JP2016151258A - Fuel injection control device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To highly accurately control a combustion state without drastically increasing cost.SOLUTION: An ECU 90 (fuel injection control device) includes cylinder pressure waveform acquisition means, reference timing calculation means, vibration waveform acquisition means, phase difference calculation means and control means. The cylinder pressure waveform acquisition means acquires a cylinder pressure waveform of a reference cylinder 11#1 from a cylinder pressure sensor 14 mounted to the reference cylinder 11#1 out of a plurality of cylinders 11#1, 11#2, 11#3, 11#4. The vibration waveform acquisition means acquires a vibration waveform caused by combustion from an acceleration sensor 13 mounted to an internal combustion engine E. The phase difference calculation means calculates a phase difference between a waveform (reference vibration waveform) corresponding to a combustion period in the reference cylinder 11#1 out of the vibration waveforms and a waveform (other vibration waveform) corresponding to a combustion period in the other cylinder. The control means controls fuel injection timing to the other cylinder on the basis of the reference timing and phase difference calculated from the cylinder pressure waveform.SELECTED DRAWING: Figure 1

Description

  The present invention relates to a fuel injection control device that controls the timing of fuel injection into each cylinder of an internal combustion engine.

  In the case of an internal combustion engine that performs self-ignition combustion compared to an internal combustion engine that performs ignition combustion, if the timing at which combustion is actually started can be detected with high accuracy, the fuel injection timing is fed back so that the detected combustion start timing becomes the target timing. You will be able to control. Techniques for detecting the combustion start timing are disclosed in Patent Documents 1 and 2.

  In Patent Document 1, an acceleration waveform attached to an internal combustion engine acquires a vibration waveform representing a time change of vibration of the internal combustion engine caused by combustion. Since this type of vibration waveform has a change that increases in amplitude with the start of combustion, the combustion start time is detected by detecting the time when this change has occurred from the vibration waveform.

  In Patent Document 2, an in-cylinder pressure waveform representing a temporal change in in-cylinder pressure is acquired by an in-cylinder pressure sensor attached to a cylinder of an internal combustion engine. Since the in-cylinder pressure rapidly rises with the start of combustion, the combustion start time is detected by detecting the time when such a sudden rise has started from the in-cylinder pressure waveform.

Japanese Patent No. 5182157 Japanese Patent No. 26099892

  However, when the acceleration sensor of Patent Document 1 is used, the change in the cylinder pressure caused by combustion is indirectly captured by the vibration waveform, so the combustion start time cannot be detected with high accuracy. For example, the time when the amplitude of the vibration waveform becomes larger than the threshold value is detected as the combustion start time, but the shape of the vibration waveform changes greatly according to the combustion state, and the optimum value of the threshold value for each shape. Is different. Therefore, the combustion start time cannot be detected with high accuracy. Moreover, since the propagation time until the vibration generated at the start of combustion propagates from the cylinder to the acceleration sensor and the vibration attenuation caused by this propagation differ from cylinder to cylinder, it is possible to detect the combustion start timing with high accuracy from the vibration waveform. It is extremely difficult. Therefore, it is difficult to control the combustion state with high accuracy so that the combustion start timing becomes the target timing.

  On the other hand, when the in-cylinder pressure sensor of Patent Document 2 is used, the change in the in-cylinder pressure generated by combustion is directly captured by the in-cylinder pressure waveform, so that the combustion start timing can be detected with high accuracy. However, in the case of Patent Document 1, it is sufficient to provide one acceleration sensor for a plurality of cylinders, whereas in Patent Document 2, an expensive in-cylinder pressure sensor is provided for each cylinder. This requires a significant cost increase.

  The present invention has been made in view of the above problems, and an object of the present invention is to provide a fuel injection control device capable of controlling the combustion state with high accuracy without causing a significant increase in cost. is there.

  The invention disclosed herein employs the following technical means to achieve the above object. Note that the reference numerals in parentheses described in the claims and in this section indicate the correspondence with the specific means described in the embodiments described later, and do not limit the technical scope of the invention. .

  One of the disclosed inventions is a cylinder attached to a reference cylinder (11 # 1) among a plurality of cylinders (11 # 1, 11 # 2, 11 # 3, 11 # 4) of the internal combustion engine (E). In-cylinder pressure waveform acquisition means (S12) for acquiring the in-cylinder pressure waveform (WP) representing the time variation of the in-cylinder pressure of the reference cylinder from the internal pressure sensor (14), and based on the acquired in-cylinder pressure waveform, A vibration representing the time change of the vibration of the internal combustion engine caused by the combustion from the reference time calculation means (S16) for calculating the reference time (TP) which is the combustion start time and the acceleration sensor (13) attached to the internal combustion engine. The vibration waveform acquisition means (S18, S20) for acquiring the waveform (WA), and the waveform corresponding to the combustion period in the reference cylinder among the vibration waveforms is called the reference vibration waveform (WA # 1), and other cylinders other than the reference cylinder (11 # 2, 11 # 3, When the waveform corresponding to the combustion period in 1 # 4) is called other vibration waveform (WA # 2, WA # 3, WA # 4), the phase difference (τ # 2, The phase difference calculating means (S21) for calculating τ # 3, τ # 4) and the fuel injection timing to the reference cylinder are controlled based on the reference timing, and the fuel injection timing to the other cylinders is controlled based on the reference timing and And control means (S23) for controlling based on the phase difference.

  According to the present invention, for the reference cylinder, the combustion start time (reference time) can be detected with high accuracy based on the in-cylinder pressure waveform by the in-cylinder pressure sensor. Therefore, the fuel injection timing to the reference cylinder can be controlled based on the reference timing detected with high accuracy, so that the combustion start timing can be controlled with high accuracy.

  As described above, since the vibration waveform by the acceleration sensor is obtained by indirectly capturing the in-cylinder pressure change caused by the combustion with the vibration waveform, the combustion start timing cannot be detected with high accuracy. However, the phase difference between the reference vibration waveform and the other vibration waveform represents the difference between the combustion start timing of the reference cylinder and the combustion start timing of the other cylinder, and this phase difference is highly accurate even with the vibration waveform. It can be detected. Since the combustion start time (reference time) can be detected with high accuracy for the reference cylinder, the combustion start time for other cylinders can be increased based on the phase difference and the reference time detected with high accuracy. It should be possible to grasp with accuracy.

  In the above invention in view of this point, the phase difference is calculated for each of the other cylinders, and the injection timing is controlled based on the phase difference and the reference timing, so that the combustion start timing can be controlled with high accuracy for the other cylinders.

  As described above, according to the above-described invention, it is possible to grasp the combustion start timing for both the reference cylinder and the other cylinders without having to provide in-cylinder pressure sensors for all the cylinders. Accordingly, it is possible to control the combustion state with high accuracy without incurring a significant cost increase.

The schematic diagram which shows the fuel-injection control apparatus which concerns on one Embodiment of this invention. The figure which shows the one aspect | mode of the vibration waveform detected with the acceleration sensor of FIG. FIG. 2 is a view showing a correspondence relationship with a vibration waveform, which is an aspect of the in-cylinder pressure waveform detected by the in-cylinder pressure sensor of FIG. 1. The enlarged view of FIG. The figure which shows the combustion vibration component contained in a vibration waveform. The figure which shows the relationship between the value of cross correlation (phi) ((tau)) and phase difference (tau). The figure which shows the effect by correct | amending injection timing based on phase difference (tau) when the value (phi) (tau) of cross correlation becomes the maximum. The flowchart which shows the procedure of control of the fuel injection timing which the microcomputer of FIG. 1 performs.

  Hereinafter, an embodiment of a fuel injection control device according to the present invention will be described with reference to the drawings. The fuel injection control device according to the present embodiment is provided by an electronic control device (ECU 90) shown in FIG. The ECU 90 controls the combustion state of the internal combustion engine E by controlling the operations of the fuel injection valve 12, the EGR valve 42, the throttle valve 22 and the like provided in the internal combustion engine E. The internal combustion engine E and the ECU 90 are mounted on a vehicle, and the vehicle runs using the output of the internal combustion engine E as a drive source.

  The internal combustion engine E is a compression self-ignition diesel engine and has a plurality of cylinders 11. A fuel injection valve 12 is attached to each cylinder 11, and the fuel injected from the fuel injection valve 12 into the cylinder 11 is compressed by a piston (not shown) and self-ignited and combusted. In the example of FIG. 1, the engine is a four-cylinder engine. In the following description, the symbols of the first cylinder, the second cylinder, the third cylinder, and the fourth cylinder are denoted by 11 # 1, 11 # 2, 11 # 3, and 11 # 4. May be written. Also, the reference numerals of the fuel injection valves 12 attached to the cylinders 11 # 1, 11 # 2, 11 # 3, and 11 # 4 are expressed as 12 # 1, 12 # 2, 12 # 3, and 12 # 4. There is.

  The internal combustion engine E includes an intake pipe 20, an intake distribution pipe 21, an exhaust pipe 30, an exhaust collecting pipe 31, and a return pipe 40. The intake pipe 20 circulates intake air (fresh air) toward the cylinder 11. The intake pipe 21 is connected to the downstream end of the intake pipe 20 and distributes intake air to the plurality of cylinders 11. The exhaust collecting pipe 31 collects the exhaust discharged from each cylinder 11. The exhaust pipe 30 distributes the collected exhaust to an exhaust purification device (not shown). The recirculation pipe 40 recirculates a part of the exhaust gas to the intake air as EGR gas (recirculation gas). The downstream end of the reflux pipe 40 is connected to the intake pipe 21 or the intake pipe 20. As a result, the EGR gas cooled by the EGR cooler 41 is mixed with fresh air, and the intake air mixed with the EGR gas and fresh air is distributed to each cylinder 11 by the intake air distribution pipe 21.

  The shape of the intake pipe 21 is set so that the intake air is evenly distributed to the cylinders 11. However, it is difficult to evenly distribute the EGR gas to the respective cylinders 11, and distribution variation occurs due to the position where the downstream end of the reflux pipe 40 is connected, the shape of the intake pipe 21, and the like. . That is, the EGR gas concentration of the intake air flowing into each cylinder 11 varies. In the following description, the cylinder with the largest amount of EGR gas contained in the distributed intake air is referred to as a multi-return cylinder. Arrows A, B, C, and D in FIG. 1 indicate the flow of EGR gas flowing into each cylinder 11. In the example of FIG. 1, the layout is such that EGR gas is most likely to flow into the first cylinder 11 # 1 as indicated by an arrow A, and the first cylinder 11 # 1 corresponds to a multi-recirculation cylinder.

  A throttle valve 22 is attached to the intake pipe 20, and the flow rate of fresh air contained in the intake air is adjusted by adjusting the opening of the throttle valve 22. An EGR cooler 41 and an EGR valve 42 are attached to the reflux pipe 40, and the flow rate of EGR gas contained in the intake air is adjusted by adjusting the opening degree of the EGR valve 42.

  Detection signals output from the acceleration sensor 13, the in-cylinder pressure sensor 14, the accelerator sensor 15, and the crank angle sensor 16 are input to the ECU 90. The in-cylinder pressure sensor 14 is attached to the first cylinder 11 # 1, which is a multi-return cylinder, and detects the pressure in the combustion chamber formed by the first cylinder 11 # 1. A microcomputer (microcomputer 91) provided in the ECU 90 obtains an in-cylinder pressure waveform WP (see the upper part of FIG. 3) that represents a change over time in the detected in-cylinder pressure.

  The acceleration sensor 13 is attached to the cylinder block 10 that forms each cylinder 11 and detects acceleration due to vibration of the cylinder block 10. In addition to vibration due to combustion, vibration due to rotation of the crankshaft, vibration due to operation of the intake / exhaust valve, vibration due to operation of the valve body of the fuel injection valve 12 and the like are superimposed on the vibration of the cylinder block 10. The acceleration sensor is attached to the cylinder block 10 at a position closest to the multi-return cylinder (first cylinder 11 # 1). For example, the acceleration sensor 13 is attached to the position where the distance from the center position of the multi-reflux cylinder to the acceleration sensor 13 is the shortest among the distances from the center position of each cylinder 11 to the acceleration sensor 13. The microcomputer 91 acquires a vibration waveform WA (see the lower part of FIG. 3) that represents a change over time of the detected vibration of the cylinder block 10.

  The microcomputer 91 calculates the amount of depression of the accelerator pedal (engine load) based on the detection signal of the accelerator sensor 15. The microcomputer 91 calculates the rotational speed (engine rotational speed) per unit time, which is the rotational speed of the output shaft (crankshaft) of the internal combustion engine E, based on the detection signal of the crank angle sensor 16.

  The microcomputer 91 calculates various target values representing the injection state of the fuel injected from the fuel injection valve 12 based on the calculated engine load and engine speed. For example, the target injection amount, the target injection timing, the number of injections in the case of performing multistage injection in which fuel is injected a plurality of times in one combustion cycle in the same cylinder 11 are calculated.

  Hereinafter, a method for calculating the target injection timing will be described in detail. The microcomputer 91 functions as various means described below when the CPU executes arithmetic processing according to a program stored in the memory. The target injection timing is calculated by these means, and the operation of the fuel injection valve 12 is controlled based on the target injection timing. That is, the microcomputer 91 functions as target combustion timing calculation means 91a, actual combustion timing calculation means 91b, phase difference calculation means 91c, and injection command signal setting means 91d.

  The target combustion timing calculation unit 91a calculates a target combustion timing based on the engine load and the engine speed. For example, in consideration of the balance between the output torque according to the engine load and the engine speed, the exhaust emission, and the combustion noise, the optimum combustion time (optimum combustion time) according to the engine load and the engine speed is tested in advance. And get it. Then, the relationship between the engine load and the engine speed and the optimum combustion timing is stored in the microcomputer 91 in the form of a map (combustion timing map). The combustion timing map is set for each cylinder 11. The target combustion timing calculation means 91a calculates the target combustion timing as the optimal combustion timing based on the engine load and the engine speed with reference to the combustion timing map.

  The injection command signal setting unit 91d calculates a target injection timing corresponding to the target combustion timing. For example, the fuel injection timing for realizing the target combustion timing is obtained by testing in advance as the target injection timing. Then, the relationship between the target combustion timing and the target injection timing is stored in the microcomputer 91 in the form of a map (injection timing map). The injection timing map is set for each cylinder 11. The injection command signal setting unit 91d sets the injection timing corresponding to the target combustion timing among the injection timings stored in the injection timing map as the target injection timing. However, the value stored in the injection timing map, that is, the target injection timing corresponding to the target combustion timing is sequentially corrected based on the calculation results by the actual combustion timing calculation means 91b and the phase difference calculation means 91c.

  Based on the in-cylinder pressure waveform WP detected by the in-cylinder pressure sensor 14, the actual combustion timing calculation unit 91b calculates a timing at which actual combustion has started in the first cylinder 11 # 1. For example, a time when the increase rate of the in-cylinder pressure becomes a predetermined value or more is calculated, and that time is set as the combustion start time (reference time TP).

  The phase difference calculation unit 91c calculates the phase difference τ of the combustion vibration waveform WB generated by combustion based on the vibration waveform WA detected by the acceleration sensor 13. The combustion vibration waveform WB is a waveform obtained by extracting a component having a predetermined frequency or less from the vibration waveform WA, and is obtained by converting the vibration waveform WA using, for example, a low-pass filter. The cut-off frequency of the low-pass filter is set to 1.5 kHz, for example. Mainly, the vibration waveform generated when the fuel injection valve 12 is driven is removed from the vibration waveform WA.

  A combustion vibration waveform WB # 1 shown in FIG. 2 is a component of a vibration waveform generated with combustion in the first cylinder 11 # 1 (reference cylinder), and a waveform that appears in the combustion process in the first cylinder 11 # 1. It is. Combustion vibration waveforms WB # 2, WB # 3, and WB # 4 are components of vibration waveforms generated by combustion in cylinders (other cylinders) other than the first cylinder 11 # 1, and each cylinder 11 # 2, 11 It is the waveform that appeared in the combustion process at # 3 and 11 # 4.

  These combustion vibration waveforms WB # 1, WB # 2, WB # 3, and WB # 4 have similar shapes. Specifically, the number of waves included in each combustion vibration waveform WB, the magnitude ratio of the plurality of waves, the pitch of the plurality of waves, and the like are similar. However, the farther the distance from the center position of the cylinder 11 to the acceleration sensor 13 is the combustion vibration waveform WB of the cylinder 11, the smaller the amplitude of the waveform.

  When the appearance time of the combustion vibration waveform WB with respect to the reference time in one combustion cycle of each cylinder 11 is called a combustion phase, the difference in the combustion phase corresponds to the phase difference τ. In the example of FIG. 2, the reference time is the top dead center time (TDC). Then, the phase difference τ # 3 of the third cylinder 11 # 3 with respect to the reference cylinder, the phase difference τ # 4 of the fourth cylinder 11 # 4, and the phase difference τ # 2 of the second cylinder 11 # 2 are converted into phase difference calculating means. 91c is calculated. In the example of FIG. 2, the first cylinder 11 # 1, the third cylinder 11 # 3, the fourth cylinder 11 # 4, and the second cylinder 11 # 2 are burned in this order.

  In general, the time shifted by the phase difference τ from the reference time TP corresponds to the combustion start time in the other cylinders. For example, it can be said that the combustion start time in the third cylinder 11 # 3 is a time obtained by subtracting the phase difference τ # 3 from the reference time TP. It can be said that the combustion start time in the fourth cylinder 11 # 4 is a time obtained by subtracting the phase difference τ # 4 from the reference time TP. It can be said that the combustion start time in the second cylinder 11 # 2 is a time obtained by subtracting the phase difference τ # 2 from the reference time TP.

  Returning to the description of the injection command signal setting unit 91d in FIG. 1, the target injection timing stored in the injection timing map is sequentially corrected based on the calculation results by the actual combustion timing calculation unit 91b and the phase difference calculation unit 91c. As described above. Specifically, for the reference cylinder, the difference between the actual combustion timing (reference timing TP) calculated by the actual combustion timing calculation means 91b and the target injection timing when the injection is executed is stored in the injection timing map. It is corrected by adding to the target injection timing. For the other cylinders, the phase differences τ # 2, τ # 3, τ # 4 calculated by the phase difference calculating means 91c are added to the target injection timing stored in the injection timing map to be corrected.

  Further, the injection command signal setting unit 91d generates an energization command signal based on the target injection timing and the target injection amount calculated as described above, and outputs them to the drive circuit 92. The energization command signal is a pulse signal that turns on a switch element included in the drive circuit 92. The pulse-on timing of the pulse signal is set based on the target injection timing, and the pulse-on period of the pulse signal is set based on the target injection amount.

  The drive circuit 92 has a switch element (not shown) that switches on / off of energization to a solenoid coil (not shown) of the fuel injection valve 12. The switch element operates based on the pulse signal generated by the injection command signal setting unit 91d, and switches on / off of voltage application to each fuel injection valve 12. As a result, the solenoid coil is energized during the pulse-on period of the pulse signal, and the fuel injection valve 12 is opened to inject fuel. Therefore, the fuel injection start timing is controlled by the pulse-on start timing. Further, the valve opening time is controlled by the length of the pulse-on period, and as a result, the amount of fuel injected by one valve opening (fuel injection amount) is controlled.

  However, the fuel is not injected by opening the valve at the same time as the pulse-on start, but the fuel injection is started by opening the valve slightly after the start of the pulse-on. Furthermore, self-ignition combustion does not occur simultaneously with the start of fuel injection, and self-ignition combustion occurs after the start of injection. Thus, the delay time from the start of injection until combustion occurs is called the ignition delay time.

  FIG. 3 shows the in-cylinder pressure waveform WP and the vibration waveform WA detected by the in-cylinder pressure sensor 14 and the acceleration sensor 13 during operation of the internal combustion engine E. Combustion in the first cylinder 11 # 1 (reference cylinder) can be acquired not only for the combustion vibration waveform WB # 1 but also for the in-cylinder pressure waveform WP for the reference cylinder, as shown in the portion surrounded by the dotted line in FIG. is there. As shown in the upper part of FIG. 3, the point at which the pressure appearing in the in-cylinder pressure waveform WP rapidly rises represents the combustion start timing (reference time TP) of the reference cylinder.

  4B and 4D are enlarged views of a portion surrounded by a dotted line in FIG. FIG. 4A shows a change in drive current flowing through the solenoid coil of the fuel injection valve 12 in accordance with the pulse signal generated by the injection command signal setting means 91d. The symbol TS in FIG. 4A indicates the time when the drive current starts to flow, that is, the injection start command time when the pulse signal is turned on, and the symbol TE indicates the time when the drive current becomes zero, that is, the pulse signal is turned off. Indicates the injection end command time. FIG. 4C shows a change in heat generated with combustion in the reference cylinder, and shows a change in the amount of heat (heat generation rate) generated per unit time. Since the combustion start time (reference time TP), the heat generation rate has increased rapidly.

  In the example of FIG. 4, the amplitude of the vibration waveform WA increases rapidly at the time TA when the time of the propagation delay td has elapsed from the injection start command time TS. The vibration whose amplitude has increased in this way is the vibration that has occurred when the valve body of the fuel injection valve is opened and propagates through the cylinder block 10 to reach the acceleration sensor 13. Thereafter, at a predetermined time after the combustion start time (reference time TP), the frequency and amplitude of the vibration waveform WA are rapidly increased. The vibration whose amplitude has increased in this way is vibration that has occurred in the combustion in the reference cylinder (combustion vibration) and has propagated through the cylinder block 10 to reach the acceleration sensor 13. The low-frequency vibration component included in the vibration waveform WA has a high correlation with the combustion vibration. Hereinafter, the low frequency vibration component extracted from the vibration waveform WA by the low-pass filter is referred to as a combustion vibration waveform WB.

  In the vibration waveform WA, vibration waveforms WA # 1, WA # 2, WA # 3, and WA # 4 corresponding to the combustion period of each cylinder sequentially appear. For example, a portion of the vibration waveform WA corresponding to the combustion period of the first cylinder 11 # 1 represents vibration caused by the combustion of the first cylinder 11 # 1.

  The upper part of FIG. 5 shows the reference vibration waveform WA # 1 and the combustion vibration waveform WB # 1 that are vibration waveforms of the reference cylinder, and the lower part of FIG. 5 shows the other vibration waveform WA # that is the vibration waveform of the third cylinder 11. 3 and the combustion vibration waveform WB # 3. The phase difference calculating unit 91c performs an operation representing the vibration (amplitude value) of the combustion vibration waveform WB # 1 as a function x (t) with the elapsed time t as a variable. The reference time of the elapsed time that is a variable of the function x (t) is set to the injection start command time TS in the reference cylinder. Further, an operation is performed to express the vibration (amplitude value) of the combustion vibration waveform WB # 3 as a function y (t) with the elapsed time t as a variable. The elapsed time reference time, which is a variable of this function y (t), is set to the injection start command time TS in the third cylinder 11 # 3.

  Therefore, if the injection start command time TS with respect to the top dead center timing TDC of the reference cylinder and the third cylinder 11 # 3 is the same, the waveforms of the function x (t) and the function y (t) should almost overlap. However, since there is actually a combustion delay time difference described below, the function x (t) and the function y (t) are out of phase by that time difference. The amount of phase shift between these two functions corresponds to the phase difference τ described in the claims.

  The combustion start delay time from the injection start command time point TS to the combustion start timing is different for each cylinder even when fuel is injected at the same injection start timing in each cylinder. The main cause of this phenomenon is the aforementioned EGR distribution variation. That is, since the self-ignitability is deteriorated in the multi-recirculation cylinder, the combustion start delay time becomes longer than in the other cylinders. Thus, the phase difference τ is caused by the fact that the combustion start delay time is different for each cylinder.

  Further, the above-described propagation delay time td is also different for each cylinder. The difference in propagation delay td between the reference cylinder and other cylinders is referred to as propagation delay time difference Δtd. In the example of FIG. 5, the difference between the propagation delay td # 1 in the reference cylinder and the propagation delay td # 3 in the third cylinder 11 corresponds to the propagation delay time difference Δtd.

  The phase difference calculating unit 91c digitizes the similarity between the function x (t) and the function y (t) with a cross-correlation function φ (τ) using the phase difference τ as a variable. FIG. 6 is a graph showing the relationship between the value (cross-correlation value) quantified by the cross-correlation function φ (τ) and the phase difference τ. The cross-correlation value is maximized at the portion indicated by the alternate long and short dash line A. That is, when the phase difference τ is set to the value of this portion, the similarity between the function x (t) and the function y (t) increases, and the waveforms of the function x (t) and the function y (t) almost overlap. Means that. Hereinafter, when simply referred to as the phase difference τ, it means a phase difference that maximizes the cross-correlation value. The time obtained by adding / subtracting the propagation delay time difference Δtd to / from the phase difference τ corresponds to the difference between the combustion start delay time in the reference cylinder and the combustion start delay time in the third cylinder 11 # 3 (combustion start delay time difference).

  The solid line in FIG. 7 shows the combustion waveform of the reference cylinder expressed by the function x (t). The dotted line in FIG. 7 shows the combustion waveform of the third cylinder 11 # 3 expressed by the function y (t). The upper part of FIG. 7 shows the phase difference τ between the function x (t) and the function y (t), and the lower part of FIG. 7 is a case where the function x (t) is corrected so as to be retarded by the phase difference τ. A function x (t) and a function y (t) are shown. Thus, if it correct | amends only the part for phase difference (tau), it will be confirmed that the function x (t) and the function y (t) almost overlap.

  FIG. 8 shows a procedure of processing repeatedly executed by the microcomputer 91 at a predetermined cycle, and is executed during operation of the internal combustion engine E. As described above, the microcomputer 91 performs the injection control based on the target injection amount, the target injection timing, the number of injections, and the like. FIG. 8 is a process for performing the injection control by setting the target injection timing according to the target combustion timing. is there.

  First, in step S10 of FIG. 8, the target combustion timing is calculated based on the engine load and the engine speed detected by the accelerator sensor 15 and the crank angle sensor 16. In a succeeding step S11, it is determined whether or not the cylinder in the combustion process performed most recently at the present time is a reference cylinder.

When an affirmative determination is made that the cylinder is the reference cylinder, detection values output from the in-cylinder pressure sensor 14 and the acceleration sensor 13 are acquired in the next step S12 (in-cylinder pressure waveform acquisition means) and step S13. A plurality of continuous detection values (voltages) of the acceleration sensor 13 correspond to the vibration waveform WA # 1 related to the combustion of the reference cylinder. In the subsequent step S14, the voltage that is the detection value of the in-cylinder pressure sensor 14 is converted into the in-cylinder pressure P (θ) according to the following equation (1). The plurality of continuous values of the converted in-cylinder pressure P (θ) correspond to the in-cylinder pressure waveform WP related to the combustion of the reference cylinder.

  K in the equation 1 is a constant that is a conversion gain of the in-cylinder pressure with respect to the voltage. VS in Equation 1 is a detection value (voltage) of the in-cylinder pressure sensor 14. In Equation 1, θ is the rotation angle of the crankshaft, and θ0 is the rotation angle (valve closing angle) when the intake valve is closed. PIm in Formula 1 is the intake pressure in the intake distribution pipe 21. In short, the in-cylinder pressure P (θ), which is a function of the variable θ, is a value obtained by adding the intake pressure in the intake distribution pipe 21 to the pressure increased after the start of closing of the intake valve (at the start of compression).

In subsequent step S15, based on the in-cylinder pressure waveform WP, a temporal change in the amount of heat generated by combustion in the reference cylinder is calculated as a heat generation rate waveform. In short, based on the in-cylinder pressure P (θ) calculated by Equation 1, the heat release rate HRR (θ) is calculated according to the following Equation 2.

  In Equation 2, Cv is a constant volume molar specific heat, and Cp is a constant pressure molar specific heat. V in Formula 2 is the cylinder volume in the cylinder. In short, the heat release rate HRR (θ), which is a function of the variable θ, is a value obtained by adding the heat change rate generated with the change in the cylinder volume and the heat change rate generated with the change in the in-cylinder pressure P (θ). .

  In the subsequent step S16 (reference time calculation means), the combustion start time (reference time TP) in the reference cylinder is calculated based on the waveform of the heat release rate. Specifically, the combustion timing MFB50 corresponding to an angle that is 50% of the total heat quantity is calculated as the combustion start timing.

  In the subsequent step S17, a deviation ΔMFB50 between the reference timing TP, which is the actual combustion timing MFB50, and the target combustion timing MFB50trg is calculated. Then, by multiplying the deviation ΔMFB50 by a predetermined feedback gain, the injection timing correction amount ΔTInj # 1 is calculated. That is, the target injection timing corresponding to the target combustion timing stored in the injection timing map is rewritten by adding / subtracting the correction amount ΔTInj # 1.

  In the subsequent step S18 (vibration waveform acquisition means), the combustion vibration waveform WB # 1 is extracted from the vibration waveform WA # 1 (reference vibration waveform) using a low-pass filter. The combustion vibration waveform WB # 1 related to this extraction is used for the calculation in step S21 described later, and is used for the phase difference τ calculation by the phase difference calculating means 91c described above.

  On the other hand, if a negative determination is made in step S11 that the cylinder is not the reference cylinder, the detection value output from the acceleration sensor 13 is acquired in the next step S19. A plurality of continuous detection values by this acquisition correspond to other vibration waveforms WA # 2, WA # 3, WA # 4 that are vibration waveforms of the corresponding other cylinders. In subsequent step S20 (vibration waveform acquisition means), combustion vibration waveforms WB # 2, WB # 3, and WB # 4 are extracted from vibration waveforms WA # 2, WA # 3, and WA # 4 using a low-pass filter.

In the subsequent step S21 (phase difference calculating means), the cross-correlation function φ (τ) is calculated according to the following equation 3 based on the combustion vibration waveform WB extracted in each of step S18 and step S20. For example, when the vibration waveform WB extracted in step S20 is the vibration waveform WB # 3 related to the combustion of the third cylinder 11, a function y (t) representing the vibration waveform WB # 3 is calculated. Also, a function x (t) representing the vibration waveform WB # 1 related to the reference cylinder is calculated. The similarity between the function y (t + τ) and the function x (t) when the function y (t) is retarded by the phase difference τ is the value of the cross-correlation function φ (τ) (cross-correlation value). It is expressed by That is, the cross correlation value is a variable of the phase difference τ.

  The phase difference τ (see A in FIG. 6) at which the cross-correlation value of the cross-correlation function φ (τ) is maximum is the phase shift amount between the vibration waveform WB # 1 and the vibration waveform WB # 3.

In the subsequent step S22, the injection amount correction amount ΔTInj of the corresponding other cylinder is calculated according to the following equation (4).

  In Equation 1, td is the propagation delay of the corresponding other cylinder, and td # 1 is the propagation delay of the reference cylinder. Therefore, (td−td # 1) in Equation 4 represents the propagation delay time difference Δtd. (Τ− (td−td # 1)) is the difference between the combustion start delay time in the reference cylinder and the combustion start delay time in the other cylinders (combustion start delay) taking into account the difference in the propagation delay time td between the cylinders. Time difference). The propagation delay time td in each cylinder is the result of a test performed in advance, and a value stored in advance in the microcomputer 91 is used.

  In Equation 1, Ne is the engine speed. Therefore, ((τ− (td−td # 1)) · 6Ne) in Equation 4 is a value obtained by converting the combustion start delay time difference into a crank angle. In short, the correction amount ΔTInj related to the other cylinders is calculated based on the correction amount ΔTINj # 1 related to the reference cylinder calculated in step S17, the phase difference τ (phase shift amount) calculated in step S21, and the propagation delay time difference Δtd. .

  The value of the target injection timing stored in the above-described injection timing map is associated with the engine speed and the engine load, and is corrected and updated by the correction amount ΔTInj # 1 related to the reference cylinder calculated in step S17. Is done. That is, learning is performed based on the detection value of the in-cylinder pressure sensor 14, and the injection timing is feedback controlled. On the other hand, the correction amount ΔTInj related to the other cylinders calculated in step S22 is stored in the injection timing correction map in association with the engine speed and the engine load.

  In the subsequent step S23 (control means), the target combustion timing calculated in step S10 and the injection timing correction amount calculated in step S17 and step S22 for the fuel injection valve 12 of the cylinder 11 in the next combustion process are set. Based on the injection control. That is, for the injection of the reference cylinder, the pulse-on timing of the pulse signal is controlled so that the injection timing corresponding to the target combustion timing is reached with reference to the injection timing map learned in step S17. On the other hand, for the injection of other cylinders, the pulse-on timing calculated with reference to the injection timing map learned as described above is corrected based on the injection timing correction map learned in step S22.

  As described above, according to the present embodiment, the effects described below are exhibited.

  (1) In this embodiment, based on the in-cylinder pressure waveform WP acquired from the in-cylinder pressure sensor 14, the combustion start time (reference time TP) in the first cylinder 11 # 1 that is the reference cylinder is calculated. Of the vibration waveform WA acquired from the acceleration sensor 13, the portion corresponding to the reference cylinder is a reference vibration waveform WA # 1, and the second cylinder 11 # 2, the third cylinder 11 # 3, and the fourth cylinder 11 # 4 ( The portions corresponding to other cylinders) are referred to as other vibration waveforms WA # 2, WA # 3, and WA # 4. Based on the phase differences τ # 2, τ # 3, τ # 4 between the reference vibration waveform WA # 1 and the other vibration waveforms WA # 2, WA # 3, WA # 4, and the reference time TP, Control fuel injection timing.

  Here, even if the fuel injection timing and the injection amount are the same, the combustion timing differs for each cylinder 11. One of the causes is EGR distribution variation. That is, in the cylinder 11 having a large EGR distribution amount, the ignitability is deteriorated, so that the ignition delay time becomes long and the combustion start timing becomes late.

  Now, since the vibration waveform WA by the acceleration sensor 13 is obtained by indirectly capturing the in-cylinder pressure change caused by the combustion by the vibration of the cylinder block 10, the combustion start timing cannot be detected with high accuracy. However, the phase differences τ # 2, τ # 3, τ # 4 between the reference vibration waveform WA # 1 and the other vibration waveforms WA # 2, WA # 3, WA # 4 are determined based on the combustion start timing of the reference cylinder and the other cylinders. The difference from the combustion start time is expressed with high accuracy. Therefore, based on the combustion start time (reference time TP) detected with high accuracy by the in-cylinder pressure sensor 14 and the phase differences τ # 2, τ # 3, τ # 4, the combustion start time can be determined with high accuracy for the other cylinders. You should be able to grasp it.

  In this embodiment in view of this point, the phase differences τ # 2, τ # 3, and τ # 4 are calculated, and the injection timing for each of the other cylinders is controlled based on the phase differences τ and the reference timing TP. Therefore, the combustion start timing can be controlled with high accuracy for the other cylinders. As described above, according to the present embodiment, it is possible to grasp the combustion start timing for both the reference cylinder and the other cylinders without providing the cylinder pressure sensors 14 in all the cylinders 11. Therefore, it is possible to control the combustion start timing for all the cylinders 11 with high accuracy without incurring a significant cost increase.

  (2) For example, the phase difference τ # 3 between the reference vibration waveform WA # 1 and the other vibration waveform WA # 3 shown in FIG. 7 can be calculated as follows, contrary to the present embodiment. That is, the amount of deviation between the peak point at which the amplitude of the reference vibration waveform WA # 1 is maximum and the peak point at which the amplitude of the other vibration waveform WA # 3 is maximum may be calculated as a phase difference. In short, it is possible to calculate the time when the characteristics of each vibration waveform appear, and to calculate the amount of deviation of these times as the phase difference.

  However, the characteristic point of the vibration waveform varies depending on the state of combustion. For example, the characteristics of the vibration waveform are different between a combustion state in which the heat generation rate suddenly increases in a short time and a combustion state in which the heat generation rate increases slowly. In addition, the characteristics of the vibration waveform are different between multi-stage injection and single-stage injection. As described above, the characteristics of the vibration waveform differ depending on the combustion state. Therefore, when the amount of deviation of the feature appearance time is calculated as the phase difference as described above, a phase difference that accurately represents the difference in the combustion start time is obtained. I can't do it.

  In this embodiment in view of this point, the phase difference calculation unit 91c uses the cross-correlation function φ (τ) to determine the similarity between the reference vibration waveform WA # 1 and the other vibration waveforms WA # 2, WA # 3, and WA # 4. By calculating numerical values, phase differences τ # 2, τ # 3, and τ # 4 are calculated. According to this, when the phase difference τ is changed, the degree of overlap of the respective vibration waveforms is digitized as a cross-correlation value, and the phase difference τ that maximizes the cross-correlation value is calculated. For this reason, it is possible to obtain a phase difference that represents the difference in the combustion start timing with higher accuracy than when the characteristics of each vibration waveform are digitized and the difference is calculated as the phase difference. In short, rather than digitizing and comparing features for each vibration waveform, the degree of overlap of the vibration waveform shape itself is digitized to calculate the phase difference, so that a highly accurate phase difference can be acquired.

  (3) The propagation time td # 1, td # 2 is different because the path and length until the combustion vibration propagates through the cylinder block 10 and reaches the acceleration sensor 13 differs depending on the mounting position of the acceleration sensor 13. , Td # 3 and td # 4 are different for each cylinder 11. Therefore, the combustion vibration waveform WB is included in the vibration waveform WA with a phase delayed by the propagation time td. However, since the propagation time td differs for each cylinder, the combustion vibration waveform WB is obtained from the cross-correlation function φ (τ). It cannot be said that the phase difference τ accurately represents the difference in the ignition delay time. For example, the propagation time td # 1 of the first cylinder 11 # 1 is different from the propagation time td # 3 of the third cylinder 11 # 3. Therefore, the phase difference τ # 3 obtained from the cross-correlation function φ (τ) includes the difference between the propagation times td # 1 and td # 3.

  In the present embodiment in view of this point, the control means (step S23) moves to the other cylinders based on the propagation time td set in advance for each of the cylinders 11 in addition to the reference time TP and the phase difference τ. The fuel injection timing is controlled. Therefore, for example, in the case of FIG. 7, the injection timing of the third cylinder 11 # 3 (other cylinders) can be controlled by subtracting the difference (td # 3-td # 1) in the propagation time td from the phase difference τ # 3. Therefore, since it is considered that the propagation time td differs between cylinders, the accuracy of grasping the combustion start time can be improved.

  (4) The more EGR gas, the longer the time (combustion delay time) from the start of energization to the fuel injection valve 12 until combustion. For this reason, in the cylinder (multiple recirculation cylinder) in which EGR gas is the largest, the degree of increase in the combustion delay time with the increase in EGR gas is large. Therefore, when the EGR gas increases, it is desirable to quickly detect that the combustion delay time has become longer and feed it back to the injection control.

  In this embodiment in view of this point, the in-cylinder pressure sensor 14 is attached to the multi-reflux cylinder. Therefore, in the multi-recirculation cylinder in which the combustion delay time tends to be long, it is possible to quickly detect that the combustion delay time is long and feed back to the injection control. Therefore, when the operating state of the internal combustion engine E suddenly changes and the EGR amount suddenly changes, it is possible to quickly realize that the actual combustion timing matches the target combustion timing.

  (5) Since the vibration due to combustion is smaller as the EGR gas is larger, the amplitude appearing in the vibration waveform WA is smaller, and the accuracy of the phase difference τ may be deteriorated. In the present embodiment in view of this point, an acceleration sensor is attached to the internal combustion engine E at a position closest to the multi-return cylinder. For this reason, the above-mentioned concern due to the small combustion vibration in the multi-recirculation cylinder can be suppressed.

(Other embodiments)
The preferred embodiments of the present invention have been described above, but the present invention is not limited to the above-described embodiments, and various modifications can be made as illustrated below. Not only combinations of parts that clearly show that combinations are possible in each embodiment, but also combinations of embodiments even if they are not explicitly stated, unless there is a problem with the combination. Is also possible.

  In the embodiment shown in FIG. 8, the phase difference τ is calculated using the cross-correlation function φ (τ). On the other hand, the amount of deviation between the peak point at which the amplitude of the reference vibration waveform WA # 1 is maximized in the predetermined period and the peak point at which the amplitude of the other vibration waveform WA # 3 is maximized in the predetermined period is calculated as a phase difference. May be.

  In step S16 of FIG. 8, the combustion timing MFB50 corresponding to an angle that is 50% of the total heat quantity is calculated as the combustion start timing. On the other hand, a time corresponding to an intermediate angle between the angle that becomes 10% of the total heat quantity and the angle that becomes 90% may be calculated as the combustion start time. Alternatively, the differential value at each time point of the in-cylinder pressure waveform WP may be calculated, and the time point when the differential value becomes larger than a predetermined value may be calculated as the combustion start time (reference time TP).

  In the embodiment shown in FIG. 1, the control device of the present invention is applied to the control of the fuel injection valve 12 of the diesel engine. On the other hand, the control device of the present invention may be applied to fuel injection control of a direct injection engine that is an ignition ignition gasoline engine and directly injects fuel into the combustion chamber.

  The presence or absence of abnormality may be determined based on the combustion start timing of the other cylinders ascertained based on the reference timing and the phase difference. For example, the combustion start delay time is calculated based on the grasped combustion start timing of the other cylinder, and if the calculated combustion start delay time is equal to or longer than a predetermined time, it is determined that an abnormality has occurred. Causes of the abnormality include use of inappropriate fuel, failure of the fuel injection valve 12 and the fuel pump, and the like.

  In the embodiment shown in FIG. 1, the acceleration sensor 13 is attached to the cylinder block 10, but may be attached to a cylinder head (not shown).

  Means and / or functions provided by the ECU 90 (control device) can be provided by software recorded in a substantial storage medium and a computer that executes the software, only software, only hardware, or a combination thereof. For example, if the controller is provided by a circuit that is hardware, it can be provided by a digital circuit including a number of logic circuits, or an analog circuit.

  11 # 1 ... 1st cylinder (reference cylinder), 11 # 2 ... 2nd cylinder (other cylinder), 11 # 3 ... 3rd cylinder (other cylinder), 11 # 4 ... 4th cylinder (other cylinder), 13 ... Acceleration sensor, 14 ... In-cylinder pressure sensor, E ... Internal combustion engine, S16 ... Reference time calculation means, S18, S20 ... Vibration waveform acquisition means, S21 ... Phase difference calculation means, S23 ... Control means, TP ... Reference time, WA # 1 ... Reference vibration waveform, WP ... In-cylinder pressure waveform, τ # 2, τ # 3, τ # 4 ... Phase difference.

Claims (5)

From an in-cylinder pressure sensor (14) attached to a reference cylinder (11 # 1) among a plurality of cylinders (11 # 1, 11 # 2, 11 # 3, 11 # 4) of the internal combustion engine (E), In-cylinder pressure waveform acquisition means (S12) for acquiring an in-cylinder pressure waveform (WP) representing a time variation of the in-cylinder pressure of the reference cylinder;
A reference timing calculation means (S16) for calculating a reference timing (TP) that is a combustion start timing in the reference cylinder based on the acquired in-cylinder pressure waveform;
Vibration waveform acquisition means (S18, S20) for acquiring, from an acceleration sensor (13) attached to the internal combustion engine, a vibration waveform (WA) representing a temporal change in vibration of the internal combustion engine caused by combustion;
Of the vibration waveforms, a waveform corresponding to the combustion period in the reference cylinder is referred to as a reference vibration waveform (WA # 1), and the other cylinders other than the reference cylinder (11 # 2, 11 # 3, 11 # 4) When the waveform corresponding to the combustion period is referred to as other vibration waveform (WA # 2, WA # 3, WA # 4), the phase difference between the reference vibration waveform and the other vibration waveform (τ # 2, τ # 3, phase difference calculating means (S21) for calculating τ # 4);
Control means (S23) for controlling the injection timing of fuel to the reference cylinder based on the reference timing, and controlling the injection timing of fuel to the other cylinder based on the reference timing and the phase difference;
A fuel injection control device comprising:   2. The fuel injection control according to claim 1, wherein the phase difference calculating unit calculates the phase difference by quantifying the similarity between the reference vibration waveform and the other vibration waveform using a cross-correlation function. apparatus. Propagating the time from the start of energization to the fuel injection valves (12 # 1, 12 # 2, 12 # 3, 12 # 4) provided in the cylinder until the vibration is propagated to the acceleration sensor In the case of calling time (td # 1, td # 2, td # 3, td # 4),
The control means controls fuel injection timing to the other cylinders based on the propagation time set in advance for each of the cylinders in addition to the reference timing and the phase difference. The fuel injection control device according to claim 1 or 2. The internal combustion engine includes a recirculation pipe (40) that recirculates a part of the exhaust gas as recirculation gas to the intake air, an intake distribution pipe (21) that distributes the intake air mixed with the recirculation gas and fresh air to the plurality of cylinders, With
Of the plurality of cylinders, when a cylinder having the largest amount of the recirculation gas contained in the distributed intake air is referred to as a multi recirculation cylinder, the in-cylinder pressure sensor is attached to the multi recirculation cylinder. The fuel injection control device according to any one of claims 1 to 3. The internal combustion engine includes a recirculation pipe (40) that recirculates a part of the exhaust gas as recirculation gas to the intake air, an intake distribution pipe (21) that distributes the intake air mixed with the recirculation gas and fresh air to the plurality of cylinders, With
In the case where the cylinder in which the recirculation gas contained in the distributed intake air is the largest among the plurality of cylinders is called a multi-recirculation cylinder, the acceleration sensor is located closest to the multi-recirculation cylinder in the internal combustion engine. The fuel injection control device according to claim 1, wherein the fuel injection control device is attached to the fuel injection control device.

Images