JP7269533B2 - Evaporative fuel processing device - Google Patents

Description

The present invention relates to an evaporative fuel processing device for releasing evaporative fuel in a fuel tank into an intake passage of an engine.

Conventionally, the evaporated fuel (evaporative gas) generated in the fuel tank is once adsorbed in the canister, and the evaporated fuel adsorbed in the canister is removed from the purge passage from the engine by opening the purge valve provided in the purge passage. Evaporative fuel processing devices are known that perform evaporative fuel purge control to purge (release) into the intake passage of the engine.

In addition, in order to suppress the deviation of the air-fuel ratio caused by the execution of the vaporized fuel purge control as described above (more specifically, the deviation occurring in the air-fuel ratio feedback control for setting the actual air-fuel ratio to the target air-fuel ratio), the vaporization During fuel purge control, learning about the concentration of evaporated fuel contained in the gas purged into the intake passage (evaporated fuel concentration learning) is performed. Then, correction of the fuel injection amount and fuel vapor purge control are performed based on the learned fuel vapor concentration (learned fuel vapor concentration value).

For example, in Patent Document 1, based on the engine stop time, outside air temperature, fuel temperature, etc., the amount of change in evaporated fuel concentration during the engine stop period is estimated, and based on this change amount, learning before engine stop Techniques for correcting the learned value of evaporated fuel concentration have been disclosed. This technique aims to suppress air-fuel ratio deviation at the start of purge after engine start-up, which is caused by a change in evaporated fuel concentration during an engine stop period. In addition, Patent Documents 2 and 3 disclose techniques related to the present invention.

JP 2009-299627 A JP-A-8-177546 JP-A-5-332208

By the way, characteristics of control valves and sensors related to fuel vapor purge control, specifically, characteristics (error characteristics) such as control errors of control valves, sensor detection errors, and response delays, are disturbances to fuel vapor concentration learning. can act as In a situation where the degree of influence of such disturbance is large, for example, in a situation where the control error of the control valve or the detection error of the sensor becomes large, the accuracy of the evaporated fuel concentration learning tends to decrease. In this situation, the execution of fuel vapor concentration learning is restricted. However, if the execution of fuel vapor concentration learning is restricted, it takes time to ensure the accuracy of the fuel vapor concentration learning value, in other words, it takes time to ensure the reliability of the fuel vapor concentration learning value. While the reliability of the evaporated fuel concentration learned value is not ensured, it is not possible to accurately estimate the concentration of the evaporated fuel introduced into the engine. Therefore, the fuel injection amount cannot be accurately corrected. As a result, the amount of fuel vapor purged is limited, and the fuel vapor cannot be processed quickly.

In particular, at the start of evaporative fuel purging, since the number of executions of evaporative fuel concentration learning is small, the reliability of the evaporative fuel concentration learning value is not ensured, so there is a tendency to limit the purge amount of evaporative fuel. If, at the start of this purge, the execution of fuel vapor concentration learning is restricted due to the disturbance as described above, the purge amount of fuel vapor will be further restricted. In order to quickly process the evaporated fuel, it is desirable to purge as much of the evaporated fuel as possible from the start of the purge of the evaporated fuel. rice field.

SUMMARY OF THE INVENTION The present invention has been made to solve the above-described problems of the prior art. It is an object of the present invention to provide an evaporative fuel processing device capable of appropriately correcting the purge amount of evaporative fuel by learning the concentration.

In order to achieve the above objects, the present invention provides an evaporated fuel processing device, which includes a purge passage extending from a fuel tank toward an intake passage of an engine for purging the evaporated fuel in the fuel tank to the intake passage. a canister provided on the purge passage for absorbing and accumulating evaporated fuel from the fuel tank; a purge valve provided on the purge passage on the downstream side of the canister for opening and closing the purge passage; and controlling at least the purge valve. and a controller configured to perform vaporized fuel purge control for purging the vaporized fuel into the intake passage, the controller configured by an air-fuel ratio sensor that detects the air-fuel ratio of the exhaust gas of the engine. Based on the difference between the detected actual air-fuel ratio and the target air-fuel ratio to be set, a feedback correction amount for correcting the fuel injection amount of the fuel injection valve of the engine in the air-fuel ratio feedback control is obtained, and is related to the evaporated fuel purge control. A variation amount indicating the degree of variation in the air-fuel ratio caused by at least one or more characteristics of one or more control valves and one or more sensors is obtained, and based on the feedback correction amount and the variation amount, An evaporative fuel concentration learning value for learning the concentration of evaporative fuel contained in the gas purged into the intake passage is obtained, and the reliability of the evaporative fuel concentration learning value is determined according to the amount of variation included in the evaporative fuel concentration learning value. The controller corrects the purge amount of gas containing fuel vapor to be purged into the intake passage, and controls the purge valve based on the corrected purge amount. It is characterized in that it is configured to perform correction so that the higher is the purge amount .

In the present invention configured as described above, at least one or more factors of one or more control valves and one or more sensors related to fuel vapor purge control that act as disturbances to fuel vapor concentration learning (Hereinafter referred to as "learning environment factor" as appropriate.) A variation amount indicating the degree of variation in the air-fuel ratio caused by the characteristics of Find the concentration learning value. By obtaining the evaporative fuel concentration learning value in this way, even in a situation where the degree of influence of disturbance on the evaporative fuel concentration learning is relatively large, that is, the amount of variation in the air-fuel ratio caused by the characteristics of the learning environment factors is relatively small. Even in a large situation, fuel vapor concentration learning can be appropriately executed by taking into consideration the amount of variation due to such learning environment factors without limiting the execution of fuel vapor concentration learning. That is, according to the present invention, even if there is variation due to learning environmental factors, the fuel vapor concentration learned value can be obtained with high accuracy by appropriately adding the amount of variation at this time. As a result, according to the present invention, the accuracy of the fuel vapor concentration learning value is improved compared to the case of limiting the execution of the fuel vapor concentration learning in a situation where the degree of influence of disturbance on the fuel vapor concentration learning is large as in the prior art. It is possible to quickly ensure the reliability of the evaporated fuel concentration learned value. Therefore, according to the present invention, the purge amount of vaporized fuel can be increased early, and the vaporized fuel can be quickly processed.

Further, according to the present invention, the purge amount from the purge passage to the intake passage is corrected based on the reliability of the evaporated fuel concentration learning value corresponding to the variation amount described above. Here, in a typical prior art, the purge amount is set according to the number of executions (learning times) of fuel vapor concentration learning. When there is less, the purge amount is decreased. In such conventional technology, even if the number of times of learning is large, if the accuracy of the fuel vapor concentration learning value is low, introduction of a large amount of purge gas may cause a large deviation in the air-fuel ratio. Even if the fuel vapor concentration learning value is small, only a small amount of purge gas can be introduced, although there is no problem even if a large amount of purge gas is introduced. In contrast, according to the present invention, the purge amount is corrected according to the reliability of the fuel vapor concentration learning value. Therefore, when the accuracy of the fuel vapor concentration learning value is low, the purge amount is appropriately set to a relatively small amount. In addition, when the accuracy of the evaporated fuel concentration learning value is high, the purge amount can be appropriately set to a relatively large amount.

In the present invention, preferably, the controller controls the purge valve to gradually increase the purge amount when starting to purge the evaporated fuel, and corrects the gradually increased purge amount based on the reliability of the evaporated fuel concentration learned value. is configured as follows.
According to the present invention configured as described above, the purge amount to be gradually increased at the start of evaporative fuel purging can be appropriately set to an amount corresponding to the reliability of the evaporative fuel concentration learning value. When the reliability of the learned value is high, the purge amount is set large so that the vaporized fuel can be quickly processed.

In the present invention, preferably, the controller is configured to perform correction such that the higher the reliability of the evaporated fuel concentration learned value, the higher the rate of change of the purge amount.
According to the present invention configured as described above, when the reliability of the evaporated fuel concentration learning value is high, the change speed of the purge amount is increased, so that the evaporated fuel can be quickly processed. On the other hand, when the reliability of the evaporated fuel concentration learning value is low, the rate of change of the purge amount is decreased, so that the deviation of the air-fuel ratio due to the introduction of the purge gas can be appropriately suppressed.

In the present invention, preferably, the controller repeatedly obtains the variation amount and the fuel vapor concentration learning value at a predetermined cycle, and based on the variation amount obtained this time, the fuel vapor concentration learning value to be obtained this time. A concentration learning candidate value is obtained, and the fuel vapor concentration learning candidate value obtained this time is weighted and added to the fuel vapor concentration learning value obtained last time to obtain the current fuel vapor concentration learning value. The reliability of the fuel vapor concentration learning value according to the amount of variation included in the fuel vapor concentration learning value obtained this time, and the fuel vapor concentration learning candidate value according to the amount of variation included in the fuel vapor concentration learning candidate value obtained this time. Based on the difference from the reliability of , a weighting coefficient to be applied to the fuel vapor concentration learning candidate value obtained this time is obtained in the weighted addition.
According to the present invention configured as described above, a weighting factor is obtained based on the difference between the reliability of the previous evaporated fuel concentration learning value and the reliability of the current evaporated fuel concentration learning candidate value, and the weighting factor is used to , weighted addition of the current fuel vapor concentration learning candidate value to the previous fuel vapor concentration learning value is performed to obtain the current fuel vapor concentration learning value. As a result, even in a state in which variations due to the characteristics of learning environment factors occur, it is possible to eliminate the effects of such variations as much as possible and obtain the evaporated fuel concentration learning value with high accuracy.

In the present invention, preferably, when the reliability of the fuel vapor concentration learning candidate value obtained this time is higher than the reliability of the fuel vapor concentration learning value obtained last time, the controller The weighting factor is made larger than when the reliability of the concentration learning candidate value is lower than the reliability of the evaporated fuel concentration learning value obtained last time.
According to the present invention configured as described above, it is possible to appropriately set the weighting coefficient according to the magnitude relationship between the reliability of the fuel vapor concentration learning candidate value of this time and the reliability of the fuel vapor concentration learning value of the previous time. can.

In the present invention, preferably, the controller uses two or more variation amounts, converts each of the two or more variation amounts so as to match the unit of the feedback correction amount, and converts the two or more after conversion The fuel vapor concentration learning value is obtained based on the amount obtained by adding up the amount of variation in and the feedback correction amount.
According to the present invention configured in this way, it is possible to obtain the evaporated fuel concentration learning value that appropriately takes into consideration all the variation amounts of two or more learning environment factors.

In the present invention, preferably, the controller uses the amount of variation caused by the characteristics of the purge valve, and is configured to obtain a larger amount of variation regarding the purge valve as the duty ratio applied in the duty control for the purge valve decreases. ing.
According to the present invention configured in this way, it is possible to appropriately obtain the variation amount of the air-fuel ratio caused by the characteristics of the purge valve.

In the present invention, the controller preferably uses the amount of variation caused by the characteristics of the fuel injection valve, and the smaller the pulse width of the control signal output to the fuel injection valve, the greater the amount of variation related to the fuel injection valve. is configured to ask for
According to the present invention configured in this way, it is possible to appropriately obtain the variation amount of the air-fuel ratio caused by the characteristics of the fuel injection valve.

In the present invention, preferably, the controller uses the amount of variation caused by the characteristics of the air-fuel ratio sensor, and is configured to obtain a larger amount of variation related to the air-fuel ratio sensor as the degree of variation in the output signal of the air-fuel ratio sensor increases. It is
According to the present invention configured in this way, it is possible to appropriately obtain the variation amount of the air-fuel ratio caused by the characteristics of the air-fuel ratio sensor.

In the present invention, preferably, the controller uses the amount of variation caused by the characteristics of the intake pressure sensor provided in the intake passage, and the smaller the intake pressure detected by the intake pressure sensor, the smaller the amount of variation related to the intake pressure sensor. Configured to seek large quantities.
According to the present invention configured in this way, it is possible to appropriately obtain the amount of air-fuel ratio variation caused by the characteristics of the intake pressure sensor.

According to the evaporative fuel processing apparatus of the present invention, evaporative fuel purge is performed by learning the concentration of evaporative fuel in consideration of variations in the air-fuel ratio caused by the characteristics of the control valves and sensors related to evaporative fuel purge control. Amounts can be corrected appropriately.

1 is a schematic configuration diagram of an engine according to an embodiment of the present invention; FIG. 1 is a block diagram showing an engine control system according to an embodiment of the present invention; FIG. FIG. 4 is an explanatory diagram of an operating range of the engine according to the embodiment of the present invention; FIG. 4 is an explanatory diagram of characteristics of a purge valve, which is one of learning environment factors; FIG. 3 is an explanatory diagram of a characteristic of a fuel injection valve, which is one of learning environment factors; FIG. 4 is an explanatory diagram of characteristics of an intake pressure sensor, which is one of learning environment factors; FIG. 2 is an explanatory diagram of characteristics of an air-fuel ratio sensor, which is one of learning environment factors; 4 is a flowchart showing fuel vapor concentration learning processing according to the embodiment of the present invention; 4 is a time chart showing an example of changes in various parameters when fuel vapor concentration learning processing is executed according to the embodiment of the present invention; 4 is a map showing a purge amount change speed suppression coefficient according to the embodiment of the present invention; 4 is a time chart showing an example of the rate of change of the purge amount when the purge amount change rate suppression coefficient according to the embodiment of the present invention is applied;

DETAILED DESCRIPTION OF THE INVENTION Evaporative fuel processing devices according to embodiments of the present invention will be described below with reference to the accompanying drawings.

<Engine configuration>
First, with reference to FIGS. 1 and 2, a specific example of an engine to which an evaporated fuel processing apparatus according to an embodiment of the present invention is applied will be described. The engine shown in FIGS. 1 and 2 is merely an example of the engine to which the present invention is applied, and application of the present invention to the engine shown in FIGS. 1 and 2 is not limited.

FIG. 1 is a diagram illustrating the configuration of an engine according to an embodiment of the invention. FIG. 2 is a block diagram showing an engine control system according to an embodiment of the present invention. Note that the intake side in FIG. 1 is on the left side of the paper, and the exhaust side is on the right side of the paper.

In this embodiment, the engine 1 is a gasoline engine that performs partial compression ignition combustion (SPCCI: SPark Controlled Compression Ignition) mounted on a four-wheeled vehicle. Specifically, the engine 1 includes a cylinder block 12 and a cylinder head 13 mounted thereon. A plurality of cylinders 11 are formed inside the cylinder block 12 . Although only one cylinder 11 is shown in FIG. 1, the engine 1 is a multi-cylinder engine in this embodiment.

A piston 3 is slidably inserted in each cylinder 11 . Piston 3 is connected to crankshaft 15 via connecting rod 14 . Piston 3 defines combustion chamber 17 together with cylinder 11 and cylinder head 13 . The term "combustion chamber" is not limited to the space formed when the piston 3 reaches compression top dead center. The term "combustion chamber" may be used broadly. In other words, the "combustion chamber" may mean the space formed by the piston 3, the cylinder 11 and the cylinder head 13 regardless of the position of the piston 3.

The geometric compression ratio of the engine 1 is set high for the purpose of improving theoretical thermal efficiency and stabilizing CI (Compression Ignition) combustion, which will be described later. Specifically, the geometric compression ratio of the engine 1 is 17 or higher. The geometric compression ratio may be 18, for example. The geometric compression ratio may be appropriately set within the range of 17 or more and 20 or less.

Two intake ports 18 ( FIG. 1 ) are formed in the cylinder head 13 for each cylinder 11 . The intake port 18 communicates with the combustion chamber 17 . An intake valve 21 is arranged in the intake port 18 . The intake valve 21 opens and closes between the combustion chamber 17 and the intake port 18 . The intake valve 21 is opened and closed at predetermined timing by an intake VVT (Variable Valve Timing) 23 (FIG. 2), which is a variable valve mechanism. The intake VVT 23 is configured to continuously change the rotation phase of the intake camshaft within a predetermined angular range. Thereby, the opening timing and the closing timing of the intake valve 21 can be changed continuously. The intake VVT 23 is configured to be driven electrically or hydraulically.

The cylinder head 13 is also formed with two exhaust ports 19 (FIG. 1) for each cylinder 11 . The exhaust port 19 communicates with the combustion chamber 17 . An exhaust valve 22 is arranged in the exhaust port 19 . The exhaust valve 22 opens and closes between the combustion chamber 17 and the exhaust port 19 . The exhaust valve 22 is opened and closed at predetermined timings by an exhaust VVT 24 (FIG. 2), which is a variable valve mechanism. The exhaust VVT 24 is configured to continuously change the rotational phase of the exhaust camshaft within a predetermined angular range. Thereby, the opening timing and the closing timing of the exhaust valve 22 can be changed continuously. It should be noted that the exhaust VVT 24 is configured to be electrically or hydraulically driven.

In this embodiment, the engine 1 can adjust the length of the overlap period between the opening of the intake valve 21 and the opening of the exhaust valve 22 using the intake VVT 23 and the exhaust VVT 24 . As a result, residual gas in the combustion chamber 17 is scavenged, and hot burned gas is confined in the combustion chamber 17 (in other words, internal EGR (Exhaust Gas Recirculation) gas is introduced into the combustion chamber 17). be able to. It should be noted that the introduction of such internal EGR gas is not limited to the VVT.

A fuel injection valve 6 is attached to the cylinder head 13 for each cylinder 11 . The fuel injection valve 6 is provided on the ceiling surface of the combustion chamber 17 so as to directly inject fuel into the combustion chamber 17 . Further, the fuel injection valve 6 is arranged such that its injection axis is along the central axis of the cylinder 11 . Note that the injection axis of the fuel injection valve 6 does not have to coincide with the central axis of the cylinder 11 . Although not shown in detail, the fuel injection valve 6 is a multi-orifice type fuel injection valve having a plurality of orifices, and injects fuel so that the fuel spray spreads radially from the center of the combustion chamber 17. It should be noted that the fuel injection valve 6 is not limited to a multi-hole type injector. The fuel injection valve 6 may employ an outward opening type injector.

Fuel is supplied to the fuel injection valve 6 from a fuel tank 63 through a fuel supply passage (not shown). This fuel supply path is provided with a fuel pump and a common rail (not shown). A fuel pump is configured to pump fuel to the common rail. The common rail is configured to store fuel pumped from the fuel pump at high fuel pressure. When the fuel injection valve 6 is opened, the fuel stored in the common rail is injected into the combustion chamber 17 from the injection port of the fuel injection valve 6 . For example, fuel with a high pressure of 30 MPa or higher (in one example, the maximum fuel pressure is about 120 MPa) is supplied to the fuel injection valve 6 . The fuel injection valve 6 is pulse-driven, and the greater the pulse width of the supplied control signal, the greater the amount of fuel to be injected.

A spark plug 25 is attached to the cylinder head 13 for each cylinder 11 . A spark plug 25 forcibly ignites the air-fuel mixture in the combustion chamber 17 . The spark plug 25 is arranged on the intake side across the center axis of the cylinder 11 in this embodiment. Also, the spark plug 25 is positioned between the two intake ports 18 . The spark plug 25 is attached to the cylinder head 13 so as to be tilted from top to bottom toward the center of the combustion chamber 17 . The electrode of the ignition plug 25 faces the inside of the combustion chamber 17 and is positioned near the ceiling surface of the combustion chamber 17 .

As shown in FIG. 1 , an intake passage 40 is connected to one side surface of the engine 1 . The intake passage 40 communicates with the intake port 18 of each cylinder 11 . The intake passage 40 is a passage through which gas to be introduced into the combustion chamber 17 flows. An air cleaner 41 for filtering fresh air is provided at the upstream end of the intake passage 40 . A surge tank 42 is arranged near the downstream end of the intake passage 40 . The intake passage 40 downstream of the surge tank 42 constitutes an independent passage that branches for each cylinder 11, although detailed illustration is omitted. A downstream end of the independent passage is connected to the intake port 18 of each cylinder 11 .

A throttle valve 43 is arranged between the air cleaner 41 and the surge tank 42 in the intake passage 40 . The throttle valve 43 is configured to adjust the amount of fresh air introduced into the combustion chamber 17 by adjusting the opening of the valve.

A supercharger 44 is also arranged in the intake passage 40 downstream of the throttle valve 43 . The supercharger 44 is configured to supercharge the gas introduced into the combustion chamber 17 . In this embodiment, the supercharger 44 is a mechanical supercharger driven by the engine 1 . The mechanical supercharger 44 may be of the Roots type, for example. The configuration of the mechanical supercharger 44 may be of any configuration. The mechanical supercharger 44 may be of Lysholm type or centrifugal type.

An electromagnetic clutch 45 is interposed between the supercharger 44 and the output shaft of the engine 1 . Between the supercharger 44 and the engine 1 , the electromagnetic clutch 45 transmits driving force from the engine 1 to the supercharger 44 or cuts off transmission of the driving force. As will be described later, the ECU 10 (FIG. 2) switches the electromagnetic clutch 45 between the connected state and the non-connected state, thereby switching the turbocharger 44 between on and off. That is, the engine 1 can switch between supercharging the gas introduced into the combustion chamber 17 by the supercharger 44 and not supercharging the gas introduced into the combustion chamber 17 by the supercharger 44 . configured to allow.

An intercooler 46 is arranged downstream of the supercharger 44 in the intake passage 40 . Intercooler 46 is configured to cool the gas compressed in supercharger 44 . The intercooler 46 may be configured to be water-cooled, for example.

A bypass passage 47 is connected to the intake passage 40 . The bypass passage 47 connects the upstream portion of the supercharger 44 and the downstream portion of the intercooler 46 in the intake passage 40 so as to bypass the supercharger 44 and the intercooler 46 . The bypass passage 47 is provided with an air bypass valve 48 as a bypass control valve. The air bypass valve 48 adjusts the flow rate of gas flowing through the bypass passage 47 .

When the supercharger 44 is turned off (that is, when the electromagnetic clutch 45 is disconnected), the air bypass valve 48 is fully opened. As a result, the gas flowing through the intake passage 40 bypasses the supercharger 44 and is introduced into the combustion chamber 17 of the engine 1 . The engine 1 operates in a non-supercharged state, that is, in a naturally aspirated state. On the other hand, when the supercharger 44 is turned on (that is, when the electromagnetic clutch 45 is connected), part of the gas that has passed through the supercharger 44 passes through the bypass passage 47 and flows upstream of the supercharger. flow back into By adjusting the degree of opening of the air bypass valve 48, the reverse flow rate can be adjusted, so the boost pressure of the gas introduced into the combustion chamber 17 can be adjusted. In this configuration example, a supercharging system 49 is configured by the supercharger 44 , the bypass passage 47 and the air bypass valve 48 .

Further, the intake passage 40 is provided with a swirl control valve (not shown) for enhancing the flow of intake air within the combustion chamber 17 . An intake passage 40 connected to each combustion chamber 17 of the engine 1 is composed of two passages extending in parallel. An intake passage 40 connected to one of the ports 18 is provided on the rear side of the engine 1 and connected to the other of the two intake ports 18 . The swirl control valve is provided in the intake passage 40 on the front side. When the opening of the swirl control valve is small, the amount of intake air flowing into the combustion chamber 17 from one intake port 18 relatively decreases, and the amount of intake air flowing into the combustion chamber 17 from the other intake port 18 relatively increases. Therefore, the swirl flow in the combustion chamber 17 becomes strong. On the other hand, when the opening of the swirl control valve is large, the amount of intake air flowing into the combustion chamber 17 from each of the two intake ports 18 becomes substantially equal, and the swirl flow in the combustion chamber 17 weakens. Such a swirl control valve is also controlled by the ECU 10 (Fig. 2).

An exhaust passage 50 is connected to the other side surface of the engine 1 . The exhaust passage 50 communicates with the exhaust port 19 of each cylinder 11 . The exhaust passage 50 is a passage through which the exhaust gas discharged from the combustion chamber 17 flows. An upstream portion of the exhaust passage 50 constitutes an independent passage that branches for each cylinder 11, although detailed illustration is omitted. The upstream end of the independent passage is connected to the exhaust port 19 of each cylinder 11 . An exhaust gas purification system having one or more catalytic converters 51 is arranged in the exhaust passage 50 . Catalytic converter 51 includes a three-way catalyst. Note that the exhaust gas purification system is not limited to including only the three-way catalyst.

An EGR passage 52 that constitutes an external EGR system is connected between the intake passage 40 and the exhaust passage 50 . The EGR passage 52 is a passage for recirculating part of the burned gas to the intake passage 40 . An upstream end of the EGR passage 52 is connected downstream of the catalytic converter 51 in the exhaust passage 50 . A downstream end of the EGR passage 52 is connected upstream of the supercharger 44 in the intake passage 40 .

A water-cooled EGR cooler 53 is arranged in the EGR passage 52 . The EGR cooler 53 is configured to cool the burned gas. An EGR valve 54 is also arranged in the EGR passage 52 . The EGR valve 54 is configured to adjust the flow rate of the burned gas flowing through the EGR passage 52 . By adjusting the opening of the EGR valve 54, the recirculation amount of the cooled burned gas, that is, the external EGR gas can be adjusted.

In this embodiment, the EGR system 55 is composed of an external EGR system including the EGR passage 52 and the EGR valve 54, and an internal EGR system including the intake VVT 23 and the exhaust VVT 24 described above. ing.

Further, as shown in FIG. 1 , the engine 1 is provided with a purge system 61 for purging vaporized fuel generated in the fuel tank 63 to the intake passage 40 . Specifically, the purge system 61 is supplied with vaporized fuel that has evaporated in a fuel tank 63, and includes a canister 64 that absorbs the vaporized fuel, an air release passage 65 that introduces air into the canister 64, and the fuel tank 63. and a purge passage 66 that connects with the intake passage 40 via a canister 64 . The purge passage 66 is connected to a position on the intake passage 40 between the throttle valve 43 and the supercharger 44 .

The vaporized fuel adsorbed in the canister 64 is desorbed from the canister 64 by the air introduced from the atmosphere release passage 65 . The vaporized fuel desorbed from the canister 64 is purged into the intake passage 40 through the purge passage 66 together with air. Hereinafter, gas containing vaporized fuel and air purged from the purge passage 66 to the intake passage 40 may be referred to as "purge gas".

A purge valve 67 for opening and closing the purge passage 66 is provided on the purge passage 66 on the downstream side of the canister 64 . The purge valve 67 is a duty control valve that is repeatedly opened and closed, and its opening is controlled by changing the duty ratio, which is the ratio of the valve opening period to the unit period of one valve opening period and one valve closing period. The degree is changed. The purge valve 67 is an electromagnetic valve, and the duty ratio is the ratio of the energized period to the unit period including one energized period and one non-energized period. The purge valve 67 is fully closed when the duty ratio is 0% and fully open when the duty ratio is 100%.

Next, as shown in FIG. 2, the engine 1 has an ECU (Power-train Control Module) 10 for operating it. The ECU 10 is a controller based on a well-known microcomputer, and includes a microprocessor 10a as a central processing unit (CPU) for executing programs, and a RAM (Random Access Memory) or ROM (Read Only). Memory 10b for storing programs and data, and an input/output bus for inputting/outputting electrical signals. The ECU 10 is an example of a controller.

Various sensors SW1 to SW15 are connected to the ECU 10 as shown in FIGS. The sensors SW1 to SW15 output detection signals to the ECU 10. FIG. The sensors include the following sensors.

That is, an air flow sensor SW1 arranged downstream of the air cleaner 41 in the intake passage 40 and detecting the flow rate of fresh air flowing through the intake passage 40, a first intake air temperature sensor SW2 detecting the temperature of the fresh air, and the intake passage. A first pressure sensor SW3 ("intake pressure sensor", and hereinafter referred to as "intake pressure sensor SW3" as appropriate) is arranged downstream of the supercharger 44 in the intake passage 40 and upstream of the connection position of the bypass passage 47, and the supercharger 44 A second pressure sensor SW5 attached to the surge tank 42 and detecting the pressure (supercharging pressure) of the gas downstream of the supercharger 44, each cylinder 11, a finger pressure sensor SW6 is attached to the cylinder head 13 and detects the pressure in each combustion chamber 17 (in-cylinder pressure); An exhaust temperature sensor SW7 for detecting, an air-fuel ratio sensor SW8 for detecting the air-fuel ratio of the exhaust gas discharged from the combustion chamber 17 (more specifically, the sensor SW8 is a linear O ₂ sensor for detecting the oxygen concentration contained in the exhaust gas ( linear A/F sensor: corresponding to LAFS), an engine speed sensor SW9 arranged near the output shaft of the engine 1 and detecting the speed of the output shaft, attached to the engine 1 and detecting the temperature of the cooling water. A water temperature sensor SW10 for detection, a crank angle sensor SW11 attached to the engine 1 and detecting the rotation angle of the crankshaft 15, and an accelerator attached to the accelerator pedal mechanism and detecting the accelerator opening corresponding to the operation amount of the accelerator pedal. An opening sensor SW12, an intake cam angle sensor SW13 attached to the engine 1 and detecting the rotation angle of the intake camshaft, an exhaust cam angle sensor SW14 attached to the engine 1 and detecting the rotation angle of the exhaust camshaft, and EGR. An EGR differential pressure sensor SW15 is arranged in the passage 52 and detects a differential pressure upstream and downstream of the EGR valve 54 .

Based on these detection signals, the ECU 10 determines the operating state of the engine 1 and calculates control amounts for each device. The ECU 10 sends a control signal related to the calculated control amount to the fuel injection valve 6, the spark plug 25, the intake VVT 23, the exhaust VVT 24, the throttle valve 43, the EGR valve 54, the electromagnetic clutch 45 of the supercharger 44, the air bypass valve 48 , and the purge valve 67 to control the engine 1 . For example, the ECU 10 adjusts the opening degree of the air bypass valve 48 based on the differential pressure across the supercharger 44 obtained from the detection signals of the first pressure sensor SW3 and the second pressure sensor SW5, thereby increasing the supercharging pressure. adjust. Further, the ECU 10 adjusts the opening degree of the EGR valve 54 based on the differential pressure across the EGR valve 54 obtained from the detection signal of the EGR differential pressure sensor SW15, thereby adjusting the amount of external EGR gas to be introduced into the combustion chamber 17. to adjust.

In this embodiment, the purge passage 66, the canister 64, the purge valve 67, and the ECU 10 mainly constitute the "evaporative fuel processing device" according to the present invention.

<Operating range>
Next, referring to FIG. 3, the operating range of the engine 1 according to the embodiment of the invention will be described. FIG. 3 shows an operation map that is applied when the engine 1 has been warmed up and is warm (for example, when "engine water temperature ≧80° C." or "intake air temperature ≧50° C.").

As shown in FIG. 3, general SI (Spark Ignition) combustion is performed in a region R1 where the engine speed is relatively low and a region R2 where the engine speed is relatively high. SI combustion is a form in which the air-fuel mixture is ignited by spark ignition using the spark plug 25, and the air-fuel mixture is forcibly burned by flame propagation that expands the combustion area from the ignition point to the surroundings. In order to realize such SI combustion, main components of the engine 1 are controlled by the ECU 10 as follows. The fuel injection valve 6 injects injection over a predetermined period that overlaps at least the intake stroke. For example, the fuel injection valve 6 injects fuel over a series of periods from the intake stroke to the compression stroke. Also, the spark plug 25 ignites the air-fuel mixture near the compression top dead center. For example, the ignition plug 25 ignites the air-fuel mixture at a timing slightly advanced from the compression top dead center. This ignition triggers SI combustion, and all of the air-fuel mixture in the combustion chamber 17 is burned by flame propagation. The opening of the EGR valve 54 is controlled so that the air-fuel ratio, which is the ratio of the amount of air in the combustion chamber 17 to the amount of fuel, becomes substantially the stoichiometric air-fuel ratio (14.7).

On the other hand, in regions R3, R4, and R5 sandwiched between the regions R1 and R2 with respect to the engine speed, partial compression ignition combustion (SPCCI combustion) in which SI combustion and CI combustion are mixed is performed. CI combustion is a mode in which an air-fuel mixture is combusted by self-ignition under an environment of high temperature and high pressure due to compression of the piston 3 . SPCCI combustion, which is a mixture of SI combustion and CI combustion, involves SI combustion of part of the air-fuel mixture in the combustion chamber 17 by spark ignition performed in an environment just before the air-fuel mixture self-ignites, and the SI combustion After that, the remaining air-fuel mixture in the combustion chamber 17 is subjected to CI combustion by self-ignition (due to further increase in temperature and pressure accompanying SI combustion).

SPCCI combustion has the property that heat release during CI combustion is steeper than heat release during SI combustion. For example, in the waveform of the heat release rate due to SPCCI combustion, the rising slope at the beginning of combustion corresponding to SI combustion is smaller than the rising slope corresponding to subsequent CI combustion. In other words, the waveform of the heat release rate during SPCCI combustion has a heat release rate portion with a relatively small rising slope formed by SI combustion and a heat release rate portion with a relatively large rising slope formed by CI combustion. are formed so as to be continuous in this order. Further, in accordance with such a tendency of the heat release rate, in SPCCI combustion, the pressure rise rate (dp/dθ) in the combustion chamber 17 that occurs during SI combustion is smaller than that during CI combustion.

When the temperature and pressure in the combustion chamber 17 increase due to SI combustion, the unburned air-fuel mixture self-ignites and CI combustion starts. At the timing of this self-ignition (that is, the timing at which CI combustion starts), the gradient of the heat release rate waveform changes from small to large. That is, the waveform of the heat release rate in SPCCI combustion has an inflection point that appears at the timing when CI combustion starts.

After the start of CI combustion, SI combustion and CI combustion are performed in parallel. In CI combustion, the combustion speed of air-fuel mixture is faster than in SI combustion, so the heat generation rate is relatively large. However, since CI combustion is performed after compression top dead center, the slope of the heat release rate waveform does not become excessive. That is, when the compression top dead center is passed, the motoring pressure drops due to the descent of the piston 3, which suppresses the increase in the heat release rate, thereby avoiding an excessive increase in dp/dθ during CI combustion. be. In this way, in SPCCI combustion, CI combustion is performed after SI combustion. ), combustion noise can be suppressed.

SPCCI combustion also ends with the end of CI combustion. Since the CI combustion has a higher combustion speed than the SI combustion, it is possible to advance the combustion end timing as compared to the simple SI combustion (when all the fuel is SI-burned). In other words, in SPCCI combustion, the end of combustion can be brought closer to compression top dead center in the expansion stroke. As a result, SPCCI combustion can improve fuel consumption performance compared to simple SI combustion.

Here, in regions R3 and R4, SPCCI combustion is performed with the air-fuel ratio in the combustion chamber 17 set to approximately the stoichiometric air-fuel ratio (14.7) or a value slightly smaller than the stoichiometric air-fuel ratio. In other words, in the regions R3 and R4, the excess air ratio λ (value obtained by dividing the actual air-fuel ratio by the stoichiometric air-fuel ratio) is approximately 1 or slightly smaller than 1, and the environment is rich in air-fuel ratio (λ≈1 or λ≦1). SPCCI combustion takes place at . In such regions R3 and R4, the main components of the engine 1 are controlled by the ECU 10 as follows. The fuel injection valve 6 advances the injection timing of at least part of the fuel to the intake stroke. For example, the fuel injection valve 6 performs the first fuel injection during the intake stroke and the second fuel injection during the compression stroke. The spark plug 25 ignites the air-fuel mixture near the compression top dead center. For example, the ignition plug 25 ignites the air-fuel mixture at a timing slightly advanced from the compression top dead center. Then, with this ignition as a trigger, SPCCI combustion is started, a part of the air-fuel mixture in the combustion chamber 17 burns due to flame propagation (SI combustion), and then the remaining air-fuel mixture burns due to self-ignition (CI combustion). .

The intake VVT 23 and the exhaust VVT 24 set the valve timings of the intake valve 21 and the exhaust valve 22 to the timing for performing internal EGR, that is, both the intake valve 21 and the exhaust valve 22 are opened across the exhaust top dead center. The timing is set so that a sufficient valve overlap period is formed. As a result, the internal EGR that causes the burned gas to remain in the combustion chamber 17 is realized, and the temperature of the combustion chamber 17 (initial temperature before compression) is increased. Specifically, in regions R3 and R4, the intake VVT 23 closes the intake valve 21 earlier than SI combustion, and the exhaust VVT 24 closes the exhaust valve 22 later than SI combustion. The throttle valve 43 is closed to a predetermined intermediate opening, and the overall air-fuel ratio in the combustion chamber 17 is set to approximately the stoichiometric air-fuel ratio or a value slightly smaller than the stoichiometric air-fuel ratio.

The opening of the EGR valve 54 is controlled so that the overall air-fuel ratio in the combustion chamber 17 becomes the target air-fuel ratio. Basically, the EGR valve 54 removes the amount of air corresponding to the target air-fuel ratio and the amount of burned gas left in the combustion chamber 17 by internal EGR from the total amount of gas introduced into the combustion chamber 17. The flow rate in the EGR passage 52 is adjusted so that the corresponding amount of gas is recirculated from the EGR passage 52 to the combustion chamber 17 as the external EGR gas.

Further, in the region R3 on the high load side of the regions R3 and R4, the supercharger 44 performs supercharging by connecting the supercharger 44 and the engine 1 by engaging the electromagnetic clutch 45 . At this time, the degree of opening of the air bypass valve 48 is controlled so that the supercharging pressure detected by the second pressure sensor SW5 matches the predetermined target pressure for each operating condition (rotational speed/load). . For example, as the degree of opening of the air bypass valve 48 increases, the flow rate of intake air flowing back to the upstream side of the supercharger 44 through the bypass passage 47 increases, resulting in the pressure of the intake air introduced into the surge tank 42, that is, the supercharging pressure. lower. The air bypass valve 48 controls the boost pressure to the target pressure by adjusting the reverse flow rate of the intake air in this manner. On the other hand, in the region R4 on the low load side, the electromagnetic clutch 45 is released to disconnect the supercharger 44 from the engine 1, and the air bypass valve 48 is fully opened. Supercharging is stopped.

On the other hand, in region R5, which is on the low speed side and low load side of region R4, SPCCI combustion is performed by setting the air-fuel ratio in combustion chamber 17 to a value greater than the stoichiometric air-fuel ratio (14.7). In other words, in region R5, SPCCI combustion is performed in an environment where the air-fuel ratio is lean (λ>1) where the excess air ratio λ is greater than 1. In one example, the excess air ratio λ is set to 2 or more. In such a region R5, the main components of the engine 1 are controlled by the ECU 10 as follows. The fuel injection valve 6 injects all or most of the fuel to be injected during one cycle during the compression stroke. For example, the fuel injection valve 6 injects the fuel in two parts from the middle to the latter part of the compression stroke. The spark plug 25 ignites the air-fuel mixture near the compression top dead center. For example, the ignition plug 25 ignites the air-fuel mixture at a timing slightly advanced from the compression top dead center. Then, with this ignition as a trigger, SPCCI combustion is started, a part of the air-fuel mixture in the combustion chamber 17 burns due to flame propagation (SI combustion), and then the remaining air-fuel mixture burns due to self-ignition (CI combustion). .

The intake VVT 23 and the exhaust VVT 24 set the valve timings of the intake valve 21 and the exhaust valve 22 to the timing for performing internal EGR. The timing is set so that a sufficient overlap period is formed. As a result, the internal EGR that causes the burned gas to remain in the combustion chamber 17 is realized, and the temperature of the combustion chamber 17 (initial temperature before compression) is increased. Specifically, in the region R5, the intake VVT 23 closes the intake valve 21 earlier than SI combustion, and the exhaust VVT 24 closes the exhaust valve 22 later than SI combustion. The throttle valve 43 is opened to an opening corresponding to full opening, and the overall air-fuel ratio in the combustion chamber 17 is set to 30-40. Further, in region R5, the electromagnetic clutch 45 is released to disconnect the supercharger 44 from the engine 1, and the air bypass valve 48 is fully opened to stop supercharging by the supercharger 44. be done.

<Evaporative fuel concentration learning>
Next, fuel vapor concentration learning according to the embodiment of the present invention will be described. This evaporated fuel concentration learning is performed to suppress the deviation of the air-fuel ratio caused by the purge of the evaporated fuel (more specifically, the deviation occurring in the air-fuel ratio feedback control for setting the actual air-fuel ratio to the target air-fuel ratio). During purge control, the concentration of evaporated fuel contained in the purge gas purged into the intake passage 40 is learned. Correction of fuel injection amount and fuel vapor purge control are performed based on the concentration of fuel vapor thus learned (learned fuel vapor concentration value).

Here, the characteristics of the control valves and sensors related to fuel vapor purge control, specifically the characteristics (error characteristics) such as the control error of the control valve, the detection error of the sensor, and the response delay, are related to the fuel vapor concentration learning. It may act as a disturbance. Naturally, it also becomes a disturbance for the air-fuel ratio feedback control that is performed based on the evaporated fuel concentration learning. In a situation where the degree of influence of such disturbance is large, for example, in a situation where the control error of the control valve or the detection error of the sensor becomes large, the accuracy of fuel vapor concentration learning will be reduced if this disturbance is not taken into account. tend to decline. Therefore, in the conventional technology, the execution of fuel vapor concentration learning is restricted (for example, prohibited) in a situation where the degree of influence of disturbance is large, but then it takes time to ensure the accuracy of the fuel vapor concentration learning value. While the accuracy of the evaporated fuel concentration learning value is not ensured, it is not possible to accurately estimate the concentration of the evaporated fuel introduced into the engine 1. Therefore, the deviation of the air-fuel ratio caused by the purge of the evaporated fuel is suppressed. Therefore, the fuel injection amount cannot be accurately corrected. Therefore, by restricting the purge amount of evaporated fuel, the evaporated fuel in the fuel tank 63 and the purge system 61 cannot be quickly processed.

Therefore, in the present embodiment, in order to deal with the above problem, even in a situation where the degree of influence of disturbance on fuel vapor concentration learning is large, the degree of influence of this disturbance is determined without limiting the execution of fuel vapor concentration learning. is taken into account to learn the evaporated fuel concentration. Specifically, the ECU 10 determines the amount of variation (corresponding to probability distribution characteristics) that indicates the degree of variation due to the characteristics of factors (learning environment factors) such as control valves and sensors that act as disturbances on the fuel vapor concentration learning. ) is obtained, and fuel vapor concentration learning is performed based on this amount of variation. The amount of variation due to learning environmental factors causes variation in the air-fuel ratio (that is, deviation from the target air-fuel ratio applied in air-fuel ratio feedback control).

Next, the learning environment factors will be specifically described with reference to FIGS. 4 to 7. FIG. FIG. 4 is an explanatory diagram of the characteristics of the purge valve 67, which is one of the learning environment factors, and FIG. 5 is an explanatory diagram of the characteristics of the fuel injection valve 6, which is one of the learning environment factors. 6 is an explanatory diagram of the characteristics of the intake pressure sensor SW3, which is one of the learning environment factors, and FIG. 7 is an explanatory diagram of the characteristics of the air-fuel ratio sensor SW8, which is one of the learning environment factors.

4, the horizontal axis indicates the duty ratio applied to the duty control of the purge valve 67, and the vertical axis indicates the flow rate error of the purge gas purged from the purge passage 66 to the intake passage 40. As shown in FIG. As shown in FIG. 4, the purge valve 67 has a characteristic that the smaller the duty ratio (for example, the region where the duty ratio is 10% or less), the larger the flow rate error of the purge gas. There is a characteristic that Variations caused by the characteristics of the purge valve 67 act as disturbances to the evaporated fuel concentration learning. Therefore, in the present embodiment, the ECU 10 takes into consideration the variation caused by the characteristics of the purge valve 67 and performs fuel vapor concentration learning. Specifically, the smaller the duty ratio applied to the purge valve 67, the greater the ECU 10 sets the variation amount indicating the degree of variation regarding the purge valve 67, and the fuel vapor concentration learning value is obtained based on this variation amount.

5, the horizontal axis represents the pulse width of the control signal (corresponding to the amount of fuel injected from the fuel injection valve 6) supplied to the fuel injection valve 6 in pulse drive, and the vertical axis represents the shows the injection amount error. As shown in FIG. 5, the smaller the pulse width of the control signal to the fuel injection valve 6, in other words, the smaller the fuel injection amount from the fuel injection valve 6 (for example, the region where the fuel injection amount is 6 mg or less), There is a characteristic that the injection amount error from the fuel injection valve 6 increases, that is, a characteristic that the variation in the fuel injection amount increases. Such variations caused by the characteristics of the fuel injection valve 6 act as disturbances on the evaporated fuel concentration learning. Therefore, in the present embodiment, the ECU 10 takes into consideration the variation caused by the characteristics of the fuel injection valve 6 and performs fuel vapor concentration learning. Specifically, the smaller the pulse width of the control signal to be output to the fuel injection valve 6, the larger the variation amount indicating the degree of variation regarding the fuel injection valve 6 is set by the ECU 10. Find the learning value.

In FIG. 6, the horizontal axis represents the intake pressure detected by the intake pressure sensor SW3 (the pressure in the intake passage 40 between the throttle valve 43 and the supercharger 44). The vertical axis represents the flow rate error of the purge gas purged into the intake passage 40 by the evaporated fuel purge control executed based on the intake pressure detected by the intake pressure sensor SW3. ing. As shown in FIG. 6, the lower the intake pressure, in other words, the lower the negative pressure (for example, in the region where the negative pressure is 5 kPa or less), the greater the purge gas flow rate error due to the fuel vapor purge control. Variation increases. This is because the intake pressure sensor SW3 has a characteristic that the smaller the intake pressure (negative pressure), the larger the detection error of the intake pressure sensor SW3. Such variations caused by the characteristics of the intake pressure sensor SW3 act as disturbances on the fuel vapor concentration learning. Therefore, in the present embodiment, the ECU 10 performs fuel vapor concentration learning in consideration of variations caused by the characteristics of the intake pressure sensor SW3. Specifically, the smaller the intake pressure detected by the intake pressure sensor SW3, the greater the ECU 10 sets the amount of variation indicating the degree of variation regarding the intake pressure sensor SW3. Ask for

FIG. 7 shows a graph schematically showing the influence of disturbance (for example, introduction of purge gas by the purge system 61) on the detection of the air-fuel ratio sensor SW8 in the upper part, and the output signal of the air-fuel ratio sensor SW8 (that is, The graph schematically shows the air-fuel ratio detected by the air-fuel ratio sensor SW8, and the lower graph schematically shows the feedback correction amount used to correct the fuel injection amount of the fuel injection valve 6 in the air-fuel ratio feedback control. , a graph is shown. The air-fuel ratio feedback control is based on the air-fuel ratio (actual air-fuel ratio) detected by the air-fuel ratio sensor SW8, so that the target air-fuel ratio (typically the theoretical air-fuel ratio) is achieved. The feedback correction amount is used to correct the fuel injection amount in this air-fuel ratio feedback control, and is basically an amount corresponding to the deviation of the actual air-fuel ratio from the target air-fuel ratio (for example, (expressed as a ratio of the difference between the target air-fuel ratio and the actual air-fuel ratio to the target air-fuel ratio).

As shown in FIG. 7, before time t11, when a disturbance such as depression of the accelerator pedal (increase in fuel injection amount) or start of introduction of purge gas occurs, the output signal of the air-fuel ratio sensor SW8 greatly fluctuates to the rich side. Along with this (see arrow A11), the feedback correction amount also fluctuates greatly (see arrow A12). After time t11, the output signal of the air-fuel ratio sensor SW8 becomes constant and stabilizes (specifically, the target air-fuel ratio is reached), and the feedback correction amount also becomes constant and stabilizes. For this reason, the air-fuel ratio does not converge to the target air-fuel ratio until time t11, and the feedback correction amount may not be accurate. It converges to the fuel ratio, and it can be said that the feedback correction amount is accurate.

Such a state in which the air-fuel ratio does not converge to the target air-fuel ratio, that is, a state in which the degree of variation in the output signal of the air-fuel ratio sensor SW8 is large, is caused by characteristics such as response delay of the air-fuel ratio sensor SW8. . The greater the degree of variation in the output signal of the air-fuel ratio sensor SW8, the greater the variation in the feedback correction amount. Therefore, variations caused by the characteristics of the air-fuel ratio sensor SW8 act as disturbances on the fuel vapor concentration learning. Therefore, in the present embodiment, the ECU 10 performs fuel vapor concentration learning in consideration of variations caused by the characteristics of the air-fuel ratio sensor SW8. Specifically, the greater the degree of variation in the output signal of the air-fuel ratio sensor SW8, the greater the ECU 10 sets the amount of variation indicating the degree of variation regarding the air-fuel ratio sensor SW8, and based on this amount of variation, the learned value of evaporated fuel concentration is calculated. Ask for

It should be noted that the ECU 10 performs fuel vapor concentration learning using the amount of variation of at least one of the four learning environment factors as described above. That is, the amount of variation in all four learning environment factors may be used, the amount of variation in two or three learning environment factors may be used, or the amount of variation in only one learning environment factor may be used. . Here, when two or more variation amounts of learning environment factors are used, the ECU 10 converts each of the two or more variation amounts into a feedback correction amount unit (for example, a ratio representing a deviation from the target air-fuel ratio). Transformed so as to be combined with each other, the evaporated fuel concentration learning is performed using the amount obtained by summing the two or more variation amounts after this transformation.

Next, fuel vapor concentration learning according to the embodiment of the present invention will be specifically described with reference to FIGS. 8 to 11. FIG. FIG. 8 is a flow chart showing fuel vapor concentration learning processing according to the embodiment of the present invention. This evaporated fuel concentration learning process is repeatedly executed by the ECU 10 at a predetermined cycle. FIG. 9 is a time chart showing an example of changes in various parameters when the evaporated fuel concentration learning process is executed according to the embodiment of the present invention. FIG. 10 is a map showing a coefficient (purge amount change speed suppression coefficient) for suppressing the change speed of the purge amount in the embodiment of the present invention, and FIG. 5 is a time chart showing an example of change speed of the purge amount when a change speed suppression coefficient is applied;

8 is started, the ECU 10 opens the purge valve 67 to start introducing purge gas from the purge passage 66 into the intake passage 40 in step S1.

Next, in step S2, the ECU 10 determines whether or not a condition (learning permitting condition) permitting execution of fuel vapor concentration learning has been established. Here, the ECU 10 determines whether or not the current operating state of the engine 1 is in a state in which it is possible to accurately perform fuel vapor concentration learning. Specifically, the ECU 10 determines whether the ratio of the purge amount of the purge gas to the intake air amount of the engine 1 (purge gas introduction ratio) is equal to or greater than a predetermined value, and/or determines whether the intake pressure detected by the intake pressure sensor SW3. It is determined whether (the negative pressure in the intake passage 40) is equal to or less than a predetermined pressure. The ECU 10 determines the purge gas introduction ratio based on the flow rate detected by the airflow sensor SW1, the opening of the purge valve 67, the negative pressure in the intake passage 40, the opening of the throttle valve 43, atmospheric pressure, and the like. The ECU 10 determines that the current operating state of the engine 1 is the fuel vapor concentration learning when the purge gas introduction ratio is equal to or higher than a predetermined value or when the intake pressure detected by the intake pressure sensor SW3 is equal to or lower than a predetermined pressure. It is determined that it is in a state where it can be performed with high accuracy, that is, that the learning permission condition is satisfied (step S2: Yes). In this case, the ECU 10 proceeds to step S3. On the other hand, when the ECU 10 determines that the learning permission condition is not satisfied (step S2: No), the series of routines shown in this flowchart is exited.

Next, in step S3, the ECU 10 acquires the feedback correction amount for the air-fuel ratio feedback control (see graph G7 in FIG. 9), and the learning environment factor (graph G1 in FIG. , G2). In this case, the ECU 10 obtains a feedback correction amount based on the difference between the air-fuel ratio (actual air-fuel ratio) detected by the air-fuel ratio sensor SW8 and the target air-fuel ratio (typically the theoretical air-fuel ratio) in the air-fuel ratio feedback control. . For example, the ECU 10 obtains the ratio of the difference between the target air-fuel ratio and the actual air-fuel ratio to the target air-fuel ratio as the feedback correction amount. The ECU 10 also controls the duty ratio applied to the purge valve 67, the pulse width of the control signal output to the fuel injection valve 6, the intake pressure detected by the intake pressure sensor SW3, and the output signal of the air-fuel ratio sensor SW8. at least one of the degree of variation of , is acquired as a learning environment factor. FIG. 9 illustrates a case where the values of two learning environment factors 1 and 2 are obtained. The values of learning environment factors 1 and 2 shown in FIG. 9 are average values of raw data containing noise.

Next, in step S4, the ECU 10 calculates the amount of variation (see graphs G3 and G4 in FIG. 9) as probability distribution characteristics of the learning environment factors acquired in step S3. Specifically, the ECU 10 obtains the amount of variation for each learning environment factor based on the characteristics (error characteristics) of each learning environment factor as shown in FIGS. For example, the smaller the duty ratio applied to the purge valve 67 , the ECU 10 obtains a larger amount as the variation amount of the purge valve 67 . In addition, the smaller the pulse width of the control signal output to the fuel injection valve 6, the greater the variation amount of the fuel injection valve 6 the ECU 10 obtains. Further, the smaller the intake pressure detected by the intake pressure sensor SW3, the greater the ECU 10 obtains as the variation amount of the intake pressure sensor SW3. Further, the greater the degree of variation in the output signal of the air-fuel ratio sensor SW8, the greater the ECU 10 obtains as the variation amount of the air-fuel ratio sensor SW8. Then, the ECU 10 converts the amount of variation for each learning environment factor obtained in this way so as to match the unit (for example, ratio) of the feedback correction amount. In one example, the ECU 10 obtains the deviation of the air-fuel ratio from the target air-fuel ratio caused by the amount of variation in each learning environment factor from a predetermined arithmetic expression, map, etc., and uses the ratio of the deviation to the target air-fuel ratio.

Next, in step S5, the ECU 10 sums up the variation amounts of the learning environment factors expressed in units of the feedback correction amount calculated in step S4 in order to obtain one variation amount in the current learning environment, A total amount of variation in learning environment factors (see graph G5 in FIG. 9) is obtained. For example, the ECU 10 obtains the total amount of variation by obtaining the square root of the sum of the squares of the amount of variation of a plurality of learning environment factors.

Next, in step S6, the ECU 10 calculates the total variation amount of the learning environment factors calculated in step S5 (see graph G5 in FIG. 9), the purge gas introduction ratio used in step S2 (see graph G6 in FIG. 9), and Based on the feedback correction amount (see graph G7 in FIG. 9) obtained in step S3, the evaporated fuel concentration learning candidate value (see graph G8 in FIG. 9) that is a candidate for the evaporated fuel concentration learning value to be obtained this time, and the evaporation A learning candidate value variation amount (see graph G9 in FIG. 9), which is a variation amount of the fuel concentration learning candidate value, is obtained. Specifically, the ECU 10 determines the amount of vaporized fuel contained in the purge gas according to the current purge gas introduction ratio based on the deviation of the air-fuel ratio from the target air-fuel ratio according to the total variation amount of the learned environmental factors and the feedback correction amount. is estimated, and this concentration is used as the fuel vapor concentration learning candidate value. Then, the ECU 10 obtains the variation amount (learned candidate value variation amount) included in the evaporated fuel concentration learning candidate value based on the total variation amount and the feedback correction amount of the learning environment factor after considering the purge gas introduction ratio. This learning candidate value variation amount indicates the reliability of the evaporated fuel concentration learning candidate value. That is, the larger the learned candidate value variation amount, the lower the reliability of the fuel vapor concentration learning candidate value, and the smaller the learned candidate value variation amount, the higher the reliability of the fuel vapor concentration learning candidate value.

Next, in step S7, the ECU 10 determines the learning value variation amount in the previous fuel vapor concentration learning value (see graph G10 in FIG. 9) and the learning candidate value variation amount in the current fuel vapor concentration learning candidate value (graph in FIG. 9). G11 (same as graph G9)) to be applied to the current evaporated fuel concentration learning candidate value in the weighted addition for obtaining the current fuel production concentration learning value (see graph G12 in FIG. 9). Ask for This weighting coefficient is a coefficient used when weighting and adding the current fuel vapor concentration learning candidate value to the previous fuel vapor concentration learning value. This is the degree of reflection of the learning candidate value (reflection rate). Also, the previous learning value variation amount is obtained together with the fuel vapor concentration learning value obtained at the time of the previous learning, and indicates the reliability of the fuel vapor concentration learning value. That is, the larger the amount of variation in the learned value, the lower the reliability of the learned value of evaporated fuel concentration, and the smaller the amount of variation in learned value, the higher the reliability of the learned value of evaporated fuel concentration.

In the example shown in FIG. 9, as indicated by an arrow A21, when the current learning candidate value variation amount is relatively smaller than the previous learning value variation amount, that is, the reliability of the current learning candidate value is lower than the previous learning value. When the reliability of the value is relatively higher than the value, the ECU 10 sets the weighting factor to, for example, 50% so that the fuel vapor concentration learning candidate value is relatively largely reflected in obtaining the current fuel vapor concentration learning value. (See graph G12). Further, as indicated by an arrow A22, while the previous learning value variation amount is a certain amount, if the current learning candidate value variation amount is 0, that is, the reliability of the current learning candidate value is very high. is high, the ECU 10 weights so that all of the fuel vapor concentration learning candidate values are reflected in determining the current fuel vapor concentration learning value, in other words, the previous fuel vapor concentration learning value is not reflected at all. Set the coefficient to 100% (see graph G12). Further, as indicated by an arrow A23, when the current learning candidate value variation amount is a certain amount, while the previous learning value variation amount is near 0, the reliability of the previous learning value is very high. is high, the ECU 10 causes the previous fuel vapor concentration learned value to be reflected considerably in obtaining the current fuel vapor concentration learning value, in other words, the current fuel vapor concentration learning candidate value is hardly reflected. , for example, the weighting factor is set to 5% (see graph G12).

Next, in step S8, the ECU 10 calculates the current evaporated fuel concentration learning candidate value (see graph G14 (same as graph G8) in FIG. 9) based on the weighting coefficient (see graph G12 in FIG. 9) obtained in step S7. By performing weighted addition, the current fuel vapor concentration learning value (see graph G13 in FIG. 9) is obtained, and the learning value variation amount (that is, reliability) of the fuel vapor concentration learning value is obtained. In the example shown in FIG. 9, while the weighting factor is set to 50%, the ECU 10 adds 50% of the previous evaporated fuel concentration learning value and 50% of the current evaporated fuel concentration learning candidate value. is obtained as the current evaporated fuel concentration learning value. Further, while the weighting coefficient is set to 100%, the ECU 10 applies the fuel vapor concentration learning candidate value of this time as it is as the fuel vapor concentration learning value of this time. Further, while the weighting factor is set to 5%, the ECU 10 adds a value obtained by adding 95% of the previous evaporated fuel concentration learning value and 5% of the current evaporated fuel concentration learning candidate value to the current evaporated fuel concentration learning value. Obtained as a fuel concentration learning value. By obtaining the evaporative fuel concentration learning value in this way, even in a situation where the degree of influence of disturbance on the evaporative fuel concentration learning is relatively large, that is, the amount of variation in the air-fuel ratio caused by the characteristics of the learning environment factors is relatively small. Even in a large situation, fuel vapor concentration learning can be properly executed in consideration of the amount of variation due to such learning environmental factors without limiting the execution of fuel vapor concentration learning. Then, the ECU 10 obtains the variation amount (learned value variation amount) included in the evaporated fuel concentration learning value obtained as described above. For example, the ECU 10 obtains the current learning value variation amount by weighting and adding the current learning candidate value variation amount to the previous learning value variation amount using a weighting factor similar to that described above.

Next, in step S9, the ECU 10 sets the rate of change of the purge amount of evaporated fuel based on the amount of variation in the learned value obtained in step S8. Specifically, the ECU 10 refers to a map such as that shown in FIG. 10, and calculates a purge amount change speed suppression coefficient for suppressing the change speed of the purge amount corresponding to the current learning value variation amount (horizontal axis). (vertical axis) is determined, and the change speed of the purge amount is corrected using this purge amount change speed suppression coefficient. As shown in FIG. 10, as the learning value variation amount increases, the purge amount change speed suppression coefficient decreases (more specifically, it approaches 0). The suppression coefficient increases (more specifically, it approaches 1). This purge amount change speed suppression coefficient is used so as to multiply the change speed of the purge amount to be set according to the purge request or the like. Therefore, by correction using the purge amount change speed suppression coefficient, the change rate of the finally applied purge amount becomes lower as the learning value variation amount increases, in other words, the change rate of the purge amount becomes smaller.

Here, FIG. 11 shows an example of temporal change of the purge amount change speed suppression coefficient according to the learning value variation amount at the start of purging of vaporized fuel. In FIG. 11, the solid line indicates the graph when the learning value variation amount is relatively small, and the broken line indicates the graph when the learning value variation amount is relatively large. Normally, at the start of purging, the reliability of the evaporated fuel concentration learning value is not ensured, so the fuel injection amount cannot be accurately corrected so as to suppress the deviation of the air-fuel ratio caused by the evaporative fuel purge. Purge valve 67 is controlled to ramp up. As shown in FIG. 11, when the amount of variation in the learning value is large at the start of purging, the rate of change in the purge amount applied to the rate of change in the gradually increasing purge amount is higher than when the amount of variation in the learning value is small. A small suppression factor is set. As a result, the rate of change in the gradually increased purge amount becomes lower, that is, the gradient of change in the purge amount becomes gentler. As a result, when the reliability of the evaporated fuel concentration learning value is not ensured, it is assumed that the fuel injection amount cannot be accurately corrected so as to suppress the deviation of the air-fuel ratio caused by the evaporated fuel purge, and the purge amount is not increased. can be properly constrained. On the other hand, when the learning value variation amount is small at the start of the purge, the purge amount change speed suppression coefficient applied to the change speed of the purge amount to be gradually increased is set larger than when the learning value variation amount is large. . As a result, the rate of change of the gradually increased purge amount becomes high. As a result, even when the purge is started, the purge amount can be quickly increased when the reliability of the evaporated fuel concentration learning value is ensured.

Returning to FIG. 8, after step S9, the ECU 10 proceeds to step S10. In step S10, the ECU 10 corrects the fuel injection amount of the fuel injection valve 6 based on the feedback correction amount obtained in step S3 and the evaporated fuel concentration learning value obtained in step S8. Specifically, the ECU 10 determines the difference between the current actual air-fuel ratio and the target air-fuel ratio according to the feedback correction amount, and the air-fuel ratio when fuel vapor having a concentration corresponding to the fuel vapor concentration learning value is introduced into the engine 1. A fuel injection amount that should be applied to achieve the target air-fuel ratio is determined based on the fluctuation of the fuel ratio. After that, the ECU 10 exits a series of routines shown in this flow chart.

<Effect>
Next, the operation and effect of the evaporated fuel processing device according to the embodiment of the present invention will be described.

According to this embodiment, variations in air-fuel ratio caused by characteristics of learned environmental factors (at least one or more of one or more control valves and one or more sensors) related to fuel vapor purge control is obtained, and the fuel vapor concentration learning value is obtained based on the variation amount and the feedback correction amount of the air-fuel ratio feedback control. By obtaining the evaporative fuel concentration learning value in this way, even in situations where the degree of influence of disturbance on the evaporative fuel concentration learning is relatively large, that is, in situations where the amount of variation in the air-fuel ratio caused by the characteristics of the learning environment factors is large. Even so, fuel vapor concentration learning can be appropriately executed in consideration of the amount of variation due to such learning environment factors without restricting the execution of fuel vapor concentration learning. That is, according to the present embodiment, even if there is variation due to learning environmental factors, the fuel vapor concentration learned value can be obtained with high accuracy by appropriately adding the amount of variation at this time. As a result, according to the present embodiment, the accuracy of the fuel vapor concentration learning value is higher than in the case of limiting the execution of the fuel vapor concentration learning in a situation where the degree of influence of disturbance on the fuel vapor concentration learning is large as in the conventional technology. can be quickly secured, that is, the reliability of the evaporated fuel concentration learning value can be quickly secured. Therefore, according to the present embodiment, the purge amount of vaporized fuel can be increased early, and the vaporized fuel in the fuel tank 63 and the purge system 61 can be quickly processed.

Further, according to the present embodiment, the purge amount from the purge passage 66 to the intake passage 40 is corrected based on the reliability of the evaporated fuel concentration learned value (that is, the amount of variation in the learned value). Here, in a typical prior art, the purge amount is set according to the number of executions (learning times) of fuel vapor concentration learning. When there is less, the purge amount is decreased. In such conventional technology, even if the number of times of learning is large, if the accuracy of the fuel vapor concentration learning value is low, introduction of a large amount of purge gas may cause a large deviation in the air-fuel ratio. Even if the fuel vapor concentration learning value is small, only a small amount of purge gas can be introduced, although there is no problem even if a large amount of purge gas is introduced. In contrast, according to the present embodiment, the purge amount is corrected according to the reliability of the fuel vapor concentration learning value. , and when the accuracy of the evaporated fuel concentration learning value is high, the purge amount can be appropriately set to a relatively large amount.

Further, according to the present embodiment, the purge amount to be gradually increased when starting to purge the evaporated fuel can be appropriately set to an amount corresponding to the reliability of the evaporated fuel concentration learned value. When the reliability of the learned value is high, the evaporated fuel can be quickly processed.

Further, according to the present embodiment, when the reliability of the evaporated fuel concentration learning value is high, the change speed of the purge amount is increased, so that the evaporated fuel can be quickly processed. On the other hand, when the reliability of the evaporated fuel concentration learning value is low, the rate of change of the purge amount is decreased, so that the deviation of the air-fuel ratio due to the introduction of the purge gas can be appropriately suppressed.

Further, according to the present embodiment, weighting is performed based on the difference between the reliability of the previous evaporated fuel concentration learning value (learning value variation) and the reliability of the current evaporated fuel concentration learning candidate value (learning candidate value variation). A coefficient is obtained, and using the weighting coefficient, the current evaporated fuel concentration learned value is weighted and added to the previous evaporated fuel concentration learned value to obtain the current evaporated fuel concentration learned value. As a result, even in a state in which variations due to the characteristics of learning environment factors occur, it is possible to eliminate the effects of such variations as much as possible and obtain the evaporated fuel concentration learning value with high accuracy.

Further, according to the present embodiment, when the reliability of the fuel vapor concentration learning candidate value of this time is higher than the reliability of the fuel vapor concentration learning value of the previous time, the reliability of the fuel vapor concentration learning candidate value of this time is Since the weighting factor is made larger than when the reliability of the fuel vapor concentration learning value is lower than the previous value, the magnitude relationship between the reliability of the fuel vapor concentration learning candidate value of this time and the reliability of the fuel vapor concentration learning value of the previous time is determined. The weighting factor can be appropriately set accordingly.

Further, according to the present embodiment, each of the variation amounts of two or more learning environment factors is converted so as to match the unit of the feedback correction amount, and the two or more variation amounts after this conversion are added together. Since the fuel vapor concentration learning value is obtained based on the above, it is possible to obtain the fuel vapor concentration learning value that appropriately takes into account all the variation amounts of the two or more learning environment factors.

Further, according to the present embodiment, the smaller the duty ratio of the purge valve 67 is, the larger the amount of variation regarding the purge valve 67 is obtained. can be done.

Further, according to the present embodiment, the smaller the pulse width of the control signal for the fuel injection valve 6, the greater the amount of variation regarding the fuel injection valve 6. Therefore, the variation in the air-fuel ratio caused by the characteristics of the fuel injection valve 6 Quantity can be determined appropriately.

Further, according to the present embodiment, the larger the degree of variation in the output signal of the air-fuel ratio sensor SW8, the larger the amount of variation related to the air-fuel ratio sensor SW8. Quantity can be determined appropriately.

Further, according to the present embodiment, the smaller the intake pressure detected by the intake pressure sensor SW3, the larger the amount of variation regarding the intake pressure sensor SW3. Quantity can be determined appropriately.

1 engine 6 fuel injection valve 10 ECU
17 combustion chamber 25 spark plug 40 intake passage 43 throttle valve 44 supercharger 50 exhaust passage 61 purge system 63 fuel tank 64 canister 65 atmospheric release passage 66 purge passage 67 purge valve SW3 first pressure sensor (intake pressure sensor)
SW8 air-fuel ratio sensor

Claims (10)

An evaporative fuel processing device,
a purge passage extending from the fuel tank toward an intake passage of the engine for purging vaporized fuel in the fuel tank to the intake passage;
a canister provided on the purge passage for absorbing and accumulating evaporated fuel from the fuel tank;
a purge valve provided on the purge passage on the downstream side of the canister to open and close the purge passage;
a controller configured to control at least the purge valve to perform evaporated fuel purge control for purging the evaporated fuel into the intake passage;
has
The controller is
Based on the difference between the actual air-fuel ratio detected by an air-fuel ratio sensor that detects the air-fuel ratio of the exhaust gas of the engine and the target air-fuel ratio to be set, the fuel injection amount of the fuel injection valve of the engine is adjusted in air-fuel ratio feedback control. Find the feedback correction amount for correction,
Obtaining a variation amount indicating the degree of variation in the air-fuel ratio caused by characteristics of at least one or more of the one or more control valves and the one or more sensors related to the fuel vapor purge control;
obtaining a fuel vapor concentration learning value for learning the concentration of the fuel vapor contained in the gas purged into the intake passage based on the feedback correction amount and the variation amount;
Based on the reliability of the fuel vapor concentration learning value corresponding to the amount of variation included in the fuel vapor concentration learning value, the purge amount of gas containing fuel vapor to be purged into the intake passage is corrected, and after the correction, controlling the purge valve based on the purge amount;
is configured as
The evaporative fuel processing apparatus according to claim 1, wherein the controller is configured to perform correction so that the purge amount increases as the reliability of the evaporative fuel concentration learning value increases. The controller controls the purge valve to gradually increase the purge amount when starting to purge the evaporated fuel, and corrects the gradually increased purge amount based on the reliability of the evaporated fuel concentration learning value. 2. The evaporative fuel treatment system of claim 1, comprising: 3. The evaporative fuel processing apparatus according to claim 1, wherein said controller is configured to perform correction such that the rate of change of said purge amount increases as said reliability of said evaporative fuel concentration learning value increases. . The controller is
Repeatedly obtaining the variation amount and the fuel vapor concentration learning value at a predetermined cycle,
determining a fuel vapor concentration learning candidate value that is a candidate for the fuel vapor concentration learning value to be determined this time based on the amount of variation determined this time;
obtaining the current fuel vapor concentration learning value by weighting and adding the fuel vapor concentration learning candidate value obtained this time to the fuel vapor concentration learning value obtained last time;
Reliability of the fuel vapor concentration learning value according to the amount of variation included in the fuel vapor concentration learning value obtained last time determining a weighting factor to be applied to the fuel vapor concentration learning candidate value obtained this time in the weighted addition based on the difference from the reliability of the fuel vapor concentration learning candidate value;
4. The evaporative fuel processing device according to any one of claims 1 to 3, which is configured as When the reliability of the fuel vapor concentration learning candidate value obtained this time is higher than the reliability of the fuel vapor concentration learning value obtained last time, the controller determines the fuel vapor concentration obtained this time. 5. Evaporative fuel processing according to claim 4, wherein the weighting factor is set larger than when the reliability of the learning candidate value is lower than the reliability of the fuel vapor concentration learned value obtained last time. Device. The controller uses two or more of the variation amounts, converts each of the two or more variation amounts so as to match the unit of the feedback correction amount, and converts the two or more variation amounts after conversion into 6. The evaporated fuel processing device according to claim 1, wherein said evaporated fuel concentration learning value is obtained based on the sum processed amount and said feedback correction amount. The controller uses the amount of variation caused by the characteristics of the purge valve, and is configured to obtain a larger amount as the amount of variation related to the purge valve as the duty ratio applied in the duty control for the purge valve decreases. 7. The evaporative fuel treatment device according to any one of claims 1 to 6. The controller uses the amount of variation resulting from the characteristics of the fuel injection valve, and the smaller the pulse width of the control signal output to the fuel injection valve, the greater the amount of variation related to the fuel injection valve. 8. An evaporative fuel treatment device according to any one of the preceding claims, adapted to determine. The controller uses the amount of variation resulting from the characteristics of the air-fuel ratio sensor, and is configured to obtain a larger amount as the amount of variation related to the air-fuel ratio sensor as the degree of variation in the output signal of the air-fuel ratio sensor increases. 9. The evaporative fuel treatment device according to any one of claims 1 to 8. The controller uses the amount of variation resulting from the characteristics of an intake pressure sensor provided in the intake passage, and the smaller the intake pressure detected by the intake pressure sensor, the greater the amount of variation related to the intake pressure sensor. 10. An evaporative fuel treatment device as claimed in any one of the preceding claims, arranged to determine a quantity.

Images