injection amount to set the actual fuel injection amount (for example, JP-A-2-245441
Reference). In the engine disclosed in Japanese Patent Laid-Open No. 2-245441, the amount of evaporated fuel to the intake system is estimated based on the difference between the feedback correction value and its neutral value. More specifically, the estimated value of the amount of evaporated fuel released to the intake system is divided by the engine speed to obtain 1
Calculate the emission amount per revolution and set the basic fuel injection amount to 1
The amount of emission per revolution is reduced and corrected.

[0009]

However, in the conventional engine in which the evaporated fuel is purged into the intake system during the feedback control of the air-fuel ratio, when the purge is treated as a disturbance, the low value in the idle region or the like is reduced. When the purge is started in the intake air amount range, the fuel injection amount is small originally, so the influence of the disturbance becomes so great that it cannot respond to the disturbance and a large disturbance occurs in the feedback control.
There arises a problem that fuel efficiency is deteriorated. On the other hand, it is conceivable to reduce the purge flow rate (flow rate of purge air containing evaporated fuel) in the low intake amount region, but in this way, evaporated fuel is accumulated in the canister in the low intake amount region. Therefore, it is necessary to increase the capacity of the canister, which causes a problem of unnecessarily increasing the size of the engine. If the capacity of the canister is not sufficient, the canister will be saturated and the evaporated fuel will be discharged into the atmosphere.

Further, in an engine in which the amount of evaporated fuel to the intake system is indirectly estimated, and the estimated amount of evaporated fuel is subtracted from the required fuel injection amount to set the actual fuel injection amount, the idle region, etc. When the purge is started in the low intake air amount range, it is not possible to accurately estimate the evaporated fuel amount to be subtracted when the estimation of the evaporated fuel amount is not completed, so feedback control is disturbed and fuel consumption performance is reduced. There is a problem such as a decrease.

The present invention has been made to solve the above-mentioned conventional problems. In the feedback region, the fuel injection amount is feedback-controlled so that the air-fuel ratio becomes the target air-fuel ratio, while the canister captures the air-fuel ratio. For an engine configured to purge the collected evaporated fuel into the intake system, without causing any disturbance in the feedback control of the air-fuel ratio,
It is an object of the present invention to provide a means capable of promptly purging the evaporated fuel collected in the canister into the intake system, reducing the size of the canister, and improving the fuel consumption performance of the engine.

[0012]

As shown in FIG. 1, a first aspect of the present invention, which has been made to achieve the above-mentioned object, is an air-fuel ratio detecting means A for detecting an air-fuel ratio and a predetermined operating range. A fuel injection amount control means B for feedback-controlling the fuel injection amount so that the air-fuel ratio detected by the air-fuel ratio detection means A becomes the target air-fuel ratio; and an evaporated fuel trap for collecting the evaporated fuel from the fuel tank. Collecting means C and, in the predetermined operation region, evaporated fuel purging means D for purging the evaporated fuel collected by the evaporated fuel collecting means C into the intake system.
And a purge flow rate control means E for controlling the purge flow rate of the evaporated fuel to the intake system by the vaporized fuel purge means D, the purge flow rate control means E is provided with the predetermined operation. When the evaporated fuel is purged in the region, the purge flow rate is gradually changed to the target purge flow rate, and the rate of change is set to be smaller in the low intake air amount range than in the high intake air amount range. It is characterized by that.

In the engine control device according to the first aspect, when the purge of the evaporated fuel is executed in the low intake air amount region where the fuel injection amount is small, the purge flow rate is gradually changed with a relatively small change rate. Can be changed,
At the initial stage of purging, the change in the amount of evaporated fuel flowing into the intake system does not increase so much, which prevents the feedback control of the air-fuel ratio from being disturbed. Then, after a while after the start of purging, the purge flow rate reaches the target purge flow rate, so that the evaporated fuel collected (adsorbed) in the canister is rapidly discharged to the intake system. Further, when the purge of the evaporated fuel is started in the high intake air amount region where the fuel injection amount is large, the purge flow rate can be changed at a relatively large rate of change, so that the evaporated fuel collected in the canister can be quickly discharged. Is discharged to the intake system.

According to a second aspect of the present invention, in the engine control device according to the first aspect, when the purge flow rate control means E starts the purge of the evaporated fuel in the predetermined operating region, The purge flow rate is gradually increased to the target purge flow rate, and the rate of increase is set to be smaller in the low intake air amount region than in the high intake air amount region.

In the engine control apparatus according to the second aspect, when the purge of the evaporated fuel is started in the low intake air amount region where the fuel injection amount is small, the purge flow rate is gradually increased with a relatively small increase rate. Since it is increased, the change in the amount of evaporated fuel flowing into the intake system does not increase so much at the initial stage of purging, and it is possible to prevent the feedback control of the air-fuel ratio from being disturbed. Then, after a while after the start of purging, the purge flow rate reaches the target purge flow rate, so that the evaporated fuel collected (adsorbed) in the canister is rapidly discharged to the intake system. Also, when the purge of the evaporated fuel is started in the high intake air amount region where the fuel injection amount is large, the purge flow rate is increased at a relatively large increase rate, so the evaporated fuel trapped in the canister is quickly It is discharged to the intake system.

A third aspect of the present invention is based on the air-fuel ratio detecting means A for detecting the air-fuel ratio and the deviation of the air-fuel ratio detected by the air-fuel ratio detecting means A from the target air-fuel ratio in a predetermined operating range. Feedback correction value setting means F for setting a feedback correction value, evaporative fuel collecting means C for collecting evaporative fuel from the fuel tank, and evaporative fuel collecting means C for collecting evaporative fuel in the predetermined operation region. Evaporative Fuel Purging Means D for Purging the Generated Evaporative Fuel into the Intake System
Based on the feedback correction value set by the purge flow rate control means E for controlling the purge flow rate of the evaporated fuel to the intake system by the evaporated fuel purge means D according to the operating state, and the feedback correction value setting means F Evaporative fuel amount calculating means G for calculating the amount of evaporated fuel released into the combustion chamber, basic fuel injection amount obtained corresponding to the intake air amount, and feedback correction value set by the feedback correction value setting means F The required fuel injection amount is obtained based on the above, and the actual fuel injection amount is set by subtracting the evaporated fuel amount calculated by the evaporated fuel amount calculation means G from the required fuel injection amount, thereby setting the air-fuel ratio detection means. A
In the engine control device provided with the fuel injection amount control means B for feedback controlling the fuel injection amount so that the air-fuel ratio detected by the target air-fuel ratio becomes the target air-fuel ratio, From the feedback correction value set by the correction value setting means F, an estimated value of the amount related to the evaporated fuel amount is calculated, and the purge flow rate control means E vaporizes in an idle state in the predetermined operating region. When the fuel is purged, the purge flow rate is gradually changed to the target purge flow rate, and the rate of change is calculated when the evaporated fuel amount calculation means G has not completed the calculation of the estimated value. It is characterized in that it is designed to be set smaller than when it is completed.

In the engine control device according to the third aspect, the air-fuel ratio feedback control is performed after the actual fuel injection amount is set by subtracting the evaporated fuel amount from the required fuel injection amount. Therefore, the stability or responsiveness of the feedback control is improved. Further, when the purge of the evaporated fuel is executed in the idle region where the fuel injection amount is small, when the calculation of the estimated value of the amount of the evaporated fuel is not completed, the purge flow rate is gradually changed with a relatively small change rate. Therefore, even if the amount related to the evaporated fuel amount (and thus the evaporated fuel amount) is not accurately grasped, it is possible to prevent the feedback control of the air-fuel ratio from being disturbed at the initial stage of purging.

According to a fourth aspect of the present invention, in the engine control device according to the third aspect, when the purge flow rate control means E starts the purge of the evaporated fuel in the predetermined operating region, The purge flow rate is gradually increased to the target purge flow rate, and the rate of increase is set to be smaller in the low intake air amount region than in the high intake air amount region.

In the engine control device according to the fourth aspect, basically, similarly to the engine control device according to the second aspect, the stability or responsiveness of the feedback control of the air-fuel ratio. It is possible to increase the purge flow rate while ensuring the above. Further, since the evaporated fuel amount is subtracted from the required fuel injection amount and the actual fuel injection amount is set before the feedback control of the air-fuel ratio is performed, the stability or responsiveness of the feedback control is further improved. . Further, when the purge of the evaporated fuel is started in the idle region where the fuel injection amount is small, the purge flow rate is gradually increased with a relatively small increase rate when the calculation of the estimated amount of the evaporated fuel amount is not completed. Therefore, even if the amount related to the evaporated fuel amount (and thus the evaporated fuel amount) is not accurately grasped, it is possible to prevent the feedback control of the air-fuel ratio from being disturbed at the initial stage of purging.

A fifth aspect of the present invention is the above third or fourth aspect.
In the engine control device according to this aspect, the amount related to the amount of evaporated fuel is the amount of collected evaporated fuel.

In the engine control device according to the fifth aspect, the amount relating to the amount of evaporated fuel is grasped by the amount of collected evaporated fuel, and is the same as in the case of the engine control device according to the third or fourth aspect. The action of will occur.

A sixth aspect of the present invention is the fourth or fifth aspect.
In the engine control device according to this aspect, the purge flow rate control means E gradually increases the purge flow rate to the target purge flow rate even when starting the purge of the evaporated fuel in the off-idle state in the predetermined operation region. In addition, the purge amount increase rate when the evaporative fuel amount calculation means G completes the calculation of the estimated value at the time of idling, It is characterized in that it is set to be smaller than the purge amount increase rate when the calculation of the estimated value is not completed.

In the engine control device according to the sixth aspect, basically the same operation as in the engine control device according to the fourth or fifth aspect occurs.
Further, when the purge of the evaporated fuel is started at the time of idle when the fuel injection amount is small, the purge flow rate is related to the evaporated fuel amount when the calculation of the amount related to the evaporated fuel amount or the estimated value of the collected amount of evaporated fuel is not completed. Amount or evaporated fuel trapped amount (and thus evaporated fuel amount) because the estimated amount of evaporated fuel or evaporated fuel is gradually increased at a smaller rate than when the engine is off-idle. Even if is not accurately grasped, it is possible to prevent the feedback control of the air-fuel ratio from being disturbed at the initial stage of purging.

According to a seventh aspect of the present invention, in the engine control device according to any one of the third to sixth aspects, the evaporated fuel amount calculating means G is controlled by the feedback correction value setting means F. The estimated value of the evaporated fuel amount is calculated based on the average value of the feedback correction values that are set, and the estimated value is updated by learning.

In the engine control device according to the seventh aspect, basically the same operation as in the engine control device according to any one of the third to sixth aspects occurs. Furthermore, since the amount related to the amount of evaporated fuel or the amount of collected evaporated fuel is estimated based on the average value of the feedback correction values, the calculation of the estimated value becomes easy. Further, since the estimated value of the amount of evaporated fuel or the estimated amount of evaporated fuel is updated by learning, the accuracy of the estimated value can be improved.

An eighth aspect of the present invention is the engine control apparatus according to any one of the first to seventh aspects, wherein the evaporated fuel purge means D is applied from the purge flow rate control means E. A purge valve whose valve element is turned on / off according to the duty ratio, and the purge flow rate control means E keeps the on time of the purge valve constant when the duty ratio to be applied to the purge valve is less than a predetermined value. In addition, the operating cycle of the purge valve is expanded and contracted according to the duty ratio.

In the engine control apparatus according to the eighth aspect, basically the same operation as in the engine control apparatus according to any one of the first to seventh aspects occurs. Further, when the duty ratio is less than a predetermined value, the on time is not increased or decreased according to the duty ratio, but the on time is kept constant and the operation cycle is expanded or contracted according to the duty ratio to correspond to the duty ratio. Since the valve opening degree (valve opening degree) is set,
The opening degree of the purge valve can be precisely adjusted. Moreover, since the operation cycle becomes long, the number of times the purge valve is opened and closed is reduced.

[0028]

DESCRIPTION OF THE PREFERRED EMBODIMENTS Embodiments of the present invention will be specifically described below. FIG. 2 is a system configuration diagram of an engine including a control device according to the present invention. As shown in FIG. 2, in each cylinder 1 (only one cylinder is shown) of the fuel injection type four-cylinder gasoline engine CE, when the intake valve 2 is opened, the air-fuel mixture is introduced from the intake port 3 into the combustion chamber 4. Is sucked in, the mixture is compressed by the piston 5, and then ignited and burned by a spark plug (not shown), and the exhaust valve 6
Combustion gas (exhaust gas) is discharged to the exhaust passage 8 through the exhaust port 7 when is opened. The exhaust passage 8 has a linear O ₂ for detecting the O ₂ concentration in the exhaust gas.
The sensor 9 is provided in the vicinity. Then, the O ₂ concentration detected by the linear O ₂ sensor 9 is input to the control unit CU, and the control unit CU calculates the air-fuel ratio of the air-fuel mixture based on this O ₂ concentration. Where it is detected by the linear O ₂ sensor 9 O ₂
Since the concentration and the air-fuel ratio calculated based on the O ₂ concentration have a unique correspondence relationship, the air-fuel ratio will be referred to as “the air-fuel ratio detected by the linear O ₂ sensor 9” or “the actual air-fuel ratio” for the sake of convenience. Fuel ratio ". Here, a λO ₂ sensor may be used instead of the linear O ₂ sensor 9. The linear O ₂ sensor 9 can detect the O ₂ concentration and thus the air-fuel ratio even in the region where the excess air ratio λ is larger than 1, but the λ O ₂ sensor basically has the excess air ratio λ larger than 1. It just detects whether or not.

An intake system 10 is provided to supply air for fuel combustion to each cylinder 1 (combustion chamber 4) of the engine CE,
The intake system 10 is provided with a common intake passage 11 whose upstream end is open to the atmosphere. Then, the common intake passage 11
Is provided with a throttle valve 12 which is opened / closed in conjunction with an accelerator pedal (not shown), and the downstream end of the common intake passage 11 is connected to a surge tank 13 for stabilizing the flow of intake air. In addition, the surge tank 13 has each cylinder 1
Independent intake passages 14 (only one is shown) for individually supplying air are connected to the respective intake ports, and the downstream ends of these independent intake passages 14 are connected to the intake ports 3 of the corresponding cylinders 1.

A fuel injection valve 15 for injecting fuel into the intake port 3 or the combustion chamber 4 is provided in each independent intake passage 14 near the intake port 3 so that the injection port faces the downstream side. . Here, the fuel injection amount (injection pulse width) and the injection timing of the fuel injection valve 15 are controlled by the control unit CU as described later.

Assist air supply means 16 (hereinafter referred to as AMI 16) for supplying assist air to each fuel injection valve 15 in order to promote vaporization and atomization of fuel is provided. Although not shown in detail, the AMI 16 is provided with an assist air introducing passage 17 whose upstream end communicates with the common intake passage 11 upstream of the throttle valve 12, and the assist air introducing passage 17 has a control unit CU. A solenoid type assist air control valve 18 that is opened and closed by the valve is provided. A bypass assist air passage 19 that bypasses the assist air control valve 18 and allows the assist air to pass is provided, and an orifice 20 that causes a predetermined pressure loss to adjust the flow rate is provided in the bypass assist air passage 19. ing.

The downstream end of the assist air introduction passage 17 is connected to the mixing chamber 21, and the mixing chamber 21 is further connected to an assist air supply passage 22. Then, the assist air supply passage 22 is branched on the downstream side into four branch assist air supply passages 23, and each of the branch assist air supply passages 23 is connected at its downstream end to the corresponding fuel injection valve 15 of the cylinder 1, Assist air is individually supplied to the corresponding fuel injection valves 15.

The engine CE has a fuel tank (not shown).
Evaporative fuel collecting means for collecting evaporated fuel (gasoline vapor) contained in the air discharged from the exhaust gas, and evaporated fuel purge for appropriately purging the evaporated fuel collected by the evaporated fuel collecting means into the intake system 10. Evaporative fuel recovery means 2 including means
4 is provided, but this evaporated fuel recovery means 24 will be described below.
Will be described. The evaporative fuel recovery means 24 has an adsorbent (for example, an adsorbent) capable of collecting (adsorbing) the evaporative fuel.
A canister 25 filled with activated carbon is provided. A relief passage 2 is provided in the canister 25, the tip of which communicates with the upper space of the fuel tank to relieve air in the upper space of the fuel tank into the canister 25.
6, an atmosphere opening passage 27 whose tip is open to the atmosphere, and a purge passage 2 whose tip is connected to the mixing chamber 21.
And 8 are connected. The tip of the atmosphere opening passage 27 may be connected to the common intake passage 11 upstream of the throttle valve 12. Instead of filling the canister 25 with an adsorbent, a phenomenon other than adsorption (eg, absorption, reaction, etc.) may be used to fill with a material that collects the evaporated fuel (however, purging with air is not necessary). Is possible). The canister 25 corresponds to the "evaporated fuel collecting means" described in the claims.

The purge passage 28 is provided with a duty solenoid type purge valve 29 capable of arbitrarily opening and closing the passage 28. The purge valve 29 is opened by the control unit CU (valve opening degree).
The duty ratio is controlled. That is, the purge valve 29 is controlled to open and close according to the drive duty ratio applied from the control unit CU,
In one drive cycle, it is fully opened for a time corresponding to the drive duty ratio, and fully closed for the remaining time. For example, when one drive cycle is 100 ms (milliseconds) (that is, the frequency is 10 Hz), when the applied drive duty ratio is 0, the drive duty ratio is 100% when the drive duty ratio is 100%. Sometimes it is fully opened (open degree 1 or 100%), and when the drive duty ratio is 30%, it is fully opened for 30 ms first and the remaining 70 ms is fully closed (open degree 0.3 or 30%). Hereinafter, the time during which the purge valve 29 is fully opened within one drive cycle is referred to as an on time, and the time during which the purge valve 29 is fully closed is referred to as an off time. The assembly including the purge passage 28 and the purge valve 29 corresponds to "evaporated fuel purge means" described in the claims.

In this way, the purge valve 29 is fully opened for the time corresponding to the drive duty ratio within one drive cycle, so if the drive cycle is kept constant, the on-time will increase or decrease in proportion to the drive duty ratio. Become. However, if the on-time is too short, the on-time cannot be precisely controlled due to the inertia of the valve body or the like. Therefore, in this embodiment, the on-time is not set to 10 ms or less. That is, when the drive duty ratio is 10% or more, the drive cycle is set to a constant value of 100 ms (frequency of 10 Hz), and the ON time is increased or decreased according to the drive duty ratio (proportionally).
On the other hand, if the drive duty ratio is less than 10%, the on-time is set to a constant value of 10 ms, and the on-time (10 ms) ratio in the drive cycle, that is, the valve opening degree is driven so as to correspond to the drive duty ratio. The cycle is expanded and contracted.

For example, as shown in FIG. 15, if the drive duty ratio is 30%, the drive cycle is set to 100 ms and the ON time is set to 30 ms. In this case, the valve opening degree is 0.3 (30 ms / 100 ms = 0.3). If the drive duty ratio is 3%, the on-time is 10
ms, and the drive cycle is 333 ms (10 ms /
0.03 = 333 ms). In this case, the degree of valve opening is 0.03 (10ms / 333ms = 0.03)

As described above, when the drive duty ratio is less than 10%, the on-time is not increased or decreased according to the drive duty ratio, but the on-time is set to a constant value of 10 ms, and the operation cycle is changed according to the drive duty ratio. Since the valve opening degree corresponding to the drive duty ratio is expanded and contracted, the valve opening degree of the purge valve 29 can be precisely controlled, and the purge flow control accuracy can be improved. Further, since the operation cycle becomes long, the number of times the purge valve 29 is opened and closed is reduced, the life of the purge valve 29 is extended, and the durability of the engine CE is improved.

In the fuel vapor recovery means 24, when the purge valve 29 is normally fully closed (driving duty ratio 0), the air in the fuel tank is relieved into the canister 25 through the relief passage 26 and then the atmosphere. Although it is discharged into the atmosphere through the open passage 27,
The evaporated fuel contained in this air is collected by the adsorbent when passing through the adsorbent layer in the canister 25, and is not discharged into the atmosphere.

On the other hand, when the purge valve 29 is opened, the air in the atmosphere is first sucked into the canister 25 through the atmosphere opening passage 27 and passes through the adsorbent layer by the negative pressure of the intake system 10, and then the purge passage is formed. 28 and A
The intake system 10 and thus the combustion chamber 4 through the MI 16 (mixing chamber 21 to branch assist air supply passage 23).
To be purged. Here, the opening degree of the purge valve 29
Of course, the flow rate of purged air (hereinafter, referred to as “purge air amount (purge flow rate)”) changes according to the drive duty ratio. Then, at this time, a part of the evaporated fuel collected in the adsorbent in the canister 25 is separated from the adsorbent, and the purged air (hereinafter, referred to as purge air) together with the intake system 10
As a result, the combustion chamber 4 is purged. Note that, hereinafter, the flow rate of the evaporated fuel purged into the intake system 10 and thus the combustion chamber 4 in this manner is referred to as "evaporated fuel purge amount (evaporated fuel amount)".

However, since the purge air or the vaporized fuel transportation path from the canister 25 to the combustion chamber 4 has a considerable capacity, the vaporized fuel discharged from the canister 25 to the purge passage 28 actually enters the combustion chamber 4. By the time it reaches, there is a transportation delay corresponding to the volume and shape of the transportation route (transport characteristics). Therefore, at a certain time, the flow rate of the evaporated fuel released from the canister 25 to the purge passage 28 (hereinafter, this is referred to as the evaporated fuel release amount).
And the flow rate of evaporated fuel that actually flows into the combustion chamber 4 (hereinafter,
This is referred to as the evaporative fuel inflow amount), but usually does not match, except in a special case where it is in a steady state. Therefore, hereinafter, the evaporated fuel purge amount will be described by distinguishing between the evaporated fuel discharge amount and the evaporated fuel inflow amount. When the volume of the purge air or the vaporized fuel transportation route from the canister 25 to the combustion chamber 4 is very small, the transportation delay can be ignored, and therefore the vaporized fuel discharge amount and the vaporized fuel inflow amount are particularly distinguished. However, even if the concept of the evaporated fuel purge amount is used, no particular problem occurs.

By the way, in the engine CE shown in FIG.
The downstream end of the purge passage 28 is connected to the mixing chamber 21, and the vaporized fuel collected in the canister 25 is AM
Although the intake system 10 is purged via I16, in the case of an engine without an AMI, the downstream end of the purge passage 28 may be branched and connected to each independent intake passage 14. As shown in FIG. 3, in the case of the engine CE ′ not provided with the AMI, the downstream end of the purge passage 28 is connected to the surge tank 13, and the evaporated fuel collected in the canister 25 is directly introduced into the intake system. If it is purged to 10, no trouble will occur. In addition, in FIG. 3, members common to those in FIG. 2 are denoted by the same reference numerals.

The control unit CU includes a "fuel injection amount control means", a "purge flow rate control means", a "feedback correction value setting means" and a "evaporated fuel amount calculation means" described in the claims. An air-fuel ratio detected by a linear O ₂ sensor 9 (or a λO ₂ sensor), which is a comprehensive control device for an engine CE (including an engine CE ′, the same applies hereinafter) which is configured by a computer.
(Actual air-fuel ratio), throttle opening detected by throttle opening sensor 31, intake air amount detected by air flow sensor 32, engine speed detected by speed sensor 33, idle output from idle switch 34 Engine CE using signals as control information
Various controls (including evaporative fuel recovery means 24), estimation of the amount of evaporative fuel collected in the canister 25 (hereinafter referred to as trap amount), calculation of evaporative fuel release amount (calculation),
The inflow amount of evaporated fuel is calculated (calculated).

However, general control of the engine CE is well known, and since such general control is not the gist of the present invention, its explanation is omitted.
In the following, related to the gist of the present invention, a control method of air-fuel ratio control (fuel injection amount control), a control method of canister purge control, a trap amount estimation method, a vaporized fuel discharge amount calculation method (calculation method) And only the calculation method (calculation method) of the inflow amount of evaporated fuel will be described.

The basic functions of the control unit CU will be described below with reference to FIG. As shown in FIG. 4, the control unit CU is functionally
An engine control block SL that performs air-fuel ratio control (fuel injection amount control) and canister purge control, a trap amount estimation block SM that estimates a trap amount, and an evaporated fuel release amount and an evaporated fuel inflow amount are calculated based on the trap amount. Evaporated fuel purge amount calculation block SN. The engine control block SL basically performs feedback control or open loop control of the injection pulse width of the fuel injection valve 15, that is, the fuel injection amount according to the operating state so that the air-fuel ratio becomes the target air-fuel ratio (air-fuel ratio Control), the canister purge is performed according to the operating state in the operating region where the canister purge should be performed (canister purge control).

In this air-fuel ratio control, the engine CE
Based on the deviation of the actual air-fuel ratio from the target air-fuel ratio (hereinafter, referred to as the air-fuel ratio deviation), if the operating state of is within a predetermined feedback area (for example, the area excluding the high load area and the high rotation area). Feedback control is performed, and if it is not within the feedback region, open loop control that is not based on the air-fuel ratio deviation is performed. The control method of the feedback control of the air-fuel ratio here is approximately as follows. That is, the basic pulse width of the fuel injection valve 15, that is, the basic fuel injection amount is calculated according to the intake air amount and the engine speed (base calculation).

On the other hand, the feedback correction value cfb is calculated based on the air-fuel ratio deviation (for example, target air-fuel ratio-actual air-fuel ratio) (step S1). Here, the feedback correction value cfb is a neutral value, that is, a neutral value that does not correct the air-fuel ratio in any direction, and when cfb> 0, the air-fuel ratio (fuel injection amount) is corrected in the rich direction. , Cfb
When <0, the air-fuel ratio (fuel injection amount) is corrected in the lean direction.

Then, based on the basic pulse width and the feedback correction value cfb, for example, by adding cfb to the basic pulse width, the basic pulse width is corrected in the direction in which the air-fuel ratio deviation is reduced and the required pulse width, that is, The required fuel injection amount is calculated (step S2). For example, when the actual air-fuel ratio is leaner than the target air-fuel ratio, cfb> 0, and accordingly, the fuel injection amount is increased and corrected, the air-fuel ratio is corrected in the rich direction, and the air-fuel ratio deviation is reduced. On the contrary, when the actual air-fuel ratio is richer than the target air-fuel ratio, cfb <0, and accordingly, the fuel injection amount is reduced and the air-fuel ratio is corrected in the lean direction to reduce the air-fuel ratio deviation. Thus, the air-fuel ratio (fuel injection amount) is feedback-controlled according to the air-fuel ratio deviation so as to eliminate the air-fuel ratio deviation.

On the other hand, when open-loop control of the air-fuel ratio is performed, the feedback correction value cfb is fixed at 0. In this case, the basic pulse width becomes the required pulse width as it is without any correction according to the air-fuel ratio deviation, so that open loop control without feedback is performed.

Further, the actual pulse width of the fuel injection valve 15 is subtracted from the required pulse width, that is, the required fuel injection amount, by subtracting the pulse width (hereinafter referred to as the purge correction pulse width) corresponding to the evaporated fuel inflow amount described later. The injection pulse width (hereinafter referred to as the actual injection pulse width), that is, the actual fuel injection amount (actual fuel injection amount) is calculated. Then, with the actual injection pulse width, that is, the actual fuel injection amount, fuel is injected from the fuel injection valve 15 at a predetermined timing. Thus, the actual air-fuel ratio is maintained at the target air-fuel ratio.

The canister purge control is performed according to the operating state of the engine CE by a well-known ordinary method when the canister purge condition is satisfied, for example, when the water temperature is equal to or higher than a predetermined value (for example, 80 ° C.). That is,
A drive duty ratio according to the operating state of the engine CE is applied to the purge valve 29, and canister purge is performed.

The trap amount estimation block SM uses step S of the engine control block SL during canister purging.
The feedback correction value cfb calculated in 1 is averaged to calculate the average feedback correction value cfbave (step S3), and the trap amount is indirectly estimated based on the average feedback correction value cfbave (step S4). ). That is, the average feedback correction value cfb
The trap amount is grasped by using ave as an index for judging whether the currently grasped trap amount (trap amount estimated value) is larger or smaller than the true trap amount.

As will be described later, the control unit CU calculates the evaporated fuel inflow amount based on the trap amount estimated value using a predetermined arithmetic expression, and further subtracts the evaporated fuel inflow amount from the required fuel injection amount. Therefore, the actual fuel injection amount is set. Here, if the trap amount estimated value is accurate, that is, if it matches the true trap amount, the evaporative fuel inflow amount is accurately calculated. Therefore, the evaporative fuel supplied to the combustion chamber 4 by the canister purge is feedback-controlled. Feedback correction value
Does not affect cfb in particular. In this case, the feedback correction value cfb only slightly fluctuates around the neutral value (that is, 0) unless there is another large disturbance, so that the average feedback correction value cfbave is almost the neutral value 0.
Becomes In other words, the average feedback correction value cfbave
If is 0, it means that the trap amount estimated value matches the true trap amount.

However, if the trap amount estimated value is larger than the true trap amount, the evaporative fuel inflow amount calculated value becomes larger than the true value accordingly, and therefore the actual fuel injection amount becomes smaller than the proper value. The fuel actually supplied to the chamber 4 becomes smaller than the required fuel amount (required fuel injection amount), and the actual air-fuel ratio becomes lean. In this case, in order to correct this leaning, the feedback correction value cfb changes in the rich direction and becomes larger than 0, and accordingly, the average feedback correction value cfbave becomes larger than 0.
In other words, if cfbave> 0, the trap amount estimated value is larger than the true trap amount.

Since the feedback correction value cfb fluctuates as described above, cfb> 0 is not always satisfied even if the trap amount estimated value is larger than the true trap amount, and therefore cfb> 0 is also satisfied. The trap amount estimated value is not always larger than the true trap amount. Therefore, when the trap amount is estimated based on the feedback correction value cfb, the estimation accuracy is considered to be extremely low. In consideration of such circumstances, in this embodiment, the trap amount is estimated based on the average feedback correction value cfbave.

On the contrary, if the trap amount estimated value is smaller than the true trap amount, the evaporative fuel inflow calculation value becomes smaller than the true value, and the actual fuel injection amount becomes larger than the proper value. The amount of fuel supplied to the combustion chamber 4 becomes larger than the required fuel amount, and the actual air-fuel ratio becomes rich. In this case, in order to correct this enrichment, the feedback correction value cfb changes in the lean direction and becomes smaller than 0, and accordingly, the average feedback correction value cfbave becomes smaller than 0. In other words, if cfbave <0, it means that the trap amount estimated value is smaller than the true trap amount.

Therefore, after setting an appropriate initial value for the trap amount estimated value, the trap amount estimated value is reduced by a predetermined correction amount σ if cfbave> 0, and the trap amount estimated if cfbave <0. By repeating the operation of increasing the value by the correction amount σ, the trap amount estimated value eventually converges (reaches) to the true trap amount, and the trap amount is grasped. Thus, the trap amount is estimated based on the average feedback correction value cfbave.

Here, whether or not the trap amount estimated value substantially matches the true trap amount, that is, whether or not the trap amount estimation is almost completed is determined by the absolute value | cfbave | of the average feedback correction value cfbave | It is preferable to determine whether or not is less than or equal to a predetermined limit value ε. This is because if │cfbave│ is very small, it is considered that the estimated trap amount is almost the same as the true trap.

In this method of estimating the trap amount, the above-mentioned correlation (correlation) is established between the trap amount or the evaporated fuel purge amount and the feedback correction value cfb or the average feedback correction value cfbave. It is assumed that Therefore, the trap amount cannot be estimated with high accuracy under such a low correlation condition or a non-correlation condition. For this reason,
It is preferable to prohibit the estimation of the trap amount in a situation where the correlation is low or there is no correlation.
Here, as the situation where the correlation is low, as will be described later, for example, when the charging efficiency and the intake pressure are very high, the charging efficiency and the intake pressure are very low, and the like. In addition, as a situation where the above correlation does not exist, for example, when canister purge is stopped, when air-fuel ratio feedback control is stopped (during open loop control)
And the like. Of course, the estimation of the trap amount may be prohibited only when there are some overlapping situations where the correlation is low or there is no correlation.

Further, in this trap amount estimating method, if the evaporative fuel inflow amount is accurately grasped, that is, if the evaporative fuel supply by the canister purge does not affect the feedback correction value cfb, the feedback correction value cfb is fed back. It is premised that the correction value cfb fluctuates around the neutral value 0, and therefore the average feedback correction value cfbave becomes the neutral value 0. By the way, generally, the air-fuel ratio learning is performed such that the control output characteristic, that is, the injection characteristic of the fuel injection valve is automatically corrected by learning so that the feedback correction value cfb becomes a neutral value 0 on average. Although an engine is widely used, when the trap amount is estimated by an engine having such an air-fuel ratio learning function, it is preferable to estimate the trap amount after the learning of the air-fuel ratio is completed. However, if the air-fuel ratio learning is completed and the canister purge has no effect, the average feedback correction value cfbave is surely set to the neutral value 0.
Because it becomes.

When the period in which the estimation of the trap amount is prohibited is continued for some time in this way, the estimated trap amount value may deviate from the true trap amount. Cancels (reset) the judgment even if the judgment is completed
Is preferred.

When the correction amount σ is large, the time required for the trap amount estimated value to converge after the estimation is started, that is, the time required for estimating the trap amount can be shortened, but the accuracy of the trap amount estimated value is high. descend. On the other hand, when the correction amount σ is small, the time required for the trap amount estimated value to converge becomes long, but the accuracy of the trap amount estimated value can be improved. Therefore, it is preferable to set the correction amount σ to an appropriate value so that the demand for the time required for convergence and the demand for the accuracy of the trap amount estimated value are compatible with each other.

The correction amount σ does not have to be a constant value and may be changed during estimation of the trap amount. For example,
It may be changed according to the progress of the trap amount estimation, or may be set according to the value of the average feedback correction value cfbave. For example, when the trap amount estimation is started, the correction amount σ may be increased to accelerate the convergence, and after the trap amount estimated value has converged to some extent, the correction amount σ may be decreased to improve the accuracy of the trap amount estimated value. Also, if the correction amount σ is increased as the average feedback correction value cfbave is larger, the convergence can be accelerated when the trap amount estimated value is far from the true trap amount, while the trap amount estimated value is the true trap amount. When it is close to, the accuracy can be increased.

The evaporative fuel purge amount calculation block SN is basically the trap amount estimation block SM in step S4.
The evaporative fuel release amount is calculated based on the trap amount estimated value estimated in step 1, the evaporative fuel inflow amount is calculated based on the evaporative fuel release amount, and the pulse of the fuel injection valve 15 corresponding to the evaporative fuel inflow amount is calculated. The width, that is, the purge correction pulse width is calculated, and this purge correction pulse width is output to the engine control block SL. That is, feedforward control
By (prospect control), the influence of the canister purge on the air-fuel ratio control is compensated without causing a time lag and thus without causing a deviation of the air-fuel ratio. More specifically, first, the differential pressure across the purge valve 29 in the purge passage 28, that is, the pressure difference between the upstream side and the downstream side of the purge valve 29 (hereinafter referred to as the differential pressure across the purge valve) is calculated ( Step S
5), the purge valve opening (purge SOL opening) is calculated from the drive duty ratio applied to the purge valve 29 (step S6), and then the purge air amount (canister) is calculated based on the differential pressure across the purge valve and the purge valve opening. The degassing Air amount) is calculated (step S7).

Here, the differential pressure across the purge valve is calculated based on the intake charging efficiency, and the reason for doing so is as follows. That is,
The intake pressure can be calculated from the charging efficiency by a well-known method, and the pressure immediately downstream of the purge valve 29 is almost the same as the intake pressure. On the other hand, the pressure immediately upstream of the purge valve 29 can be regarded as a substantially constant value (atmospheric pressure). The differential pressure before and after the purge valve is the purge valve 29.
Is the difference between the pressure on the upstream side and the pressure on the downstream side, that is, the difference between the atmospheric pressure and the intake pressure. Therefore, the differential pressure across the purge valve can be obtained by performing a predetermined calculation process on the charging efficiency. By doing so, it is not necessary to provide an intake pressure sensor, and the intake system 10 is simplified. Of course, the pressure immediately downstream of the purge valve 29 may be detected using an intake pressure sensor. Further, a differential pressure sensor that directly detects the differential pressure across the purge valve may be provided.

The purge air amount is calculated by a well-known method based on the differential pressure across the purge valve and the purge valve opening. That is, generally, there is a constant value which is well known in the field of fluid dynamics between the pressure difference ΔP, that is, the pressure loss across the device, which is provided in a gas closed passage, and the flow velocity u of the gas flowing through the device. A functional relationship is established (for example, ΔP
= K · u ² ). Therefore, the flow velocity of gas in the device can be calculated based on the differential pressure across the device. Then, by multiplying this flow velocity by the flow cross-sectional area of the device, it is possible to obtain the volumetric flow rate of the gas flowing through the flow path. Therefore, in view of such a general principle, in this embodiment, since the flow cross-sectional area of the purge valve 29 can be easily obtained from the purge valve opening degree, the differential pressure across the purge valve and the purge valve opening degree (driving duty ratio) can be calculated. Based on this, the purge air amount (volume flow rate) can be obtained. A flow rate detection sensor that can directly detect the purge air amount may be provided and the flow rate detection sensor may detect the purge air amount.

Then, the evaporated fuel discharge amount (purge gas mass flow rate, that is, evaporated fuel mass flow rate) is calculated based on the purge air amount calculated in step S7 and the trap amount estimated value (step S8). Next, the engine speed is measured (step S9), and the purge gas ratio is calculated based on this engine speed and the evaporated fuel discharge amount calculated in step S8 (step S10). Here, the purge gas ratio is the ratio of the evaporated fuel released from the canister 25 to the purge passage 28 in the total required fuel (required fuel injection amount), and the contribution ratio of the evaporated fuel to combustion. Is represented.

After this, the transportation delay characteristic (intake air amount model) of the purge air or vapor fuel transportation route (hereinafter referred to as vapor fuel transportation route) from the canister 25 to the combustion chamber 4 is set (step S11). , Then step S
The net purge gas ratio is calculated based on the purge gas ratio calculated in 10 and the intake air amount model set in step S11 (step S12). This net purge gas ratio is a ratio of the evaporated fuel flowing into the combustion chamber 4 in all the required fuel (required fuel injection amount), that is, a ratio of the evaporated fuel inflow amount in the required fuel injection amount.
Therefore, the amount of fuel to be injected from the fuel injection valve 15 is
It is the required fuel injection amount multiplied by (1-net purge gas ratio).

Then, the pulse width of the fuel injection valve 15 corresponding to the net purge gas ratio (evaporated fuel inflow amount) calculated in step S12, that is, the purge correction pulse width is calculated.
(Step S13), the purge correction pulse width is output to the engine control block SL.

Hereinafter, the control method of various controls by the control unit CU or the operation method of various operations will be described with reference to FIGS. 2 to 4 in accordance with the flow charts shown in FIGS. First, the overall flow of the control method or the calculation method, that is, the main routine will be described with reference to the flowchart shown in FIG. In this main routine, first, initialization is performed in step T1. Specifically, the trap amount estimation value trap, the air-fuel ratio learning completion flag xlrnd, and the trap amount estimation completion flag xtrap
lrn and the trap amount estimation possible flag xtlex are set to 0 as initial values, respectively. Here, the air-fuel ratio learning completion flag xlrnd is a flag that is set to 1 when the air-fuel ratio learning is completed. Trap amount estimation completion flag xtraplrn
Is set to 1 when the estimation of the trap amount is completed,
This is a flag that is reset to 0 when prohibition of trap amount estimation is continued for a predetermined time or longer. The trap amount estimable flag xtlex is a flag that is set to 1 when the trap amount estimation condition is satisfied and is reset to 0 when the trap amount estimation condition is not satisfied.

Next, at step T2, the engine speed ne is calculated, and then at step T3, the charging efficiency ce is calculated. Here, the charging efficiency ce is calculated by a well-known method based on the intake air amount, the engine speed ne, the intake temperature, and the like.

Thereafter, steps T4 to T8 are sequentially executed. Here, steps T4 to T8 are
It is executed using various subroutines described later. Specifically, in step T4, the fuel injection amount is calculated using the subroutine whose flowchart is shown in FIG. In step T5, the purge execution determination is performed using the subroutine whose flowchart is shown in FIG. In step T6, the purge amount is calculated using the subroutine whose flowchart is shown in FIG. In step T7, the execution determination of the trap amount estimation is performed using the subroutine whose flowchart is shown in FIG. In step T8,
The trap amount is calculated by using the subroutine whose flowchart is shown in FIG. Then, the process returns to step T2.

The subroutines will be specifically described below. First, referring to FIG. 2 to FIG. 4 according to the flow chart shown in FIG. 6, step T of the main routine
A trap amount calculation subroutine executed (started) in 8, that is, a concrete estimation method of the trap amount by the control unit CU will be described.

In this subroutine, first, at step # 2, whether or not a predetermined trap amount estimation condition is satisfied, that is, the operating state of the engine CE is in a state in which the trap amount can be estimated with high accuracy. It is determined whether or not. Here, when all of the following four conditions are satisfied, it is determined that the trap amount estimation condition is satisfied. (1) Canister purge is performed (2) Air-fuel ratio feedback control is performed (3) Filling efficiency is less than a predetermined value (see step # 33 in FIG. 7) (4) The air-fuel ratio learning is completed (see step # 34 in FIG. 7). In addition to the above (1) to (4), when the intake pressure is below a predetermined value, estimation of the trap amount can be prohibited. In addition, the trapping amount estimation condition is that the charging efficiency capable of obtaining the intake pressure or the intake pressure equivalent is equal to or higher than the lower limit value (second predetermined value) lower than the predetermined value (upper limit value) of (3) above. Is also good.

In other words, when the canister purge is stopped, the air-fuel ratio feedback control is stopped, the charging efficiency is equal to or higher than a predetermined value, or the air-fuel ratio learning is not completed, a trap is generated. The amount estimation is prohibited. The reason for doing this is as follows. That is, as described above, when the canister purge or the feedback control is stopped, there is no correlation between the trap amount and the feedback correction value or the average feedback correction value, and therefore the trap amount cannot be estimated. Therefore, the estimation of the trap amount is prohibited. When the charging efficiency or the intake pressure is very high, the differential pressure across the purge valve becomes very small and the intake pulsation becomes severe, and the feedback correction value fluctuates, so that the purge air amount cannot be calculated with high accuracy. The estimation of the trap amount is prohibited. Further, when the charging efficiency or the intake pressure is very low, the differential pressure across the purge valve becomes too large to calculate the purge air amount with high accuracy, so the estimation of the trap amount is prohibited. Further, as described above, the trap amount is estimated after the air-fuel ratio learning is completed.
This is because the estimation accuracy of the trap amount becomes very high after the air-fuel ratio learning is completed.

Thus, when it is determined in step # 2 that the trap amount estimation condition is satisfied (YES), the trap amount estimation possible flag xtlex is set to 1 and the trap amount estimation is prohibited in step # 3. The initial value ct ₀ is set in the counter ct. Trap amount estimation prohibition counter ct
Is a counter for counting the period (time) during which the trap amount estimation condition is not satisfied and the trap amount estimation is prohibited.

Then, in step # 4, the average feedback correction value cfbave is calculated by the following equation 1, and the calculation number counter P for counting the number of calculation times of the average feedback correction value cfbave is incremented by one. (P = P + 1). That is, in this engine using the linear O ₂ sensor 9, the arithmetic mean value of the feedback correction values for every fixed time is the average feedback correction value cf.
It is said to be bave.

[Formula 1] cfbave = Σ (k = 0 → n−1) [cfb (i−k)] / n ……………………………… Equation 1 cfbave: Average feedback correction value cfb (i): Feedback correction value this time cfb (i−k): Feedback correction value k times before n: Number of cfb samples to be averaged Note that in equation 1, Σ (k = α → β) [f (k)] Is the function f
It is assumed that the sigma operation (Σ) from k = α to k = β in (k) is represented. Next, in step # 5, it is determined whether cfbave is 0 or less. Here, if it is determined that cfbave ≦ 0 (YES), the trap amount estimated value trap is increased by the correction amount σ in step # 6 (trap = trap + σ),
On the other hand, if it is determined that cfbave> 0 (NO), the trap amount estimated value trap is reduced by the correction amount σ in step # 7 (trap = trap−σ).

Next, at step # 8, the number-of-operations counter P
Is determined to be greater than or equal to a predetermined set value P ₀ , and P <
When it is determined to be P ₀ (NO), all the following steps are skipped, and the process returns to step # 2.

On the other hand, if it is determined in step # 8 that P ≧ P ₀ (YES), then in step # 9 whether the absolute value | cfbave | of the average feedback correction value cfbave is equal to or greater than a predetermined limit value ε. It is determined whether or not. Here, │cfbave
If │ <ε, it is determined that the trap amount estimation is completed, and in this case, the trap amount estimated value is considered to be almost the same as the true trap amount, so the trap amount estimation completion flag xtraplrn I am trying to add 1 to it.
For example, as schematically shown in FIG. 11, when the true trap amount is a ₂ , the trap amount estimation is completed when the trap amount estimated value trap falls within the range of a _{1 to} a _3. It will be judged that there is. In FIG. 11, the graph G ₁ represents | cfbave |. Also, cfbave> 0 in the range of trap> a ₂ and cf in the range of trap <a _2.
bave <0. On the other hand, if | cfbave | ≧ ε, the trap amount estimation completion flag xtraplrn is not changed.

Specifically, in step # 9, │cfbave│ <
If it is determined to be ε (NO), the trap amount estimation completion flag xtraplrn is set to 1 in step # 10, and then the process returns to step # 2. On the other hand, in step # 6 │
If it is determined that cfbave│ ≧ ε (YES), the process immediately returns to step # 2.

If it is determined in step # 2 that the trap amount estimation condition is not satisfied (N
O), the trap amount estimation prohibition counter ct in step # 11
Is decremented by 1 and counting of the period (time) during which estimation of the trap amount is prohibited is started ((c
t = ct-1), the operation counter P is reset to 0
(P = 0).

Next, at step # 12, it is judged whether or not the trap amount estimation prohibition counter ct is 0 or less, that is, whether or not a predetermined period ct ₀ has elapsed after the trap amount estimation is prohibited, and ct If it is determined that ≦ 0 (YE
S), since ct ₀ has already passed, the trap amount estimation completion flag xtraplrn is reset to 0 in step # 13. In this case, as described above, the trap amount estimated value tr
Since ap may deviate from the true trap amount,
The trap amount estimation completion flag xtraplrn is reset. On the other hand, if it is determined in step # 12 that ct> 0 (NO), ct ₀ has not yet elapsed, so step # 13 is skipped.

FIG. 12 shows the drive duty ratio dpg (graph G ₂ ) applied to the purge valve 29 and the feedback correction when the canister purge is started at time t ₁ when the trap amount estimating routine is executed. value
An example of changes over time in cfb (graph G ₃ ), average feedback correction value cfbave (graph G ₄ ) and trap amount estimated value trap (graph G ₅ ) is shown. As is clear from FIG. 12, the estimation of the trap amount estimated value trap starts at time t ₁ and soon converges to a constant value. In this way, the trap amount is estimated with high accuracy.

In the trap amount calculation subroutine shown in FIG. 6, in step # 4, the arithmetic average of the feedback correction values cfb at constant time intervals is set as the average feedback correction value cfbave as described above. Not using the feedback correction value cf
Calculate the weighted average value, that is, the smoothed value, with weighting b,
This weighted average value may be used as the average feedback correction value cfbave.

[Formula 2] cfbave = α ・ cfb (i) ＋ (1-α) ・ cfb (i-1) ………………………………………………………………………………………………………………………………………………………………………………………………………………………………. ): Current feedback correction value cfb (i-1): Previous feedback correction value

If the control unit CU has an air-fuel ratio learning function, it is preferable to estimate the trap amount after the air-fuel ratio learning is completed, as described above. The advantage of estimating the trap amount after the air-fuel ratio learning is completed is as described above.

A trap amount estimation execution determination subroutine executed (started) in step T7 of the main routine, that is, an air-fuel ratio learning function is provided according to the flowchart shown in FIG. 7 and with reference to FIGS. A method for estimating a preferable trap amount in the case of the above will be described. In this subroutine, basically, the air-fuel ratio learning is performed when the predetermined air-fuel ratio learning condition is satisfied at the idle time, while the predetermined trap amount estimation condition is satisfied after the air-fuel ratio learning is completed. The amount of traps is estimated when there is.

Specifically, as shown in FIG. 7, first, at step # 22, it is judged if the idle judgment flag xidl is "1". The idle determination flag xidl is the engine C
This flag is set to 1 when E is in the idle state and reset to 0 when it is in the off-idle state (non-idle state). If it is determined that xidl = 1 (YES), step # 23 to step #
The routine for idle time for performing 29 air-fuel ratio learning is executed. On the other hand, if it is determined that xidl ≠ 1 (xidl = 0) (NO), the air-fuel ratio learning execution counter clrn and the idle duration time counter tidl are both 0 in step # 30.
After being reset to step # 31, step # 31 described later will be executed.

When the idle routine is executed, first, in steps # 23 and # 24,
It is determined whether the air-fuel ratio feedback control execution flag xfb is 1 and whether the idle duration time counter tidl exceeds a predetermined value α. Here, the air-fuel ratio feedback control execution flag xfb is set to 1 when feedback control of the air-fuel ratio is being performed, and when feedback control is not being performed (during open loop control)
Is a flag that is reset to 0. Further, the idle duration time counter tidl is a counter for counting the elapsed time after the start of idle operation at the time of idling.

In this trap amount estimation execution determination subroutine, it is considered that the engine CE has not reached the stable idle state when the idle operation duration time is the predetermined period α or less during the feedback control of the air-fuel ratio at the idle time. Therefore, the air-fuel ratio learning is not performed. Thus, in step # 23, xfb ≠ 1 (xf
If it is determined that b = 0) (NO), the air-fuel ratio learning cannot be performed, so that steps # 24 to #
29 is not executed, and after the air-fuel ratio learning execution counter clrn and the idle duration time counter tidl are reset to 0 in step # 30, step # 31 is executed.

Further, if it is determined in step # 23 that xfb = 1 (YES), that is, if it is determined in step # 24 that tidl ≦ α even during feedback control (NO). However, since the air-fuel ratio learning cannot be performed again, after the idle duration time counter tidl is incremented by 1 in step # 28 (tidl = tidl
+1), step # 31 is executed.

On the other hand, if it is determined in step # 24 that tidl> α (YES), it is determined in step # 25 whether the air-fuel ratio learning execution counter clrn is less than the predetermined value β. The air-fuel ratio learning execution counter clrn is a counter for counting the number of times the air-fuel ratio learning has been executed after the start of idle operation. Here, it is determined that the air-fuel ratio learning is completed when the number of executions of the air-fuel ratio learning becomes equal to or larger than the predetermined value β.

Thus, if it is determined at step # 25 that clrn <β (YES), the air-fuel ratio learning is not yet completed, so the air-fuel ratio learning is continued at step # 26, and then step # 27. With air-fuel ratio learning execution counter cl
rn is incremented by 1 (clrn = clrn + 1), and then step # 31 is executed. The air-fuel ratio learning is
When the air-fuel ratio deviation is 0, the feedback correction value cfb is changed to the neutral value 0 on average, and the injection characteristic of the fuel injection valve 15 is changed by an ordinary method.

On the other hand, if it is determined in step # 25 that clrn ≧ β (NO), the air-fuel ratio learning has been completed, and therefore the air-fuel ratio learning completion flag xlrnd is set to 1 in step # 29. After this, step # 31 is executed.

Thus, step # 31 to step # 3
In 4, it is determined whether the trap estimation condition is satisfied. Here, when the following four conditions are all satisfied, it is determined that the trap amount estimation condition is satisfied, and if any one of these conditions is not satisfied, the trap amount estimation condition is not satisfied. I try to judge that there is. (1) Canister purge is being performed (2) Air-fuel ratio feedback control is being performed (3) Filling efficiency (or intake pressure) is less than a prescribed value (4) Air-fuel ratio learning is complete (Ended) The reason for providing these four conditions (1) to (4) is the same as in step # 2 of the trap amount calculation subroutine shown in the flowchart of FIG.

Specifically, step # 31 to step #
In 4 steps of 34, purge execution flag xpg in order
Is 1, that is, whether or not canister purge is being performed, and the air-fuel ratio feedback control execution flag xf
Whether b is 1 or not, that is, whether the air-fuel ratio feedback control is being performed, whether or not the charging efficiency ce is less than a predetermined value γ, and whether or not the air-fuel ratio learning completion flag xlrnd is 1 It is determined whether the air-fuel ratio learning is completed.

Where xpg = 1 and xfb = 1,
When it is determined that ce <γ and xlrnd = 1 (steps # 31 to # 34 are all YES),
In step # 35, it is determined that the trap amount estimating condition is satisfied, and in this case, the trap amount is estimated by the trap amount calculating subroutine whose flow chart is shown in FIG.

On the other hand, when it is judged that xpg = 0, xfb = 0, ce ≧ γ, or xlrnd = 0 (one of step # 31 to step # 34). If the trap amount estimation condition is not satisfied, step # 35 is skipped. As shown in FIG.
According to this trap amount estimation method whose flow chart is shown in FIG. 1, the trap amount is estimated after the air-fuel ratio learning is completed, so that the estimation accuracy of the trap amount is significantly improved.

The fuel injection amount calculation subroutine executed (started) in step T4 of the main routine, that is, the purge gas ratio (evaporation by the control unit CU), will be described below with reference to FIGS. The calculation method of the fuel release amount) and the net purge gas ratio (evaporated fuel inflow amount) and the control method of the air-fuel ratio control (fuel injection amount control) will be specifically described. In this fuel injection amount calculation subroutine, step # 41 to step # 47 are sequentially executed. In step # 41, the differential pressure table before and after the purge valve dp is searched based on the intake charging efficiency ce by searching the differential pressure table before and after the purge valve.
Is calculated. Where sipol (table1, ce) represents a given functional relationship with ce as the independent variable and dp as the dependent variable.
In table 1, it means dp corresponding to an arbitrary ce.
The purge valve front-rear differential pressure table table1 is a table showing a functional relationship between the intake charging efficiency ce and the purge valve front-rear differential pressure dp. The reason why the differential pressure dp before and after the purge valve can be calculated based on the intake charging efficiency ce is as described above. The differential pressure across the purge valve dp
Instead of using a table lookup like this, ce and dp
Alternatively, a function f ₁ (ce) representing the relation of may be used for direct calculation (dp = f ₁ (ce)).

At step # 42, the purge air amount map
By searching map1, the differential pressure before and after the purge valve dp
And the drive duty ratio dpg applied to the purge valve 29
The purge air amount qpg is calculated based on and. Where sma
p (map1, dpg, dp) means qpg corresponding to an arbitrary dpg and dp in map1, which represents a predetermined functional relationship in which dpg and dp are independent variables and qpg is a dependent variable. The purge air amount map1 is a map showing a functional relationship among the drive duty ratio dpg, the purge valve front-rear differential pressure dp, and the purge air amount qpg. The purge air amount qpg can be calculated based on the drive duty ratio dpg and the purge valve front-rear differential pressure dp as described above. It should be noted that the purge air amount qpg is not based on such a map search, but is a function that expresses the interrelationship of dpg, dp and qpg.
f ₂ (dpg, dp) may be used for direct calculation
(qpg = f ₂ (dpg, dp)).

In step # 43, the evaporated fuel discharge amount map map2 is searched to calculate the evaporated fuel discharge amount gpg based on the purge air amount qpg and the trap amount trap. Here, smap (map2, qpg, trap) is a gpg corresponding to an arbitrary qpg and trap in map2 that represents a predetermined functional relationship with qpg and trap as independent variables and gpg as a dependent variable.
Means The purge air amount map2 is the purge air amount qpg,
It is a map showing the functional relationship between the trap amount trap and the evaporated fuel discharge amount gpg. Evaporative fuel emission amount gp
instead of g with qpg, tr
The function f ₃ (gpg, trap), which expresses the mutual relationship between ap and gpg, may be used for direct calculation (gpg ＝ f ₃ (qpg, t)
rap)).

At step # 44, the purge gas ratio cpgo is calculated by the following equation (3).

[Equation 3] cpgo = Ys · [120 / (γ ₀ · Vc)] · (gpg / ne) …………………………………………………………………………………… 3 cpgo: Purge gas ratio Ys: To convert the intake air amount to the fuel injection amount Conversion coefficient of γ ₀ : Density Vc: Effective cylinder volume gpg: Evaporative fuel emission amount ne: Engine speed [rpm] In formula 3, 120 / (γ ₀ · Vc · ne) is per unit time (second) Is the reciprocal of the intake air amount (mass flow rate) into the combustion chamber 4, and thus Ys · 120 / (γ ₀ · Vc · ne) multiplied by the conversion coefficient Ys is the reciprocal of the required fuel injection amount per unit time. Is. Therefore, the purge gas ratio
cpgo is a value obtained by dividing the evaporated fuel release amount by the required fuel injection amount, that is, the ratio of the evaporated fuel release amount to the total fuel flow rate.

At step # 45, the net purge gas ratio cpg is calculated by the following equation (4).

[Equation 4] cpg = λ · cpg + (1-λ) · cpgo ………………………………………………………………………………………………………………………………………………………………………………………… 4. cpgo: purge gas ratio Equation 4 is a model equation representing the transport delay characteristic of the evaporated fuel transport path. By properly setting the primary filter coefficient λ according to the shapes of the intake system 10, the AMI 16 and the purge passage 28 of the engine CE, the net purge gas ratio cpg (evaporated fuel inflow amount) can be accurately calculated using the equation (4). it can.

At step # 46, the actual pulse width ta (actual fuel injection amount), that is, the fuel injection amount to be actually injected from the fuel injection valve 15 is calculated by the following equation 5.

[Equation 5] ta = K · (ce · ctotal−cpg) ……………………………………………………………………………………………… Ta: Actual pulse width K: Conversion factor ce: Intake charging efficiency ctotal: Correction In the equation (5), K · ce · ctotal represents the required pulse width (required fuel injection amount), that is, the amount of fuel actually required in the combustion chamber 4. Further, k · cpg is the amount of fuel supplied by the purge converted to the injection pulse width. Therefore, in Equation 5, the injection pulse width corresponding to the fuel amount (actual fuel injection amount) to be actually injected from the fuel injection valve 15, that is, the actual pulse width ta is calculated.

In step # 47, fuel is injected from the fuel injection valve 15 with the actual pulse width ta calculated in step # 46, and then the process returns to step # 41. In this way, even when the canister purge is performed, the required amount of fuel is accurately supplied to the combustion chamber 4, and the control accuracy of the air-fuel ratio control (fuel injection amount control) is improved. The actual air-fuel ratio is increased and the target value is maintained. In this case, the process of determining the required fuel injection amount according to the operating state, that is, the air-fuel ratio control main body is feedback control, but the process of subtracting the evaporated fuel inflow amount from the required fuel injection amount to eliminate the influence of the canister purge is the feed control. It is forward control. Therefore, there is no time lag in the calculation of the net purge gas ratio or the evaporated fuel inflow amount. Therefore, the deviation of the actual air-fuel ratio from the target value due to the canister purge does not occur.

As described above, in the engine CE, the trap amount is estimated, the evaporated fuel inflow amount (net purge gas ratio) is accurately calculated based on the trap amount, and the evaporated fuel inflow amount is subtracted from the required fuel injection amount. Since the actual fuel injection amount is set by the canister purge, the evaporated fuel flowing into the intake system 10 or the combustion chamber 4 does not become a disturbance of the air-fuel ratio feedback control. Therefore, when the estimation of the trap amount is completed, the canister purge does not cause the air-fuel ratio to deviate from the target value.

By the way, as described in the above-mentioned "Problems to be solved by the invention", when the canister purge is started in the low intake air amount region such as the idle region where the fuel injection amount is small, the feedback of the air-fuel ratio is performed. Control is likely to be disturbed. If the canister purge is started when the trap amount estimation is not completed, that is, when the trap amount estimation completion flag xtraplrn is 0, the net purge gas ratio or the evaporated fuel inflow amount is not accurately grasped. Disturbances easily occur in the air-fuel ratio control. Therefore, in this embodiment, when the canister purge is started, the canister purge is controlled or regulated as described below to prevent the feedback control of the air-fuel ratio from being disturbed.

That is, as shown in FIG. 13, basically, when the canister purge is started in the feedback control region of the air-fuel ratio, the purge amount is gradually increased to the target purge amount. Then, the increase rate is set to be smaller in the idle region (low intake air amount region) than in the off-idle region (high intake air amount region). Further, in both the idle region and the off-idle region, when the estimation of the trap amount is not completed (learning incomplete), the above increase rate is higher than that when the estimation is completed (learning completed). It is set to a small value.
Further, the purge amount increase rate when the estimation of the trap amount is completed in the idle region is set to a value smaller than the purge amount increase rate when the estimation is not completed in the off idle region.

In FIG. 13, graph J ₁ and graph J ₂
3A and 3B show changes with time of the drive duty ratio (canister purge amount), that is, the increase rate of the canister purge amount when the estimation of the trap amount is completed and when the estimation of the trap amount is not completed during off-idle. Graphs J ₃ and J ₄ respectively show changes with time of the drive duty ratio when the trap amount estimation is completed and when it is not completed at the time of idling. It should be noted that d ₁ and d ₂ indicate the target drive duty ratio (target canister purge amount) at the time of off idle and at the time of idle, respectively.

As described above, when the canister purge is started, the canister purge amount is gradually increased. Therefore, the inflow amount of the evaporated fuel into the intake system 10 does not increase so much at the beginning of the canister purge, and the air-fuel ratio feedback control is performed. Is prevented, and the stability or responsiveness of the control is improved. Then, after a while after the start of the canister purge, the canister purge amount reaches the target canister purge amount, so the evaporated fuel collected (adsorbed) in the canister 25 is quickly discharged to the intake system 10. Therefore, the canister can be downsized. That is, the canister purge amount can be increased while ensuring the stability or responsiveness of the feedback control of the air-fuel ratio, the canister 25 can be downsized, and the fuel efficiency can be improved. Furthermore, when the estimation of the trap amount has not been completed, the increase rate of the canister purge amount is set to a small value, so even if the trapped amount of fuel vapor, and thus the canister purge amount, is not accurately grasped, the canister purge amount is not initially calculated. Disturbance is prevented from occurring in the feedback control of the fuel ratio, and the control stability is further improved.

The purge amount calculation subroutine executed (started) in step T6 of the main routine, that is, the drive duty ratio (control) at the start of the canister purge, according to the flowchart shown in FIG. 9 and with reference to FIGS. A method of controlling the increase rate of the drive duty ratio when the duty ratio) is gradually increased will be described. In this purge amount calculation subroutine, first in step # 51, it is judged whether or not the purge execution flag xpg is 1, and if xpg ≠ 1 (xpg = 0) (NO), step # 52 and step # 53. Then, the idle purge correction value cmodi and the off-idle purge correction value cmodn are set to 0, respectively, and thereafter, the process returns to step # 51.

Both of the correction values cmodi and cmodn are 0 or more and 1 or less for gradually increasing the drive duty ratio toward the target drive duty ratio after the canister purge is started at the time of idling or off idling. The correction value, which is the target drive duty ratio and the correction value cmodi, c
The product of modn and drive duty ratio of purge valve 29 is dpg
becomes φ. Both correction values cmodi and cmodn are set to 0 before the start of the canister purge, and are gradually increased by the increment SP after the start of the canister purge, so that the purge air amount is gradually increased. After both correction values cmodi and cmodn have reached 1, they are maintained at 1 without further increase. When the correction values cmodi and cmodn are 0, the canister purge is stopped regardless of the value of the target drive duty ratio.

On the other hand, if it is determined in step # 51 that xpg = 1 (YES), it is determined in step # 54 whether or not the idle determination flag xidl is 1, that is, whether or not the engine is idle. To be judged. Where xidl ≠ 1 (xid
If it is determined that (l = 0) (NO), that is, at the time of off-idle, the routine for off-idle at step # 64 to step # 71 is executed. In this case, first, at step # 64, the idle purge correction value cmodi is set to 0. In short, the idle purge correction value cmodi is not used.

Next, at step # 65, it is judged if the off-idle purge correction value cmodn is 1. If it is determined that cmodn ≠ 1 (that is, cmodn <1) (NO), it means that the off-idle purge correction value cmodn should be gradually increased after the start of the canister purge, so step # 66. At step # 69, the off-idle purge correction value cmodn is gradually increased in consideration of whether the trap amount estimation is completed.

Specifically, in step # 66, it is determined whether the trap amount estimation completion flag xtraplrn is 1, that is, whether the trap amount estimation is completed.
If it is determined that xtraplrn = 1, then (YE
S), a value KM with a relatively large increment SP in step # 68
N1, and then step # 69 is executed. In this case, since the trap amount has been estimated, the net purge gas ratio or the evaporated fuel inflow amount can be accurately calculated. Therefore, the influence of the canister purge can be reliably compensated by the feedforward control. That is, even if the canister purge is started abruptly to some extent, the effect is sufficiently compensated and the air-fuel ratio control is not disturbed. Therefore, the increment SP is increased, that is, the increase rate of the off-idle purge correction value cmodn is increased, and the canister purge is performed at the target drive duty ratio early (see graph J _{1 in} FIG. 13).

On the other hand, in step # 66, xtraplrn ≠ 1 (xtr
If it is determined that aplrn = 0) (NO), in step # 67, the increment SP has a relatively small value KMN2 (KMN2).
<KMN1), and then step # 69 is executed. In this case, since the estimation of the trap amount has not been completed yet, the net purge gas ratio or the evaporated fuel inflow amount cannot be accurately calculated. Therefore, since the feedforward function does not work sufficiently, if the canister purge is suddenly started, its effect is not sufficiently compensated, and the air-fuel ratio control is disturbed. Therefore, the increment SP is made relatively small so that the increase rate of the off-idle purge correction value cmodn is made relatively small (see the graph J _{2 in} FIG. 13).

At step # 69, SP is added to the previous off idle purge correction value cmodn to calculate the current off idle purge correction value cmodn. If the off-idle purge correction value cmodn of this time exceeds 1 as a result of the addition, it is kept at 1. Where addclip (cmodn, S
P, 1) represents an arithmetic process in which SP is added to cmodn but the upper limit value is set to 1. In this way,
The off-idle purge correction value cmodn is gradually increased.

Next, at step # 70, the drive duty ratio dpgφ of the purge valve 29 is calculated by the following equation (6).

[Equation 6] dpgφ = cmodn ・ smap (map3, ne, ce) ……………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………… This 6 , ne, ce): Target drive duty ratio [%] ne: Engine speed ce: Intake charging efficiency In Equation 6, smap (map3, ne, ce) is ne and ce.
Where d is the independent variable and dpgφ is the dependent variable.
Means dpgφ corresponding to e. The duty ratio map3 is
5 is a map showing a functional relationship among an engine speed ne, an intake charging efficiency ce, and a drive duty ratio dpgφ.
In this way, when the canister purge is started, the drive duty ratio dpgφ and thus the purge air amount is gradually increased.

Incidentally, in the above step # 65, cmodn = 1
If it is determined to be (YES), it means that cmodn has already reached 1 after the start of the canister purge. Therefore, step # 66 to step # 69 are skipped, and in step # 70, the idle purge is started. Correction value cmodn
The drive duty ratio dpgφ is calculated with 1 as 1.

After step # 70 is executed, drive frequency ppreq of purge valve 29 is fixed to a constant value of 10 Hz in step # 71. That is, the driving cycle is 100
ms. In this case, as described above, the on time of the purge valve 29 changes according to the drive duty ratio. Thereafter, step # 63 described later is executed.

By the way, in step # 54, xidl
When it is determined that = 1 (YES), that is, at the time of idling, the routine for idling at steps # 55 to # 63 is executed. In this case, first, in step # 55, the off-idle purge correction value cmodn is set to 0. In short, the off-idle purge correction value cmodn
Is not used. Note that # 52, # 53, # 55, # 6
Although the off-idle page correction value cmodn of 4 is set to 0, it is set to 0.1 in order to speed up the rising of the purge.
You may give it a little as 25. Next, step #
At 56, it is determined whether the idle purge correction value cmodi is 1. Where cmodi ≠ 1 (ie cmodi <
If it is determined to be 1) (NO), since the idle purge correction value cmodi should be gradually increased after the start of the canister purge, the trap amount estimation is completed in step # 57 to step # 60. The idle purge correction value cmodi is gradually increased in consideration of whether or not it is.

Specifically, in step # 57, it is determined whether the trap amount estimation completion flag xtraplrn is 1, that is, whether the trap amount estimation is completed.
If it is determined that xtraplrn = 1, then (YE
S), a value KM with a relatively large increment SP in step # 58
I1 is set, and then step # 60 is executed. The KMI1 is set to a value smaller than the KMN2. In this case, since the trap amount has been estimated, the net purge gas ratio or the evaporated fuel inflow amount can be accurately calculated. Therefore, the influence of the canister purge can be reliably compensated by the feedforward control. That is, even if the canister purge is started abruptly to some extent, the effect is sufficiently compensated and the air-fuel ratio control is not disturbed. Therefore, the increment SP is made relatively large, that is, the increase rate of the idle purge correction value cmodi is made relatively large, and the canister purge is performed at the target drive duty ratio at an early stage. However, since the fuel injection amount is originally small at the time of idling, and therefore the feedback control of the air-fuel ratio is likely to be disturbed, the KMI1 when the estimation of the trap amount is completed at this idling is KMN when the estimation of
The value is smaller than 2 (see FIG. 13).

On the other hand, in step # 57, xtraplrn ≠ 1 (xtr
When it is determined that aplrn = 0) (NO), the increment SP is a relatively small value KMI2 (KMI2 <KMI1 <KMN2 <K) in step # 59 compared with the above KMI1.
MN1), after which step # 57 is executed.
In this case, since the estimation of the trap amount has not been completed yet, the net purge gas ratio or the evaporated fuel inflow amount cannot be accurately calculated. Therefore, since the feedforward function does not work sufficiently, if the canister purge is suddenly started, its effect is not sufficiently compensated and the air-fuel ratio control is disturbed. Therefore, the increment SP is reduced, that is, the increase rate of the idle purge correction value cmodi is reduced.

At step # 60, SP is added to the previous idle purge correction value cmodi to calculate the current idle purge correction value cmodi. If the current idle purge correction value cmodi exceeds 1 as a result of the addition, it is kept at 1. Here, the meaning of addclip (cmodi, SP, 1) is the same as in step # 69. In this way, the idle purge correction value cmodi is gradually increased.

Next, in step # 61, the drive duty ratio dpgφ of the purge valve 29 is calculated by the following equation 7.

[Equation 7] dpgφ = cmodi · α ……………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………… Drive duty ratio [%]

Incidentally, in the above step # 56, cmodi = 1
If YES is determined (YES), it means that cmodi has already reached 1 after the start of the canister purge, so steps # 57 to # 60 are skipped and the idle purge correction value is set in step # 61. cmodi 1
The drive duty ratio dpgφ is calculated as

After step # 61 is executed, the drive frequency pfreq of the purge valve 29 is calculated by the following equation 8 in step # 62, and then step # 63 is executed.

[Formula 8] pfreq = clip (dpgφ / 1ms, 2Hz, 10Hz) Equation 8 pfreq: drive frequency dpgφ: drive duty ratio Note that in equation 8, clip (dpgφ / 1ms, 2Hz,
10Hz) basically sets the value of dpgφ to pfreq, but sets the upper limit to 10Hz and the lower limit to 2Hz.
Means that For example, when dpgφ is 0 to 2%, pfreq is 2 Hz, and when dpgφ is 3%, pfreq is
Is 3Hz, and if dpgφ is 20%, pfreq is 10H
z.

At step # 63, the on-time dpg of the purge valve 29 is calculated by the following equation 9.

[Equation 9] dpg = max (dpgφ × 1 ms, 10 ms) ……………………………………………………………………………………………………………………………………………… 9 9 9 max (dpgφ × 1ms, 10ms)
Means that the maximum of dpgφ and 10 ms is dpg.

Thus, the drive duty ratio dpgφ is 10
If it is more than%, the driving frequency pfreq is set to a constant value of 10 Hz (driving cycle is 100 ms), and the on-time dpg (ms)
Becomes the same value as the drive duty ratio dpgφ (%). On the other hand,
If the drive duty ratio dpgφ is less than 10%, the on-time will be a constant value of 10 ms, and the drive frequency pfreq (Hz)
Becomes the same value as the drive duty ratio dpgφ (%), but the lower limit of the drive frequency pfreq is 2 Hz. For example, when the drive duty ratio dpgφ is 30%, the drive frequency pfreq is 1
It is 0 Hz (driving cycle 100 ms), and the on-time dpg is 30 ms. Also, the drive duty ratio dpgφ is 3%
In the case of, the drive frequency pfreq is 3 Hz (drive cycle 333 ms), and the on-time is 10 ms.

It should be noted that in FIG. 14, the drive duty ratio dpgφ (graph J ₅ ) is maintained at a constant value of 2% during the period from time θ ₁ to time θ ₂ by such control, and thereafter at a constant increase rate. An example of a temporal change (graph J ₆ ) of the drive frequency pfreq in the case where the driving frequency pfreq increases and reaches 10% at the time θ ₃ and further increases and is then held at a constant value (greater than 10%) is shown.

The purge execution determination subroutine executed (started) in step T5 of the main routine will be described below with reference to the flowchart shown in FIG. In this purge execution determination subroutine, it is first determined in step # 81 whether or not purge execution is possible. If purge execution is not possible (NO), 0 is set in the purge execution flag xpg in step # 85.
After this, the process returns to step # 81. It should be noted that here, when the water temperature is 80 ° C. or higher and the operating state of the engine CE is in the feedback zone or the enrichment zone, it is possible to execute the purge.

On the other hand, if it is determined in step # 81 that the purge can be executed (YES), it is determined in step # 82 whether or not the idle determination flag xidl is 1, that is, whether or not the engine is idle. If xidl ≠ 1 (xidl = 0) (NO), that is, if the engine is not idle, the purge execution flag xpg is set to 1 in step # 84.

If it is determined in step # 82 that xidl = 1 (YES), it is determined in step # 83 whether the air-fuel ratio learning completion flag xlrnd is 1. Here, the purge execution is permitted only when xlrnd = 1, that is, when the air-fuel ratio learning is completed at the time of idling (the purge execution flag xpg is set to 1).
Set).

Thus, if it is determined in step # 83 that xlrnd = 1 (YES), 1 is set in the purge execution flag xpg in step # 84. On the other hand, xlrnd
Is 0 (NO), the purge execution flag xpg is set to 0 in step # 85.

In this way, even when idle,
The canister purge can be performed without causing disturbance in the air-fuel ratio control, that is, without causing a deviation of the air-fuel ratio from the target value.

In the above embodiment, the trap amount is estimated based on the average feedback correction value. However, a trap amount detecting sensor for directly detecting the trap amount is provided, and the trap detected by this trap amount detecting sensor is provided. The evaporated fuel discharge amount (purge gas ratio) or the evaporated fuel inflow amount (net purge gas ratio) may be calculated based on the amount. In this case, as the trap amount detection sensor, a sensor that detects the trap amount based on the electrostatic capacity of the adsorbent in the canister 25, an HC sensor, or the like can be used.

[0135]

In the engine control apparatus according to the first aspect of the present invention, when the purge of the evaporated fuel is executed in the low intake air amount range, the purge flow rate gradually changes at a relatively small change rate. The change in the inflow of evaporated fuel into the intake system does not increase so much at the beginning of purging.
Disturbance is prevented from occurring in the feedback control of the air-fuel ratio, and the stability or responsiveness of the control is improved. Then, after a while after the start of purging, the purge flow rate reaches the target purge flow rate, so that the evaporated fuel collected (adsorbed) in the canister is rapidly discharged to the intake system. Further, when the purge of the evaporated fuel is executed in the high intake air amount region where the fuel injection amount is large, the purge flow rate is changed at a relatively large rate of change, so that the evaporated fuel trapped in the canister is quickly discharged. Is discharged to the intake system. Therefore, the canister can be downsized.

In the engine control system according to the second aspect of the present invention, when the purge of the evaporated fuel is started in the low intake air amount range, the purge flow rate is gradually increased at a relatively small increase rate, At the initial stage of purging, the change in the amount of evaporated fuel flowing into the intake system does not increase so much, so that the feedback control of the air-fuel ratio is prevented from being disturbed, and the stability or responsiveness of the control is improved. Then, after a while after the start of purging, the purge flow rate reaches the target purge flow rate, so that the evaporated fuel collected (adsorbed) in the canister is rapidly discharged to the intake system. Also, when the purge of the evaporated fuel is started in the high intake air amount region where the fuel injection amount is large, the purge flow rate is increased at a relatively large increase rate, so the evaporated fuel trapped in the canister is quickly It is discharged to the intake system. Therefore, the canister can be downsized. That is, the purge flow rate can be increased while ensuring the stability or responsiveness of the feedback control of the air-fuel ratio, the canister can be downsized, and the fuel efficiency can be improved.

In the engine control system according to the third aspect of the present invention, the actual fuel injection amount is set by subtracting the evaporated fuel amount from the required fuel injection amount and then the air-fuel ratio feedback control is performed. Therefore, the stability or responsiveness of the feedback control is improved. Further, in the case where the purge of the evaporated fuel is executed in the idle region where the fuel injection amount is small, the purge flow rate gradually changes at a relatively small change rate when the calculation of the estimated value of the evaporated fuel amount is not completed. Therefore, even when the amount related to the amount of evaporated fuel (and thus the amount of evaporated fuel) is not accurately grasped, the feedback control of the air-fuel ratio is prevented from being disturbed at the initial stage of purging, and the control stability is further improved.

In the engine control device according to the fourth aspect of the present invention, basically, as in the case of the engine control device according to the second aspect, the stability of feedback control of the air-fuel ratio or The purge flow rate can be increased while ensuring the responsiveness, the canister can be downsized, and the fuel efficiency can be improved.
Further, since the evaporated fuel amount is subtracted from the required fuel injection amount and the actual fuel injection amount is set before the feedback control of the air-fuel ratio is performed, the stability or responsiveness of the feedback control is further improved. . Further, when the purge of the evaporated fuel is started in the idle region where the fuel injection amount is small, the purge flow rate is gradually increased at a relatively small increase rate when the calculation of the estimated value of the evaporated fuel amount is not completed. Therefore, even if the amount related to the evaporated fuel amount (and thus the evaporated fuel amount) is not accurately grasped, the feedback control of the air-fuel ratio is prevented from being disturbed at the initial stage of the purge, and the control stability is further improved.

In the engine control device according to the fifth aspect of the present invention, the amount relating to the amount of evaporated fuel is grasped by the amount of collected evaporated fuel, and in the case of the engine control device according to the third or fourth aspect, The same effect as will be obtained.

In the engine control device according to the sixth aspect of the present invention, basically the same effect as in the engine control device according to the fourth or fifth aspect can be obtained. Further, when the purge of the evaporated fuel is started at the time of idle when the fuel injection amount is small, the purge flow rate is related to the evaporated fuel amount when the calculation of the amount related to the evaporated fuel amount or the estimated value of the evaporated fuel trap amount is not completed. Amount or evaporated fuel trapped amount (and thus evaporated fuel amount) as the estimated amount of evaporated fuel or the amount of evaporated fuel is gradually increased at a smaller rate than when the engine is off-idle. Is not accurately grasped, it is prevented that the feedback control of the air-fuel ratio is disturbed at the initial stage of purge, and the control stability is further improved.

In the engine control device according to the seventh aspect of the present invention, basically, the same effect as in the engine control device according to any one of the third to sixth aspects can be obtained. To be Further, since the amount related to the evaporated fuel amount or the evaporated fuel trapped amount is estimated based on the average value of the feedback correction values, the calculation of the estimated value becomes easy, and therefore the amount related to the evaporated fuel amount or the evaporated fuel amount is increased. It is possible to quickly calculate the collected amount and thus the evaporated fuel amount. In addition, since the estimated value of the amount of evaporated fuel or the estimated amount of evaporated fuel is updated by learning, the accuracy of the estimated value is improved, and the accuracy of the air-fuel ratio control is improved.

In the engine control device according to the eighth aspect of the present invention, basically, the same effect as that of the engine control device according to any one of the first to seventh aspects is obtained. To be Further, when the duty ratio is less than the predetermined value, the on-time is not increased or decreased according to the duty ratio, but the on-time is made constant and then the operation cycle is expanded or contracted according to the duty ratio to respond to the duty ratio. Since the valve opening degree (valve opening degree) is set, the valve opening degree of the purge valve can be precisely adjusted, and the purge flow control accuracy is improved. Further, since the operation cycle becomes long, the number of times the purge valve is opened and closed is reduced, the life of the purge valve is extended, and the durability of the control device is enhanced.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a configuration of first to eighth aspects of the present invention corresponding to claims 1 to 8.

FIG. 2 is a system configuration diagram of an engine including a control device according to the present invention.

FIG. 3 is a system configuration diagram showing a modified example of an engine including a control device according to the present invention.

FIG. 4 is a block diagram showing a functionalized configuration of a control unit.

FIG. 5 is a flowchart of a main routine of engine control by the control unit.

FIG. 6 is a flowchart of a trap amount calculation subroutine.

FIG. 7 is a flowchart showing a trap amount estimation execution determination subroutine.

FIG. 8 is a flowchart showing a fuel injection amount calculation subroutine.

FIG. 9 is a flowchart showing a purge amount calculation subroutine.

FIG. 10 is a flowchart showing a purge execution determination subroutine.

FIG. 11 is a diagram showing a relationship between an average feedback correction value and an evaporated fuel trap amount.

FIG. 12 is a diagram showing an example of changes over time of a drive duty ratio, a feedback correction value, an average feedback correction value, and a trap amount estimated value.

FIG. 13 is a diagram showing a change with time of a drive duty ratio at the start of canister purging.

FIG. 14 is a diagram showing a change over time in the drive frequency of the purge valve when the drive duty ratio changes.

FIG. 15 is a diagram showing a relationship between a drive duty ratio and a drive frequency.

[Explanation of symbols]

CE, CE '... Engine CU ... Control unit 1 ... Cylinder 4 ... Combustion chamber 9 ... Linear O ₂ sensor 10 ... Intake system 15 ... Fuel injection valve 25 ... Canister 28 ... Purge passage 29 ... Purge valve

Claims (8)

[Claims] 1. An air-fuel ratio detecting means for detecting an air-fuel ratio, and a fuel injection for feedback controlling an amount of fuel injection so that the air-fuel ratio detected by the air-fuel ratio detecting means becomes a target air-fuel ratio in a predetermined operating range. Amount control means, evaporated fuel collecting means for collecting evaporated fuel from the fuel tank, and evaporated fuel for purging the evaporated fuel collected by the evaporated fuel collecting means into the intake system in the predetermined operation region. In an engine control device provided with a purge means and a purge flow rate control means for controlling a purge flow rate of the vaporized fuel to the intake system by the vaporized fuel purge means, the purge flow rate control means has the predetermined operating range. When the purge of the evaporated fuel is executed, the purge flow rate is gradually changed to the target purge flow rate, and the rate of change is increased in the low intake air amount range. The engine control apparatus which is characterized in that is adapted to set smaller than the air amount region. 2. The engine control device according to claim 1, wherein when the purge flow rate control means starts the purge of the evaporated fuel in the predetermined operation region, the purge flow rate is gradually increased to a target purge flow rate. The engine control device is characterized in that the increasing rate is set to be smaller in the low intake air amount range than in the high intake air amount range. 3. An air-fuel ratio detecting means for detecting an air-fuel ratio, and a feedback correction for setting a feedback correction value based on a deviation of the air-fuel ratio detected by the air-fuel ratio detecting means from a target air-fuel ratio in a predetermined operating range. Value setting means, evaporated fuel collecting means for collecting evaporated fuel from the fuel tank, and evaporated fuel for purging the evaporated fuel collected by the evaporated fuel collecting means into the intake system in the predetermined operation region. Purge means, purge flow rate control means for controlling the purge flow rate of the evaporated fuel to the intake system by the evaporated fuel purge means according to the operating state, and combustion based on the feedback correction value set by the feedback correction value setting means. Evaporative fuel amount calculation means for calculating the amount of evaporated fuel released into the room, and the basic fuel injection amount obtained corresponding to the intake air amount The required fuel injection amount is obtained based on the feedback correction value set by the feedback correction value setting means,
By setting the actual fuel injection amount by subtracting the evaporated fuel amount calculated by the evaporated fuel amount calculation means from the required fuel injection amount, the air-fuel ratio detected by the air-fuel ratio detection means becomes the target air-fuel ratio. Therefore, in the engine control device provided with the fuel injection amount control means for feedback-controlling the fuel injection amount, the evaporated fuel amount calculation means is based on the feedback correction value set by the feedback correction value setting means. , An estimated value relating to the amount of evaporated fuel is calculated, and when the purge flow rate control means executes the purge of evaporated fuel in an idle state in the predetermined operating region, the purge flow rate is targeted. While gradually changing to the purge flow rate, the rate of change is calculated by the evaporative fuel amount calculation means after the calculation of the estimated value is completed. When not, the engine control apparatus is characterized in that is adapted to set smaller than when you have completed the calculated and out. 4. The engine control device according to claim 3, wherein when the purge flow rate control means starts the purge of the evaporated fuel in the predetermined operating region, the purge flow rate is gradually increased to a target purge flow rate. The engine control device is characterized in that the increasing rate is set to be smaller in the low intake air amount range than in the high intake air amount range. 5. The engine control device according to claim 3 or 4, wherein the amount related to the amount of evaporated fuel is an amount of collected evaporated fuel. 6. The engine control device according to claim 4 or 5, wherein when the purge flow rate control means starts the purge of evaporated fuel in an off-idle state in the predetermined operating region, The purge flow rate is gradually increased up to the target purge flow rate, and the purge amount increase rate when the evaporated fuel amount calculation means completes the calculation of the estimated value during idle is In the control device of the engine, the evaporated fuel amount calculation means is set to be smaller than the purge amount increase rate when the calculation of the estimated value is not completed. 7. The engine control device according to any one of claims 3 to 6, wherein the evaporated fuel amount calculation means is an average of feedback correction values set by the feedback correction value setting means. An engine control device, wherein an estimated value of the evaporated fuel amount is calculated based on the value, and the estimated value is updated by learning. 8. The engine control device according to any one of claims 1 to 7, wherein the evaporated fuel purge means is a valve body according to a duty ratio applied from the purge flow rate control means. When the duty ratio to be applied to the purge valve is less than a predetermined value, the purge flow rate control means keeps the on time of the purge valve constant and then changes the duty ratio to the duty ratio. According to another aspect of the present invention, the engine control device is characterized in that the operating cycle of the purge valve is expanded and contracted.