Estimated fuel temperature generation method and common rail type fuel injection controller

Abstract

To provide an estimated fuel temperature which has less variance than the prior art and is usable as a substitute value for a detected temperature by a fuel temperature sensor.SOLUTION: Map generation which enables an estimated fuel temperature to be output for input of an engine speed and a discharge correction quantity learning value is performed based upon a first basic learning value as correlation data between a fuel temperature based upon a cooling water temperature regarded as a fuel temperature and acquired immediately before a vehicle starts and a discharge correction quantity learning value acquired through learning processing in advance, a second basic learning value as correlation data between a fuel temperature after complete warming-up and the discharge correction quantity learning value, and a third basic learning value found through extrapolation operation on the first and second basic learning values. When the required engine speed and discharge correction quantity learning value are input, a corresponding estimated fuel temperature can be output.SELECTED DRAWING: Figure 7

Description

Technical Field

[0001] The present invention relates to a method for generating an estimated value of fuel temperature used for operation control of a common rail type fuel injection control device, and particularly relates to improving the accuracy of the estimated value and reducing the fitting work in the manufacturing process.

Background Art

[0002] A so-called common rail type fuel injection control device pressurizes fuel by a high-pressure pump and pumps it to a common rail which is a pressure accumulator to accumulate the pressure, and supplies the accumulated high-pressure fuel to an injector, thereby enabling injection of high-pressure fuel into an internal combustion engine by the injector, and is well known as being excellent in fuel consumption, emission characteristics, etc. (see, for example, Patent Document 1, etc.). In such a common rail type fuel injection control device, various control processes such as fuel injection amount control processing are executed, and various parameters are taken into account when executing the control process. The fuel temperature is one of the parameters that are relatively often used.

[0003] Normally, the fuel temperature is detected using a fuel temperature sensor. In recent years, in common rail type fuel injection control devices for emerging countries, those without a fuel temperature sensor have been put into practical use due to cost reduction requirements. In such a device without a fuel temperature sensor, for example, a configuration is adopted in which an alternative value calculated based on a simple method is used as the fuel temperature. As a means for generating the alternative value, for example, there is a method of detecting a change in the resistance value of an electromagnetic coil provided in an electromagnetic metering valve that adjusts the amount of fuel supplied to a high-pressure pump, and calculating an estimated value of the fuel temperature based on the detected value.

Prior Art Documents

Patent Documents

[0004]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0005] However, in the case of the method of obtaining the estimated fuel temperature based on the change in the resistance value of the electromagnetic coil described above, since a plurality of variation factors can be considered, the accuracy of the obtained estimated value may, in some cases, vary by about ±20°C with respect to the original fuel temperature. Therefore, there is a problem that its use is inevitably limited. On the other hand, among the various control processes executed in the common rail type fuel injection control device, there are some that cannot perform the desired function with an estimated value having a variation of ±20°C as described above. However, if the variation is about half of that, it does not have a significant impact on the ultimately required function.

[0006] The present invention has been made in view of the above actual situation, and provides an estimated fuel temperature generation method and a common rail type fuel injection control device that are less variable than conventional ones and are suitable for use in control processes that enable the use of an approximate estimated fuel temperature, and can be used as an alternative value for the detected temperature by a fuel temperature sensor.

Means for Solving the Problems

[0007] To achieve the object of the present invention described above, the estimated fuel temperature generation method according to the present invention is Fuel in the fuel tank (9) is pressurized and pumped to the common rail (1) by the high-pressure pump (7), and high-pressure fuel injection into the engine (3) is enabled through the fuel injection valves (2-1 to 2-n) connected to the common rail (1). At least a metering valve (6) is provided upstream of the high-pressure pump (7), and while the rail pressure of the common rail (1) can be controlled by driving and controlling at least the metering valve (6), a correction amount for correcting the variation in the discharge amount of the high-pressure pump (7) is obtained and updated by a discharge correction amount learning process, which is a learning process, as a discharge correction amount learning value. It is an estimated fuel temperature generation method in a common rail type fuel injection control device configured as described above. As for acquisition and update of the first basic learning value, For each required engine rotation speed, Immediately before starting the vehicle, the coolant water temperature is regarded as the fuel temperature and acquired. Immediately after starting the vehicle, acquisition and update of the fuel injection correction amount learning value for a plurality of preset specified fuel temperatures are performed based on the fuel injection correction amount learning value selected from the fuel injection correction amount learning values acquired in advance by the fuel injection correction amount learning process and the coolant water temperature. As for acquisition and update of the second basic learning value, For each required engine rotation speed, When it is determined that the engine is in a fully warmed-up state, the fuel temperature is acquired as a default value, and the fuel injection correction amount learning value corresponding to the engine rotation speed is selected from the fuel injection correction amount learning values acquired in advance by the fuel injection correction amount learning process. The oldest fuel injection correction amount learning value in the time series is overwritten and updated among the acquired specified number of fuel injection correction amount learning values by the selected fuel injection correction amount learning value. Then, the average value of the sum of the specified number of fuel injection correction amount learning values and the initial value is calculated, and the average value is set as the fuel injection correction amount learning value for the fuel temperature of the default value, thereby performing acquisition and update of the fuel injection correction amount learning value for the fuel temperature. As for acquisition and update of the third basic learning value, For each required engine rotation speed, Based on the first basic learning value and the second basic learning value, acquisition and update of the fuel injection correction amount learning value for the assumed minimum value and maximum value of the fuel temperature are performed by extrapolation calculation. Based on the first to third basic learning values, a map is generated that can output the fuel temperature corresponding to the input fuel injection correction amount learning value as the estimated fuel temperature for the input of the desired engine rotation speed and the fuel injection correction amount learning value corresponding to the desired engine rotation speed. When the engine rotation speed and the fuel injection correction amount learning value corresponding to the engine rotation speed are input, it is configured to output the estimated fuel temperature corresponding to the input. Further, in order to achieve the object of the present invention, the common rail type fuel injection control device according to the present invention is, Fuel in the fuel tank (9) is pressurized and pumped to the common rail (1) by the high-pressure pump (7), enabling injection of high-pressure fuel into the engine (3) via the fuel injection valves (2-1 to 2-n) connected to the common rail (1). At least a metering valve (6) is provided upstream of the high-pressure pump (7), and the electronic control unit (4) can control the rail pressure of the common rail (1) by driving and controlling at least the metering valve (6). On the other hand, a correction amount for correcting variations in the discharge amount of the high-pressure pump (7) is acquired and updated by a discharge correction amount learning process, which is a learning process, as a discharge correction amount learning value. This is a common rail type fuel injection control device configured as follows: The electronic control unit (4) is:[[]] Execute the acquisition and update of the first basic learning value, Execute the acquisition and update of the second basic learning value, Execute the acquisition and update of the third basic learning value, respectively, After that, Based on the first to third basic learning values, generate a map that can output the fuel temperature corresponding to the input discharge correction amount learning value as the estimated fuel temperature for a desired engine speed and the input of the discharge correction amount learning value corresponding to the desired engine speed. When the engine speed and the discharge correction amount learning value corresponding to the engine speed are input, output the estimated fuel temperature corresponding to the input, Regarding the acquisition and update of the first basic learning value, For each required engine rotation, Obtain the coolant temperature immediately before starting the vehicle as the fuel temperature. Immediately after starting the vehicle, based on the discharge correction amount learning value selected from the discharge correction amount learning values previously obtained by the discharge correction amount learning process and the coolant temperature, execute the acquisition and update of the discharge correction amount learning value for a plurality of preset specified fuel temperatures, Regarding the acquisition and update of the second basic learning value, For each required engine rotation, When it is determined that the engine is in a fully warmed-up state, the fuel temperature is acquired as a default value, and a fuel injection correction amount learning value corresponding to the engine speed is selected from the fuel injection correction amount learning values acquired in advance by the fuel injection correction amount learning process. The oldest fuel injection correction amount learning value in the obtained specified number of fuel injection correction amount learning values is overwritten and updated by the selected fuel injection correction amount learning value. Then, an average value of the sum of the specified number of fuel injection correction amount learning values and the initial value is calculated, and the average value is set as the fuel injection correction amount learning value for the fuel temperature corresponding to the default value of the fuel temperature, thereby acquiring and updating the fuel injection correction amount learning value for the fuel temperature. As for the acquisition and update of the third basic learning value, For each required engine speed, Based on the first basic learning value and the second basic learning value, it is configured to execute acquisition and update of fuel injection correction amount learning values for the assumed minimum and maximum fuel temperature values by extrapolation calculation.

Advantages of the Invention

[0008] According to the present invention, by configuring to obtain an estimated fuel temperature through a learning process and an interpolation operation, it is possible to provide a practical estimated fuel temperature as an alternative value to the detected temperature by a fuel temperature sensor, with less variation compared to the prior art and without requiring a complicated adaptation operation.

Brief Description of the Drawings

[0009]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Figure 8

Figure 9

Figure 10

Embodiments for Carrying Out the Invention

[0010] Hereinafter, embodiments of the present invention will be described with reference to FIGS. 1 to 10. Note that the members, arrangements, etc. described below do not limit the present invention, and various modifications can be made within the scope of the gist of the present invention. First, a configuration example of a common rail type fuel injection control device in an embodiment of the present invention will be described with reference to FIG. 1. In the common rail type fuel injection control device according to the embodiment of the present invention, a high-pressure pump device 50 for pumping high-pressure fuel, a common rail 1 for storing the high-pressure fuel pumped by the high-pressure pump device 50, and a plurality of fuel injection valves (injectors) 2-1 to 2-n for injecting and supplying the high-pressure fuel supplied from the common rail 1 to the cylinders of a diesel engine (hereinafter referred to as "engine") 3, and an electronic control unit (denoted as "ECU" in FIG. 1) 4 that executes fuel injection control processing and estimated fuel temperature generation processing in the embodiment of the present invention described later are mainly configured as components. Such a configuration itself is the same as the basic configuration of this type of common rail type fuel injection control device that has been well known in the art.

[0011] The high-pressure pump device 50 has a known and well-known configuration mainly composed of a feed pump 5, a metering valve 6, and a high-pressure pump 7 as a fuel supply pump. In such a configuration, the fuel in the fuel tank 9 is pumped up by the feed pump 5 and supplied to the high-pressure pump 7 via the metering valve 6. An electromagnetic proportional control valve is used for the metering valve 6, and by controlling the energization amount thereof by the electronic control unit 4, the flow rate of the fuel supplied to the high-pressure pump 7, in other words, the discharge amount of the high-pressure pump 7 is adjusted.

[0012] A return valve 8 is provided between the output side of the feed pump 5 and the fuel tank 9 so that the surplus fuel on the output side of the feed pump 5 can be returned to the fuel tank 9. The fuel injection valves (injectors) 2-1 to 2-n are provided for each cylinder of the engine 3, each receiving the supply of high-pressure fuel from the common rail 1 and being provided to enable fuel injection by injection control by the electronic control unit 4.

[0013] The electronic control unit 4 has, for example, a microcomputer (not shown) having a known and well-known configuration as the center, an EEPROM 20-1 (see FIG. 7) described later, and It has a memory element such as a RAM (not shown), and is configured with main components including a drive circuit (not shown) for driving the fuel injection valves 2-1 to 2-n, and an energization circuit (not shown) for energizing the metering valve 6 and the pressure control valve 12. Various detection signals and data such as the engine speed Ne and the accelerator opening Acc are input to such an electronic control unit 4, and are used for the operation control of the engine 3, the fuel injection control, and further the estimated fuel temperature generation process in the embodiment of the present invention.

[0014] Next, the estimated fuel temperature generation process in the embodiment of the present invention will be described with reference to FIGS. 2 to 10. First, the estimated fuel temperature generation process in the embodiment of the present invention will be generally described with reference to the flowchart shown in FIG. 2 and the functional block diagram shown in FIG. 7. FIG. 2 is a subroutine flowchart showing the schematic procedure of the estimated fuel temperature generation process in the embodiment of the present invention. Further, FIG. 7 is a functional block diagram representing the functions required when the electronic control unit 4 executes the estimated fuel temperature generation process in the embodiment of the present invention by functional blocks.

[0015] First, in a common rail type fuel injection control device, it is well known that a correction value for correcting the deviation of the discharge amount of a high-pressure pump is conventionally obtained by so-called learning value processing, and the desired discharge amount can be achieved by using it for discharge amount correction. The estimated fuel temperature generation process in the embodiment of the present invention focuses on the fact that the correction value for discharge amount correction (hereinafter referred to as "discharge correction amount learning value" for convenience of explanation) obtained by the above learning process has a correlation with the fuel temperature, and as will be described in detail later, the estimated fuel temperature can be generated.

[0016] To generally explain the execution procedure of this estimated fuel temperature generation process, first, at step S510, the electronic control unit 4 acquires and updates the estimated fuel temperature learning value. Next, at step S520, the electronic control unit 4 executes an acquisition update completion determination. That is, the electronic control unit 4 determines whether the acquisition and update of the required estimated fuel temperature learning value have been completed.

[0017] If, at step S520, the electronic control unit 4 determines that the acquisition and update of the required estimated fuel temperature learning value have been completed (YES), the process proceeds to step S530. On the other hand, if, at step S520, the electronic control unit 4 determines that the acquisition and update of the required estimated fuel temperature learning value have not been completed (NO), the acquisition and update of the estimated fuel temperature learning value continue.

[0018] At step S530, the electronic control unit 4 performs map conversion. That is, based on the estimated fuel temperature learning value, the electronic control unit 4 generates an estimated fuel temperature map that can output the estimated fuel temperature for the input of the injection correction amount learning value and the engine speed.

[0019] Next, regarding the outline of the estimated fuel temperature generation process in the embodiment of the present invention, focusing on the functions required for the electronic control unit 4, it will be described with reference to FIG. 7. First, the injection correction amount learning value is basically acquired by the same method as in the prior art as described later, and the learning value is stored, for example, in the EEPROM 20-1 provided in the electronic control unit 4.

[0020] This fuel injection correction amount learning value is associated with temperature data and the like appropriately selected by the first and second software switches 20-2 and 20-3 (details will be described later), and is stored as an estimated fuel temperature learning value in the basic learning value storage area (denoted as "SA" in FIG. 7) 20-4 in the electronic control unit 4. At the same time, generation and update of the estimated fuel temperature learning value are performed through necessary interpolation calculations and the like (refer to step S510 in FIG. 2). Next, based on the estimated fuel temperature learning value, map conversion is performed in the map generation area (denoted as "CONV" in FIG. 7) 20-5 so that the corresponding estimated fuel temperature can be output for the input of the fuel injection correction amount learning value and the engine speed, and an estimated fuel temperature map is generated (refer to step S530 in FIG. 2). The estimated fuel temperature map generated in the map generation area 20-5 is stored in the map storage area 20-6 and is used for various processes in which the estimated fuel temperature can be used in the common rail type fuel injection control device.

[0021] Next, the procedure of the acquisition process of the fuel injection correction amount learning value used in the execution of the estimated fuel temperature generation process in the embodiment of the invention will be described with reference to the subroutine flowchart shown in FIG. 3. The acquisition of this fuel injection correction amount learning value is to obtain the correction amount necessary for the correction by a so-called learning process in order to correct the variation in the discharge amount of the high-pressure pump 7, and is basically the same as that conventionally performed.

[0022] First, when the process by the electronic control unit 4 is started, first, in step S110, a determination of whether the learning condition is satisfied is executed. That is, in step S110, the electronic control unit 4 determines whether the driving state of the vehicle is in a state suitable for acquiring the fuel injection correction amount learning value. The criteria for this determination consist of multiple conditions. Regarding each individual condition, although some may vary depending on the individual specifications of the vehicle, etc., generally common criteria include that the vehicle is in a normal driving state, the target rail pressure is stable, the indicated injection quantity is equal to or greater than a certain value, the I term (integral term) of the PID control in rail pressure control is in a stable state within a desired range, and so on.

[0023] In addition, it is necessary that the state in which the above-mentioned determination criteria are satisfied continues for a predetermined period of about several dozen tasks. Here, since the appropriate value for setting the predetermined period varies depending on the specifications of each vehicle, etc., it is preferable to determine it based on test results and simulation results while considering the specifications of each vehicle.

[0024] The process of step S110 is repeated until it is determined that the above-mentioned determination criteria are satisfied. When it is determined that the above-mentioned determination criteria are satisfied (in the case of YES), in step S120, the electronic control unit 4 will execute the calculation of the I term average value. That is, the average value of the I term of the PID control in the predetermined period when it is determined that the above-mentioned learning conditions are satisfied is calculated by the electronic control unit 4. Next, in step S130, the electronic control unit 4 executes learning value storage.

[0025] That is, the average value of the I term calculated as described above is stored by the electronic control unit 4, for example, in the EEPROM 20-1 as one of the injection correction amount learning values. Note that this injection correction amount learning value is stored in the EEPROM 20-1 using the engine speed when the I term is obtained as described above as an index. After the execution of learning value storage, once it returns to the main routine (not shown), after other required processes are executed, the above-described series of processes will be executed again.

[0026] Next, the procedure for obtaining and updating the estimated fuel temperature learning value will be specifically described. In the embodiment of the present invention, the estimated fuel temperature learning value is obtained by a combination of temperature information selected by switching the first and second software switches 20-2 and 20-3 (see FIG. 7) and the injection correction amount learning value. FIG. 4 shows the switching procedure of the first and second software switches 20-2 and 20-3 as a subroutine flowchart. Hereinafter, the switching procedure of the first and second software switches 20-2 and 20-3 will be described with reference to the figure.

[0027] When the process is started by the electronic control unit 4, first, in step 210, an engine start determination is executed. That is, it is determined by the electronic control unit 4 whether the engine 3 (see FIG. 1) of the vehicle has been started, or in other words, whether the vehicle has been started.

[0028] In step S210, if it is determined by the electronic control unit 4 that the engine 3 has been started (YES), the process proceeds to step S220 described later. On the other hand, if it is determined that the engine 3 has not been started (NO), the process proceeds to the process of step S270.

[0029] In step S270, the index temperature is read by the electronic control unit 4. First, in the embodiment of the present invention, the acquisition of the estimated fuel temperature learning value is performed in two vehicle operating states, when the vehicle is started and after it is fully warmed up. Based on the estimated fuel temperature learning values obtained in these two vehicle operating states, it is obtained by interpolation calculation (details will be described later). And the index temperature is an index used to discriminate between the two vehicle operating states. In step S270, this index temperature is read by the electronic control unit 4. In the embodiments of the present invention, such an index temperature is specifically the cooling water temperature, the engine oil temperature, and the intake air temperature. Note that these temperatures are not used only in the present invention. Usually, they are used in the operation control of a vehicle, and sensors for that purpose are installed, so it is possible to use their detected values.

[0030] Next, in step S280, the electronic control unit 4 executes an index temperature matching determination. That is, the electronic control unit 4 determines whether the cooling water temperature, the engine oil temperature, and the intake air temperature are substantially the same. This index temperature matching determination is executed to determine whether it is before engine start and a sufficient time has elapsed since the most recent engine stop. The reason for making such a determination is that when the engine 3 is stopped and a sufficient time has elapsed, each index temperature can be regarded as being in a substantially identical state. Normally, in reality, a temperature difference occurs between the respective temperatures. However, the appropriate value for determining how much of a temperature difference is allowable for the temperatures to be considered the same varies depending on the vehicle specifications and the like. Therefore, it is preferable to determine it based on test results, simulation results, etc. in consideration of the specific vehicle specifications and the like.

[0031] Thus, in step S280, when the electronic control unit 4 determines that the cooling water temperature, the engine oil temperature, and the intake air temperature are substantially the same (YES case), the process proceeds to the process of step S290. In step S290, the electronic control unit 4 executes a first input state selection. That is, the electronic control unit 4 outputs a signal corresponding to the first input state selection as the input state selection signal ST-SIG (see FIG. 7) to the first and second software switches 20-2 and 20-3, and the first and second software switches 20-2 and 20-3 are set to the first input state.

[0032] Here, the switching operation of the first and second software switches 20-2 and 20-3 with respect to the input state selection signal ST-SIG will be described with reference to FIG. 7. First, in the functional block diagram shown in FIG. 7, the first and second software switches 20-2 and 20-3 are both shown as switches equivalent to single-pole three-pole switches with the same configuration. That is, in the first software switch 20-2, the switching common terminal 21 is configured to be selectively connected to any one of the three individual selection terminals 21a to 21c. Also, the three individual selection terminals 21a to 21c side is the input side of the data (data A) to be described later.

[0033] In the second software switch 20-3, the switching common terminal 22 is configured to be selectively connected to any one of the three individual selection terminals 22a to 22c. And the three individual selection terminals 22a to 22c side is the input side of the data (data B, data C) to be described later. In FIG. 7, the individual selection terminals 21a and 22a are denoted as "0", the individual selection terminals 21b and 22b are denoted as "1", and the individual selection terminals 21c and 22c are denoted as "2" for distinction.

[0034] Also, in the embodiment of the present invention, the input state selection signal ST-SIG is simultaneously input to both the first and second software switches 20-2 and 20-3, and therefore, the same switching operation occurs in both. Thus, in step S290 of FIG. 4 above, when the input state selection signal ST-SIG corresponding to the first input state selection is applied by the electronic control unit 4 to the first and second software switches 20-2 and 20-3, the switching common terminals 21 and 22 are each set to "1". That is, the switching common terminal 21 is connected to the individual selection terminal 21b, and the switching common terminal 22 is connected to the individual selection terminal 22b. Note that after the execution of step S290, once it returns to the main routine (not shown), after other required processing is executed, the above-described series of processes is executed again.

[0035] Also, when the input state selection signal ST-SIG is a signal corresponding to the standby state selection described later by the electronic control unit 4 for the first and second software switches 20-2 and 20-3, the switching common terminals 21 and 22 are each set to "0". That is, the switching common terminal 21 is connected to the individual selection terminal 21a, and the switching common terminal 22 is connected to the individual selection terminal 22a.

[0036] Furthermore, when the input state selection signal ST-SIG is a signal corresponding to the second input state selection described later by the electronic control unit 4 for the first and second software switches 20-2 and 20-3, the switching common terminals 21 and 22 are each set to "2". That is, the switching common terminal 21 is connected to the individual selection terminal 21c, and the switching common terminal 22 is connected to the individual selection terminal 22c.

[0037] Here, returning to FIG. 4 again, the description of the figure will be given. In step S290, as described above, when the switching common terminals 21 and 22 of the first and second software switches 20-2 and 20-3 are both set to "1", the acquisition and update of the estimated fuel temperature learning value are performed as described later. On the other hand, in the previous step S280, when it is determined by the electronic control unit 4 that the coolant temperature, engine oil temperature, and intake air temperature are not substantially the same (in the case of NO), in step S300, the electronic control unit 4 executes standby state selection. That is, the electronic control unit 4 outputs a signal corresponding to standby state selection as an input state selection signal ST-SIG (see FIG. 7) to the first and second software switches 20-2 and 20-3. As a result, the first and second software switches 20-2 and 20-3 are set to the standby state.

[0038] Therefore, in the first and second software switches 20-2 and 20-3, the switching common terminals 21 and 22 are both set to "0", and the switching common terminal 21 is connected to the individual selection terminal 21a, and the switching common terminal 22 is connected to the individual selection terminal 22a, respectively. In the embodiment of the present invention, since both of the individual selection terminals 21a and 22a are in the non-input state, in this case, the estimated fuel temperature learning value is not acquired and updated via the first and second software switches 20-2 and 20-3, so to speak, it is in the standby state. After the execution of step S300, once it returns to the main routine (not shown), after other required processes are executed, the above-described series of processes is executed again.

[0039] On the other hand, in the previous step S210, when it is determined by the electronic control unit 4 that the engine 3 has been started (in the case of YES), in step S220, the electronic control unit 4 executes learning value acquisition condition determination. That is, the electronic control unit 4 determines whether the acquisition and update of the second basic learning value described later are completed.

[0040] In step S220, when it is determined by the electronic control unit 4 that the update of the second basic learning value is completed (in the case of YES), the process proceeds to step S230 described below. On the other hand, in step S220, when it is determined by the electronic control unit 4 that the second basic learning value update has not been completed (in the case of NO), the process proceeds to step S260. Similar to the process of step S300 described above, the first and second software switches 20-2 and 20-3 are set to the standby state.

[0041] In step S230, the electronic control unit 4 executes the target temperature reading. Since this target temperature reading is the same as the process in the previous step S270, a detailed description here will be omitted. Next, in step S240, the electronic control unit 4 executes the target temperature determination. That is, in the embodiments of the present invention, for each target temperature, whether each of the cooling water temperature, engine oil temperature, and intake air temperature exceeds its respective threshold value is determined by the electronic control unit 4. For the convenience of the following description, the threshold value for the cooling water temperature will be referred to as the "cooling threshold value", the threshold value for the engine oil temperature will be referred to as the "engine oil threshold value", and the threshold value for the intake air temperature will be referred to as the "intake threshold value", respectively.

[0042] The determination of whether each of these target temperatures exceeds its respective threshold value means determining whether the engine is in a fully warmed-up state. That is, after full warm-up, generally, the fuel temperature tends to approach a certain value. However, in the embodiments of the present invention, when each of the above-mentioned target temperatures exceeds its respective threshold value as described above, it is regarded as being in the state after full warm-up.

[0043] In step S240, when it is determined by the electronic control unit 4 that any of the target temperatures exceeds its respective threshold value (in the case of YES), in step S250, the electronic control unit 4 will execute the second input state selection. That is, the electronic control unit 4 outputs a signal corresponding to the second input state selection as the input state selection signal ST-SIG (see FIG. 7) to the first and second software switches 20-2 and 20-3. As a result, the first and second software switches 20-2 and 20-3 are set to the second input state. That is, in the first and second software switches 20-2 and 20-3, the switching common terminals 21 and 22 are both set to "2". Therefore, the switching common terminal 21 is connected to the individual selection terminal 21c, and the switching common terminal 22 is connected to the individual selection terminal 22c. Note that after the execution of step S250, once it returns to the main routine (not shown), after other required processes are executed, the above-described series of processes are executed again.

[0044] Next, a procedure for acquiring and updating the estimated fuel temperature learning value according to the switching operation of the first and second software switches 20-2 and 20-3 described above will be described with reference to the flowchart shown in FIG. 5. When the processing is started by the electronic control unit 4, first, in step 410, the first input state determination is executed. That is, in the first and second software switches 20-2 and 20-3, it is determined by the electronic control unit 4 whether the switching common terminals 21 and 22 are set to "1".

[0045] In step S410, when it is determined by the electronic control unit 4 that the switching common terminals 21 and 22 in the first and second software switches 20-2 and 20-3 are set to "1", in other words, when it is determined that they are set to the first input state (YES), the process proceeds to the process of step S420. On the other hand, when it is determined by the electronic control unit 4 that the switching common terminals 21 and 22 are not set to "1" (NO), the process proceeds to the process of step S460.

[0046] In step S420, the first basic learning value update is executed by the electronic control unit 4. The acquisition and update of the estimated fuel temperature learning value are respectively performed when the first and second software switches 20-2 and 20-3 are in the first input state and the second input state, as described above with reference to FIG. 7. Based on the estimated fuel temperature learning values obtained in these input states, they are obtained by interpolation calculation (details will be described later). In the following description, for the sake of convenience of explanation, the acquisition and update of the estimated fuel temperature learning value in the first input state will be referred to as "first basic learning value update", and the acquisition and update of the estimated fuel temperature learning value in the second input state will be referred to as "second basic learning value update", respectively.

[0047] First, as described above, the first basic learning value update is executed when the switching common terminals 21 and 22 are set to "1" respectively in the first and second software switches 20-2 and 20-3 in the first input state. This first input state is a state determined that each index temperature (cooling water temperature, engine oil temperature, and intake air temperature) is approximately the same before engine start. Usually, even if the engine starts after this, it can be considered that the fuel temperature hardly rises within about 1 minute after start. Therefore, by acquiring the fuel temperature and the injection correction amount learning value in this state as the first basic learning value, it becomes possible to construct the correlation between the so-called cold fuel temperature and the injection correction amount learning value.

[0048] In the first input state, paying attention to this point, when each index temperature is approximately the same, the index temperature at that time is regarded as the fuel temperature, and the cooling water temperature is regarded as the fuel temperature as the representative value of the index temperature. Together with the injection correction amount learning value from the previous EEPROM 20-1, it is stored in the basic learning value storage area 20-4 as the estimated fuel temperature learning value as described below.

[0049] That is, in the first input state, in the first software switch 20-2 where the switching common terminal 21 is set to "1", the injection correction amount learning value and the engine speed are input as data A from the EEPROM 20-1 (see FIG. 7). At the same time, for the second software switch 20-3 in which the switching common terminal 22 is set to "1", the cooling water temperature at this point is input as data B as the fuel temperature (see Fig. 7). In this case, among the fuel injection correction amount learning values read from the EEPROM 20-1, one corresponding to the engine speed within about one minute after engine start is selected. This is because in the embodiment of the present invention, the cooling water temperature read as data B above is the one before engine start, and it is regarded as almost unchanged until about one minute after engine start and is used as the fuel temperature. Thus, in the embodiment of the present invention, the cooling water temperature is regarded as the fuel temperature at immediately after vehicle start during the period of about one minute after engine start when the cooling water temperature hardly changes from before engine start, and is used in a simulated manner.

[0050] The fuel injection correction amount learning value and the engine speed taken in via the first software switch 20-2, and the cooling water temperature taken in via the second software switch 20-3 are stored in the basic learning value storage area 20-4 as a set of first basic learning values, that is, in other words, as a set of estimated fuel temperature learning values in the following manner.

[0051] Fig. 8 schematically shows an example of storing the estimated fuel temperature learning value in the basic learning value storage area 20-4. Hereinafter, with reference to the same figure, an example of storing the estimated fuel temperature learning value in the first input state will be described. Basically, the estimated fuel temperature learning value is stored divided by engine speed. The engine speed is the engine speed read together with the fuel injection correction amount learning value when the fuel injection correction amount learning value is read from the EEPROM 20-1.

[0052] In FIG. 8, an example of storing the estimated fuel temperature learning values obtained for each of the engine speeds Ne of 1000 rpm, 2000 rpm, and 3000 rpm is shown. The engine speeds shown in FIG. 8 are merely examples and are not limited to these three engine speeds. Which engine speed to set depends on the vehicle specifications and the like, and the appropriate set value varies. Therefore, it is preferable to determine it based on test results, simulation results, etc. in consideration of the specifications of each vehicle.

[0053] In FIG. 8, the estimated fuel temperature learning value for one engine speed is stored in the form of a frequency distribution table. That is, in the figure, "Tf" means the fuel temperature, and each numerical value in the horizontal row with this notation represents the fuel temperature. Also, in FIG. 8, "LV" means the injection correction amount learning value, and each numerical value in the horizontal row with this notation is the injection correction amount learning value obtained as described below.

[0054] The area where the estimated fuel temperature learning value is stored in the form of a frequency distribution table is roughly divided into three areas: the area where the learning value obtained in the first input state is mainly stored (the part surrounded by a thick line frame and marked with symbol a in FIG. 8), the area where the learning value obtained in the second input state is stored (the part marked with symbol b located at the right end of the area marked with symbol a in FIG. 8), and the area where the learning value obtained by extrapolation calculation based on these learning values is stored (the parts marked with symbols c1 and c2 in FIG. 8). Note that initial values are stored in each of the above-mentioned areas in advance. This initial value is the so-called central characteristic of a new high-pressure pump selected at the development stage of this device and having operating characteristics that almost match the specifications. When the device is configured using this high-pressure pump, the injection correction amount learning value LV for the measured fuel temperature Tf is used.

[0055] And in the first input state, learning values are acquired and updated for each engine speed. For example, when acquiring the learning value at an engine speed of 1000 revolutions, the fuel injection correction amount learning value for 1000 engine revolutions is read from the EEPROM 20-1, and together with the cooling water temperature as the fuel temperature at this time, it is read into the basic learning value storage area 20-4. Then, based on the read fuel injection correction amount learning value and the cooling water temperature, the acquisition and update of the estimated fuel temperature learning value are performed by extrapolation calculation as described later.

[0056] First, the fuel temperature for the learning value acquired in the first input state is predetermined. That is, in the embodiment of the present invention, the fuel temperature is set in five levels from -10°C to 30°C at intervals of 10°C (see FIG. 8). By thus predetermining the fuel temperature at intervals of about 10°C, the present invention aims to provide an estimated fuel temperature that can be used for control processing that can suppress variations compared to the conventional estimated temperature with variations of about ±20°C and can also withstand a certain degree of temperature variation without causing problems. Note that the above temperature intervals and each set temperature are merely examples and are not limited thereto and can be arbitrarily determined. However, considering the accuracy of the estimated fuel temperature obtained by the estimated fuel temperature generation process of the present invention, it is considered that around 10°C is appropriate for at least the temperature interval.

[0057] Next, taking the acquisition and update of the learning value for 1000 engine revolutions as an example, the acquisition and update of the estimated fuel temperature learning value by extrapolation calculation will be specifically described. First, assuming that the previously acquired cooling water temperature is, for example, 5°C, which is other than the temperature preset as described above (hereinafter referred to as "prescribed fuel temperature" for convenience of explanation), the fuel injection correction amount learning value for the prescribed fuel temperature is acquired and updated by interpolation processing as described below using the fuel injection correction amount learning value taken from the EEPROM 20-1 corresponding to 5°C.

[0058] FIG. 9 shows a schematic diagram for explaining a method of obtaining and updating the learning value of the injection correction amount with respect to the specified fuel temperature based on the injection correction amount learning value captured from the EEPROM20-1 as described above. Hereinafter, the content will be described with reference to this figure. First, in FIG. 9, the horizontal axis represents the fuel temperature Tf, and the vertical axis represents the learning value LV of the injection correction amount. In FIG. 9, the white circle marked with the symbol LV1 is assumed to indicate the learning value of the injection correction amount corresponding to Tfb = 5°C when the above-described cooling water temperature is taken as the fuel temperature Tfb. Since this 5°C is between 0°C and 10°C of the specified fuel temperature, first, the update of the learning value of the injection correction amount corresponding to 0°C and 10°C is performed as follows based on the learning value of the injection correction amount corresponding to 5°C.

[0059] That is, first, in FIG. 9, the circles marked with the symbols LV2 and LV3 and shaded are assumed to be the learning values of the injection correction amount obtained in the most recent process. That is, in the case of this example, the learning value of the injection correction amount indicated by the symbol LV2 is the learning value obtained in the most recent process corresponding to Tfa = 0°C, and the learning value of the injection correction amount indicated by the symbol LV3 is the learning value obtained in the most recent process corresponding to Tfa = 10°C. Also, in FIG. 9, d1 is the difference between the fuel temperature Tfb corresponding to the learning value of the injection correction amount marked with the symbol LV1 and the fuel temperature Tfa corresponding to the learning value of the injection correction amount marked with the symbol LV2, and d2 is the difference between the fuel temperature Tfb corresponding to the learning value of the injection correction amount marked with the symbol LV1 and the fuel temperature Tfc of the learning value of the injection correction amount marked with the symbol LV3. Hereinafter, for the sake of convenience of explanation, LV1, LV2, and LV3 are assumed to be the above-described learning values of the injection correction amount, respectively.

[0060] Furthermore, let the difference between the injection correction amount at the point where the virtual line (two-dot chain line) connecting the most recent learning values LV2 and LV3 intersects the vertical line at the fuel temperature Tfb and LV1 (hereinafter referred to as "learning value difference" for convenience of explanation) be Vdif. Additionally, if the update value for LV2 is LV4 and the update value for LV3 is LV5, then LV4 and LV5 are obtained by the following equations (1) and (2).

[0061] LV4 = LV2 + K × Vdif × {d2 2 / (d1 2 + d2 2 )} ··· Equation (1)

[0062] LV5 = LV3 + K × Vdif × {d1 2 / (d1 2 + d2 2 )} ··· Equation (2)

[0063] Here, K × Vdif in Equations (1) and (2) is a weighting factor, and hereinafter, for convenience of explanation, it is referred to as "learning value difference weighting". Note that in Equations (1) and (2), K is an arithmetic coefficient, but its appropriate value varies depending on vehicle specifications, etc. Therefore, it is preferable to determine it based on test results and simulation results while considering such specifications. Also, {d2 2 / (d1 2 + d2 2 )} in Equation (1) and {d1 2 / (d1 2 + d2 2 )} in Equation (2) are coefficients for inversely proportionally distributing the previous learning value difference weighting with respect to the fuel temperature difference represented by d1 and d2 described above, and hereinafter, for convenience of explanation, they are referred to as "distribution coefficients". The calculation of the update values LV4 and LV5 of such injection correction amount learning values can be summarized as follows. That is, first, in a quadratic plane where the fuel temperature is on the horizontal axis and the learned value of the discharge correction amount is on the vertical axis, a virtual line (dashed line) connecting the most recent learned values LV2 and LV3 intersects the vertical line at the fuel temperature Tfb for the most recently acquired learned value of the discharge correction amount LV1. The learning value difference Vdif between the discharge correction amount corresponding to the intersection point and LV1 is obtained.

[0064] Next, the product of the weighting coefficient inversely proportional to the respective temperature differences d1 and d2 between the fuel temperature Tfb for the learned value of the discharge correction amount LV1 and the fuel temperatures Tfa and Tfc for the most recent learned values LV2 and LV3, and the learning value difference Vdif is calculated. By adding the calculation result to the most recent learned values LV2 and LV3, updated values LV4 and LV5 of the learned value of the discharge correction amount are obtained. Thus, in the frequency distribution table shown in FIG. 8, the updated value LV4 is written and stored as the updated value of the learned value of the discharge correction amount corresponding to the fuel temperature Tf = 0°C, and the updated value LV5 is written and stored as the updated value of the learned value of the discharge correction amount corresponding to the fuel temperature Tf = 10°C.

[0065] On the other hand, even when the cooling water temperature matches one of the specified fuel temperatures, for example, 0°C, instead of directly saving the learned value of the discharge correction amount obtained from the EEPROM20-1 as the updated value, an updated value that reflects only the weighting in the above-described extrapolation calculation is calculated as follows. The calculated value is written and stored as the new learned value of the discharge correction amount. That is, using the example shown in FIG. 9, in this case, it can be interpreted that the temperature Tfb matches Tfa or Tfc of the specified fuel temperature. For example, when it is interpreted that the temperature Tfb matches the specified fuel temperature Tfa, the temperature difference d1 becomes d1 = 0. Therefore, in Equation 1, when d1 = 0, the distribution coefficient becomes 1, so Equation 1 becomes an equation that reflects only the weighting as described below.

[0066] LV4 = LV2 + K × Vdif ··· Equation 1A

[0067] For example, at the above-mentioned cooling water temperature of 0°C, if the learned value of the discharge correction amount obtained from the EEPROM 20-1, that is, the most recently acquired learned value of the discharge correction amount LV1, is assumed to be 700. And if the nearest learned value LV2 to LV1 is 600 as illustrated in FIG. 8, then the difference Vdif between the most recently acquired learned value of the discharge correction amount LV1 and the nearest learned value LV2 is Vdif = 700 - 600. Then, the updated value LV4 reflecting the weighting for LV1 is obtained as follows by Equation 1A.

[0068] LV4 = 600 + K × 100

[0069] Here, if it is assumed that K = 0.2, then LV4 = 600 + 0.2 × 100 = 620, and instead of 600, 620 is written and stored as the new updated value of the learned value in the region marked with symbol a in FIG. 8 as the learned value of the discharge correction amount corresponding to the fuel temperature Tf = 0°C.

[0070] In the above description, the case where the temperature Tfb is interpreted as being equal to the specified fuel temperature Tfa has been described. However, if the temperature Tfb is interpreted as being equal to the specified fuel temperature Tfc, the following similar results will be obtained. That is, first, in this case, the updated value LV5 reflecting the weighting for LV1 is calculated using the previous Equation 2. Since the temperature difference d2 is zero, the distribution coefficient is 1. Therefore, Equation 2 becomes as follows.

[0071] LV5 = LV3 + K × Vdif ··· Equation 2A

[0072] In this case, the learned value difference Vdif is defined as Vdif = LV1 - LV3. This Equation 2A is substantially the same as the previous Equation 1A, and even when Equation 2A is used in the previous numerical example, the calculation results are the same. In this way, when the update of the fuel injection correction value learning value for 0°C and 10°C, or the update of the fuel injection correction value learning value for 0°C is executed, then, in step S430, the electronic control unit 4 executes the first update completion determination (see FIG. 5).

[0073] That is, the electronic control unit 4 determines whether the required first basic learning value update has been completed. Here, in the case of the example shown in FIG. 8, the first basic learning value update is considered to be completed when the updates of the fuel injection correction value learning values for all five specified fuel temperatures have been performed. In other words, the update for one cycle is considered completed. Therefore, at the stage where the updates of the fuel injection correction value learning values for 0°C and 10°C are executed as described above, in step S430, the electronic control unit 4 determines that the first update has not been completed.

[0074] Thus, after it is determined in step S430 by the electronic control unit 4 that the first update has not been completed, the process once returns to the main routine (not shown), and after other required processes are executed, the process shown in FIG. 5 is executed again. Therefore, in this case, since it is the first input state, the process proceeds from the process of step S410 to step S420, and the electronic control unit 4 executes the first basic learning value update again.

[0075] That is, in the same manner as the update of the fuel injection correction value learning values for 0°C and 10°C described above, based on the fuel injection correction value learning value corresponding to 5°C, the updates of the fuel injection correction value learning values for -10°C and 20°C are executed. Subsequently, in the same manner, the electronic control unit 4 executes the updates of the fuel injection correction value learning values for all the specified fuel temperatures. As described above, when the updates of the fuel injection correction value learning values for the specified fuel temperatures above and below the actually acquired fuel temperature of 5°C are actually executed, in this example, only the update of the fuel injection correction value learning value for the specified fuel temperature of 30°C remains at the end. In that case, the update of the fuel injection correction value learning value will be performed as described below.

[0076] That is, in that case, as the temperature on the low-temperature side with respect to 5°C, the lowest specified fuel temperature, that is, -10°C in the above example, is arbitrarily selected, and as the update of the fuel injection correction amount learning value with respect to -10°C and 30°C centered on 5°C, the update of the fuel injection correction amount learning value is performed as described above. After the necessary first basic learning value update is performed as described above, in step S430, the electronic control unit 4 executes a first update completion determination, determines that the first update is completed, and proceeds to the process of step S440. In step S440, the electronic control unit 4 executes a second update completion determination. The second update completion determination is a determination as to whether or not the second update, which is the acquisition and update (details will be described later) of the estimated fuel temperature learning value performed in the second input state, is completed. A more specific explanation of the determination of the second update completion will be made together with the explanation of the specific content of the second update described later.

[0077] Thus, when in the stage immediately after the completion of the first basic learning value update described above, in step S440, it is determined that the second update is not yet completed (in the case of NO), once returns to a main routine (not shown), after other required processes are executed, step S410 is executed again. When step S410 is executed, assuming that the first and second software switches 20-2 and 20-3 are set to the second input state, the electronic control unit 4 determines that it is not in the first input state, and proceeds to the process of step S460.

[0078] In step S460, the electronic control unit 4 executes a second input state determination. That is, the electronic control unit 4 determines whether or not the first and second software switches 20-2 and 20-3 are set to the second input state. Assuming that the first and second software switches 20-2 and 20-3 are set to the second input state as described above, in step S460, the electronic control unit 4 determines that the first and second software switches 20-2 and 20-3 are set to the second input state (if YES), and proceeds to the process of step S470.

[0079] In step S470, the electronic control unit 4 executes the second basic learning value update. This second basic learning value update is performed in the second input state, but the second input state is conditional on being in the state after complete warm-up as described above. As will be explained below, the second basic learning value update is intended to obtain and update the learning value at the fuel temperature after complete warm-up. That is, in the second input state, via the first software switch 20-2, the injection correction amount learning value from the EEPROM 20-1 and the engine speed are read as data A, and via the second software switch 20-3, the fuel temperature after complete warm-up is read as data C, and the second basic learning value update is executed by storing them as the second basic learning value in the basic learning value storage area 20-4 as follows.

[0080] Also in this second basic learning value update, similar to the case of the first basic learning value update, the learning values are stored separately for each engine speed. That is, in FIG. 8, the portion marked with the symbol b is the target of the second basic learning value update. In the example of the figure, the italic numerical value "40" at the corresponding position in the row marked with "Tf" is an example of the fuel temperature captured and stored via the second software switch 20-3. In the embodiment of the present invention, the fuel temperature obtained after complete warm-up is used as the default value of the fuel temperature for the estimated fuel temperature learning value.

[0081] Also, in FIG. 8, the learned value of the injection correction amount for a fuel temperature of 40° C. is stored at the corresponding position in the row marked with "LV". This learned value of the injection correction amount is the learned value of the injection correction amount corresponding to the engine speed of the vehicle when the fuel temperature is acquired via the second software switch 20-3. In view of the fact that the fuel temperature in this second basic learning value update is the default value as described above, the corresponding learned value of the injection correction amount is set as the average value of a plurality of learned values of the injection correction amount, aiming to ensure the reliability and stability as the default value. Note that in FIG. 8, the plurality of numerical values shown at the storage location of the learned value of the injection correction amount for a fuel temperature of 40° C. schematically represent examples of the learned values of the injection correction amount acquired for the above-mentioned average value calculation.

[0082] The procedure for obtaining the average value of the learned value of the injection correction amount in this second basic learning value update will be described with reference to the subroutine flowchart shown in FIG. 6. First, to generally explain the procedure for obtaining the average value of the learned value of the injection correction amount, when a fuel temperature of 40° C. is input via the second software switch 20-3 and each time a learned value of the injection correction amount is read from the EEPROM 20-1, an average value calculation is performed with the previously acquired learned value of the injection correction amount, and the calculation result becomes the new average value.

[0083] Specifically described below with reference to FIG. 6, when the processing by the electronic control unit 4 is started, first, in step S610, the acquisition of the learned value of the injection correction amount is executed. That is, by the electronic control unit 4, as described above, a fuel temperature of 40° C. is input via the second software switch 20-3, and at the same time, the learned value of the injection correction amount is read from the EEPROM 20-1 to the basic learning value storage area 20-4.

[0084] Next, in step S620, the processing time determination is executed by the electronic control unit 4. That is, the electronic control unit 4 determines whether or not the ejection correction amount learning value read from the above-described EEPROM 20-1 to the basic learning value storage area 20-4 is within the number of times corresponding to a prescribed number (hereinafter referred to as the "start-up period" for convenience of explanation) after the start of use of this device. The reason for making such a determination is that, as generally described above, the average value of the ejection correction amount learning values in the embodiment of the present invention is basically calculated as the average of the prescribed number of ejection correction amount learning values and the initial value. However, until the prescribed number of ejection correction amount learning values are acquired, such a calculation process cannot be adopted. Therefore, it is for switching to the calculation process as described later.

[0085] In step S620, when the electronic control unit 4 determines that the reading of the ejection correction amount learning value from the EEPROM 20-1 to the basic learning value storage area 20-4 is within the start-up period (YES case), the process proceeds to the process of step S630. On the other hand, in step S620, when the electronic control unit 4 determines that the reading of the ejection correction amount learning value from the EEPROM 20-1 to the basic learning value storage area 20-4 is not within the start-up period (NO case), the process proceeds to the process of step S640.

[0086] In step S630, normal storage is executed by the electronic control unit 4. That is, the ejection correction amount learning value read from the EEPROM 20-1 to the basic learning value storage area 20-4 is written into a predetermined area as described below, and thus is stored and saved. First, in the basic learning value storage area 20-4, a storage area for storing and saving data used for calculating the average value (hereinafter referred to as the "average data storage area" for convenience of explanation) is secured separately from the storage and saving area schematically shown in FIG. 8 above for calculating the average value of the ejection correction amount learning value.

[0087] This average data storage area (not shown) has a capacity capable of storing and saving a predetermined number of fuel injection correction amount learning values and the initial value of the fuel injection correction amount learning value with respect to the default value of the fuel temperature described above with reference to FIG. 8. Here, the predetermined number is preferably about 20 to 30, for example. However, since the specific set number varies depending on the individual specifications of the vehicle and the like, it is preferable to determine it based on test results and simulation results while considering such specifications. The second basic learning value update usually acquires a set of fuel injection correction amount learning values from the EEPROM 20-1 and a set of learning values of the engine speed and fuel temperature in one so-called driving cycle. Therefore, in the starting period, in order to acquire the predetermined number of fuel injection correction amount learning values, it is necessary to go through the same number of driving cycles as that predetermined number. For example, when the predetermined number is 20, a total of 21 pieces of data, including the initial value and 20 fuel injection correction amount learning values, are stored and saved in the average data storage area.

[0088] Thus, during the starting period, the fuel injection correction amount learning value read from the EEPROM 20-1 is sequentially written into the area where the data in the average data storage area has not been written yet, and is stored and saved (normal storage). Then, in step S650, the electronic control unit 4 calculates an average value. During the starting period, the average value is calculated as the average value of the data stored and saved in the average data storage area at the time of average value calculation.

[0089] That is, for example, assuming that the average value calculation is the first time after the start of use of the device, in this case, since there are two pieces of data in the average data storage area, namely the initial value described above and one fuel injection correction amount learning value stored and saved in the process of step S610, the average value is the value obtained by dividing the sum of these two pieces of data by the number of data, which is 2. Thereafter, until the number of the learning values of the ejection correction amount in the average data storage area reaches the specified number, the value obtained by dividing the sum of the total learning values of the ejection correction amount existing in the average data storage area and the initial value by both numbers will be the average value. The average value calculated as described above will be stored in the format described in FIG. 8 above as the average value of the learning value of the ejection correction amount at this time with respect to the default fuel temperature (for example, 40°C).

[0090] On the other hand, in step S640, the electronic control unit 4 performs overwriting storage. That is, when the learning value of the ejection correction amount obtained by executing step S610 becomes the (specified number + 1)th, for example, the 21st when the specified number is 20 after the start of use of the device, this 21st learning value of the ejection correction amount will overwrite and be stored as the oldest learning value of the ejection correction amount in time series among the 20 learning values of the ejection correction amount stored in the average data storage area. Thereafter, every time the learning value of the ejection correction amount is obtained by executing step S610 in the electronic control unit 4, the obtained learning value of the ejection correction amount will be sequentially overwritten and stored in the order of the oldest in time series among the learning values of the ejection correction amount stored in the average data storage area as described above.

[0091] Next, after the overwriting storage is performed as described above, the electronic control unit 4 calculates the average value in step S650. The calculation of the average value after the above overwriting storage is obtained as the value obtained by dividing the sum of the specified number of learning values of the ejection correction amount and the initial value stored in the average data storage area by the specified number + 1. That is, when the specified number is 20, the divisor for calculating the average value is 21. The calculated average value will be stored in the format described in FIG. 8 as described above.

[0092] After the average value calculation is executed as described above, the second update is completed, and the process proceeds from the process of the previous step S470 (see FIG. 5) to the process of step S480. Even when an average value based on a number of data less than the specified number is calculated as described above, by assuming that one second basic learning update has been completed, compared with the case where the second update completion is determined only when an average value based on the specified number of data is calculated, the execution of the third basic learning update described later is performed earlier, so that it is possible to obtain an output of the estimated fuel temperature at an early stage after the apparatus is started.

[0093] Then, in step S480, the electronic control unit 4 executes a second update completion determination. That is, the electronic control unit 4 determines whether or not the second basic learning value update described above has been completed. As described above, in the second basic learning value update, every time an average value is calculated, the second update is considered completed. Therefore, in step S480, the electronic control unit 4 determines that the second update has been completed, and the process proceeds to the process of step S490. In step S480, when the electronic control unit 4 determines that the second update has not been completed (NO), the process once returns to a main routine (not shown), and after other necessary processes are executed, through the switching process of the first and second software switches 20-2 and 20-3 shown in FIG. 4, the process starts again from step S410.

[0094] In step S490, the electronic control unit 4 executes a first update completion determination. That is, the electronic control unit 4 determines whether or not the update of the first basic learning value described above has been completed. In step S490, when the electronic control unit 4 determines that the update of the first basic learning value has been completed, that is, the first update has been completed (YES), the process proceeds to the process of step S450. On the other hand, in step S490, when the electronic control unit 4 determines that the update of the first basic learning value has not been completed yet (NO), the process once returns to a main routine (not shown), and after other necessary processes are executed, the process starts again from step S410.

[0095] However, when the above-described first basic learning update and second basic learning update are completed, in step S450, the electronic control unit 4 executes a third basic learning value update. This third basic learning value update acquires and updates the learning value based on the first basic learning value and the second basic learning value, and will be specifically described below. As described above with reference to FIG. 8, the first basic learning value and the second basic learning value are stored in the portions marked with reference signs a and b, respectively, in FIG. 8. The third basic learning value is stored in the portions marked with reference signs c1 and c2 in FIG. 8, and the learning value is obtained by extrapolation processing based on the first basic learning value and the second basic learning value, as described below.

[0096] This third basic learning value is obtained by performing an extrapolation operation on the fuel injection correction amount learning values for the minimum and maximum values of the assumed fuel temperature (hereinafter referred to as "assumed fuel temperature" for convenience of explanation) with respect to the first basic learning value and the second basic learning value. Note that such a third basic learning value is obtained for each engine speed, similarly to the first basic learning value and the second basic learning value. In the example shown in FIG. 8, -25°C is set as the minimum value of the assumed fuel temperature, and 80°C is set as the maximum value.

[0097] The fuel injection correction amount learning values for the assumed fuel temperature of -25°C and the fuel injection correction amount learning value for the assumed fuel temperature of 80°C are obtained by extrapolation operation based on the respective fuel injection correction amount learning values from the specified fuel temperature of -10°C to 40°C. Since the extrapolation operation itself is a well-known method, detailed description thereof will be omitted here. In FIG. 8, the portions of the fuel injection correction amount learning values for the minimum and maximum values of the assumed fuel temperature obtained by the extrapolation operation as described above are indicated by the US marks ※ for convenience.

[0098] Note that the minimum and maximum values of the assumed fuel temperature described above are merely examples and do not necessarily have to be limited to the above examples. These values should be set to the required values respectively in consideration of the individual specifications of the vehicle and the like. After the third basic learning value update is executed as described above, once the process returns to a main routine (not shown), after other required processes are executed, the above-described series of processes will be executed again.

[0099] Thus, the mutual relationship of the estimated fuel temperature learning values, which is a set of the first to third basic learning values obtained as described above, is a relationship that can be approximated by a linear function, as shown in the schematic diagram in FIG. 10 with respect to the fuel temperature. In FIG. 10, "Tf" on the horizontal axis indicates the fuel temperature, and "LV" on the vertical axis indicates the learning value.

[0100] When the acquisition and update of the above-described estimated fuel temperature learning value in the basic learning value storage area 20-4 are completed, the electronic control unit 4 reads the estimated fuel temperature learning value in the basic learning value storage area 20-4 into the map generation area 20-5, and map conversion is executed (see step S530 in FIG. 2). That is, the read estimated fuel temperature learning value is used by the electronic control unit 4 to generate a map that can output the fuel temperature paired with the fuel injection correction amount learning value as the estimated fuel temperature Tf(es) with respect to the input of the fuel injection correction amount learning value and the engine speed. This map is then stored in the map storage area 20-6 as an estimated fuel temperature map.

[0101] In the electronic control unit 4, when the engine speed Ne and the fuel injection correction amount learning value LV from the EEPROM 20-1 are input to the map storage area 20-6, the corresponding estimated fuel temperature Tf(es) is output based on the stored estimated fuel temperature map.

Industrial Applicability

[0102] It can be applied to a common rail type fuel injection control device that is less variable than conventional ones and for which an estimated fuel temperature that can be used as an alternative value to the detected temperature by a fuel temperature sensor is desired.

Explanation of symbols

[0103] 1...Common rail 3...Engine 4...Electronic control unit 6...Fuel metering valve 7...High-pressure pump

Claims

1. Fuel in a fuel tank (9) is pressurized and pumped to a common rail (1) by a high-pressure pump (7), and high-pressure fuel injection into an engine (3) is enabled via fuel injection valves (2-1 to 2-n) connected to the common rail (1). A metering valve (6) is provided at least on the upstream side of the high-pressure pump (7), and the rail pressure of the common rail (1) can be controlled by driving and controlling at least the metering valve (6). On the other hand, a correction amount for correcting variations in the discharge amount of the high-pressure pump (7) is obtained and updated by a discharge correction amount learning process, which is a learning process, as a discharge correction amount learning value. In an estimated fuel temperature generation method in a common rail type fuel injection control device configured as described above, As for obtaining and updating the first basic learning value, For each required engine speed, The coolant temperature immediately before starting the vehicle is regarded as the fuel temperature and obtained. Immediately after starting the vehicle, based on the discharge correction amount learning value selected from the discharge correction amount learning values previously obtained by the discharge correction amount learning process and the coolant temperature, the acquisition and update of the discharge correction amount learning values for a plurality of preset specified fuel temperatures are performed. As for obtaining and updating the second basic learning value, For each required engine speed, When it is determined that the engine is in a fully warmed-up state, the fuel temperature is obtained as a default value, and the discharge correction amount learning value corresponding to the engine speed is selected from the discharge correction amount learning values previously obtained by the discharge correction amount learning process. The oldest discharge correction amount learning value in a preset number of obtained discharge correction amount learning values is overwritten and updated by the selected discharge correction amount learning value. Then, the average value of the sum of the preset number of discharge correction amount learning values and the initial value is calculated, and the average value is used as the discharge correction amount learning value for the fuel temperature corresponding to the default value, thereby obtaining and updating the discharge correction amount learning value for the fuel temperature. As for obtaining and updating the third basic learning value, For each required engine speed, Based on the first basic learning value and the second basic learning value, the acquisition and update of the discharge correction amount learning values for the assumed minimum and maximum fuel temperatures are performed by extrapolation calculation. Based on the first to third basic learning values, for the input of a desired engine speed and a fuel injection correction amount learning value corresponding to the desired engine speed, a map is generated that can output the fuel temperature corresponding to the input fuel injection correction amount learning value as an estimated fuel temperature. When the engine speed and the fuel injection correction amount learning value corresponding to the engine speed are input, the estimated fuel temperature corresponding to the input is outputtable. A method for generating an estimated fuel temperature in a common rail type fuel injection control device, characterized in that.

2. In the acquisition and update of the first basic learning value The coolant temperature immediately before starting the vehicle is regarded as the fuel temperature and acquired. Immediately after starting the vehicle, the acquisition and update of the fuel injection correction amount learning value for a plurality of preset specified fuel temperatures based on the fuel injection correction amount learning value selected from the fuel injection correction amount learning values acquired in advance by the fuel injection correction amount learning process and the coolant temperature are as follows: The coolant temperature immediately before starting the vehicle is regarded as the fuel temperature and acquired. Then, immediately after starting the vehicle, from among the fuel injection correction amount learning values acquired in advance by the fuel injection correction amount learning process, a fuel injection correction amount learning value corresponding to the engine speed at that time is selected. A learning value difference weighting is calculated based on the learning value difference, which is the difference between the selected fuel injection correction amount learning value and the fuel injection correction amount learning value acquired in the most recent process. Then, a distribution coefficient inversely proportional to the fuel temperature difference is calculated based on the fuel temperature difference between the coolant temperature and the specified fuel temperatures above and below the coolant temperature. The multiplication result of multiplying the learning value difference weighting by the distribution coefficient is added to the fuel injection correction amount learning value acquired in the most recent process, and the addition result is used as a new fuel injection correction amount learning value to repeat the update of the learning value. The acquisition and update of the fuel injection correction amount learning value for the required specified fuel temperature are performed. A method for generating an estimated fuel temperature in a common rail type fuel injection control device according to claim 1, characterized in that.

3. Fuel in the fuel tank (9) is pressurized and pumped to the common rail (1) by the high-pressure pump (7), and high-pressure fuel injection into the engine (3) is enabled via the fuel injection valves (2-1 to 2-n) connected to the common rail (1). Further, a metering valve (6) is provided at least on the upstream side of the high-pressure pump (7), and the rail pressure of the common rail (1) can be controlled by driving and controlling at least the metering valve (6) by the electronic control unit (4). On the other hand, a correction amount for correcting the variation in the discharge amount of the high-pressure pump (7) is obtained and updated by a discharge correction amount learning process, which is a learning process, as a discharge correction amount learning value. A common rail type fuel injection control device is configured as follows: The electronic control unit (4) is configured to: acquire and update a first basic learning value, acquire and update a second basic learning value, acquire and update a third basic learning value, respectively, and then, based on the first to third basic learning values, generate a map capable of outputting, as an estimated fuel temperature, the fuel temperature corresponding to the input discharge correction amount learning value for a desired engine speed and the input discharge correction amount learning value corresponding to the desired engine speed. When the engine speed and the discharge correction amount learning value corresponding to the engine speed are input, the estimated fuel temperature corresponding to the input can be output, Regarding the acquisition and update of the first basic learning value, for each required engine rotation, the coolant temperature immediately before starting the vehicle is regarded as the fuel temperature and acquired. Immediately after starting the vehicle, acquisition and update of the discharge correction amount learning value for a plurality of preset specified fuel temperatures are executed based on the discharge correction amount learning value selected from the discharge correction amount learning values previously obtained by the discharge correction amount learning process and the coolant temperature, Regarding the acquisition and update of the second basic learning value, for each required engine rotation, When it is determined that the engine is in a fully warmed-up state, obtain the fuel temperature as a default value, and select a fuel injection correction amount learning value corresponding to the engine speed from among the fuel injection correction amount learning values previously obtained by the fuel injection correction amount learning process. Overwrite and update the oldest fuel injection correction amount learning value in the time series among the obtained specified number of fuel injection correction amount learning values with the selected fuel injection correction amount learning value. Then, calculate the average value of the sum of the specified number of fuel injection correction amount learning values and the initial value, and use this average value as the fuel injection correction amount learning value for the fuel temperature corresponding to the default value of the fuel temperature, thereby executing the acquisition and update of the fuel injection correction amount learning value for the fuel temperature. Regarding the acquisition and update of the third basic learning value, For each required engine speed, A common rail type fuel injection control device, characterized in that it is configured to execute acquisition and update of fuel injection correction amount learning values for the assumed minimum and maximum fuel temperatures by extrapolation calculation based on the first basic learning value and the second basic learning value.

4. The electronic control unit (4) Regarding the acquisition and update of the first basic learning value, Obtain the coolant temperature immediately before starting the vehicle as the fuel temperature, and after starting the vehicle immediately, execute the acquisition and update of the fuel injection correction amount learning values for a plurality of preset specified fuel temperatures based on the fuel injection correction amount learning value selected from among the fuel injection correction amount learning values previously obtained by the fuel injection correction amount learning process and the coolant temperature. Obtain the coolant temperature immediately before starting the vehicle as the fuel temperature, and then, after starting the vehicle immediately, select a fuel injection correction amount learning value corresponding to the engine speed at that time from among the fuel injection correction amount learning values previously obtained by the fuel injection correction amount learning process. Calculate a learning value weight based on the learning value difference, which is the difference between the selected fuel injection correction amount learning value and the fuel injection correction amount learning value obtained in the most recent process. Then, calculate a distribution coefficient inversely proportional to the fuel temperature difference based on the fuel temperature difference between the coolant temperature and the specified fuel temperatures above and below the coolant temperature. Multiply the learning value weight by the distribution coefficient and add the multiplication result to the fuel injection correction amount learning value obtained in the most recent process. Repeat the process of updating the learning value with the addition result as the new fuel injection correction amount learning value, and execute the acquisition and update of the fuel injection correction amount learning values for the required specified fuel temperatures. The common rail type fuel injection control device according to claim 3, characterized in that it is configured as such.

Images