WO2013132300A2

WO2013132300A2 - Fuel injection characteristic learning device for internal combustion engine

Info

Publication number: WO2013132300A2
Application number: PCT/IB2013/000210
Authority: WO
Inventors: Ryo Suzuki; Yoshiyasu Ito; Takeshi Miyaura
Original assignee: Toyota Jidosha Kabushiki Kaisha
Priority date: 2012-03-09
Filing date: 2013-02-19
Publication date: 2013-09-12
Also published as: JP2013185548A; CN104160136B; WO2013132300A3; CN104160136A; JP5938955B2; DE112013001352T5; WO2013132300A8

Abstract

A fuel injection characteristic learning device includes a pressure sensor (51) that detects a fuel pressure inside a fuel supply system including a fuel injection valve (20), and the device executes a learning processing of learning an operation characteristic of the fuel injection valve (20) on the basis of the fuel pressure detected by the pressure sensor (51). A main processing unit (42) executes a determination processing of determining whether or not the learning processing is able to be executed, on a basis of an engine parameter. A subsidiary processing unit (41) receives a determination result of the determination processing from the main processing unit (42), and executes the learning processing on the basis of the received determination result.

Description

FUEL INJECTION CHARACTERISTIC LEARNING DEVICE FOR INTERNAL

COMBUSTION ENGINE

BACKGROUND OF THE INVENTION

1. Field of the Invention

[0001] The invention relates to a fuel injection characteristic learning device for an internal combustion engine, which learns an operation characteristic of a fuel injection valve on the basis of a fuel pressure in a fuel supply system.

2. Description of Related Art

[0002] An internal combustion engine is provided with a fuel supply system that is constituted by a supply passage to which pressurized fuel is supplied, a fuel injection valve that is connected to the supply passage, and the like. In recent years, there have been proposed devices that are each provided with a pressure sensor for detecting a fuel pressure inside such a fuel supply system and learn an operation characteristic of a fuel injection valve on the basis of the fuel pressure detected by the pressure sensor (see Japanese Patent Application Publication No. 2009-57928 (JP-2009-57928 A)). In carrying out fuel injection, the fuel pressure in the fuel supply system fluctuates, that is, the fuel pressure falls as the fuel injection valve starts to be opened, and then rises as the fuel injection valve is closed. In the device described in Japanese Patent Application Publication No. 2009-57928 (JP-2009-57928 A), the operation characteristic of the fuel injection valve is estimated and learned on the basis of a manner of such fluctuations in the fuel pressure inside the fuel supply system.

[0003] Meanwhile, in the case where the operation characteristic of the fuel injection valve is learned on the basis of the fuel pressure detected by the pressure sensor, the interval of detecting a fuel pressure is shortened to make it possible to finely grasp the manner of fluctuations in the fuel pressure. Therefore, the operation characteristic of the fuel injection valve can be accurately learned. In such a case, however, a large arithmetic load is applied to a processing unit in executing a processing regarding the learning of the operation characteristic of the fuel injection valve.

[0004] Besides, in the case where the operation characteristic of the fuel injection valve is estimated simply on the basis of the manner of fluctuations in fuel pressure, the accuracy of learning the operation characteristic may become extremely low when an internal combustion engine is in a certain operation state, for example, when the fuel pressure abruptly changes as a result of a change in engine operation state. In order to suppress such a deterioration in learning accuracy, it is conceivable to determine, on the basis of an engine parameter grasped by the processing unit, whether or not the learning of the aforementioned operation characteristic can be executed. In this device, although the accuracy of learning the aforementioned operation characteristic is restrained from deteriorating, further concentration of the arithmetic load on the processing unit is caused. Such concentration of the arithmetic load leads to a decrease in the degree of freedom (flexibility) in setting the arithmetic processing, and hence is undesirable.

SUMMARY OF THE INVENTION

[0005] The invention provides a fuel injection characteristic learning device for an internal combustion engine, which suppresses both the concentration of an arithmetic load on a processing' unit and a deterioration in the accuracy of learning an operation characteristic of a fuel injection valve.

[0006] An aspect of the invention relates to a fuel injectio characteristic learning device for an internal combustion engine, the fuel injection characteristic learning device including a pressure sensor that detects a fuel pressure inside a fuel supply system including a fuel injection valve. The fuel injection characteristic learning device executes a learning processing of learning an operation characteristic of the fuel injection valve on a basis of the fuel pressure detected by the pressure sensor. The fuel injection characteristic learning device includes a first processing unit that executes a determination processing of determining whether or not the learning processing is able to be executed, on a basis of an engine parameter; and a second processing unit that receives a determination result of the determination processing from the first processing unit, and executes the learning processing on a basis of the received determination result.

[0007] In the aforementioned device, when the learning of the operation characteristic of the fuel injection valve is executed, the determination processing and the learning processing are executed by the separate processing units. Therefore, the arithmetic load is restrained from concentrating on a specific processing unit. Moreover, if it is determined that the learning processing is not able to be executed on the basis of the engine parameter, in the determination processing executed by the first processing unit, the learning processing is not executed by the second processing unit. If it is determined that the learning processing is able to be executed in the determination processing, the learning processing is executed by the second processing unit. Thus, when the internal combustion engine is in an engine operation state in which the learning accuracy of the learning processing may become low enough to cause a problem, the learning processing is prohibited from being executed. As a result, the accuracy of learning the operation characteristic of the fuel injection valve is restrained from deteriorating.

[0008] In the fuel injection characteristic learning device according to the above-described aspect, the second processing unit may execute, as the learning processing, a processing of forming a fluctuation waveform of the fuel pressure that is detected by the pressure sensor when the fuel injection valve is operated to be opened, and learning a learning correction term on a basis of a relationship between the fluctuation waveform and a predetermined basic waveform.

[0009] In the device with the above configuration, the processing of forming the fluctuation waveform of the fuel pressure is executed on the basis of the fuel pressure detected by the pressure sensor. Therefore, when executing the learning processing, the arithmetic load that is applied to the processing unit is likely to become large, as compared to a device that executes a learning processing simply on the basis of one of detection values of a pressure sensor. In the aforementioned device, the arithmetic load is restrained from concentrating on a specific one of the processing units. Therefore, it is possible to suppress a decrease in the degree of freedom (flexibility) in setting the arithmetic processing regarding the learning processing or the determination processing.

[0010] In the fuel injection characteristic learning device according to the above-described aspect, the internal combustion engine may include a plurality of cylinders, the fuel injection valves may be provided for the respective cylinders of the internal combustion engine, and the second processing unit may separately execute the learning processing for each of the fuel injection valves that are provided for the respective cylinders of the internal combustion engine.

[0011] In the device with the above configuration, the learning processing is executed for each of the plurality of the fuel injection valves. Therefore, in the device in which the arithmetic load applied to the processing units is likely to become large when executing the learning processing, the arithmetic load is restrained from concentrating on a specific one of the processing units.

[0012] In the fuel injection characteristic learning device according to the above-described aspect, the pressure sensor may be formed integrally with each of the fuel injection valves that are provided for the respective cylinders.

[0013] In the device with the above configuration, a fuel pressure in a region close to an injection hole of the fuel injection valve can be detected, as compared to a device in which a fuel pressure at a position spaced apart from a fuel injection valve is detected by a pressure sensor. Moreover, since a fuel pressure at a position spaced apart from a fuel injection valve of another cylinder can be detected, the influence of pressure fluctuations resulting from the opening/closing operation of the fuel injection valve of the other cylinder is small. In consequence, changes in the fuel pressure inside the fuel injection valve due to the opening/closing operation of the fuel injection valve can be detected with high accuracy by the pressure sensor that is integrally fitted to the fuel injection valve. The operation characteristic of the fuel injection valve can be learned with high accuracy on the basis of the pressure.

[0014] In the learning processing, when the amount of fluctuations in the fuel pressure resulting from a factor other than the opening/closing operation of the fuel injection valve (i.e., resulting from a disturbance) is large, the accuracy of learning the operation characteristic of the fuel injection valve through the learning processing is highly likely to deteriorate.

[0015] In the device according to the above-described aspect, the engine parameter may include the fuel pressure. In this device, the determination processing is executed on the basis of the fuel pressure. Therefore, it is possible to properly determine whether or not the learning processing is able to be executed, on the basis of the magnitude of the amount of fluctuations in fuel pressure resulting from the disturbance, in the determination processing.

BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Features, advantages, and technical and industrial significance of an exemplary embodiment of the invention will be described below with reference to the accompanying drawings, in which like numerals denote like elements, and wherein:

FIG. 1 is a schematic diagram showing an overall configuration of a fuel injection characteristic learning device for an internal combustion engine according to the embodiment of the invention;

FIG. 2 is a sectional view showing a sectional structure of each of fuel injection valves;

FIG. 3 is a time chart showing an example of a basic time waveform;

FIG. is a schematic diagram showing how an electronic control unit is connected to the fuel injection valves;

FIG. 5 is a flowchart showing a procedure of executing a determination processing; FIG. 6 is a flowchart showing a procedure of executing a characteristic learning control processing; and

FIG. 7 is a timing chart showing an example of a mode of executing the determination processing.

DETAILED DESCRIPTION OF EMBODIMENT [0017] A fuel injection characteristic learning device for an internal combustion engine according to one embodiment of the invention will be described hereinafter. As shown in FIG. 1, an intake passage 12 is connected to cylinders 11 of an internal combustion engine 10. Air is sucked into the cylinders 11 of the internal combustion engine 10 via the intake passage 12. A diesel engine including a plurality of cylinders 11 (four cylinders (#1, #2, #3, and #4) in this embodiment of the invention) is adopted as this internal combustion engine 10. Direct injection fuel injection valves 20 are provided for the respective cylinders 11 (#1 to #4) of the internal combustion engine 10. The fuel injection valves 20 directly inject fuel into the cylinders 11 respectively. Fuel that has been injected by opening each of these fuel injection valves 20 comes into contact with intake air that has been compressed and heated in a corresponding one of the cylinders 11 of the internal combustion engine 10, ignites, and burns. Then in the internal combustion engine 10, each of pistons 13 is pressed down due to energy that is generated as the fuel in a corresponding one of the cylinders 11 burns, and a crankshaft 14 as an engine output shaft is forcibly caused to rotate. The combustion gas that has burned in the cylinders 11 of the internal combustion engine 10 is discharged, as exhaust gas, to an exhaust passage 15 of the internal combustion engine 10.

[0018] Each of the fuel injection valves 20 is separately connected to a common rail 34 via a corresponding one of branch passages 31a. The common rail 34 is connected to a fuel tank 32 via a supply passage 31b. This supply passage 31b is provided with a fuel pump 33 that force-feeds fuel. In this embodiment of the invention, fuel that has been pressurized by being force-fed by the fuel pump 33 is accumulated in the common rail 34, and is supplied into the respective fuel injection valves 20. In this embodiment of the invention, the fuel injection valves 20, the branch passages 31a, the supply passage 31b, the fuel pump 33, and the common rail 34 function as a fuel supply system.

[0019] Besides, return passages 35 are connected to the fuel injection valves 20 respectively. Each of the return passages 35 is connected to the fuel tank 32. Part of the fuel inside the fuel injection valves 20 is returned to the fuel tank 32 via these return passages 35 respectively.

[0020] The internal structure of each of the fuel injection valves 20 will be described hereinafter. As shown in FIG. 2, a needle valve 22 is provided inside a housing 21 of the fuel injection valve 20. This needle valve 22 is provided in the housing 21 to be movable in a reciprocating manner (movable vertically in FIG. 2). A spring 24 that constantly urges the aforementioned needle valve 22 toward injection holes 23 (downward in FIG. 2) is provided inside the housing 21. Besides, a nozzle chamber 25 is formed inside the housing 21, at a position located on one side (on a lower side in FIG. 2) with respect to the aforementioned needle valve 22, and a pressure chamber 26 is formed inside the housing 21 at a position located on the other side (on an upper side in FIG. 2) with respect to the aforementioned needle valve 22.

[0021] The injection holes 23, which establish communication between the interior of the nozzle chamber 25 and the exterior of the housing 21, are formed through the nozzle chamber 25. Fuel is supplied to the nozzle chamber 25 from a corresponding one of the branch passages 31a (the common rail 34) via an introduction passage 27. The aforementioned nozzle chamber 25 and the aforementioned branch passage 31a (the common rail 34) are connected to the pressure chamber 26 via a communication channel 28. Besides, the pressure chamber 26 is connected to a corresponding one of the return passages 35 (the fuel tank 32) via a discharge channel 30.

[0022] An electrically driven valve is adopted as the aforementioned fuel injection valve 20. More specifically, a piezoelectric actuator 29 that is obtained by stacking piezoelectric elements (e.g., piezo elements), which expand and contract when a driving signal is input thereto, is provided inside the housing 21 of the fuel injection valve 20. This piezoelectric actuator 29 is provided with a valve body 29a. This valve body 29a is provided inside the pressure chamber 26. Then, as the valve body 29a

C

moves through the operation of the piezoelectric actuator 29, one of the communication channel 28 (the nozzle chamber 25) and the discharge channel 30 (the return passage 35) is selectively brought into communication with the pressure chamber 26.

[0023] In this fuel injection valve 20, when a valve-closing signal is input to the piezoelectric actuator 29, the piezoelectric actuator 29 contracts to move the valve body 29, whereby the communication channel 28 and the pressure chamber 26 are brought into communication with each other, and the communication between the return passage 35 and the pressure chamber 26 is interrupted. Thus, the fuel in the pressure chamber 26 is prohibited from being discharged to the return passage 35 (the fuel tank 32), and the nozzle chamber 25 and the pressure chamber 26 are brought into communication with each other. As a result, the pressure difference between the nozzle chamber 25 and the pressure chamber 26 becomes considerably small, and the needle valve 22 moves to such a position as to block up the injection holes 23 due to an urging force of the spring 24. Therefore, the fuel injection valve 20 is brought to a state in which fuel is not injected (a closed valve state) at this time.

[0024] On the other hand, when a valve-opening signal is input to the piezoelectric actuator 29, the piezoelectric actuator 29 expands to move the valve body 29a, whereby the communication between the communication channel 28 and the pressure chamber 26 is interrupted, and the return passage 35 and the pressure chamber 26 are brought into communication with each other. Thus, fuel is prohibited from flowing out from the nozzle chamber 25 to the pressure chamber 26, and part of the fuel in the pressure chamber 26 is returned to the fuel tank 32 via the return passage 35. As a result, the pressure of the fuel in the pressure chamber 26 falls, the pressure difference between the pressure chamber 26 and the nozzle chamber 25 increases, and the needle valve 22 moves away from the injection holes 23 against the urging force of the spring 24 due to the pressure difference. Therefore, the fuel injection valve 20 is brought to a state in which fuel is injected (an open valve state) at this time.

[0025] A pressure sensor 51 that detects a fuel pressure PQ inside the aforementioned introduction passage 27 is integrally fitted to the fuel injection valve 20 (in other words, the pressure sensor 51 is formed integrally with the fuel injection valve 20). Thus, a fuel pressure in a region that is close to the injection holes 23 of the fuel injection valve 20 can be detected, as compared to a device in which a fuel pressure at a position spaced apart from the fuel injection valve 20 is detected, for example, a fuel pressure in the common rail 34 (see FIG. 1) is detected. As a result, changes in the fuel pressure inside the fuel injection valve 20 due to the opening operation of the fuel injection valve 20 can be accurately detected. This pressure sensor 51 includes a sensor body 51 A that outputs a signal corresponding to a fuel pressure, and a memory 5 IB that stores a detection value of the sensor body 51 A. One pressure sensor 51 is provided for each of the fuel injection valves 20, namely, for each of the cylinders 11 (#1 to #4) of the internal combustion engine 10.

[0026] As shown in FIG. 1, the internal combustion engine 10 is provided, as peripheral devices thereof, with various sensors for detecting an operation state. As these sensors, for example, an intake air amount sensor 52 for detecting an amount of air passing through the intake passage 12 (a passage air amount GA), and a crank sensor 53 for detecting a rotational speed of the crankshaft 14 (an engine rotational speed NE) are provided in addition to the aforementioned pressure sensor 51. Moreover, an accelerator sensor 54 for detecting an operation amount (an accelerator operation amount ACC) of an accelerator operation member (e.g., an accelerator pedal), a fuel temperature sensor 55 for detecting a temperature THQ of fuel, a coolant temperature sensor 56 for detecting a temperature THW of coolant, and the like are also provided.

[0Q27] Besides, an electronic control unit 40 configured to include processing units, and the like are also provided as peripheral devices of the internal combustion engine 10. This electronic control unit 40 takes in output signals of the various sensors, executes various arithmetic operations on the basis of the output signals, and performs various kinds of control regarding the operation of the internal combustion engine 10, such as operation control for the fuel injection valves 20 (injection amount control) and operation control for the fuel pump 33 (injection pressure control) on the basis of results of the arithmetic operations. The electronic control unit 40 includes two processing units, namely, a subsidiary processing unit 41 and a main processing unit 42. The functions of these processing units, namely, the subsidiary processing unit 41 and the main processing unit 42 will be described later in detail.

[0028] In this embodiment of the invention, injection amount control is performed as follows. That is, first of all, an injection pattern is selected and various control target values are calculated as to each injection according to the injection pattern, on the basis of values that are correlated with the operation state of the internal combustion engine 10 (so-called engine parameters), such as the passage air amount GA, the engine rotational speed NE, and the accelerator operation amount ACC. In this embodiment of the invention, a plurality of injection patterns obtained by combining main injection, pilot injection, after injection and the like are set in advance, and one of these injection patterns is selected when performing injection amount control. Besides, as the various control target values, target values regarding fuel injection amounts of the respective kinds of injection such as main injection, pilot injection and after injection (target injection amounts), a target value regarding an injection timing of main injection (a target injection timing), an interval between main injection and pilot injection (a pilot interval), and an interval between main injection and after injection (an after interval) are calculated. In this embodiment of the invention, relationships between the engine operation state determined by the aforementioned engine parameters and the respective control target values suited for the operation state, and a relationship between the aforementioned engine operation state and an injection pattern suited for the operation state are obtained in advance on the basis of a result of an experiment or a simulation, and are stored into the main processing unit 42 of the electronic control unit 40. Then, the main processing unit 42 separately sets various control target values and an injection pattern by using the aforementioned relationships, on the basis of the engine parameters at each time.

[0029] Then, a control target value regarding the open valve period of the fuel injection valves 20 (a target injection period TAU) is set by using a model formula, on the basis of the aforementioned target injection amount and the fuel pressure PQ. In this embodiment of the invention, a physical model is structured by modeling the fuel supply system including the common rail 34, the branch passages 31a, the fuel injection valves 20, and the like, and the aforementioned target injection period TAU is calculated through the physical model. More specifically, a model formula including a target injection amount, the fuel pressure PQ, a later-described learning correction term, and the like as variables is determined and stored into the main processing unit 42 in advance, and the target injection period TAU is calculated through the model formula.

[0030] Then, a driving signal is output from the electronic control unit 40 in accordance with a target injection timing and the target injection period TAU, and the fuel injection valves 20 are separately operated to be opened on the basis of the driving signal that is input. Thus, an amount of fuel matching an engine operation state at each time is injected from each of the fuel injection valves 20 according to an injection pattern suited for the engine operation state, and is supplied into a corresponding one of the cylinders 11 of the internal combustion engine 10. Therefore, a rotational torque matching the engine operation state is imparted to the crankshaft 14.

[0031] In this embodiment of the invention, a learning processing of learning the operation characteristics of the fuel injection valves 20 is executed on the basis of the fuel pressures PQ detected by the pressure sensors 51. In this learning processing, first of all, a basic time waveform regarding a fuel injection rate is calculated on the basis of various calculation parameters such as the target injection amount, the target injection timing, and the fuel pressure PQ. In this embodiment of the invention, a relationship between an engine operation range determined by those calculation parameters and a basic time waveform suited for the operation range is obtained in advance on the basis of results of various experiments and simulations, and is stored into the subsidiary processing unit 41 of the electronic control unit 40. Then, the subsidiary processing unit 41 calculates a basic time waveform by using the aforementioned relationship, on the basis of the various calculation parameters.

[0032] FIG. 3 shows an example of the aforementioned basic time waveform. As indicated by a solid line in FIG. 3, as the basic time waveform, a trapezoidal waveform is set. The trapezoidal waveform is defined by a timing at which an opening operation of the fuel injection valve 20 is started (a valve-opening operation start timing To), a speed of increase in the fuel injection rate after the start of the opening operation of the fuel injection valve 20 (an injection rate increasing speed Vo), a timing at which a closing operation of the fuel injection valve 20 is started (a valve-closing operation start timing Tc), a speed Vc of decrease in the fuel injection rate after the start of the closing operation of the fuel injection valve 20, and a maximum value of the fuel injection rate (a maximum fuel injection rate Rm). In this embodiment of the invention, the basic time waveform functions as a predetermined basic waveform.

[0033] On the other hand, a time waveform of an actual fuel injection rate (a detection time waveform) is formed on the basis of the fuel pressure PQ detected by the pressure sensor 51. More specifically* first of all, a valve-opening operation start timing Tor of the fuel injection valve 20, an injection rate increasing speed Vor, a valve-closing operation start timing Tcr, an injection rate decreasing speed Vcr, and a maximum injection rate Rmr are determined on the basis of a change in the fuel pressure PQ. The fuel pressure inside the fuel injection valve 20 (more specifically, the nozzle chamber 25) falls with increase in lift amount when the fuel injection valve 20 is operated to be opened, and then rises with decrease in lift amount when the fuel injection valve 20 is operated to be closed. In this embodiment of the invention, the valve-opening operation start timing Tor, the injection rate increasing speed Vor, the valve-closing operation start timing Tcr, the injection rate decreasing speed Vcr, and the maximum injection rate Rmr as mentioned above are accurately determined on the basis of the change in the fuel pressure inside the fuel injection valve 20 (more specifically, the fuel pressure PQ). Then, as indicated by an alternate long and short dash line in FIG. 3, a time waveform of an actual fuel injection rate (a detection time waveform) is formed by the determined values. In this embodiment of the invention, the detection time waveform functions as a fluctuation waveform of a fuel pressure detected by a pressure sensor at the time when a fuel injection valve is operated to be opened.

[0034] In the learning processing, a learning correction term is learned on the basis of a relationship between the detection time waveform and the basic time waveform. That is, first of all, the detection time waveform and the basic time waveform are compared with each other during the operation of the internal combustion engine 10, and differences in the parameters (waveform parameters) between those waveforms (i.e., differences between values of the waveform parameters of the detection time waveform and values of the waveform parameters of the basic time waveform) are sequentially calculated. More specifically, as the differences in the waveform parameters (the differences between the values of the waveform parameters of the detection time waveform and the values of the waveform parameters of the basic time waveform), a difference ATog (= To-Tor) in the valve-opening operation start timing, a difference AVog (=Vo-Vor) in the injection rate increasing speed, a difference ATcg (= Tc-Tcr) in the valve-closing operation start timing, a difference AVcg (= Vc-Vcr) in the injection rate decreasing speed, and a difference ARmg (= Rm-Rmr) in the maximum injection rate are calculated. Then, these differences ATog, AVog, ATcg, AVcg, and ARmg are stored into the subsidiary processing unit 41 as learning correction terms for compensating for a variation of the operation characteristic among the fuel injection valves 20.

[0035] In this embodiment of the invention, these learning correction terms (ATog, AVog, ATcg, AVcg, and ARmg) are used as calculation parameters for calculating the target injection period TAU on the basis of the above-stated model formula. By thus calculating the target injection period TAU, an influential factor of the variation of the operation characteristic among the fuel injection valves 20 is compensated for. In this embodiment of the invention, a processing of calculating learning correction terms on the basis of the fuel pressure PQ is executed for each of the cylinders 11 (#1 to #4) of the internal combustion engine 10 on the basis of an output signal of a corresponding one of the pressure sensors 51. Besides, in the device according to this embodiment of the invention, a plurality of learning ranges that are defined by the fuel pressure PQ and the fuel injection amount (more specifically, the target injection amount) are determined, and learning correction terms are learned and stored for each of those ranges.

[0036] In this embodiment of the invention, injection pressure control is performed as follows. That is, first of all, a control target value regarding the fuel pressure in the common rail 34 (a target fuel pressure) is calculated on the basis of the passage air amount GA and the engine rotational speed NE, and an operation amount of the fuel pump 33 (a fuel force-feed amount or a fuel return amount) is adjusted such that the actual fuel pressure becomes equal to the target fuel pressure. Through this adjustment of the operation amount of the fuel pump 33, the fuel pressure in the common rail 34, in other words, the fuel injection pressure of each fuel injection valve 20 is adjusted to a pressure corresponding to the engine operation state.

[0037] As shown in FIG. 4, all the pressure sensors 51 that are provided for the cylinders 11 (#1 to #4) of the internal combustion engine 10 respectively are connected to the subsidiary processing unit 41. This subsidiary processing unit 41 executes the learning processing.

[0038] Two (#1 and #4) of the pressure sensors 51 that are provided for the cylinders 11 (#1 to #4) of the internal combustion engine 10 respectively are connected to the main processing unit 42. As indicated by a blank arrow in FIG. 4, the device according to this embodiment of the invention is configured such that the subsidiary processing unit 41 of the electronic control unit 40 and the main processing unit 42 of the electronic control unit 40 are connected to each other by a signal line, and that data can be transferred between the subsidiary processing unit 41 and the main processing unit 42.

[0039] When calculating the target injection period TAU, the main processing unit 42 executes an arithmetic processing of reading learning correction terms from the subsidiary processing unit 41, and an arithmetic processing of calculating the target injection period TAU from a model formula on the basis of the learning correction terms.

[0040] Besides, the main processing unit 42 executes processings regarding injection pressure control, such as an arithmetic processing of calculating a target fuel pressure on the basis of engine parameters, and an arithmetic processing of adjusting the operation amount of the fuel pump 33 such that the target fuel pressure and the actual fuel pressure PQ (more specifically, the higher value of the fuel pressures PQ detected by the two pressure sensors 51 (#1 and #4) that are connected to the main processing unit 42) coincide with each other.

[0041] It should be noted herein that a processing of forming a detection time waveform on the basis of the fuel pressure PQ detected by each pressure sensor 51 is executed in the device according to this embodiment of the invention. Thus, when executing the learning processing, the arithmetic load that is applied to the electronic control unit 40 (more specifically, the subsidiary processing unit 41 and the main processing unit 42) is likely to become large, as compared to a device that learns the operation characteristic of the fuel injection valve 20 simply on the basis of one of detection values of the pressure sensor 51. Besides, in the device according to this embodiment of the invention, the learning processing is separately executed for each of the plurality of the fuel injection valves 20. Therefore, from this standpoint as well, it can be said that the arithmetic load applied to the electronic control unit 40 is likely to become large.

[0042] Besides, in the device according to this embodiment of the invention, if the learning processing is executed simply on the basis of the fuel pressures PQ detected by the pressure sensors 51, the accuracy of learning the operation characteristics of the fuel injection valves 20 in the learning processing may become extremely low when the internal combustion engine 10 is in a certain operation state, for example, when the fuel pressures PQ abruptly change as a result of a change in the engine operation state. In order to suppress such a deterioration in the learning accuracy, it is desirable to execute a determination processing of determining, on the basis of engine parameters, whether or not the learning processing is able to be executed. By executing this determination processing, the learning accuracy in the learning processing is restrained from deteriorating, but further concentration of the arithmetic load on the electronic control unit 40 is caused. In addition, such concentration of the arithmetic load leads to a decrease in the degree of freedom (flexibility) in setting the various arithmetic processings and hence is undesirable.

[0043] In view of the above, in the device according to this embodiment of the invention, the main processing unit 42 executes the aforementioned determination processing, namely, an arithmetic processing of determining, on the basis of the engine parameters (more specifically, the engine rotational speed NE, the target injection amount, the accelerator operation amount ACC, and the fuel pressures PQ), whether or not the learning processing is able to be executed. On the other hand, the subsidiary processing unit 41 receives and reads a determination result of the aforementioned determination processing from the main processing unit 42 through data communication, and executes the learning processing on the basis of the determination result.

[0044] Thus, when learning the operation characteristics of the fuel injection valves 20, the determination processing and the learning processing are executed by the separate processing units 41 and 42. Therefore, the arithmetic load is distributed to the respective processing units 41 and 42, and thus is restrained from concentrating on one of the processing units. Moreover, in the case where the result of the determination processing executed by the main processing unit 42 on the basis of the engine parameters is that the aforementioned learning processing is not able to be executed, the subsidiary processing unit does not execute the learning processing. In the case where the result of the determination processing is that the aforementioned learning processing is able to be executed, the subsidiary processing unit 41 executes the learning processing. Thus, when the internal combustion engine 10 is in an engine operation state in which the learning accuracy of the learning processing may become low enough to cause a problem, the learning processing can be prohibited from being executed, and the accuracy of learning the operation characteristics of the fuel injection valves 20 is restrained from deteriorating.

[0045] The determination processing and the learning processing will be described hereinafter in detail. The determination processing will now be described first. FIG. 5 shows a procedure of executing the determination processing. A series of processes shown in a flowchart of FIG. 5 is executed by the main processing unit 42 as interruption handling at intervals of a predetermined period.

[0046] As shown in FIG. 5, first of all in this processing, it is determined whether or not all the following "Conditions 1 to 4" are fulfilled (steps S10 to SI 3). The "Condition 1" is that the components constituting the fuel supply system normally function (step S10). In the process of step S10, more specifically, if there is no fuel leaking out from the interior of the fuel supply system and the fuel injection valves 20, the pressure sensors 51 (#1 and #4), and the fuel pump 33 are in normal operation, it is determined that the "Condition 1" is fulfilled. The "Condition 2" is that the coolant temperature THW is within a predetermined temperature range AR1 (step Sl l). In the device according to this embodiment of the invention, a temperature range regarding the coolant temperature THW (the predetermined temperature range AR1), in which the learning processing is executed with high accuracy, is obtained in advance on the basis of results of various experiments and simulations, and is stored into the main processing unit 42. The "Condition 3" is that a fuel temperature THQ is within a predetermined temperature range AR2 (step S12). In the device according to this embodiment of the invention, a temperature range regarding the fuel temperature THQ (the predetermined temperature range AR2), in which the learning processing is executed with high accuracy, is obtained in advance on the basis of results of various experiments and simulations, and is stored into the main processing unit 42. The "Condition 4" is that the operation state of the internal combustion engine 10 that is determined by the engine parameters (the engine rotational speed NE, the target injection amount, the accelerator operation amount ACC, and the fuel pressures PQ) is other than an operation state that is rapidly changing to such an extent that a deterioration in the learning accuracy of the aforementioned learning processing causes a problem (i.e., the internal combustion engine 10 is in an operation state other than a transitional (changing) operation state, in other words, the internal combustion engine 10 is not in a transitional (changing) operation state) (step SI 3). In the device according to this embodiment of the invention, a relationship between the engine operation state determined by the aforementioned engine parameters and the aforementioned transitional operation state is obtained in advance on the basis of results of various experiments and simulations, and is stored into the main arithmetic operation unit 42. In the process of step SI 3, it is determined whether or not the internal combustion engine 10 is in a transitional operation state (i.e., it is determined whether or not the condition that the internal combustion engine 10 is in an operation state other than a transitional operation state is fulfilled, in other words, it is determined whether or not the condition that the internal combustion engine 10 is not in a transitional operation state is fulfilled), by using the aforementioned relationship, on the basis of the engine parameters.

[0047] Then, if it is determined that all the "Conditions 1 to 4" are fulfilled ("YES" in all steps S10 to SI 3), a learning execution flag is turned on (step S14). On the other hand, if it is determined that any one of the "Conditions 1 to 4" is not fulfilled ("NO" in any one of steps S10 to S13), the learning execution flag is turned off (step S15). After the learning execution flag is thus operated, the present processing is ended.

[0048] Next, a procedure for executing the learning processing, more specifically, a processing that includes the learning processing and relates to the control of learning the operation characteristics of the fuel injection valves 20 (a characteristic learning control processing) will be described. FIG. 6 shows the procedure for executing the characteristic learning control processing. A series of processes shown in a flowchart of FIG. 6 is executed by the subsidiary processing unit 41 as interruption handling at intervals of a predetermined period.

[0049] As shown in FIG. 6, first of all in this processing, a determination result of the determination processing (see FIG. 5) (more specifically, an operation state of the learning execution flag) is read from the main processing unit 42 through data communication, and it is determined whether or not the learning execution flag has been turned on (step S20).

[0050] Then, if the learning execution flag has been turned on (i.e., if the determination result is that the learning processing is able to be executed) (step S20: YES), the above-mentioned learning processing is executed (step S21). On the other hand, if the learning execution flag has been turned off (i.e., if the determination result is that the learning processing is not able to be executed) (step S20: NO), the learning processing is not executed (the process of step S21 is skipped). After the learning processing is thus executed on the basis of the operation state of the learning execution flag, the present processing is ended.

[0051] In the device according to this embodiment of the invention, when the amount of fluctuations in the fuel pressures PQ resulting from a factor other than the opening/closing operation of the fuel injection valves 20 (i.e., resulting from a disturbance) is large, the accuracy of learning the operation characteristics of the fuel injection valves 20 through the aforementioned learning processing is highly likely to deteriorate. In the aforementioned determination processing (FIG. 5), since the "Condition 4" (the process of step SI 3) is set, the determination processing is executed on the basis of the engine parameters including the fuel pressures PQ as mentioned above. Thus, it is possible to properly determine whether or not the learning processing is able to be executed, on the basis of the magnitude of the amount of fluctuations in the fuel pressures PQ resulting from the disturbance, in the determination processing. More specifically, if the amount of fluctuations in the fuel pressures PQ resulting from the disturbance is large and the learning accuracy may become low enough to cause a problem, the learning execution flag is turned off to prohibit the learning processing from being executed. On the other hand, if the amount of fluctuations in the fuel pressure PQ resulting from the disturbance is small and the learning accuracy is relatively unlikely to deteriorate, the learning execution flag is turned on to allow the learning processing to be executed.

[0052] The operation of the device according to this embodiment of the invention will be described hereinafter. In an example shown in FIG. 7, when the accelerator operation member is operated at a time Tl and the accelerator operation amount ACC ((a) in FIG. 7) rapidly increases, it is determined, through the determination processing executed by the main processing unit 42, that the internal combustion engine 10 is in a transitional operation state over a subsequent predetermined period (from a time T2 to a time T3), and the learning execution flag ((c) in FIG. 7) is turned off.

[0053] Besides, when the accelerator operation member is operated at a time T4 and the accelerator operation amount ACC rapidly decreases, it is determined, through the determination processing executed by the main processing unit 42, that the internal combustion engine 10 is in a transitional operation state over a subsequent predetermined period (from the time T4 to a time T5), and the learning execution flag is turned off.

[0054] In the device according to this embodiment of the invention, as is apparent from FIG. 7, a target fuel pressure in injection pressure control (indicated by an alternate long and short dash line at (b) of FIG. 7) and the actual fuel pressure PQ (indicated by a solid line at (b) of FIG. 7) greatly deviate from each other in the periods in which it is determined that the internal combustion engine 10 is in a transitional operation state in the determination processing (the aforementioned predetermined periods T2 to T3 and T4 to T5). Thus, if the learning processing is executed in the periods, the learning accuracy is highly likely to deteriorate. In the device according to this embodiment of the invention, the learning execution flag is turned off in those predetermined periods to prohibit the learning processing from being executed by the subsidiary processing unit 41. Thus, the learning accuracy in the learning processing is restrained from deteriorating.

[0055] As described above, according to this embodiment of the invention, effects described below are obtained. (1) The main processing unit 42 executes the determination processing, and the subsidiary processing unit 41 receives and reads a determination result of the determination processing from the aforementioned main processing unit 42 through data communication, and executes the learning processing on the basis of the determination result. Thus, it is possible to suppress both the concentration of the arithmetic load on one of the processing units 41 and 42 and a deterioration in the accuracy of learning the operation characteristics of the fuel injection valves 20.

[0056] (2) Each pressure sensor 51 for detecting the fuel pressure PQ inside the introduction passage 27 is integrally fitted to the fuel injection valve 20. Thus, changes in the fuel pressures inside the fuel injection valves 20 resulting from the opening of the fuel injection valves 20 can be detected with high accuracy, and the operation characteristics of the fuel injection valves 20 based on the fuel pressures can be learned with high accuracy.

[0057] (3) The determination processing is executed on the basis of the engine parameters including the fuel pressures PQ. Therefore, it is possible to properly determine whether or not the learning processing is able to be executed, on the basis of the magnitude of the amount of fluctuations in the fuel pressures PQ resulting from the disturbance, in the determination processing.

[0058] The foregoing embodiment of the invention may be implemented after being modified as follows. The procedure for executing the determination processing (FIG. 5) can be arbitrarily changed as long as it is possible to appropriately determine whether there is a possibility that the learning accuracy in the learning processing may become low enough to cause a problem. More specifically, at least one of the "Condition 1" (the process of step S10), the "Condition 2" (the process of step Sl l), and the "Condition 3" (the process of step SI 2) in the determination processing may be omitted.

[0059] The engine parameters used in the determination processing are not limited to the engine rotational speed NE, the accelerator operation amount ACC, the target injection amount, and the fuel pressures PQ. Any engine parameters can be selected and adopted. More specifically, for example, only some of the engine rotational speed NE, the accelerator operation amount ACC, the target injection amount, and the fuel pressures PQ may be adopted, or engine parameters other than those (e.g., a target injection timing, a target fuel pressure and the like) may be adopted.

[0060] The procedure for executing the learning processing can be arbitrarily changed as long as it is possible to learn the operation characteristics of the fuel injection valves 20 with high accuracy on the basis of the fuel pressures PQ detected by the pressure sensors 51. For example, a "Concrete Example 1" or a "Concrete Example 2" can be mentioned as the execution procedure. In the "Concrete Example 1", an actual fuel injection amount is calculated on the basis of a manner of fluctuations in the fuel pressures PQ, and a learning correction term for correcting a target injection amount is learned on the basis of a difference between the fuel injection amount and the target injection amount. In the "Concrete Example 2", a learning correction term for compensating for an influence of the variation of the operation characteristic among the fuel injection valves 20 is learned on the basis of the actual fuel pressures PQ at a specific timing when the fuel injection valves 20 are operated to be opened.

[0061] If it is possible to properly detect a pressure serving as an index of the fuel pressure inside the fuel injection valve 20 (more specifically, in the nozzle chamber 25), in other words, a fuel pressure that changes as the fuel pressure inside the fuel injection valve 20 changes, the pressure sensor 51 does not necessarily need to be directly fitted to the fuel injection valve 20. The manner in which the pressure sensor 51 is fitted can be arbitrarily changed. More specifically, the pressure sensor 51 may be fitted to a region between the common rail 34 and the fuel injection valve 20 in the fuel supply passage (the branch passage 31a), or may be fitted to the common rail 34.

[0062] Instead of each of the fuel injection valves 20 that is operated by the piezoelectric actuator 29, for example, a fuel injection valve that is operated by an electromagnetic actuator equipped with a solenoid coil or the like can also be adopted.

[0063] The invention is not limited to an internal combustion engine including four cylinders, but is also applicable to an internal combustion engine including one to three cylinders, or to an internal combustion engine including five or more cylinders. The invention is not limited to a diesel engine, but is also applicable to a gasoline engine in which gasoline fuel is used or a natural gas engine in which natural gas fuel is used.

Claims

1. A fuel injection characteristic learning device for an internal combustion engine, the fuel injection characteristic learning device including a pressure sensor (51) that detects a fuel pressure inside a fuel supply system including a fuel injection valve (20), wherein the fuel injection characteristic learning device executes a learning processing of learning an operation characteristic of the fuel injection valve (20) on a basis of the fuel pressure detected by the pressure sensor (51), the fuel injection characteristic learning device being characterized by comprising:

a first processing unit (42) that executes a determination processing of determining whether or not the learning processing is able to be executed, on a basis of an engine parameter; and

a second processing unit (41) that receives a determination result of the determination processing from the first processing unit (42), and executes the learning processing on a basis of the received determination result.

2. The fuel injection characteristic learning device according to claim 1, wherein the second processing unit (41) executes, as the learning processing, a processing of forming a fluctuation waveform of the fuel pressure that is detected by the pressure sensor (51) when the fuel injection valve (20) is operated to be opened, and learning a learning correction term on a basis of a relationship between the fluctuation waveform and a predetermined basic waveform.

3. The fuel injection characteristic learning device according to claim 1 or 2, wherein the internal combustion engine includes a plurality of cylinders, the fuel injection valves (20) are provided for the respective cylinders of the internal combustion engine, and the second processing unit (41) separately executes the learning processing for each of the fuel injection valves (20) that are provided for the respective cylinders of the internal combustion engine.

4. The fuel injection characteristic learning device according to claim 3, wherein the pressure sensor (51) is formed integrally with each of the fuel injection valves (20) that are provided for the respective cylinders.

5. The fuel injection characteristic learning device according to any one of claims 1 to 4, wherein the engine parameter includes the fuel pressure.

6. The fuel injection characteristic learning device according to claim 1, wherein if the determination result is that the learning processing is able to be executed, the second processing unit (41) executes the learning processing, and if the determination result is that the learning processing is not able to be executed, the second processing unit (41) does not execute the learning processing.

7. The fuel injection characteristic learning device according to claim 2, wherein the second processing unit (41) forms the fluctuation waveform of the fuel pressure that is detected by the pressure sensor (51) when the fuel injection valve (20) is operated to be opened and then operated to be closed, in the learning processing.

8. The fuel injection characteristic learning device according to claim 7, wherein the learning correction term is a difference between a value of a waveform parameter of the fluctuation waveform and a value of the waveform parameter of the predetermined basic waveform.

9. The fuel injection characteristic learning device according to claim 8, wherein the waveform parameter includes a timing at which an opening operation of the fuel injection valve (20) is started, a speed of increase in a fuel injection rate after start of the opening operation of the fuel injection valve (20), a timing at which a closing operation of the fuel injection valve (20) is started, a speed of decrease in the fuel injection rate after start of the closing operation of the fuel injection valve (20), and a maximum value of the fuel injection rate.