Control device for internal combustion engine

The present invention relates to a control device for an internal combustion engine, and more particularly to a control device for an internal combustion engine provided with an external EGR device.

In recent years, regulations concerning fuel consumption of automobiles (hereinafter referred to as fuel consumption) and hazardous exhaust gas have been strengthened and will become severer in the future. Regarding the fuel consumption, in particular, low fuel consumption is strongly required due to the influence of discharged carbon dioxide on global warming. To advance low fuel consumption, it is likely recently to provide a supercharger, such as a turbocharger, to increase the compression ratio. However, in a case where the internal combustion engine provides a relatively high output, a possibility of occurrence of abnormal combustion, such as knocking, increases. To avoid abnormal combustion, ignition timing is retarded.

Retarding ignition timing, however, requires a large amount of fuel to maintain the output, thus deteriorating the fuel consumption. Therefore, the fuel consumption is improved by reintroducing a recirculation exhaust gas (hereinafter referred to as EGR gas) into the combustion chamber by the external EGR device to prevent knocking and set an appropriate value for the ignition timing. For this reason, it is necessary to accurately determine the amount of EGR introduced into the combustion chamber or the EGR rate. Here, the EGR rate refers to a value defined by EGR gas amount/(air flow rate+EGR gas amount).

As a method of estimating the EGR amount (rate), an external EGR device described in, for example, JP 2001-280202 A (PTL 1) or JP 2013-108464 A (PTL 2) has been known. PTL 1 discloses a method of accurately estimating the EGR amount (rate) from the differential pressure across the EGR valve and the EGR valve opening degree. PTL 2 discloses a method of accurately estimating the EGR amount (rate) using an intake air amount sensor and an oxygen concentration sensor provided in the intake system.

Patent Literature

PTL 1: JP 2001-280202 A

PTL 2: JP 2013-108464 A

Technical Problem

The external EGR device may include a system called a low pressure EGR system in which the exhaust gas is taken out from the downstream side of the turbine of the supercharger provided in the exhaust pipe and the EGR gas is introduced to the upstream of the compressor of the supercharger provided in the intake pipe. In the low pressure EGR system, a relatively long passage distance is provided between the point where the EGR gas is introduced into the intake pipe and the combustion chamber. When the mixed gas of air and EGR gas moves through the long passage distance, the mixed gas passes through a plurality of control valves for controlling the flow rate and pressure of the mixed gas and a cooler for reducing the temperature of the mixed gas.

For this reason, in the case of using only the EGR valve opening degree and the differential pressure as in PTL 1, the estimation accuracy of the EGR amount (rate) may deteriorate during the transient state in which the operation region of the internal combustion engine changes. Further, in the case of providing an EGR sensor such as an oxygen sensor for detecting the EGR amount (rate) in the intake pipe as in PTL 2, the EGR sensor is used to detect the EGR amount (rate) with high accuracy if the internal combustion engine operates in a constant operation region. However, in the transient state in which the operating region of the internal combustion engine varies, the response characteristic of the EGR sensor is often slower than the speed at which the EGR amount (rate) changes, thus making it difficult to appropriately estimate the EGR amount (rate) during the transient state.

Therefore, an appropriate amount of EGR gas may not be introduced into the combustion chamber, causing unstable combustion of the combustible air-fuel mixture in the combustion chamber and deteriorating the fuel consumption.

An object of the present invention is to provide a novel control device for an internal combustion engine capable of highly accurately estimating the EGR amount (rate) in the transient state.

Solution to Problem

In the present invention, a first EGR rate is determined using, as an input, a detection signal of an EGR sensor provided on the downstream side of a throttle valve which adjusts a flow rate of a mixed gas of air and EGR gas flowing through an intake pipe, a second EGR rate is estimated using, as an input, one or more detection signals of at least an air flow sensor, an EGR valve opening degree sensor, and a throttle valve opening degree sensor, a third EGR rate is determined by carrying out delay processing on the second EGR rate corresponding to a response delay of the EGR sensor, and the second EGR rate is subjected to learning correction by reflecting a difference between the third EGR rate and the first EGR rate.

Advantageous Effects of Invention

According to the present invention, the EGR rate can be estimated with high accuracy even in the transient state. As a result, an appropriate amount of EGR gas is introduced into the combustion chamber to avoid abnormal combustion, whereby deterioration of fuel consumption can be prevented.

FIG. 1 is a configuration diagram of a system configuration of an internal combustion engine to which the present invention is applied.

FIG. 2 is a system block diagram illustrating a configuration of the control unit of FIG. 1 .

FIG. 3 is a flowchart illustrating physical model learning for estimating an EGR rate according to a first embodiment of the present invention.

FIG. 4 is an explanatory diagram of an operation of the physical model learning for estimating the EGR rate according to the first embodiment.

FIG. 5A is an explanatory diagram for explaining a state before the learning according to the first embodiment.

FIG. 5B is an explanatory diagram for describing a state after the learning according to the first embodiment.

FIG. 6 is a flowchart of the physical model learning for estimating the EGR rate by considering dew condensation according to a second embodiment of the present invention.

FIG. 7A is an explanatory diagram for explaining a gas composition in an intake pipe when the gas is air.

FIG. 7B is an explanatory diagram for explaining a gas composition in the intake pipe when the gas is a mixed gas of air and exhaust gas.

FIG. 7C is an explanatory diagram for explaining a gas composition in the intake pipe when the gas is a mixed gas of air, exhaust gas, and moisture.

FIG. 8 is a configuration diagram illustrating a system configuration of another internal combustion engine to which the present invention is applied.

FIG. 9 is a flowchart illustrating the physical model learning for estimating the EGR rate according to a third embodiment of the present invention.

FIG. 10 is a flowchart illustrating physical model learning for estimating an EGR rate by considering dew condensation according to a fourth embodiment of the present invention.

Embodiments of the present invention will be described in detail below with reference to the accompanying drawings. The present invention, however, is not limited to the following embodiments, and various modifications and application examples within the technical concept of the present invention also fall within that range.

First Embodiment

FIG. 1 illustrates a system configuration of an internal combustion engine according to a first embodiment of the present invention. An internal combustion engine Eng is a spark ignition-type internal combustion engine.

The internal combustion engine Eng includes the following components that are disposed appropriately about an intake pipe 10 : an air flow (or mass flow) sensor 3 that measures an air flow rate at an inlet of an intake pipe serving as an intake system, an atmospheric humidity sensor 24 that measures humidity of the intake air, an intake pipe pressure sensor 8 that measures an intake pipe pressure, a mixed gas humidity sensor 9 that detects a humidity of the mixed gas of air and exhaust gas immediately before the gas enters a combustion chamber 14 , a first throttle valve 4 that adjusts the pressure of intake mixed gas, a second throttle valve 7 that adjusts the amount of the intake mixed gas, a compressor 5 a of a supercharger that compresses the intake mixed gas, and an intercooler 6 that cools the intake mixed gas to decrease the temperature.

The internal combustion engine Eng also includes a fuel injection valve 12 injecting fuel into a combustion chamber 14 and an ignition plug 13 that supplies ignition energy which are disposed in individual appropriate positions of the internal combustion engine Eng. Also provided are an intake-side movable valve mechanism 11 a that controls opening and closing phases of an intake valve and an exhaust-side movable valve mechanism 11 b that controls opening and closing phases of an exhaust valve. The movable valve mechanism may also be referred to as a valve timing control system (VTC).

Further, a turbine 5 b that drives a compressor 5 a using the exhaust gas energy, a three-way catalyst 17 that purifies exhaust gas, and an air-fuel ratio sensor 16 which is one type of an air-fuel ratio detector detecting an air-fuel ratio of the exhaust gas on the upstream side of the three-way catalyst 17 are disposed appropriately about an exhaust pipe 15 serving as an exhaust system. The air-fuel ratio sensor 16 may be an oxygen concentration sensor. An exhaust pipe 18 is connected in the downstream of the three-way catalyst 17 , and a three-way catalyst 23 is provided at an appropriate position in the exhaust pipe 18 .

An EGR pipe 22 taking out EGR gas branches in the upstream of the three-way catalyst 23 of the exhaust pipe 18 . An EGR cooler 21 cooling the EGR gas, an EGR valve 20 adjusting the EGR amount, and a differential pressure sensor 19 measuring the pressure before and after the EGR valve are disposed appropriately about the EGR pipe 22 .

Detection signals are sent to a control unit (hereinafter referred to as an ECU) 1 serving as a control unit, the detection signals including a detection signal Ss 3 obtained from the air flow sensor 3 , detection signals Ss 7 and Ss 4 obtained respectively from the opening degree sensors of the first and second throttle valves 4 and 7 , a detection signal Ss 8 obtained from the intake pipe pressure sensor 8 , a detection signal Ss 9 obtained from the mixed gas humidity sensor 9 , detection signals Ss 11 a and Ss 11 b obtained respectively from the intake/exhaust movable valve mechanisms 11 a and 11 b , a detection signal Ss 16 obtained from the air-fuel ratio sensor 16 , a detection signal Ss 19 obtained from the differential pressure sensor 19 , a detection signal Ss 20 obtained from the EGR valve 20 , and the detection signal Ss 24 obtained from the atmospheric humidity sensor 24 . A signal Ss 2 obtained from an accelerator opening degree sensor 2 detecting a depression amount of the accelerator pedal, that is, an accelerator opening degree, is also sent to the ECU 1 .

The ECU 1 calculates a required torque on the basis of the output signal Ss 2 of the accelerator opening degree sensor 2 and various sensor signals. That is, the accelerator opening degree sensor 2 is used as a required torque detection sensor detecting the required torque to the internal combustion engine Eng. The ECU 1 optimally calculates major operation amounts of the internal combustion engine Eng on the basis of the operation state of the internal combustion engine Eng obtained from each output of the various sensors, the major operation amounts including the opening degree of the first throttle valve 4 , the opening degree of the second throttle valve 7 , an injection pulse period of the fuel injection valve 12 , an ignition timing of the ignition plug 13 , opening and closing timings of the intake-side movable valve mechanism 11 a and the exhaust-side movable valve mechanism 11 b , the opening degree of the EGR valve 20 , and the like.

The fuel injection pulse period calculated by the ECU 1 is converted into a fuel-injection-valve open-valve-pulse signal Ds 12 and sent to the fuel injection valve 12 . The opening degree of the first throttle valve 4 calculated by the ECU 1 is sent to the first throttle valve 4 as a first throttle valve drive signal Ds 4 . Similarly, the opening degree of the second throttle valve 7 is sent to the second throttle valve 7 as a second throttle valve drive signal Ds 7 . An ignition plug drive signal Ds 13 is sent to the ignition plug 13 , and the fuel injection valve drive signal Ds 12 is sent to the fuel injection valve 12 . An EGR valve drive signal Ds 20 is sent to the EGR valve 20 .

In the above configuration, a mixed gas flows into the combustion chamber 14 from the intake pipe 10 via the intake valve. To the mixed gas, fuel is injected from a fuel injection valve 12 through a fuel pump from a fuel tank, which is not illustrated, to generate a combustible air-fuel mixture. The combustible air-fuel mixture is ignited and burned by a spark generated from the ignition plug 13 at predetermined ignition timing, and a resulting combustion pressure pushes down the piston to generate a rotational driving force.

After combustion, the exhaust gas passes through the exhaust valve, the exhaust pipe 15 , and the turbine 5 b to the three-way catalyst 17 where components, such as NOx, CO, and HC, are purified in the three-way catalyst 17 , purified again in the three-way catalyst 23 , and discharged externally through the exhaust pipe 18 . Meanwhile, the exhaust gas is partially reintroduced into the intake pipe as the EGR gas via the EGR pipe 22 , the EGR cooler 21 , and the EGR valve 20 .

FIG. 2 illustrates system blocks of the configuration of the ECU 1 . Signals output from the accelerator opening degree sensor 2 , the air flow sensor 3 , the intake pipe pressure sensor 8 , the mixed gas humidity sensor 9 , and the air-fuel ratio sensor 16 are input to an input circuit 30 a of the ECU 1 . The input signals may include other signals. Signals input from the sensors are sent to an input port of an input/output port 30 b.

Values of the input signals sent to the input/output port 30 b are stored in a random access memory (RAM) 30 c and receive arithmetic processing by a CPU 30 e . At this time, analog signals of the input signals sent to the input circuit 30 a are converted into digital signals by an A/D converter provided in the input circuit 30 a.

A control program describing details of the arithmetic processing is written in a read only memory (ROM) 30 d in advance. Values indicating the operation amount of each actuator calculated according to the control program are stored in the RAM 30 c , sent to the output port of the input/output port 30 b , and sent to each actuator via each drive circuit.

Drive circuits of the present embodiment are, for example, a first throttle drive circuit 30 f , a second throttle drive circuit 30 g , a movable-valve-mechanism drive circuit 30 h , a fuel-injection-valve drive circuit 30 i , an ignition output circuit 30 j , an EGR valve drive circuit 30 k , and the like. Respective drive circuits control the first throttle valve 4 , the second throttle valve 7 , the movable valve mechanisms 11 a and 11 b , the fuel injection valve 12 , the ignition plug 13 , and the EGR valve 20 . Although the ECU 1 of the present embodiment includes drive circuits in the ECU 1 , the ECU 1 is not limited thereto and any or all of the drive circuits may be provided outside the ECU 1 .

FIG. 3 illustrates a control flow relating to the learning correction of the EGR rate in the first embodiment of the present invention.

First, in step S 10 , the atmospheric humidity sensor 24 detects the atmospheric humidity RHAMB of the intake air. Next, in step S 11 , the mixed gas humidity sensor 9 detects the mixed gas humidity RHIMANI of the mixed gas flowing through the intake pipe 10 in the downstream of the second throttle valve 7 . After the humidity is detected from the humidity sensors 9 and 24 in steps S 10 and S 11 , the process proceeds to step S 12 . The humidity sensors 24 and 9 function as EGR sensors.

In step S 12 , an EGR rate RVAP in the downstream of the second throttle valve 7 in the intake pipe is calculated from the atmospheric humidity RHATM and the mixed gas humidity RHIMANI detected in steps S 10 and S 11 . Since the EGR gas is a mixed gas, the mixed gas humidity RHIMANI changes its value more than the value of the atmospheric humidity RHATM. Therefore, the “EGR rate RVAP” can be calculated from each humidity by a known method. After the “EGR rate RVAP” is calculated, the process proceeds to step S 13 .

In step S 13 , “EGR rate RMODEL” of the mixed gas in the downstream of the second throttle valve 7 is calculated by a plurality of mathematical equations (which is so-called a physical model) describing the behavior of the mixed gas on the basis of one or more pieces of information of the air flow rate information of the air flow sensor 3 , the opening degree information of the EGR valve 20 , the opening degree information of the second throttle valve 7 , and other sensor information as necessary. The physical model can be constituted by combining a state equation or a motion equation of the gas. This physical model is stored in the ROM 30 d of the ECU 1 . When the “EGR rate RMODEL” of the mixed gas is determined by the physical model, the process proceeds to step S 14 .

In step S 14 , the “EGR rate RMODELF” after response delay processing is calculated by reflecting a measurement response delay of the design of the mixed gas humidity sensor 9 as a first order lag on the “EGR rate RMODEL,” which has been estimated according to the physical model. Thus, the detection delay of the humidity sensor 9 is reflected on the “EGR rate RMODEL” estimated with the physical model. After the “EGR rate RMODELF” is subjected to the response delay processing, the process proceeds to step S 15 .

In step S 15 , “difference DREGR” between the “EGR rate RVAP” based on the humidity sensor and the “EGR rate RMODELF” subjected to response delay processing is calculated. This is determined by calculating “difference DREGR”=“EGR rate RMODELF”−“EGR rate RVAP”. After the “difference DREGR” is obtained, the process proceeds to step S 16 .

In step S 16 , it is determined whether the “difference DREGR” of the EGR rate is equal to or greater than a predetermined “threshold SLEGR.” If the “difference DREGR” is equal to or greater than the “threshold SLEGR,” the process proceeds to step S 17 . If the “difference DREGR” is smaller than the “threshold SLEGR,” the process proceeds to the end. If the “difference DREGR” is equal to or greater than the predetermined “threshold SLEGR,” the opening degree of the EGR valve of the physical model is regarded as inappropriate. If the “difference DREGR” is smaller than the predetermined “threshold SLEGR,” the opening degree of the EGR valve of the physical model is regarded as falling within an appropriate range. When the “difference DREGR” is equal to or greater than the “threshold SLEGR,” the process proceeds to step S 17 .

In step S 17 , the “difference DREGR” and “correction coefficient KAEGRV” of the EGR valve opening area, which corresponds to the opening degree of the EGR valve 20 at the time when the “difference DREGR” is obtained, are read and calculated from the table stored in the ROM 30 d . The table is set such that larger “difference DREGR” corresponds to larger “correction coefficient KAEGRV.” When the “correction coefficient KAEGRV” is obtained, the process proceeds to step S 18 .

In step S 18 , a calculation for reflecting the “correction coefficient KAEGRV,” which corresponds to the “difference DREGR,” on the opening area of the EGR valve 20 , which is used in the equations constituting the physical model for calculating the “EGR rate RMODEL” calculated in the previous step S 13 , is executed. This allows execution of learning correction of the “EGR rate RMODEL.”

Although the “correction coefficient KAEGRV” is reflected by multiplying it by the opening area of the EGR valve 20 in the present embodiment, the “correction opening area” may be added to and reflected on the opening area of the EGR valve 20 . If the “correction coefficient KAEGRV” is reflected on the opening area of the EGR valve 20 , the process goes to the end and waits for the next activation timing.

Temporal changes in the estimated EGR rate, the difference of the EGR rate, the correction coefficient, and the like, which are obtained by the above-described control flow, are briefly described by referring to FIG. 4 .

In FIG. 4 , in the transient state in which the operating state of the internal combustion engine changes, for example, when the EGR valve 20 is largely opened at time T 1 , the EGR gas amount increases and the estimated EGR rate greatly changes. At this time, the “EGR rate RVAP” based on the humidity sensor is obtained from step S 10 to step S 12 .

Meanwhile, the “EGR rate RMODEL” of the mixed gas is obtained by the physical model on the basis of the air flow rate of the air flow sensor 3 , the opening degree of the second throttle valve 7 , and the opening degree of the EGR valve 20 in step S 13 . Further, in step S 14 , the “EGR rate RMODEL” is subjected to response delay processing to obtain “EGR rate RMODELF.” As illustrated in FIG. 4 , the “EGR rate RMODEL” is subjected to the delay processing to the “EGR rate RMODELF.”

Then, as the control flow proceeds to step S 15 in which the “difference DREFR” between “EGR rate RVAP” and the “EGR rate RMODELF” is calculated. In step S 16 , if it is determined that the “difference DREFR” is greater than the “threshold SLEGR” at time T 2 , that is, if it is determined that “EGR rate RVAP” and “EGR rate RMODELF” are largely divergent from each other, the “correction coefficient KAEGRV” is reflected (i.e., learning correction) on the calculation result of the opening area of the EGR valve 20 of the physical model.

In the case of FIG. 4 , therefore, the “EGR rate RMODELF” is greater than the “EGR rate RVAP,” so that the learning to decrease the “correction coefficient KAEGRV” of the opening area of the EGR valve 20 is carried out. Naturally, therefore, when the “EGR rate RMODELF” is smaller than the “EGR rate RVAP,” the learning to increase the “correction coefficient KAEGRV” of the opening area of the EGR valve 20 is carried out.

FIGS. 5A and 5B are graphs illustrating changing states of the actual EGR rate in the intake pipe 10 , the “EGR rate RMODEL” according to the physical model, the “EGR rate RMODELF” subjected to the response delay processing, and the “EGR rate RVAP” from the humidity sensors 9 and 24 , when the EGR valve 20 is largely opened.

FIG. 5A illustrates the EGR rates described above before carrying out the correction in a case where predetermined “clogging” is given artificially to the EGR valve 20 . The actual EGR rate rises stepwise and the “EGR rate RMODEL” estimated by the physical model resembles the actual EGR rate in responsiveness during rising, but a predetermined error ΔEGR occurs in the case of transition to the steady state. The “EGR rate RVAP” calculated by the humidity sensors 9 and 24 and the actual EGR rate accurately match in the steady state, but the responsiveness delays and the measurement accuracy decreases in the transient state. The “EGR rate RMODELF” that does not reflect the response delay of the mixed gas humidity sensor 9 on the “EGR rate RMODEL” estimated by the physical model is relatively close to the “EGR rate RVAP,” but in the steady state, a deviation occurs against the actual EGR rate and the deviation increases from the actual EGR rate with the lapse of time. This proves the necessity of the correction of the EGR rate.

Meanwhile, FIG. 5B illustrates a case where the “correction coefficient KAEGRV” is reflected by considering the “clogging” in the opening area of the EGR valve 20 of the physical model according to the present embodiment. It can be seen that “EGR rate RMODELF” and “EGR rate RVAP” substantially match. It is also seen that, in this state, the actual EGR rate and the “EGR rate RMODEL” estimated by the physical model well aligned.

As described above, in the present embodiment, the EGR rate that is calculated on the basis of the value of the humidity sensor (i.e., EGR sensor) attached to the intake pipe is compared with the EGR rate that is estimated from the air flow sensor or the like according to the physical model and subjected to the response delay processing. Then, on the basis of the deviation of both EGR rates, the learning correction is carried out on the opening area of the EGR valve of the physical model to improve the estimation accuracy of the EGR rate by the physical model.

Second Embodiment

Next, a second embodiment of the present invention is described. The present embodiment is different from the first embodiment in that compensating control for the variation of the EGR rate that occurs due to the influence of dew condensation adhering to the intake pipe is added. Therefore, description including the operations and effects of the first embodiment is not repeated below.

FIG. 6 illustrates a control flow of the second embodiment of the present invention. Steps S 20 to S 24 are the same as steps S 10 to S 14 of FIG. 3 and the description thereof is not repeated. When the process of step S 24 is completed, the process proceeds to step S 25 .

In step S 25 , it is estimated if dew condensation occurs in the intake pipe 10 . Specifically, “saturated water vapor pressure PWIMANI” is calculated from “temperature TIMANI” of the intake pipe 10 . A relationship between temperature and saturated water vapor pressure may be stored in the table in the ROM 30 d or calculated using “Tetens equation” given as equation (1) below.

[MATH. 1]

PWIMANI=6.1078×10 (7.5×TIMANI/(237.3+TIMANI))   (1)

Next, a water vapor partial pressure in the intake pipe 10 is calculated from the “EGR rate RMODEL” estimated by the physical model. Using “water vapor volume fraction H2OCMB” in the exhaust gas generated by combustion, “water vapor volume fraction H2OATM” calculated from the atmospheric humidity, and the “EGR rate RMODEL” obtained by the physical model, “water vapor volume fraction H2OIMANI” in the intake pipe 10 is calculated by equation (2).

[MATH. 2]

H2OIMANI=RMODEL×H2OCMB+H2OATM  (2)

Next, the “water vapor partial pressure PWIMANIEST” in the intake pipe 10 is calculated from the “water vapor volume fraction H2OIMANI” in the intake pipe 10 and the “intake pressure PIMANI” in the intake pipe 10 by equation (3).

[MATH. 3]

H2OIMANI=RMODEL×H2OCMB+H2OATM  (2)

If the “water vapor partial pressure PWIMNIEST” is greater than “saturated water vapor pressure PWIMANI,” the presence of dew condensation is determined. If no dew condensation is determined, the process proceeds to step S 34 , and the physical model learning of the EGR valve 20 is carried out according to steps S 15 to S 18 of FIG. 3 , and the process proceeds to step S 35 . When the learning of the physical model in the case of no dew condensation is executed, an execution flag is set in step S 35 and used for determination in step S 30 which is described later. Meanwhile, if the presence of dew condensation is determined, the process proceeds to step S 26 .

In step S 26 , “partial pressure difference DPW” between the “saturated water vapor pressure PWIMANI” and “water vapor partial pressure PWIMNIEST.” This is the “partial pressure difference DPW” corresponding to the dew condensation. When “partial pressure difference DPW” is obtained, the process proceeds to step S 27 .

In step S 27 , “EGR rate difference DEGRPW” corresponding to the dew condensation in the intake pipe 10 , which corresponds to the calculated “partial pressure difference DPW”, is estimated and calculated. Since the ratio of CO2 and H2O in the exhaust gas is substantially constant in the “partial pressure difference DPW” and the “EGR rate difference DEGRPW”, the relationship between these differences can be stored in the table of the ROM 30 d . After the “EGR rate difference DEGRPW” is estimated, the process proceeds to step S 28 .

In step S 28 , “estimated EGR rate RMODELFCOND” which reflects the detection deviation of the mixed gas humidity sensor due to the dew condensation and the detection delay of the mixed gas humidity sensor 9 is calculated by subtracting the “EGR rate difference DEGRPW” from the “EGR rate RMODEF” subjected to the delay processing. The actual ratio of H2O and CO2 changes, as illustrated in FIG. 7 , when the dew condensation occurs, but the ratio is constant in the humidity sensor, so that it is possible to simulate the behavior of the humidity sensor by subtracting the “EGR rate difference DEGRPW” from the “EGR rate RMODEF.”

FIGS. 7A to 7C schematically illustrate the composition of the mixed gas in the intake pipe 10 . Note that the area of each component is changed in the drawings for convenience of explanation, so that the area ratio is different from the actual ratio. FIG. 7A illustrates a case where only air is present, FIG. 7B illustrates a case where air and the EGR gas are mixed, and FIG. 7C illustrates a case where the dew condensation is present in addition to air and the EGR gas.

As illustrated in FIG. 7B , moisture in the air and moisture generated as a result of combustion are present in the mixed gas including the EGR gas. The ratio of H2O to CO2 in the EGR gas is substantially constant. Therefore, it is possible to estimate the EGR amount on the basis of the difference between the humidity in the air and the humidity inside the intake pipe. However, when the dew condensation occurs, as illustrated in FIG. 7C , only H2O decreases and the actual ratio of H2O and CO2 changes. Therefore, when the EGR rate is calculated on the basis of the moisture content measured by the mixed gas humidity sensor 9 , the EGR rate is underestimated, as it is determined that CO2 is also decreasing. In step S 608 , therefore, the EGR amount including CO2 is subtracted in addition to the amount of moisture. After the “estimated EGR rate RMODELFCOND” is calculated, the process proceeds to step S 29 .

In step S 29 , the “difference DREGR” between the “EGR rate RVAP” and the “estimated EGR rate RMODELFCOND” is calculated. Note that the “difference DREGR” corresponds to the dew condensation amount. After the “difference DREGR” is obtained, the process proceeds to step S 30 .

In step S 30 , it is determined whether the learning correction of the “correction coefficient KAEGRV” of the EGR valve opening area has already been carried out in a predetermined period in the past. This determination can be made by referring to the above-described execution flag. If the learning correction has been carried out, the process proceeds to step S 31 . If the learning correction has not been carried out, the process proceeds to step S 33 .

In step S 31 , “dew condensation amount DCOND” corresponding to the “difference DREGR” is calculated. The “dew condensation amount DCOND” corresponding to the “difference DREGR” needs to be stored in advance in the table of the ROM 30 d . After the “dew condensation amount DCOND” is obtained, the process proceeds to step S 32 .

In step S 32 , the learning correction in which “cooling correction coefficient KHEATEGR” corresponding to the “dew condensation amount DCOND” is applied to the heat transfer parameter relating to the intercooler 6 of the physical model, thus improving the estimation accuracy of the EGR rate by the physical model.

If no learning correction of the EGR valve is performed in step S 30 , the EGR valve is fully closed or the opening degree is controlled to decrease in step S 33 , so that excessive EGR gas supply is suppressed and, simultaneously, the ignition timing is retarded to suppress occurrence of the abnormal combustion.

As described above, estimating the dew condensation amount in the physical model makes it possible to improve the estimation accuracy of the EGR rate even under a condition in which the dew condensation is likely to occur, thereby improving fuel consumption. Further, when the learning condition is not satisfied, the opening degree of the EGR valve is decreased and the ignition timing is retarded, so that excessive EGR is introduced to suppress misfire and abnormal combustion.

Third Embodiment

Next, a third embodiment of the present invention is described. Although the humidity sensor is used as the EGR sensor in the first embodiment, the present embodiment is different from the first embodiment in that the intake-side oxygen concentration sensor is used as the EGR sensor. The intake-side oxygen concentration sensor can measure the oxygen concentration in the mixed gas in the intake pipe 10 , whereby the EGR rate can be measured.

FIG. 8 illustrates a configuration of the internal combustion engine system according to the present embodiment, and an intake-side oxygen concentration sensor 25 is disposed in the intake pipe 10 . The intake-side oxygen concentration sensor 25 is capable of outputting an output signal that is linear with respect to the oxygen concentration, whereby the EGR rate of the mixed gas can be obtained. Other portions of the configuration are the same as those illustrated in FIG. 1 and are not described in detail.

FIG. 9 illustrates a control flow relating to the learning correction of the EGR rate in the third embodiment of the present invention. The control steps of the steps S 41 to S 46 are the same as the control of the steps S 13 to S 18 of FIG. 3 and are not described in detail.

In step S 40 of FIG. 9 , “EGR rate RIMO2” in the intake pipe 10 is calculated from the oxygen concentration detected by the intake-side oxygen concentration sensor 25 . Since various methods for calculating the EGR rate from the oxygen concentration have been proposed, they may be obtained by these known methods. When “EGR rate RIMO2” in the intake pipe 10 is obtained, the process proceeds to step s 1 . Thereafter, the “EGR rate RVAP” in FIG. 3 is replaced by the “EGR rate RIMO2,” and the same control steps as in FIG. 3 are executed.

As described above, by comparing the EGR rate, which is calculated on the basis of the value of the intake-side oxygen concentration sensor (i.e., EGR sensor) attached to the intake pipe, and the EGR rate, which is estimated from the air flow sensor or the like on the basis of the physical model and subjected to the response delay processing, and executing the learning correction on the opening area of the EGR valve of the physical model on the basis of the deviation between the EGR rates, the estimation accuracy of the EGR rate according to the physical model is improved.

Fourth Embodiment

Next, a fourth embodiment of the present invention is described. Although a humidity sensor is used as the EGR sensor in the second embodiment, this embodiment is different from the second embodiment in that an intake-side oxygen concentration sensor is used as the EGR sensor. As described above, the intake-side oxygen concentration sensor can measure the oxygen concentration in the mixed gas in the intake pipe 10 , whereby the EGR rate can be measured.

FIG. 10 illustrates a control flow relating to the learning correction of the EGR rate according to the fourth embodiment of the present invention. The control steps of the steps S 51 to S 62 are the same as the control of the steps S 23 to S 35 of FIG. 6 and are not described in detail.

In step S 50 of FIG. 10 , the “EGR rate RIMO2” in the intake pipe 10 is calculated from the oxygen concentration detected by the intake-side oxygen concentration sensor 25 . As in other embodiments, various methods have been proposed to calculate the EGR rate from the oxygen concentration in the present embodiment, and the EGR rate needs to be determined by any of those known methods. When “EGR rate RIMO2” in the intake pipe 10 is obtained, the process goes to step S 51 . Thereafter, the same control steps as those in FIG. 6 are executed.

As described above, estimating the dew condensation amount in the physical model makes it possible to improve the estimation accuracy of the EGR rate even under a condition in which the dew condensation is likely to occur, thereby improving fuel consumption. Further, when the learning condition is not satisfied, the opening degree of the EGR valve is decreased and the ignition timing is retarded, so that excessive EGR is introduced to suppress misfire and abnormal combustion.

As described above, in the present invention, the first EGR rate is obtained using, as an input, the detection signal of the EGR sensor, which is provided on the downstream side of the throttle valve adjusting the flow rate of the mixed gas of air and the EGR gas flowing through the intake pipe, the second EGR rate is estimated by carrying out predetermined calculations using, as an input, the detection signals of at least the air flow sensor and the EGR valve opening degree sensor, the third EGR rate is determined by carrying out the delay processing on the second EGR rate corresponding to the response delay of the EGR sensor, and the learning correction is carried out on the second EGR rate by reflecting the difference between the third EGR rate and the first EGR rate.

Thus, it is possible to estimate the EGR rate with high accuracy even in the transient state. As a result, an appropriate amount of EGR gas is introduced into the combustion chamber to avoid abnormal combustion, whereby deterioration of fuel consumption can be prevented.

The present invention is not limited to the embodiments described above, and may include various other modifications. For example, the above embodiments have been described in detail to facilitate the understanding of the present invention, and all constituent components are not necessarily provided. Some of the constituent components of a certain embodiment can be partially replaced by those of other embodiments, or the constituent components of other embodiments can be added to a certain embodiment. Further, some of the constituent components of each embodiment may be added, deleted, or substituted for by other constituent components.

1 ECU

2 accelerator opening degree sensor

3 air flow sensor

4 first throttle valve

5 supercharger

6 intercooler

7 second throttle valve

8 intake pipe pressure sensor

9 mixed gas humidity sensor

12 fuel injection valve

13 ignition plug

19 differential pressure sensor

20 EGR valve

21 EGR cooler

24 atmospheric humidity sensor

25 intake-side oxygen concentration sensor