JP2020002940A - Control device of compression ignition type engine - Google Patents

Abstract

To improve estimation accuracy in self-ignition timing of an unburned air-fuel mixture, in an engine performing SPCCI combustion.SOLUTION: A part of an air-fuel mixture starts combustion with flame propagation by forcible ignition of an ignition portion, and then the remaining unburned air-fuel mixture is burned by self-ignition. A control portion (ECU10) outputs an ignition signal to the ignition portion before a target timing so that the unburned air-fuel mixture is self-ignited at the target timing. The control portion has a correction portion for estimating a timing when a cylinder internal pressure parameter is over a threshold value as the self-ignition timing of the unburned air-fuel mixture, on the basis of the cylinder internal pressure parameter relating to a pressure in a combustion chamber measured by a measurement portion, and delaying the estimated timing when a speed of the combustion with flame propagation is high.SELECTED DRAWING: Figure 23

Description

  The technology disclosed herein relates to a control device for a compression ignition engine.

  Combustion by self-ignition, in which the air-fuel mixture burns at once without going through the flame propagation, has a minimum combustion period. When the air-fuel mixture burns by self-ignition at an appropriate time, the fuel efficiency of the engine can be maximized.

  For example, the engine control device described in Patent Literature 1 calculates the timing at which the mass combustion ratio becomes 50% based on the output signal of the in-cylinder pressure sensor in order to adjust the timing of combustion by self-ignition.

Japanese Patent No. 3873580

  Unlike the technique described in Patent Literature 1, the present applicant has proposed SPCCI (SPark Controlled Compression Ignition) combustion that combines SI (Spark Ignition) combustion and CI (Compression Ignition) combustion. SI combustion is combustion with flame propagation that is initiated by forcibly igniting an air-fuel mixture in a combustion chamber. CI combustion is combustion started by the compression ignition of the air-fuel mixture in the combustion chamber. In SPCCI combustion, when the air-fuel mixture in the combustion chamber is forcibly ignited to start combustion by flame propagation, the unburned air-fuel mixture in the combustion chamber is generated by the heat generation of SI combustion and the pressure increase by flame propagation. Is a form in which CI combustion occurs. SPCCI combustion is a form of "combustion by compression ignition" because it includes CI combustion.

  CI combustion occurs when the in-cylinder temperature reaches an ignition temperature determined by the composition of the air-fuel mixture. If the in-cylinder temperature reaches the ignition temperature near the compression top dead center and CI combustion occurs, the fuel efficiency of SPCCI combustion can be maximized.

  The in-cylinder temperature increases as the in-cylinder pressure increases. The in-cylinder pressure in SPCCI combustion is the result of two pressure increases: a pressure increase due to the compression work of the piston during the compression stroke and a pressure increase resulting from the heat generated by SI combustion.

  On the other hand, in the SPCCI combustion, if the CI combustion occurs near the compression top dead center, the in-cylinder pressure may increase excessively and the combustion noise may become excessive. In this case, if the ignition timing is retarded, CI combustion occurs at a time when the piston is considerably lowered in the expansion stroke, so that combustion noise can be suppressed. However, the fuel efficiency of the engine decreases.

  In order to achieve both suppression of combustion noise and improvement of fuel efficiency in an engine that performs SPCCI combustion, it is necessary to control SPCCI combustion so that the combustion waveform that changes with the progress of the crank angle becomes an appropriate combustion waveform. No.

  In order to control the SPCCI combustion, for example, it is conceivable to use the SI ratio as a parameter representing the characteristics of the SPCCI combustion. The SI rate can be defined as an index related to the ratio of the amount of heat generated by SI combustion to the total amount of heat generated by SPCCI combustion. When the SI ratio is high, the ratio of SI combustion in SPCCI combustion is high, and when the SI ratio is low, the ratio of CI combustion in SPCCI combustion is high. A high ratio of SI combustion in SPCCI combustion is advantageous for suppressing combustion noise. A high ratio of CI combustion in SPCCI combustion is advantageous for improving the fuel efficiency of the engine.

  Further, for example, it is conceivable to use the CI combustion start timing θci as a parameter representing the characteristics of SPCCI combustion. The CI combustion start timing θci is a timing at which the unburned air-fuel mixture self-ignites. When the SI ratio changes, θci also changes.

  Here, if the actual θci is advanced more than the target θci due to the variation of the flame propagation speed of the SI combustion of the SPCCI, the CI combustion occurs at a timing close to the compression top dead center, and the combustion noise is reduced. It gets bigger. In order to suppress the combustion noise, the control unit needs to know the actual θci. If the actual θci can be estimated, the control unit can bring the actual θci closer to the target θci by adjusting the ignition timing according to the difference between the actual θci and the target θci. For example, when the actual θci is more advanced than the target θci, the control unit retards the ignition timing, so that the actual θci is retarded, so that the combustion noise can be suppressed.

  The actual θci can be used not only to suppress combustion noise but also to calculate the SI ratio described above and to estimate the temperature in the combustion chamber.

  By accurately estimating the actual θci, the engine can be operated so as to improve the fuel consumption performance while suppressing the combustion noise of the SPCCI combustion. Therefore, in the control of an engine that performs SPCCI combustion, there is a need to accurately estimate the actual θci.

  The inventors of the present application first accurately estimate the CI combustion start timing θci based on a value having a frequency equal to or higher than the first frequency and equal to or lower than the second frequency in the measurement signal of the in-cylinder pressure sensor that measures the pressure in the combustion chamber. Has completed and proposed a new method. This estimation method is based on the case where the present inventors burn all the air-fuel mixture in the combustion chamber by flame propagation accompanying forced ignition and the case where all the air-fuel mixture in the combustion chamber is compressed and burned. Are found to differ greatly in a specific frequency band between the first frequency and the second frequency. Combustion by flame propagation has the characteristic that the heat release rate (dQ / dθ) at the start of combustion is lower than the heat release rate of compression ignition combustion. It is believed that the intensity is different in certain frequency bands.

  More specifically, the present inventors set the timing at which the measurement signal passing through a band-pass filter having a pass band equal to or higher than the first frequency and equal to or lower than the second frequency among the measurement signals of the in-cylinder pressure sensor has a minimum value. And CI combustion start timing θci. Therefore, in the above proposal, when the minimum value of the measurement signal passing through the band-pass filter exceeds a predetermined threshold, the control unit determines the timing of the minimum value as the timing at which the unburned mixture self-ignites (that is, , CI combustion start timing θci).

  However, depending on the operating state of the engine, the speed of SI combustion in SPCCI combustion may increase. In this case, the inventors of the present application newly found that the above-described estimation method lowers the estimation accuracy of the CI combustion start timing θci.

  The technology disclosed herein increases the accuracy of estimating the timing at which the unburned air-fuel mixture self-ignites in an engine that performs SPCCI combustion.

  Specifically, the technology disclosed herein relates to a control device for a compression ignition engine. The control device of the compression ignition type engine is attached to the cylinder head, a combustion chamber of the engine formed by a cylinder, a piston reciprocating in the cylinder, and a cylinder head closing one end of the cylinder, A fuel injection unit for injecting fuel to be supplied into the combustion chamber, an ignition unit disposed facing the combustion chamber and igniting an air-fuel mixture in the combustion chamber, and measuring parameters related to operation of the engine. The measurement unit, the ignition unit, the fuel injection unit, and the measurement unit are connected to each other, receive the measurement signal from the measurement unit and perform calculations, the ignition unit and the fuel injection unit And a control unit that outputs a signal to the air-fuel mixture, wherein the air-fuel mixture has started combustion accompanied by flame propagation by a part of the air-fuel mixture due to forced ignition of the ignition unit. , The rest of the unburned air-fuel mixture is combusted by self-ignition.

  The control unit is configured to measure the ignition control unit that outputs an ignition signal to the ignition unit and the measurement unit before the target timing so that the unburned mixture self-ignites at the target timing. An ignition timing estimation unit that estimates a timing at which the in-cylinder pressure parameter exceeds a threshold value as a timing at which the unburned air-fuel mixture self-ignites, based on an in-cylinder pressure parameter related to the pressure in the combustion chamber, The control unit further includes a correction unit that delays the estimated timing when the speed of the combustion involving the flame propagation is high.

  According to this configuration, the ignition control section outputs an ignition signal to the ignition section, so that the ignition section ignites the air-fuel mixture. Combustion with flame propagation starts. After the start of SI combustion, the remaining unburned air-fuel mixture burns by self-ignition. This compression ignition type engine performs SPCCI combustion. The ignition control unit adjusts the timing at which the ignition unit performs forced ignition so that the unburned air-fuel mixture self-ignites at the target timing.

  The measuring unit measures an in-cylinder pressure parameter related to the pressure in the combustion chamber during the combustion of the air-fuel mixture. The ignition timing estimation unit estimates the timing at which the in-cylinder pressure parameter exceeds the threshold value as the timing at which the unburned air-fuel mixture self-ignites (that is, the CI combustion start timing θci).

  As described above, when the speed of SI combustion increases in SPCCI combustion, the accuracy of estimating the CI combustion start timing θci may decrease. Specifically, when the speed of SI combustion increases, the timing at which the in-cylinder pressure parameter exceeds the threshold value is advanced from the actual CI combustion start timing θci.

  On the other hand, in the configuration described above, when the speed of the SI combustion is high, the correction unit performs the correction to retard the estimated timing, so that the estimated θci approaches the actual θci. The accuracy of estimating θci can be improved. By increasing the accuracy of estimating θci, the ignition control unit can adjust the start of CI combustion in SPCCI combustion to an appropriate timing. Therefore, the above configuration can improve the fuel efficiency of the engine while suppressing the combustion noise of the SPCCI combustion.

  The measurement unit may be provided facing the combustion chamber and include an in-cylinder pressure sensor that measures a pressure in the combustion chamber.

  The control unit can grasp the speed of the combustion accompanied by the flame propagation based on the measurement value of the in-cylinder pressure sensor, and can estimate the CI combustion start timing.

  The control unit has a first bandpass filter that passes a signal having a frequency equal to or higher than a first frequency and equal to or lower than a second frequency among the measurement signals of the in-cylinder pressure sensor, and the ignition timing estimating unit includes: The timing at which the measurement signal passing through the one band-pass filter exceeds the first threshold value is estimated as the timing at which the unburned air-fuel mixture self-ignites, and the first frequency and the second frequency are 0.5 kHz or more and 4 kHz or more. The frequency may be set to be equal to or lower than 0.0 kHz.

  According to the study of the present inventors, a measurement signal (in-cylinder pressure parameter) passing through a band-pass filter having a pass band equal to or higher than a first frequency and equal to or lower than a second frequency set to a range of 0.5 kHz or more and 4.0 kHz or less is set. ), It is possible to accurately estimate the CI combustion start timing θci.

  The control unit calculates a heat release rate (dQ / dθ) for a predetermined period from the forced ignition based on the parameters measured by the measurement unit, and the ignition timing estimating unit calculates when the heat release rate is high. Alternatively, the correction amount in the retard direction may be made larger than when it is low.

  When the speed of SI combustion in SPCCI combustion increases, the rate of heat generation during a predetermined period from forced ignition increases. When the heat release rate during a predetermined period from the forced ignition increases, the accuracy of estimating the CI combustion start timing θci decreases.

  The ignition timing estimator sets a larger correction amount in the retard direction when the heat generation rate during the predetermined period from the forced ignition is higher than when it is low. When the heat generation rate during the predetermined period from the forced ignition is high, the ignition timing estimation unit erroneously estimates θci to a more advanced side. By increasing the correction amount in the retard direction, the ignition timing estimating unit can increase the accuracy of estimating θci.

  The control unit calculates a crank angle period until the mass combustion ratio reaches a predetermined value, based on the parameter measured by the measurement unit, and the ignition timing estimating unit sets a long crank angle period when the crank angle period is short. The amount of correction in the retard direction may be made larger than at the time.

  As the speed of SI combustion in SPCCI combustion increases, the crank angle period until the mass combustion ratio reaches a predetermined value decreases. The ignition timing estimator sets a larger correction amount in the retard direction when the crank angle period until the mass combustion ratio reaches the predetermined value is shorter than when the crank angle period is longer. The ignition timing estimator can improve the accuracy of estimating the CI combustion start timing θci, as described above.

  The control unit has a first band-pass filter that passes a signal having a frequency equal to or higher than a first frequency and equal to or lower than a second frequency from a measurement signal of the in-cylinder pressure sensor. A first ignition timing estimator for estimating a timing at which the measurement signal passing through the pass filter exceeds a first threshold as a timing at which the unburned air-fuel mixture self-ignites, wherein the first frequency and the second frequency are: The control unit is also configured to pass a component having a frequency equal to or higher than a third frequency and equal to or lower than a fourth frequency from a measurement signal of the in-cylinder pressure sensor. The ignition timing estimating unit has a timing at which the unburned mixture self-ignites at a timing when the measurement signal passing through the second bandpass filter exceeds a second threshold value. A second ignition timing estimating unit for estimating that the first ignition is performed, wherein the third frequency and the fourth frequency are set in a range of 5.5 kHz or more and 8.0 kHz or less, and the control unit further includes the first ignition A timing selection unit that selects one of the timings estimated by the second ignition timing estimation unit; the selection unit includes the first ignition timing estimation unit and the second ignition timing estimation unit; When only one of the two estimates the timing, the timing is estimated to be the timing at which the unburned air-fuel mixture self-ignites, and the selection unit also determines the first ignition timing estimation unit and the second ignition timing estimation unit. When both of them estimate the timing, the timing on the relatively advanced side may be estimated as the timing at which the unburned air-fuel mixture self-ignites.

  The pass band of the second band pass filter has a higher frequency than the pass band of the first band pass filter. The pass band of the second bandpass filter corresponds to the frequency of a pressure wave (standing wave) generated in the combustion chamber when knocking occurs in SI combustion. The present inventors have found that the frequency of the pressure wave generated in the combustion chamber at the start of CI combustion of SPCCI combustion is close to the frequency of the pressure wave generated in the combustion chamber when knocking of SI combustion occurs. The timing at which the measurement signal (corresponding to the in-cylinder pressure parameter) passing through the second bandpass filter exceeds the second threshold value corresponds to the CI combustion start timing θci. The ignition timing estimator can estimate the CI combustion start timing θci based on the measurement signal that has passed through the second bandpass filter.

  The first ignition timing estimating unit estimates the crank angle timing at which the unburned air-fuel mixture self-ignites using the first bandpass filter, and the second ignition timing estimating unit estimates the unburned mixture using the second bandpass filter. Estimating the crank angle timing at which the gas self-ignites corresponds to estimating the CI combustion start timing θci by different methods. By estimating the CI combustion start timing θci by each of the two methods, the estimation accuracy of the CI combustion start timing θci increases.

  Specifically, when the timing can be estimated by only one of the methods, the selecting unit can estimate that the timing is the CI combustion start timing θci.

  In addition, when the timing can be estimated by each of the two types of methods, the selection unit estimates that the relatively advanced timing is the CI combustion start timing θci.

  As described above, the estimated CI combustion start timing θci can be used to suppress combustion noise. If the estimated CI combustion start timing θci is more advanced than the target CI combustion start timing θci, combustion noise may increase. Therefore, the control unit delays the ignition timing such that the combustion noise is reduced by delaying the timing at which the unburned air-fuel mixture self-ignites.

  When the selection unit estimates the relatively advanced timing as the CI combustion start timing θci among the two timings estimated by each of the two types of techniques, the estimated CI combustion start timing θci becomes the target CI combustion timing. With respect to the start timing θci, it becomes easier to shift to the advance side than to shift to the retard side. The control unit delays the ignition timing so that the timing of the self-ignition of the unburned air-fuel mixture is retarded, in other words, the ignition timing is retarded so as to reduce the combustion noise. Therefore, the control unit effectively suppresses the combustion noise of the SPCCI combustion. be able to.

  A control device for a compression ignition engine disclosed herein includes a combustion chamber of an engine formed by a cylinder, a piston that reciprocates in the cylinder, and a cylinder head that closes one end of the cylinder, and attached to the cylinder head. A fuel injection unit configured to inject fuel to be supplied into the combustion chamber, an ignition unit disposed facing the combustion chamber and igniting an air-fuel mixture in the combustion chamber, and disposed facing the combustion chamber. And a measuring unit that includes at least an in-cylinder pressure sensor that measures the pressure in the combustion chamber, and that measures a parameter related to the operation of the engine, and the ignition unit, the fuel injection unit, and the measurement unit, respectively. Is connected, performs a calculation by receiving a measurement signal from the measurement unit, and outputs a signal to the ignition unit and the fuel injection unit. And a control unit, wherein the air-fuel mixture, the air-fuel mixture portion by forced ignition of the igniter is after starting combustion with flame propagation, the remaining unburned air-fuel mixture is combusted by self-ignition.

  The control unit is configured to store a target timing at which the unburned air-fuel mixture self-ignites, and that the unburned air-fuel mixture is self-timed at the target timing based on the target timing of the target timing storage unit. An ignition control unit that outputs an ignition signal to the ignition unit before the target timing so as to ignite, the control unit also includes a specific frequency band among the measurement signals of the in-cylinder pressure sensor. A band-pass filter that passes the signal of the above, a threshold storage unit that stores a threshold value, and a timing at which the value of the measurement signal that has passed through the band-pass filter exceeds the threshold value. An ignition timing estimating unit for estimating the self-ignition timing, and a correction for correcting the timing estimated by the ignition timing estimating unit based on a measurement signal from the measuring unit. When have, the correction unit, when the speed of the combustion with the flame propagation is fast, retarding the timing of the estimated.

  As described above, the control device for a compression ignition engine can increase the accuracy of estimating the crank angle timing at which the unburned mixture self-ignites in an engine that performs SPCCI combustion.

FIG. 1 is a diagram illustrating the configuration of the engine. FIG. 2 is a view exemplifying a configuration of the combustion chamber. The upper figure is a plan view equivalent view of the combustion chamber, and the lower figure is a sectional view taken along the line II-II. FIG. 3 is a plan view illustrating the configuration of the combustion chamber and the intake system. FIG. 4 is a block diagram illustrating the configuration of the engine control device. FIG. 5 is a diagram illustrating a waveform of SPCCI combustion. FIG. 6 is a diagram illustrating a map when the engine is warm. FIG. 7A is a block diagram illustrating a part of the functional configuration of the control unit of the engine. FIG. 7B is a block diagram illustrating a part of the functional configuration of the control unit of the engine. FIG. 8 is a diagram illustrating a result of frequency analysis of a measurement value of the in-cylinder pressure sensor. FIG. 9 is a diagram illustrating a waveform of each parameter when CI combustion appropriately occurs after SI combustion. FIG. 10 is a diagram illustrating a waveform of each parameter when CI combustion does not occur after SI combustion. FIG. 11 is a flowchart illustrating a procedure for estimating the CI combustion start timing by the first ignition timing estimator. FIG. 12 is a diagram illustrating a relationship between the engine speed and the passband of the first bandpass filter. FIG. 13 is a flowchart illustrating a procedure for calculating the mass combustion ratio. FIG. 14 is a flowchart illustrating a procedure for calculating the offset amount. FIG. 15 is a block diagram illustrating a functional configuration related to estimation of the CI combustion start timing in the control unit. 16 is a diagram illustrating the relationship between dQ / dθ and the first threshold, and the lower diagram is a diagram illustrating the relationship between the period of 10-50% of the mass combustion ratio and the first threshold. is there. FIG. 17 is a diagram illustrating a relationship between the first threshold value and the crank angle of the mass combustion ratio of 10% to 50%, the mass combustion ratio of 10%. FIG. 18 is a block diagram illustrating a modified example related to the estimation of the CI combustion start timing by the first ignition timing estimation unit. FIG. 19 is a diagram illustrating a waveform of each parameter when erroneous estimation of the CI combustion start timing occurs. FIG. 20 is a diagram comparing simulation results obtained by frequency analysis of changes in in-cylinder pressure during SPCCI combustion with respect to the level of engine speed. FIG. 21 is a diagram illustrating a relationship between the engine speed and the second threshold value. FIG. 22 is a flowchart illustrating a procedure of estimating the CI combustion start timing by the second ignition timing estimator. FIG. 23 is a block diagram illustrating a functional configuration of the control unit for correcting the estimated CI combustion start timing. 24 is a diagram illustrating the relationship between dQ / dθ and the correction value of the CI combustion start timing, and the lower diagram is the relationship between the mass combustion ratio of 10-50% and the correction value of the CI combustion start timing. FIG. FIG. 25 is a diagram exemplifying a relationship between a 10-50% mass combustion period, a crank angle of a 10% mass combustion ratio, and a correction value of the CI combustion start timing.

  Hereinafter, embodiments of a control device for a compression ignition engine will be described in detail with reference to the drawings. The following description is an example of an engine and an engine control device.

  FIG. 1 is a diagram illustrating a configuration of a compression ignition type engine. FIG. 2 is a diagram illustrating a configuration of a combustion chamber of the engine. FIG. 3 is a diagram illustrating the configuration of the combustion chamber and the intake system. In FIG. 1, the intake side is on the left side of the paper, and the exhaust side is on the right side of the paper. 2 and 3, the intake side is on the right side of the paper, and the exhaust side is on the left side of the paper. FIG. 4 is a block diagram illustrating the configuration of the engine control device.

  The engine 1 is a four-stroke engine in which the combustion chamber 17 operates by repeating an intake stroke, a compression stroke, an expansion stroke, and an exhaust stroke. The engine 1 is mounted on a four-wheeled vehicle. When the engine 1 operates, the automobile runs. The fuel of the engine 1 is gasoline in this configuration example. The fuel may be a liquid fuel containing at least gasoline. The fuel may be gasoline containing, for example, bioethanol.

(Engine configuration)
The engine 1 includes a cylinder block 12 and a cylinder head 13 mounted thereon. A plurality of cylinders 11 are formed inside the cylinder block 12. 1 and 2, only one cylinder 11 is shown. The engine 1 is a multi-cylinder engine.

  The piston 3 is slidably inserted in each cylinder 11. The piston 3 is connected to a crankshaft 15 via a connecting rod 14. The piston 3 defines a combustion chamber 17 together with the cylinder 11 and the cylinder head 13. The “combustion chamber” may be used in a broad sense. That is, the “combustion chamber” may mean a space formed by the piston 3, the cylinder 11, and the cylinder head 13 regardless of the position of the piston 3.

  The lower surface of the cylinder head 13, that is, the ceiling surface of the combustion chamber 17 is constituted by an inclined surface 1311 and an inclined surface 1312 as shown in the lower diagram of FIG. The inclined surface 1311 is inclined upward from the intake side toward the injection axis X2 of the injector 6 described later. The slope 1312 has an upward slope from the exhaust side toward the injection axis X2. The ceiling surface of the combustion chamber 17 has a so-called pent roof shape.

  The upper surface of the piston 3 protrudes toward the ceiling surface of the combustion chamber 17. A cavity 31 is formed on the upper surface of the piston 3. The cavity 31 is recessed from the upper surface of the piston 3. The cavity 31 has a shallow dish shape in this configuration example. The center of the cavity 31 is displaced from the central axis X1 of the cylinder 11 to the exhaust side.

  The geometric compression ratio of the engine 1 is set to 10 or more and 30 or less. As described later, the engine 1 performs SPCCI combustion that combines SI combustion and CI combustion in a part of the operating range. SPCCI combustion controls CI combustion by utilizing heat generation and pressure rise due to SI combustion. The engine 1 is a compression ignition type engine. However, in the engine 1, it is not necessary to increase the temperature of the combustion chamber 17 when the piston 3 reaches the compression top dead center (that is, the compression end temperature). The engine 1 can set the geometric compression ratio relatively low. Lowering the geometric compression ratio is advantageous for reducing cooling loss and mechanical loss. The geometric compression ratio of the engine 1 is 14 to 17 in a regular specification (a low octane fuel having an octane number of about 91), and 15 in a high octane specification (a high octane fuel having an octane number of about 96). To 18 may be used.

  An intake port 18 is formed in the cylinder head 13 for each cylinder 11. The intake port 18 has a first intake port 181 and a second intake port 182, as shown in FIG. The intake port 18 communicates with the combustion chamber 17. Although not shown in detail, the intake port 18 is a so-called tumble port. That is, the intake port 18 has a shape such that a tumble flow is formed in the combustion chamber 17.

  The intake port 18 is provided with an intake valve 21. The intake valve 21 opens and closes between the combustion chamber 17 and the intake port 18. The intake valve 21 opens and closes at a predetermined timing by a valve operating mechanism. The valve operating mechanism may be a variable valve operating mechanism that changes valve timing and / or valve lift. In this configuration example, as shown in FIG. 4, the variable valve mechanism has an intake electric S-VT (Sequential-Valve Timing) 23. The intake electric S-VT 23 continuously changes the rotation phase of the intake camshaft within a predetermined angle range. The valve opening timing and the valve closing timing of the intake valve 21 change continuously. The intake valve mechanism may have a hydraulic S-VT instead of the electric S-VT.

  The cylinder head 13 is also provided with an exhaust port 19 for each cylinder 11. The exhaust port 19 also has a first exhaust port 191 and a second exhaust port 192, as shown in FIG. The exhaust port 19 communicates with the combustion chamber 17.

  The exhaust port 19 is provided with an exhaust valve 22. The exhaust valve 22 opens and closes between the combustion chamber 17 and the exhaust port 19. The exhaust valve 22 is opened and closed at a predetermined timing by a valve operating mechanism. This valve operating mechanism may be a variable valve operating mechanism that changes valve timing and / or valve lift. In this configuration example, as shown in FIG. 4, the variable valve mechanism has an exhaust electric S-VT24. The exhaust electric S-VT 24 continuously changes the rotation phase of the exhaust camshaft within a predetermined angle range. The valve opening timing and the valve closing timing of the exhaust valve 22 change continuously. The exhaust valve mechanism may have a hydraulic S-VT instead of the electric S-VT.

  The intake electric S-VT 23 and the exhaust electric S-VT 24 adjust the length of the overlap period in which both the intake valve 21 and the exhaust valve 22 are opened. If the length of the overlap period is increased, the residual gas in the combustion chamber 17 can be scavenged. Further, by adjusting the length of the overlap period, the internal EGR (Exhaust Gas Recirculation) gas can be introduced into the combustion chamber 17. The intake electric S-VT 23 and the exhaust electric S-VT 24 constitute an internal EGR system. Note that the internal EGR system is not always constituted by the S-VT.

  The injector 6 is attached to the cylinder head 13 for each cylinder 11. The injector 6 injects fuel directly into the combustion chamber 17. The injector 6 is an example of a fuel injection unit. The injector 6 is provided at a valley of the pent roof where the inclined surfaces 1311 and 1312 intersect. As shown in FIG. 2, the injection axis X2 of the injector 6 is located closer to the exhaust side than the center axis X1 of the cylinder 11. The injection axis X2 of the injector 6 is parallel to the central axis X1. The injection axis X2 of the injector 6 coincides with the center of the cavity 31. The injector 6 faces the cavity 31. Note that the injection axis X2 of the injector 6 may coincide with the center axis X1 of the cylinder 11. In the case of the configuration, the injection axis X2 of the injector 6 may be coincident with the center of the cavity 31.

  Although not shown in detail, the injector 6 is configured by a multi-injection type fuel injection valve having a plurality of injection ports. The injector 6 injects the fuel such that the fuel spray spreads radially from the center of the combustion chamber 17 as shown by a two-dot chain line in FIG. In the present configuration example, the injector 6 has ten injection holes, and the injection holes are arranged at equal angles in the circumferential direction.

  A fuel supply system 61 is connected to the injector 6. The fuel supply system 61 includes a fuel tank 63 configured to store fuel, and a fuel supply path 62 that connects the fuel tank 63 and the injector 6 to each other. A fuel pump 65 and a common rail 64 are interposed in the fuel supply path 62. The fuel pump 65 pumps fuel to the common rail 64. The fuel pump 65 is a plunger pump driven by the crankshaft 15 in this configuration example. The common rail 64 stores the fuel pumped from the fuel pump 65 at a high fuel pressure. When the injector 6 opens, the fuel stored in the common rail 64 is injected into the combustion chamber 17 from the injection port of the injector 6. The fuel supply system 61 can supply the fuel at a high pressure of 30 MPa or more to the injector 6. The pressure of the fuel supplied to the injector 6 may be changed according to the operating state of the engine 1. Note that the configuration of the fuel supply system 61 is not limited to the above configuration.

  An ignition plug 25 is attached to the cylinder head 13 for each cylinder 11. The ignition plug 25 forcibly ignites the mixture in the combustion chamber 17. In this configuration example, the spark plug 25 is disposed closer to the intake side than the center axis X1 of the cylinder 11. The spark plug 25 is located between the two intake ports 18. The ignition plug 25 is attached to the cylinder head 13 so as to be inclined from the upper side to the lower side in a direction approaching the center of the combustion chamber 17. As shown in FIG. 2, the electrode of the spark plug 25 faces the inside of the combustion chamber 17 and is located near the ceiling surface of the combustion chamber 17. Note that the ignition plug 25 may be disposed on the exhaust side of the center axis X1 of the cylinder 11. Further, the spark plug 25 may be arranged on the central axis X1 of the cylinder 11.

  An intake passage 40 is connected to one side of the engine 1. The intake passage 40 communicates with the intake port 18 of each cylinder 11. The gas introduced into the combustion chamber 17 flows through the intake passage 40. An air cleaner 41 is provided at an upstream end of the intake passage 40. The air cleaner 41 filters fresh air. A surge tank 42 is provided near the downstream end of the intake passage 40. The intake passage 40 downstream of the surge tank 42 forms an independent passage that branches off for each cylinder 11. The downstream end of the independent passage is connected to the intake port 18 of each cylinder 11.

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

  In the intake passage 40, a supercharger 44 is disposed downstream of the throttle valve 43. The supercharger 44 supercharges the gas introduced into the combustion chamber 17. In this configuration example, the supercharger 44 is a mechanical supercharger driven by the engine 1. The mechanical supercharger 44 may be of a Roots type, a Riesholm type, a vane type, or a centrifugal type.

  An electromagnetic clutch 45 is interposed between the supercharger 44 and the engine 1. The electromagnetic clutch 45 transmits a driving force from the engine 1 to the supercharger 44 between the supercharger 44 and the engine 1 or interrupts the transmission of the driving force. As will be described later, the supercharger 44 is switched between on and off by the ECU 10 switching between disconnection and connection of the electromagnetic clutch 45.

  An intercooler 46 is provided downstream of the supercharger 44 in the intake passage 40. The intercooler 46 cools the gas compressed in the supercharger 44. The intercooler 46 may be configured, for example, as a water-cooled or oil-cooled type.

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

  The ECU 10 fully opens the air bypass valve 48 when the supercharger 44 is turned off (that is, when the electromagnetic clutch 45 is shut off). The gas flowing through the intake passage 40 bypasses the supercharger 44 and is introduced into the combustion chamber 17 of the engine 1. The engine 1 operates in a non-supercharged state, that is, in a state of natural intake.

  When the turbocharger 44 is turned on, the engine 1 operates in a supercharged state. The ECU 10 adjusts the opening of the air bypass valve 48 when the supercharger 44 is turned on (that is, when the electromagnetic clutch 45 is connected). Part of the gas that has passed through the supercharger 44 flows back to the upstream of the supercharger 44 through the bypass passage 47. When the ECU 10 adjusts the opening of the air bypass valve 48, the supercharging pressure of the gas introduced into the combustion chamber 17 changes. The supercharging time is defined as a time when the pressure in the surge tank 42 exceeds the atmospheric pressure, and the non-supercharging time is defined as a time when the pressure in the surge tank 42 becomes equal to or less than the atmospheric pressure. Is also good.

  In this configuration example, a supercharging system 49 is configured by the supercharger 44, the bypass passage 47, and the air bypass valve 48.

  The engine 1 has a swirl generator that generates a swirl flow in the combustion chamber 17. The swirl generator has a swirl control valve 56 attached to the intake passage 40, as shown in FIG. The swirl control valve 56 is provided in the secondary passage 402 of the primary passage 401 connected to the first intake port 181 and the secondary passage 402 connected to the second intake port 182. The swirl control valve 56 is an opening control valve that can narrow the cross section of the secondary passage 402. When the opening degree of the swirl control valve 56 is small, the intake flow rate flowing into the combustion chamber 17 from the first intake port 181 is relatively large, and the intake flow rate flowing into the combustion chamber 17 from the second intake port 182 is relatively large. Since it is small, the swirl flow in the combustion chamber 17 becomes strong. When the opening degree of the swirl control valve 56 is large, the flow rate of the intake air flowing into the combustion chamber 17 from each of the first intake port 181 and the second intake port 182 becomes substantially equal, so that the swirl flow in the combustion chamber 17 is weak. Become. When the swirl control valve 56 is fully opened, no swirl flow occurs. The swirl flow circulates in the counterclockwise direction in FIG. 3 as indicated by the white arrow (see also the white arrow in FIG. 2).

  An exhaust passage 50 is connected to the other side of the engine 1. The exhaust passage 50 communicates with the exhaust port 19 of each cylinder 11. The exhaust passage 50 is a passage through which the exhaust gas discharged from the combustion chamber 17 flows. Although not shown in detail, the upstream portion of the exhaust passage 50 constitutes an independent passage branched for each cylinder 11. The upstream end of the independent passage is connected to the exhaust port 19 of each cylinder 11.

  An exhaust gas purification system having a plurality of catalytic converters is provided in the exhaust passage 50. Although not shown, the upstream catalytic converter is disposed in the engine room. The upstream catalytic converter has a three-way catalyst 511 and a GPF (Gasoline Particulate Filter) 512. The downstream catalytic converter is located outside the engine room. The downstream catalytic converter has a three-way catalyst 513. Note that the exhaust gas purification system is not limited to the configuration shown in the figure. For example, the GPF may be omitted. Further, the catalytic converter is not limited to one having a three-way catalyst. Furthermore, the arrangement order of the three-way catalyst and the GPF may be appropriately changed.

  An EGR passage 52 constituting an external EGR system is connected between the intake passage 40 and the exhaust passage 50. The EGR passage 52 is a passage for returning a part of the exhaust gas to the intake passage 40. The upstream end of the EGR passage 52 is connected between the upstream catalytic converter and the downstream catalytic converter in the exhaust passage 50. The downstream end of the EGR passage 52 is connected to an upstream portion of the supercharger 44 in the intake passage 40. The EGR gas flowing through the EGR passage 52 enters the intake passage 40 upstream of the supercharger 44 without passing through the air bypass valve 48 of the bypass passage 47.

  A water-cooled EGR cooler 53 is provided in the EGR passage 52. The EGR cooler 53 cools the exhaust gas. An EGR valve 54 is also provided in the EGR passage 52. The EGR valve 54 adjusts the flow rate of the exhaust gas flowing through the EGR passage 52. By adjusting the opening of the EGR valve 54, it is possible to adjust the recirculation amount of the cooled exhaust gas, that is, the external EGR gas.

  In this configuration example, the EGR system 55 includes an external EGR system and an internal EGR system. The external EGR system can supply exhaust gas having a lower temperature than the internal EGR system to the combustion chamber 17.

  The control device of the compression ignition type engine includes an ECU (Engine Control Unit) 10 for operating the engine 1. The ECU 10 is a control unit based on a known microcomputer. As shown in FIG. 4, the ECU 10 includes a microcomputer 101 including a central processing unit (CPU) for executing a program, and a RAM (Random Access Memory), for example. The memory 102 includes a memory 102 configured by an access memory (ROM) or a read only memory (ROM) and stores programs and data, and an I / F circuit 103 that inputs and outputs electric signals.

  Various sensors SW1 to SW17 are connected to the ECU 10, as shown in FIGS. The sensors SW1 to SW17 output signals to the ECU 10. The sensors include the following sensors.

Airflow sensor SW1: disposed downstream of air cleaner 41 in intake passage 40 and measures the flow rate of fresh air flowing through intake passage 40. First intake air temperature sensor SW2: disposed downstream of air cleaner 41 in intake passage 40; , Which measures the temperature of fresh air flowing through the intake passage 40. The first pressure sensor SW <b> 3 is disposed downstream of the connection position of the EGR passage 52 in the intake passage 40 and upstream of the supercharger 44, and The second intake air temperature sensor SW4 is disposed downstream of the supercharger 44 in the intake passage 40 and upstream of the connection position of the bypass passage 47, and flows out of the supercharger 44. Measuring gas temperature Intake pressure sensor SW5: attached to surge tank 42 and measures gas pressure downstream of supercharger 44 In-cylinder pressure sensor SW6: attached to cylinder head 13 corresponding to each cylinder 11 and measures the pressure in each combustion chamber 17 Exhaust temperature sensor SW7: disposed in exhaust passage 50 and exhausted from combustion chamber 17 linear O ₂ sensor SW8 for measuring the temperature of the gas: is located upstream of the catalytic converter upstream of the exhaust passage 50 and the lambda O ₂ sensor SW9 for measuring the oxygen concentration in the exhaust gas: ternary upstream of the catalytic converter A water temperature sensor SW10 that is disposed downstream of the catalyst 511 and measures the oxygen concentration in the exhaust gas. The water temperature sensor SW10 is mounted on the engine 1 and measures the temperature of the cooling water. A crank angle sensor SW11 is mounted on the engine 1 and has a crankshaft. Accelerator opening sensor SW12 for measuring rotation angle of 15: accelerator pedal machine The intake cam angle sensor SW13: attached to the engine 1 and measures the rotation angle of the intake camshaft. The exhaust cam angle sensor SW14: attached to the engine 1 and measures the accelerator opening corresponding to the operation amount of the accelerator pedal. Attached and measures the rotation angle of the exhaust camshaft EGR differential pressure sensor SW15: disposed in the EGR passage 52 and measures differential pressure upstream and downstream of the EGR valve 54 Fuel pressure sensor SW16: common rail of the fuel supply system 61 The third intake air temperature sensor SW17 is attached to the surge tank 42 and measures the pressure of the fuel supplied to the injector 6. The third intake air temperature sensor SW17 is attached to the surge tank 42 and is introduced into the combustion chamber 17 in other words, the temperature of the gas in the surge tank 42. Measure the temperature of the intake air.

  The ECU 10 determines the operating state of the engine 1 based on the signals of the sensors SW1 to SW17, and calculates the control amount of each device according to a predetermined control logic. The control logic is stored in the memory 102. The control logic includes calculating a target amount and / or a control amount using the map stored in the memory 102.

  The ECU 10 sends an electric signal related to the calculated control amount to the injector 6, the ignition plug 25, the intake electric S-VT23, the exhaust electric S-VT24, the fuel supply system 61, the throttle valve 43, the EGR valve 54, and the supercharger 44. To the electromagnetic clutch 45, the air bypass valve 48, and the swirl control valve 56.

  For example, the ECU 10 sets the target torque of the engine 1 and determines the target supercharging pressure based on the signal of the accelerator opening sensor SW12 and the map. Then, the ECU 10 adjusts the opening degree of the air bypass valve 48 based on the target supercharging pressure and the differential pressure across the supercharger 44 obtained from the signals of the first pressure sensor SW3 and the intake pressure sensor SW5. By performing the control, the supercharging pressure is set to the target supercharging pressure.

  Further, the ECU 10 sets a target EGR rate (that is, a ratio of EGR gas to all gases in the combustion chamber 17) based on the operating state of the engine 1 and the map. The ECU 10 determines the target EGR gas amount based on the target EGR rate and the intake air amount based on the signal from the accelerator opening sensor SW12, and determines the differential pressure across the EGR valve 54 obtained from the signal from the EGR differential pressure sensor SW15. By performing feedback control for adjusting the opening of the EGR valve 54 on the basis of the above, the amount of external EGR gas introduced into the combustion chamber 17 is made equal to the target EGR gas amount.

Further, the ECU 10 executes the air-fuel ratio feedback control when a predetermined control condition is satisfied. Specifically ECU10 includes a linear O ₂ sensor SW8, and, based on the oxygen concentration in the exhaust gas lambda O ₂ sensor SW9 is measured, so that the air-fuel ratio of the mixture has a desired value, the fuel injection of the injector 6 Adjust the volume.

  The details of the control of the engine 1 by the ECU 10 will be described later.

(SPCCI combustion concept)
The engine 1 performs combustion by compression self-ignition in a predetermined operating state with a primary purpose of improving fuel efficiency and exhaust gas performance. In the combustion by self-ignition, when the temperature in the combustion chamber 17 before the start of compression varies, the timing of self-ignition greatly changes. Therefore, the engine 1 performs SPCCI combustion that combines SI combustion and CI combustion.

  In the SPCCI combustion, the spark plug 25 forcibly ignites the air-fuel mixture in the combustion chamber 17 so that the air-fuel mixture performs SI combustion by flame propagation and generates heat in the combustion chamber 17 by the heat generated by the SI combustion. The unburned air-fuel mixture performs CI combustion by self-ignition by increasing the temperature and increasing the pressure in the combustion chamber 17 by flame propagation.

  By adjusting the calorific value of the SI combustion, it is possible to absorb the temperature variation in the combustion chamber 17 before the start of compression. By adjusting the ignition timing by the ECU 10, the air-fuel mixture can self-ignite at the target timing.

  In SPCCI combustion, heat generation during SI combustion is milder than heat generation during CI combustion. As shown in FIG. 5, the waveform of the heat release rate (dQ / dθ) in the SPCCI combustion has a rising slope smaller than the rising slope in the CI combustion waveform. Also, the pressure fluctuation (dp / dθ) in the combustion chamber 17 becomes gentler during SI combustion than during CI combustion.

  When the unburned air-fuel mixture self-ignites after the start of SI combustion, the slope of the waveform of the heat generation rate may change from small to large at the timing of self-ignition. The waveform of the heat release rate may have an inflection point X at the timing when the CI combustion starts.

  After the start of CI combustion, SI combustion and CI combustion are performed in parallel. Since the CI combustion generates more heat than the SI combustion, the heat generation rate is relatively large. However, since the CI combustion is performed after the compression top dead center, the slope of the waveform of the heat release rate is prevented from becoming too large. The pressure fluctuation (dp / dθ) during CI combustion also becomes relatively moderate.

  The pressure fluctuation (dp / dθ) can be used as an index representing combustion noise. As described above, in the SPCCI combustion, since the pressure fluctuation (dp / dθ) can be reduced, it is possible to prevent the combustion noise from becoming too large. The combustion noise of the engine 1 is suppressed to an allowable level or less.

  When the CI combustion ends, the SPCCI combustion ends. The combustion period of CI combustion is shorter than that of SI combustion. SPCCI combustion has an earlier combustion end timing than SI combustion.

The waveform of the heat generation rate in SPCCI combustion, a first heat generation rate portion Q _SI formed by SI combustion, and the second heat generating section Q _CI formed by CI combustion, but so as to be continuous in this order Is formed.

  Here, the SI ratio is defined as a parameter indicating the characteristics of SPCCI combustion. The present application defines the SI rate as an index relating to the ratio of the amount of heat generated by SI combustion to the total amount of heat generated by SPCCI combustion. The SI ratio is a ratio of the amount of heat generated by two combustions having different combustion modes. When the SI ratio is high, the ratio of SI combustion is high, and when the SI ratio is low, the ratio of CI combustion is high. A high ratio of SI combustion in SPCCI combustion is advantageous for suppressing combustion noise. A high ratio of CI combustion in SPCCI combustion is advantageous for improving the fuel efficiency of the engine.

The SI rate may be defined as the ratio of the amount of heat generated by SI combustion to the amount of heat generated by CI combustion. That is, in SPCCI combustion, the crank angle at which the CI combustion starts as CI combustion start time Shitaci, in the waveform 801 shown in FIG. 5, the area _{Q SI} the SI combustion also advance side than Shitaci, slow including Shitaci angle from the area _{Q CI} of CI combustion is a side, it may be SI index ₌ Q SI _{/ Q CI.}

  The engine 1 may generate a strong swirl flow in the combustion chamber 17 when performing SPCCI combustion. A strong swirl flow may be defined as a flow having a swirl ratio of 4 or more, for example. The swirl ratio can be defined as a value obtained by dividing the value obtained by measuring and integrating the intake flow lateral angular velocity for each valve lift by the engine angular velocity. Although not shown, the intake flow lateral angular velocity can be determined based on measurement using a known rig tester.

  When a strong swirl flow is generated in the combustion chamber 17, the outer peripheral portion of the combustion chamber 17 becomes a strong swirl flow, while the swirl flow in the central portion becomes relatively weak. The turbulence caused by the velocity gradient at the boundary between the central portion and the outer peripheral portion increases the turbulent energy in the central portion. When the ignition plug 25 ignites the air-fuel mixture in the center, the combustion speed of SI combustion increases due to high turbulent energy.

  The flame of the SI combustion rides on a strong swirl flow in the combustion chamber 17 and propagates in the circumferential direction. In the CI combustion, the CI combustion is performed from the outer peripheral portion to the central portion in the combustion chamber 17.

  When a strong swirl flow is generated in the combustion chamber 17, the SI combustion can be sufficiently performed before the start of the CI combustion. The generation of combustion noise can be suppressed, and the variation in torque between cycles can be suppressed.

(Engine operating area)
FIG. 6 illustrates a map 501 relating to the control of the engine 1. The map 501 is stored in the memory 102 of the ECU 10. The map 501 is a map when the engine 1 is warm. The map 501 is defined by the load and the rotation speed of the engine 1. The map 501 is roughly divided into three regions according to the level of the load and the level of the rotation speed. More specifically, the three regions include a low-load region A1 including idle operation and extending to low-rotation and medium-rotation regions, and a middle-high load region A2 having a higher load than the low-load region A1, as indicated by the solid line. , A3, A4, and a low-load region A1, and a high-rotation region A5 having a higher rotation speed than the middle-high load regions A2, A3, and A4.

  Here, the low rotation region, the middle rotation region, and the high rotation region are obtained by dividing the entire operation region of the engine 1 into approximately three equal parts of the low rotation region, the middle rotation region, and the high rotation region in the rotation speed direction. The low rotation region, the middle rotation region, and the high rotation region may be used. In the example of FIG. 6, the rotation speed is low when the rotation speed is less than N1, high rotation is when the rotation speed is N2 or more, and is medium rotation when the rotation speed is N1 or more and less than N2. The rotation speed N1 may be, for example, about 1200 rpm, and the rotation speed N2 may be, for example, about 4000 rpm.

  The low load region may be a region including a light load operation state, the high load region may be a region including a fully open load operation state, and the medium load region may be a region between a low load region and a high load region. In addition, the low load region, the medium load region, and the high load region are respectively obtained by dividing the entire operation region of the engine 1 in the load direction into approximately three equal parts of the low load region, the medium load region, and the high load region. It may be a low load area, a medium load area, and a high load area.

  The map 501 shows the state and combustion state of the air-fuel mixture in each region, the opening degree of the swirl control valve 56 in each region, and the driving region and the non-driving region of the supercharger 44. The engine 1 performs SPCCI combustion in the low load area A1, the medium load area A2, the high load medium rotation area A3, and the high load low rotation area A4. The engine 1 also performs SI combustion in the other high-speed region A5. Hereinafter, the operation of the engine 1 in each region will be described in detail.

(Operation of engine in low load range)
When the engine 1 is operating in the low load region A1, the engine 1 performs SPCCI combustion.

  The EGR system 55 introduces EGR gas into the combustion chamber 17 to improve the fuel efficiency of the engine 1. Specifically, the intake electric S-VT 23 and the exhaust electric S-VT 24 provide a positive overlap period in which both the intake valve 21 and the exhaust valve 22 are opened near the exhaust top dead center. Part of the exhaust gas discharged from the combustion chamber 17 to the intake port 18 and the exhaust port 19 is re-introduced into the combustion chamber 17. Since hot exhaust gas is introduced into the combustion chamber 17, the temperature inside the combustion chamber 17 increases. This is advantageous for stabilizing SPCCI combustion. The intake electric S-VT 23 and the exhaust electric S-VT 24 may have a negative overlap period in which both the intake valve 21 and the exhaust valve 22 are closed.

  The swirl generator forms a strong swirl flow in the combustion chamber 17. The swirl ratio is, for example, 4 or more. The swirl control valve 56 has a predetermined opening degree on the fully closed or closed side. As described above, since the intake port 18 is a tumble port, an oblique swirl flow having a tumble component and a swirl component is formed in the combustion chamber 17.

  The injector 6 injects fuel into the combustion chamber 17 several times during the intake stroke. The mixture is stratified by the multiple fuel injections and the swirl flow in the combustion chamber 17.

  The fuel concentration of the air-fuel mixture at the center of the combustion chamber 17 is higher than the fuel concentration at the outer periphery. Specifically, the A / F of the air-fuel mixture at the center is 20 or more and 30 or less, and the A / F of the air-fuel mixture at the outer periphery is 35 or more. The value of the air-fuel ratio is the value of the air-fuel ratio at the time of ignition, and is the same in the following description. By setting the A / F of the air-fuel mixture close to the spark plug 25 to 20 or more and 30 or less, generation of RawNOx during SI combustion can be suppressed. Further, by setting the A / F of the air-fuel mixture at the outer peripheral portion to 35 or more, CI combustion is stabilized.

  The air-fuel ratio (A / F) of the air-fuel mixture is leaner than the stoichiometric air-fuel ratio in the entire combustion chamber 17 (that is, the excess air ratio λ> 1). More specifically, the A / F of the air-fuel mixture in the entire combustion chamber 17 is 30 or more. By doing so, the generation of RawNOx can be suppressed, and the exhaust gas performance can be improved.

  After the end of the fuel injection, at a predetermined timing before the compression top dead center, the spark plug 25 ignites the air-fuel mixture at the center of the combustion chamber 17. The ignition timing may be the end of the compression stroke. The end of the compression stroke may be the end of the compression stroke when the compression stroke is divided into three parts: an initial stage, a middle stage, and an end stage.

  As described above, the fuel mixture in the central portion has a relatively high fuel concentration, so that the ignitability is improved and the SI combustion due to the flame propagation is stabilized. By stabilizing SI combustion, CI combustion starts at an appropriate timing. In SPCCI combustion, controllability of CI combustion is improved. Generation of combustion noise is suppressed. Further, by performing the SPCCI combustion with the A / F of the air-fuel mixture leaner than the stoichiometric air-fuel ratio, the fuel efficiency of the engine 1 can be significantly improved.

(Operation of engine in middle and high load range)
Even when the engine 1 is operating in the middle and high load regions A2, A3, and A4, the engine 1 performs SPCCI combustion similarly to the low load region A1.

  The EGR system 55 introduces EGR gas into the combustion chamber 17. Specifically, the intake electric S-VT 23 and the exhaust electric S-VT 24 provide a positive overlap period in which both the intake valve 21 and the exhaust valve 22 are opened near the exhaust top dead center. Internal EGR gas is introduced into the combustion chamber 17. Further, the EGR system 55 introduces the exhaust gas cooled by the EGR cooler 53 into the combustion chamber 17 through the EGR passage 52. External EGR gas having a lower temperature than the internal EGR gas is introduced into the combustion chamber 17. The external EGR gas adjusts the temperature in the combustion chamber 17 to an appropriate temperature. The EGR system 55 reduces the amount of EGR gas as the load on the engine 1 increases. The EGR system 55 may set the EGR gas including the internal EGR gas and the external EGR gas to zero at the full load.

  In the middle load region A2 and the high load middle rotation region A3, the swirl control valve 56 has a predetermined degree of opening on the fully closed or closed side. A strong swirl flow having a swirl ratio of 4 or more is formed in the combustion chamber 17. On the other hand, in the high-load low-rotation region A4, the swirl control valve 56 is open.

  The air-fuel ratio (A / F) of the air-fuel mixture is the stoichiometric air-fuel ratio (A / F ≒ 14.7) in the entire combustion chamber 17. Since the three-way catalysts 511 and 513 purify the exhaust gas discharged from the combustion chamber 17, the exhaust gas performance of the engine 1 is improved. The A / F of the air-fuel mixture may be within the purification window of the three-way catalyst. The excess air ratio λ of the air-fuel mixture may be 1.0 ± 0.2. When the engine 1 is operating in the high-load medium rotation region A3 including the full-open load (that is, the maximum load), the A / F of the air-fuel mixture is the stoichiometric air-fuel ratio or the stoichiometric air-fuel ratio in the entire combustion chamber 17. (That is, the excess air ratio λ of the air-fuel mixture is λ ≦ 1).

  Since the EGR gas is introduced into the combustion chamber 17, the G / F, which is the weight ratio between the total gas and the fuel in the combustion chamber 17, is leaner than the stoichiometric air-fuel ratio. The G / F of the mixture may be 18 or more. By doing so, the occurrence of so-called knocking can be avoided. G / F may be set at 18 or more and 30 or less. Further, the G / F may be set in the range of 18 to 50.

  The injector 6 performs one or more fuel injections during the intake stroke.

  After the fuel is injected, the spark plug 25 ignites the mixture at a predetermined timing near the compression top dead center. The spark plug 25 may ignite before the compression top dead center. The ignition plug 25 may ignite after the compression top dead center.

  By performing the SPCCI combustion of the air-fuel mixture at the stoichiometric air-fuel ratio, the exhaust gas discharged from the combustion chamber 17 can be purified using the three-way catalysts 511 and 513. Further, by introducing the EGR gas into the combustion chamber 17 to dilute the air-fuel mixture, the fuel efficiency of the engine 1 is improved.

(Operation of the turbocharger)
Here, as shown in the map 501, in a part of the low load area A1 and a part of the middle load area A2, the supercharger 44 is off (see S / C OFF). Specifically, the supercharger 44 is off in the low-rotation-side region of the low-load region A1. In the high-rotation-side region of the low-load region A1, the supercharger 44 is turned on in order to secure a necessary intake air filling amount in response to an increase in the rotation speed of the engine 1. Further, the supercharger 44 is off in a part of the middle load region A2 on the low-load low-rotation side. In the region on the high load side in the medium load region A2, the supercharger 44 is turned on in order to secure a necessary intake charge amount in response to an increase in the fuel injection amount. Also, the supercharger 44 is on in the high-rotation side region in the middle load region A2.

  In each of the high load medium rotation area A3, the high load low rotation area A4, and the high rotation area A5, the supercharger 44 is on in all the areas (see S / CON).

(Operation of engine in high speed range)
If the rotation speed of the engine 1 is high, the time required for the crank angle to change by 1 ° becomes short. It becomes difficult to stratify the mixture in the combustion chamber 17. When the number of revolutions of the engine 1 increases, it becomes difficult to perform SPCCI combustion.

  Therefore, when the engine 1 is operating in the high rotation region A5, the engine 1 performs SI combustion instead of SPCCI combustion. The high rotation region A5 extends from the low load to the high load in the entire load direction.

  The EGR system 55 introduces EGR gas into the combustion chamber 17. The EGR system 55 reduces the amount of EGR gas as the load increases. The EGR system 55 may set the EGR gas to zero at the full load.

  The swirl control valve 56 is fully open. No swirl flow is generated in the combustion chamber 17 and only a tumble flow is generated. By fully opening the swirl control valve 56, the filling efficiency can be increased and the pump loss can be reduced.

  The air-fuel ratio (A / F) of the air-fuel mixture is basically the stoichiometric air-fuel ratio (A / F ≒ 14.7) in the entire combustion chamber 17. The excess air ratio λ of the air-fuel mixture may be 1.0 ± 0.2. Note that when the engine 1 is operating near the full load, the excess air ratio λ of the air-fuel mixture may be less than 1.

  The injector 6 starts fuel injection during the intake stroke. The injector 6 injects fuel in a lump. By starting the fuel injection during the intake stroke, a homogeneous or substantially homogeneous mixture is formed in the combustion chamber 17. Further, a long fuel vaporization time can be ensured, so that the unburned loss can be reduced.

  The ignition plug 25 ignites the air-fuel mixture at an appropriate timing before the compression top dead center after the fuel injection is completed.

(Engine control logic)
7A and 7B are block diagrams illustrating the functional configuration of the ECU 10 that executes the control logic of the engine 1. The ECU 10 operates the engine 1 according to the control logic stored in the memory 102. Specifically, the ECU 10 determines the operating state of the engine 1 based on the signals of the sensors SW1 to SW17, and performs the combustion so that the combustion in the combustion chamber 17 becomes the combustion at the SI rate according to the operating state. Calculations for adjusting the state quantity in the chamber 17, adjusting the injection amount, adjusting the injection timing, and adjusting the ignition timing are performed.

  The ECU 10 controls SPCCI combustion using two parameters, the SI ratio and θci. Specifically, the ECU 10 determines a target SI ratio and a target θci corresponding to the operation state of the engine 1, and sets the combustion chamber 17 so that the actual SI ratio matches the target SI ratio and the actual θci becomes the target θci. The internal temperature and the ignition timing are adjusted. The temperature in the combustion chamber 17 is adjusted by adjusting the temperature and / or the amount of the exhaust gas introduced into the combustion chamber 17.

  The ECU 10 first reads signals from the sensors SW1 to SW17 through the I / F circuit 103. Next, the target SI ratio / target θci setting unit 101a in the microcomputer 101 of the ECU 10 determines the operating state of the engine 1 based on the signals of the sensors SW1 to SW17, and also sets the target SI ratio (that is, the target heat amount ratio). And the target CI combustion start timing θci. The target SI ratio is determined according to the operating state of the engine 1. The target SI ratio is stored in the target SI ratio storage unit 1021 of the memory 102. The target SI ratio / target θci setting unit 101a sets the target SI ratio low when the load on the engine 1 is low, and sets the target SI ratio high when the load on the engine 1 is high. When the load of the engine 1 is low, the suppression of the combustion noise and the improvement of the fuel economy are compatible with each other by increasing the ratio of the CI combustion in the SPCCI combustion. When the load of the engine 1 is high, increasing the ratio of SI combustion in SPCCI combustion is advantageous for suppressing combustion noise.

  θci means the crank angle timing at which the CI combustion starts in the SPCCI combustion as described above (see FIG. 5). The target θci is also determined according to the operating state of the engine 1. The target θci is stored in the target θci storage unit 1022 of the memory 102. The target θci storage unit 1022. It is an example of a target timing storage unit. If θci is on the retard side, combustion noise is reduced. If θci is on the advance side, the fuel efficiency of the engine 1 is improved. The target θci is set on the advance side as much as possible in a range where the combustion noise can be suppressed to an allowable level or less.

  The target in-cylinder state quantity setting unit 101b sets a target in-cylinder state quantity for realizing the set target SI ratio and target θci based on the model stored in the memory 102. Specifically, the target in-cylinder state quantity setting unit 101b sets a target temperature and a target pressure in the combustion chamber 17 and a target state quantity.

  The in-cylinder state quantity control unit 101c is configured to open the EGR valve 54, the throttle valve 43, the air bypass valve 48, and the swirl control valve 56 necessary to achieve the target in-cylinder state quantity. The degree and the phase angle of the intake electric S-VT 23 and the exhaust electric S-VT 24 (that is, the valve timing of the intake valve 21 and the valve timing of the exhaust valve 22) are set. The in-cylinder state quantity control unit 101c sets the control amounts of these devices based on the map stored in the memory 102. The in-cylinder state quantity control unit 101c is configured to control the EGR valve 54, the throttle valve 43, the air bypass valve 48, the swirl control valve (SCV) 56, the intake electric S-VT 23, and the exhaust electric S based on the set control amount. A control signal is output to VT24. When each device operates based on a signal from the ECU 10, the state quantity in the combustion chamber 17 becomes the target state quantity.

  The in-cylinder state quantity control unit 101c further calculates a predicted value of the state quantity in the combustion chamber 17 and an estimated value of the state quantity based on the set control amounts of the respective devices. The state quantity prediction value is a value that predicts a state quantity in the combustion chamber 17 before the intake valve 21 closes. The state quantity predicted value is used for setting the fuel injection amount in the intake stroke, as described later. The state quantity estimation value is a value obtained by estimating the state quantity in the combustion chamber 17 after the intake valve 21 is closed. As will be described later, the state quantity estimation value is used for setting the fuel injection amount in the compression stroke and setting the ignition timing.

  The first injection amount setting unit 101d sets the fuel injection amount during the intake stroke based on the state amount prediction value. When performing split injection during the intake stroke, the injection amount of each injection is set. When the fuel injection is not performed during the intake stroke, the first injection amount setting unit 101d sets the fuel injection amount to zero. The first injection control unit 101e outputs a control signal to the injector 6 so that the injector 6 injects fuel into the combustion chamber 17 at a predetermined injection timing. The first injection control unit 101e also outputs a fuel injection result during the intake stroke.

  The second injection amount setting unit 101f sets the fuel injection amount during the compression stroke based on the estimated state amount and the fuel injection result during the intake stroke. When the fuel injection is not performed during the compression stroke, the second injection amount setting unit 101f sets the fuel injection amount to zero. The second injection control unit 101g outputs a control signal to the injector 6 so that the injector 6 injects fuel into the combustion chamber 17 at an injection timing based on a preset map. The second injection control unit 101g also outputs a fuel injection result during the compression stroke.

  The ignition timing setting unit 101h sets the ignition timing based on the estimated state quantity and the fuel injection result during the compression stroke. The ignition control unit 101i outputs a control signal to the ignition plug 25 so that the ignition plug 25 ignites the air-fuel mixture in the combustion chamber 17 at the set ignition timing.

  Here, when the ignition timing setting unit 101h predicts that the temperature in the combustion chamber 17 will be lower than the target temperature based on the estimated state value, the ignition stroke setting unit 101h sets the compression stroke so that the ignition timing can be advanced. The middle injection timing is advanced more than the injection timing based on the map. When the ignition timing setting unit 101h predicts that the temperature in the combustion chamber 17 will be higher than the target temperature based on the estimated state amount, the ignition timing setting unit 101h sets the ignition timing during the compression stroke so that the ignition timing can be retarded. The injection timing is retarded from the injection timing based on the map.

  In other words, when the temperature in the combustion chamber 17 is low, the timing at which the unburned air-fuel mixture self-ignites (CI combustion start timing θci) is delayed after the start of SI combustion by spark ignition, and the SI rate becomes lower than the target SI ratio. It deviates from the SI rate. In this case, an increase in unburned fuel and a decrease in exhaust gas performance are caused.

  Therefore, when the temperature in the combustion chamber 17 is predicted to be lower than the target temperature, the first injection control unit 101e and / or the second injection control unit 101g advances the injection timing and sets the ignition timing setting unit 101h. Advances the ignition timing. Since the early start of the SI combustion enables sufficient heat generation by the SI combustion, it is possible to prevent the timing θci of the self-ignition of the unburned mixture from being delayed when the temperature in the combustion chamber 17 is low. Can be. As a result, θci approaches the target θci, and the SI ratio approaches the target SI ratio.

  When the temperature in the combustion chamber 17 is high, the unburned air-fuel mixture self-ignites immediately after SI combustion starts due to spark ignition, and the SI ratio deviates from the target SI ratio. In this case, combustion noise increases.

  Therefore, when it is predicted that the temperature in the combustion chamber 17 becomes higher than the target temperature, the first injection control unit 101e and / or the second injection control unit 101g delays the injection timing and sets the ignition timing setting unit 101h. Retards the ignition timing. Since the start of the SI combustion is delayed, it is possible to prevent the timing θci of the self-ignition of the unburned mixture from being advanced when the temperature in the combustion chamber 17 is high. As a result, θci approaches the target θci, and the SI ratio approaches the target SI ratio.

  When the ignition plug 25 ignites the air-fuel mixture, SI combustion or SPCCI combustion is performed in the combustion chamber 17. As shown in FIG. 7B, the in-cylinder pressure sensor SW6 measures a change in pressure in the combustion chamber 17.

  The measurement signal of the in-cylinder pressure sensor SW6 is input to a first low-pass filter (LPF) 1031 of the I / F circuit 103. The first low-pass filter 1031 outputs only a signal having a frequency equal to or lower than a predetermined frequency. The first low-pass filter 1031 removes high-frequency electrical noise (so-called white noise) from the measurement signal of the in-cylinder pressure sensor SW6. The A / D converter 101j of the microcomputer 101 changes the measurement signal of the in-cylinder pressure sensor SW6 that has passed through the first low-pass filter 1031 into a digital signal. The A / D converter 101j converts the measurement signal of the in-cylinder pressure sensor SW6 into a digital signal at a sampling frequency of, for example, 50 kHz.

  The sensor signal storage unit 1023 of the memory 102 stores the measurement signal of the in-cylinder pressure sensor SW6 converted into a digital signal.

  The θci deviation calculating unit 101k calculates a deviation between the CI combustion start timing θci estimated based on the measurement signal of the in-cylinder pressure sensor SW6 converted into a digital signal and the target θci. The θci shift calculating unit 101k outputs the calculated θci shift to the target in-cylinder state quantity setting unit 101b. The target in-cylinder state quantity setting unit 101b corrects the model based on the θci shift. The target in-cylinder state quantity setting unit 101b sets the target in-cylinder state quantity using the corrected model in the next and subsequent cycles. Details of the estimation of θci will be described later.

  The control logic of the engine 1 is based on an SI ratio and θci by a state quantity setting device including a throttle valve 43, an EGR valve 54, an air bypass valve 48, a swirl control valve 56, an intake electric S-VT23, and an exhaust electric S-VT24. It is configured to adjust. By adjusting the state quantity in the combustion chamber 17, the SI ratio can be roughly adjusted. The control logic of the engine 1 is also configured to adjust the SI rate and θci by adjusting the fuel injection timing and the ignition timing. By adjusting the injection timing and the ignition timing, for example, correction between cylinders can be performed, or fine adjustment of the self-ignition timing can be performed. By performing the adjustment of the SI ratio in two stages, the engine 1 can accurately achieve the target SPCCI combustion corresponding to the operating state.

(Estimation of CI combustion start timing θci)
The ECU 10 controls the operation of the engine 1 according to the control logic of the engine 1 described above. The ECU 10 estimates the CI combustion start timing θci during the operation of the engine 1 by performing a calculation based on the measured value of the pressure fluctuation in the combustion chamber 17 measured by the in-cylinder pressure sensor SW6. The ECU 10 corrects the ignition timing and the like based on the estimated θci. By doing so, since the actual θci approaches the target θci, the combustion noise of the engine 1 can be suppressed and the fuel efficiency of the engine 1 can be improved.

  The ECU 10 performs the estimation of θci using two types of methods, a first estimation method and a second estimation method, in order to increase the estimation accuracy of θci. Then, an appropriate θci is selected from θci1 estimated by the first estimation method and θci2 estimated by the second estimation method.

  Hereinafter, the signal processing of the in-cylinder pressure sensor SW6 will be described, and the first estimation method for estimating θci and the second estimation method will be sequentially described.

(Signal processing of in-cylinder pressure sensor)
The signal processing of the in-cylinder pressure sensor SW6 in the ECU 10 will be described with reference to FIG. 7B. The I / F circuit 103 of the ECU 10 is provided with the first low-pass filter (LPF) 1031 as described above. The microcomputer 101 of the ECU 10 includes an A / D converter 101j, a first band-pass filter (BPF) 101l, a second band-pass filter (BPF) 101m, and a second low-pass filter (LPF) 101n. At the same time, it has respective functional blocks of a first ignition timing estimation unit 101o, a second ignition timing estimation unit 101p, a selection unit 101q, an angle synchronization processing unit 101r, a pressure conversion unit 101s, and a combustion parameter calculation unit 101t. .

  As described above, the sensor signal storage unit 1023 stores the signal of the in-cylinder pressure sensor SW6 converted into a digital signal. The signal of the in-cylinder pressure sensor SW6 stored in the sensor signal storage unit 1023 is input to the first ignition timing estimation unit 101o via the first bandpass filter 101l. The signal of the in-cylinder pressure sensor SW6 is input to the second ignition timing estimator 101p via the second bandpass filter 101m.

  Although the details will be described later, the CI combustion start timing θci1 estimated by the first ignition timing estimating unit 101o by the first estimating method, and the CI combustion estimated by the second ignition timing estimating unit 101p by the second estimating method. The start timing θci2 is sent to the selection unit 101q. The selection unit 101q sets the CI combustion start timing θci based on θci1 and / or θci2. The memory 102 stores θci.

  Further, the signal of the in-cylinder pressure sensor SW6 stored in the sensor signal storage unit 1023 is input to the angle synchronization processing unit 101r via the second low-pass filter 101n. Then, it is sent to the combustion parameter calculation unit 101t via the pressure conversion unit 101s. The combustion parameter calculation unit 101t calculates a parameter representing a combustion state based on the input information. In this configuration example, the combustion parameter calculation unit 101t calculates the heat release rate dQ / dθ, the crank angle θmfb10 at which the mass combustion ratio becomes 10%, and the crank angle period until the mass combustion ratio becomes 10% to 50%. At least a certain θmfb10-50 period is calculated.

(First Method for Estimating θci)
The inventors of the present application have earnestly studied the estimation of θci, and have obtained the following findings. FIG. 8 is a diagram showing a result of frequency analysis of the in-cylinder pressure measured by the in-cylinder pressure sensor SW6. FIG. 8 shows a result when only SI combustion is performed (dashed line) and a result when only CI combustion is performed (solid line) at a predetermined rotation speed of the engine 1 and a predetermined load of the engine 1. The comparison is shown. The present inventors have found that in the first specific frequency band S1 equal to or higher than the first frequency f1 and equal to or lower than the second frequency f2, the spectrum of the in-cylinder pressure is clearly different between during SI combustion and during CI combustion.

  Then, the inventors of the present invention determine the value of the component of the first specific frequency band S1 included in the time waveform of the in-cylinder pressure, that is, the time when the value of the in-cylinder pressure in the first specific frequency band S1 is minimum, and the time when the CI combustion starts It was found that θci was almost the same.

  FIG. 9 is a diagram exemplifying this, and shows a waveform (time change) of each parameter when CI combustion appropriately occurs after SI combustion at a predetermined rotation speed of the engine 1 and a predetermined load of the engine 1. Is shown. FIG. 9 shows, in order from the top, a waveform 91 of the in-cylinder pressure, a waveform 92 of the first specific frequency band S1 included in the waveform of the in-cylinder pressure (only the waveform of the first specific frequency band S1 is extracted from the waveform of the in-cylinder pressure). ), A heat generation amount waveform 93, and a heat generation rate waveform 94.

  When CI combustion occurs after SI combustion, an inflection point X occurs in the waveform of the heat release rate as described above. Specifically, when CI combustion occurs after SI combustion, the heat generation rate sharply increases with the start of CI combustion during combustion (after the heat generation rate rises from around 0). The timing at which the heat generation rate rises rapidly (timing of the inflection point X) is θci. Then, the value of the component of the in-cylinder pressure in the first specific frequency band S1 (hereinafter, referred to as a first specific frequency output value) is minimum near θci.

  In addition, the inventors of the present application have found that the minimum value of the first specific frequency output value differs between when CI combustion properly occurs after SI combustion and when CI combustion does not properly occur. Specifically, the present inventors have found that the minimum value of the first specific frequency output value is smaller when CI combustion properly occurs after SI combustion than when CI combustion does not properly occur after SI combustion. I figured out.

  FIG. 10 shows a cylinder in the case where CI combustion does not occur after SI combustion at the same rotation speed of the engine 1 and the load of the engine 1 as in FIG. 9, that is, when only SI combustion is performed in the combustion chamber 17. A waveform 96 of the internal pressure, a waveform 97 of the first specific frequency output value, a waveform 98 of the heat generation amount, and a waveform 99 of the heat generation rate are shown. As is clear from the comparison between FIGS. 9 and 10, when the CI combustion occurs properly after the SI combustion, the first specific frequency output is smaller than when the CI combustion does not properly occur after the SI combustion. The minimum value of the value Cmin_ci becomes smaller. That is, when CI combustion appropriately occurs after SI combustion, the minimum value Cmin_ci of the first specific frequency output value becomes smaller than the first threshold value Cj1, and CI combustion does not appropriately occur after SI combustion. At this time, the minimum value Cmin_ci of the first specific frequency output value becomes larger than the first threshold value Cj1.

  In addition, the inventors of the present application have found that the first specific frequency S1 changes according to the rotation speed of the engine 1. Specifically, the higher the rotation speed of the engine 1, the higher the first specific frequency S1.

  The first ignition timing estimating unit 101o estimates θci by the first technique. Specifically, the first method is to pass the measurement value of the in-cylinder pressure sensor SW6 through a first band-pass filter 101l (see FIG. 7B) in which the first specific frequency S1 is set to a pass band, and When the minimum value Cmin_ci of the first specific frequency output value, which is the output value of the bandpass filter 101l, is smaller than the first threshold value Cj1, the crank angle θmin that becomes the Cmin_ci is estimated as the CI combustion start timing θci.

  FIG. 11 is a flowchart illustrating a procedure in which the ECU 10 estimates θci by the first technique. In step S1, the ECU 10 reads the signal of the in-cylinder pressure sensor SW6 stored in the sensor signal storage unit 1023 (that is, the voltage signal of the in-cylinder pressure sensor SW6 after being converted into a digital signal).

  Next, in step S2, the ECU 10 passes the signal of the in-cylinder pressure sensor SW6 read in step S1 through the first band-pass filter 101l. The first bandpass filter 101l is a filter that allows only the signal of the first frequency band S1 to pass. In step S2, a first specific frequency output value is extracted from the measurement signal of the in-cylinder pressure sensor SW6. The first specific frequency output value output from the first bandpass filter 101l is sent to the first ignition timing estimating unit 101o.

  The ECU 10 also changes the first specific frequency band S1 according to the rotation speed of the engine 1. Specifically, as shown in FIG. 12, the first specific frequency band S1 is set such that the higher the rotation speed of the engine 1, the higher the frequency side. In the example shown in FIG. 12, the first frequency f1 is set to 0.5 kHz and the second frequency f2 is set to 1.5 kHz when the rotation speed of the engine 1 is 1000 rpm, and the first frequency f1 is set to 1.5 kHz when the rotation speed of the engine 1 is 2000 rpm. 1 kHz, the second frequency f2 is set to 2 kHz, the first frequency f1 is set to 1.25 kHz, the second frequency f2 is set to 2.25 kHz when the rotation speed of the engine 1 is 3000 rpm, and the rotation speed of the engine 1 is set to 4000 rpm. , The first frequency f1 is set to 1.5 kHz, the second frequency f2 is set to 2.5 kHz, and when the rotation speed of the engine 1 is 5000 rpm, the first frequency f1 is set to 1.75 kHz and the second frequency f2 is set to 2.75 kHz. Is set. The first specific frequency band S1 may be set so as to be included in a range from 0.5 kHz to 4 kHz. The ECU 10 extracts the first specific frequency band S1 corresponding to the current rotational speed of the engine 1 from the map of FIG. 12 and applies the first specific frequency band S1 to the first bandpass filter 1011.

  Next, in step S3, the ECU 10 (the first ignition timing estimating unit 101o) obtains the minimum value of the extracted first specific frequency output value. The minimum value of the first specific frequency output value also includes a minimum value (hereinafter, referred to as a minimum specific frequency output value) Cmin_ci. Specifically, when the output value of the in-cylinder pressure sensor SW6 passes through the band-pass filter 123, a waveform 92 as shown in the second from the top in FIG. 9 is obtained. This waveform 92 is a waveform obtained by synthesizing a waveform of each frequency in the first specific frequency band included in the waveform of the in-cylinder pressure. Then, as indicated by a white circle in FIG. 9, a value at which the values (pressure and voltage) become minimum in this waveform is a minimum value. The ECU 10 extracts a plurality of minimum values in step S3. The ECU 10 may extract, for example, up to three local minimum values.

  Next, in step S4, the ECU 10 (first ignition timing estimating unit 101o) obtains the crank angle of each of the minimum values extracted in step S3 as the CI ignition timing candidate θmin.

  Specifically, the signal of the crank angle sensor SW11 is stored in the sensor signal storage unit 1023 in association with the signal of the in-cylinder pressure sensor SW6 sampled at 50 kHz, and the ECU 10 (first ignition timing estimation unit 101o) Based on these signals, the crank angle at which the first specific frequency output value becomes minimum is obtained.

  Next, in step S5, the ECU 10 (first ignition timing estimation unit 101o) estimates the CI combustion start timing. Specifically, the ECU 10 determines the minimum first specific frequency output value Cmin_ci from the minimum values of the first specific frequency output values calculated in step S3. Then, the ECU 10 determines whether or not the minimum specific frequency output value Cmin_ci is less than the first threshold value Cj1. The first threshold value Cj1 is stored in the threshold value storage unit 1024 of the memory 102 (see FIG. 7B). If the minimum specific frequency output value Cmin_ci is less than the first threshold value Cj1, the ECU 10 determines the crank angle θmin at which the minimum specific frequency output value Cmin_ci becomes the CI combustion start timing θci. On the other hand, if the minimum specific frequency output value Cmin_ci is equal to or greater than the first threshold value Cj1, it is assumed that the ECU 10 cannot estimate θci.

  In subsequent step S6, ECU 10 determines whether or not θci was successfully estimated in step S5. If the determination in step S6 is YES, the process proceeds to step S7, and ECU 10 (first ignition timing estimation unit 101o). Outputs the θci determined in step S5 to the selection unit 101q as θci1.

  On the other hand, if the determination in step S6 is NO, the process ends without proceeding to step S7. The first ignition timing estimating unit 101o does not output θci1.

(Calculation of mass combustion ratio)
Next, a calculation procedure of the mass combustion ratio performed by the combustion parameter calculation unit 101t of the ECU 10 will be described with reference to a flowchart of FIG. First, in step S31, the ECU 10 reads the signal of the in-cylinder pressure sensor SW6 stored in the sensor signal storage unit 1023.

  Next, in step S32, the ECU 10 passes the signal of the in-cylinder pressure sensor SW6 read in step S31 through the second low-pass filter 101n. The second low-pass filter 101n is a filter that can remove a signal of a predetermined frequency. The second low-pass filter 101n is configured to remove a signal of a relatively high frequency which is a frequency of a waveform of the in-cylinder pressure when knocking occurs and which is set in advance. The second low-pass filter 101n removes a knocking signal from the signal of the in-cylinder pressure sensor SW6. The signal of the in-cylinder pressure sensor SW6 output from the second low-pass filter 101n is sent to the angle synchronization processing unit 101r.

  Next, in step S33, the ECU 10 (the angle synchronization processing unit 101r) associates the signal sampled at 50 kHz, which is the measurement signal of the in-cylinder pressure sensor SW6 output from the second low-pass filter 101n, with this signal. Using the stored signal of the crank angle sensor SW11, the signal is converted into a signal for each predetermined crank angle. In this configuration example, the ECU 10 (angle synchronization processing unit 101r) converts the signal of the in-cylinder pressure sensor SW6 into a signal for each 3 ° CA in step S33. The signal from the in-cylinder pressure sensor SW6 is sent to the pressure converter 101s.

  Next, in step S34, the ECU 10 (pressure conversion unit 101s) converts the signal of the in-cylinder pressure sensor SW6 input from the angle synchronization processing unit 101r into the absolute pressure of the in-cylinder pressure. That is, the signal output from the angle synchronization processing unit 101r is still a voltage value, and the pressure conversion unit 101s converts this signal into the absolute pressure of the in-cylinder pressure for the first time.

  In this configuration example, the absolute pressure Pcps of the in-cylinder pressure can be calculated by Pcps = K × Vcps + OFFSET where the voltage of the in-cylinder pressure sensor SW6 is Vcps. The ECU 10 (pressure conversion unit 101s) converts the output value (voltage value) of the in-cylinder pressure sensor SW6 into an absolute pressure using this equation.

  The coefficient K is a value predetermined for each in-cylinder pressure sensor SW6, and is stored in the memory 102 of the ECU 10. On the other hand, the coefficient OFFSET (hereinafter, this coefficient is appropriately referred to as an offset amount) is not set in advance, and in this configuration example, the ECU 10 (the pressure conversion unit 101s) calculates using the value of the intake pressure sensor SW5. The procedure for calculating the offset amount OFFSET will be described later.

  The output value of the in-cylinder pressure sensor SW6 converted into the absolute pressure of the in-cylinder pressure is input to the combustion parameter calculation unit 101t.

  Next, in step S35, the ECU 10 (combustion parameter calculation unit 101t) calculates and calculates the heat release rate dQ for each predetermined crank angle using the output value (absolute pressure) P of the in-cylinder pressure sensor SW6. The heat generation amount Q (θ) at each crank angle is calculated by integrating dQ.

  Next, in step S36, the ECU 10 (combustion parameter calculation unit 101t) calculates the minimum value Qmin and the corresponding crank angle Qmf0 in the heat release amount Q (θ) calculated in step S35. Then, in the subsequent step S37, the ECU 10 (combustion parameter calculation unit 101t) corrects the heat generation amount Q (θ) so that the minimum value Qmin calculated in step S36 becomes zero [J]. This suppresses the occurrence of an error when calculating the mass combustion ratio, which will be described later.

  In step S38, the ECU 10 (combustion parameter calculation unit 101t) calculates the maximum heat generation amount Qmax of the heat generation amount Q (θ), and in the following step S39, the heat generation amount Q10 that is 10% of the maximum heat generation amount Qmax. , And a heat release amount Q50 that is 50% of the maximum heat release amount Qmax.

  Then, in step S310, the ECU 10 (combustion parameter calculation unit 101t) determines the crank angle at which the 10% heat generation amount Q10 of the maximum heat generation amount Qmax becomes θmfb10, and the 50% heat generation amount of the maximum heat generation amount Qmax. The crank angle to be Q50 is determined to be θmfb50.

  In the last step S311, the ECU 10 (combustion parameter calculation unit 101t) calculates a θmfb10-50 period that is a crank angle period from θmfb10 to θmfb50.

  As described above, the combustion parameter calculating unit 101t calculates at least dQ / dθ, θmfb10-50 period, and θmfb10.

(Calculation of offset amount)
Next, a procedure for calculating the offset amount OFFSET required to convert the voltage output from the in-cylinder pressure sensor SW6 into an absolute pressure will be described with reference to the flow of FIG.

  First, in step S41, the pressure conversion unit 101s (see FIG. 7B) reads the intake valve closing timing IVC, which is the timing when the intake valve 21 is closed. In detail, the intake valve closing timing IVC of the cylinder 11 provided with the in-cylinder pressure sensor SW6 to be converted and the intake valve closing timing IVC in the combustion cycle to be converted are read.

  Next, in step S42, the pressure conversion unit 101s performs a period from a timing that is a predetermined crank angle before the intake valve closing timing IVC to the intake valve closing timing IVC (a predetermined period, hereinafter referred to as an averaging process period as appropriate). ), A plurality of intake pressures detected by the intake pressure sensor SW5 are read, and a plurality of voltage values output from the in-cylinder pressure sensor SW6 during the averaging process are read from the memory 102. The average processing period is set to, for example, 12 ° CA (crank angle).

  Next, in step S43, the pressure conversion unit 101s calculates the average value Pim_ave of the plurality of intake pressures read in step S42, that is, the average value Pim_ave of the intake pressures during the average processing period.

  In step S44, the pressure conversion unit 101s determines the average value Vcps_ave of the output values (voltage values) of the plurality of in-cylinder pressure sensors SW6 read in step S42, that is, the output value (voltage) of the in-cylinder pressure sensor SW6 during the average processing period. ) Is calculated.

  Next, in step S45, the pressure conversion unit 101s sets a value obtained by multiplying the average value Vcps_ave of the output value (voltage value) of the in-cylinder pressure sensor SW6 calculated in step S44 by a coefficient K as the in-cylinder pressure before offset correction. calculate. That is, in step S45, the cylinder pressure before the offset correction is set as Pcps_of, which is calculated by Pcps_of = K × Vcps_ave.

  Next, in step S46, the pressure conversion unit 101s calculates, as the offset amount, a value obtained by subtracting the in-cylinder pressure before offset correction Pcps_of calculated in step S45 from the average value Pim_ave of the intake pressure calculated in step S43. That is, the offset amount is set as OFFSET, and this is calculated by OFFSET = Pim_ave-Pcps0.

(Suppression of erroneous estimation in the first method)
By the way, depending on the operation state of the engine 1, the heat generation rate of the SI combustion in the SPCCI combustion may increase as shown by a dashed line in the heat generation rate waveform 94 of FIG. That is, the speed of the initial combustion in the SPCCI combustion increases, and the rising angle in the heat generation rate waveform 94 may become steep. In this case, as shown by a dashed line in the output value 92 of the first specific frequency band S1 in FIG. 9, the first specific frequency output value at a timing advanced from the actual θci is smaller than the first threshold value Cj1. It will be a small, minimum value Cmin. When the first ignition timing estimating unit 101o attempts to estimate θci by the first method described above, the first ignition timing estimating unit 101o makes an erroneous estimation.

  The first ignition timing estimation unit 101o is configured to suppress erroneous estimation in the first estimation method. There are five methods of suppressing erroneous estimation, from “Part 1” to “Part 5”. The first ignition timing estimating unit 101o suppresses the erroneous estimation by using any one of the methods from “No. 1” to “No. 5” or a combination of a plurality of methods. Hereinafter, five methods for suppressing erroneous estimation in the first estimation method will be sequentially described.

(Method to suppress erroneous estimation (Part 1))
FIG. 15 shows a functional configuration of the ECU 10 related to the estimation of θci. Specifically, the first ignition timing estimating unit 101o includes an adjusting unit 101u. The adjustment unit 101u adjusts the first threshold value Cj1 stored in the threshold value storage unit 1024.

The upper diagram of FIG. 16 illustrates an adjustment map 161 of the first threshold value Cj1. The adjustment map 161 of the first threshold value Cj1 defines the relationship between the heat release rate dQ / dθ and the first threshold value Cj1. The vertical axis of the adjustment map 161 indicates the absolute value of the first threshold value Cj1.
The horizontal axis of the adjustment map 161 indicates the heat release rate dQ / dθ during a predetermined period from ignition. The adjustment map 161 of the first threshold value Cj1 is stored in the threshold value storage unit 1024.

  The combustion parameter calculation unit 101t outputs the calculated value of the heat release rate dQ / dθ to the adjustment unit 101u. The adjusting unit 101u sets the first threshold value Cj1 based on the value of the heat release rate dQ / dθ during a predetermined period after the ignition of the spark plug 25 and the adjustment map 161 of the first threshold value Cj1. Specifically, the adjusting unit 101u increases the absolute value of the first threshold value Cj1 when the heat release rate dQ / dθ in the predetermined period is high, and increases the first value when the heat release rate dQ / dθ in the predetermined period is low. The absolute value of the threshold value Cj1 is reduced.

  When the rising angle of the heat release rate of the SI combustion becomes steep, the heat release rate dQ / dθ in the predetermined period after the ignition of the spark plug 25 becomes high as shown by the dashed line in the heat release rate waveform 94 of FIG. Become. When the absolute value of the first threshold value Cj1 is increased, the first threshold value Cj1 falls below the plane of the paper, as indicated by the dashed arrow in the waveform 92 of FIG. It is possible to prevent the minimum value Cmin of the first specific frequency output value at a timing advanced from the actual CI combustion start timing θci from falling below the first threshold value Cj1.

  That is, in the example shown by the dashed line in FIG. 9, the adjustment unit 101u adjusts the first threshold value Cj1, so that the first ignition timing estimation unit 101o determines the first ignition timing in step S5 of the flow of FIG. The determination that the minimum value Cmin of the specific frequency output value is less than the corrected first threshold value Cj1 is suppressed. The first ignition timing estimating unit 101o may erroneously estimate the CI combustion start timing θci based on the minimum value Cmin of the first specific frequency output value advanced from the actual CI combustion start timing θci. Is suppressed.

  In this case, the minimum value Cmin_ci of the first specific frequency output value corresponding to the actual θci also does not fall below the first threshold value Cj1, so that the first ignition timing estimation unit 101o can estimate θci1. May disappear. However, as described above, in the configuration example disclosed herein, the ECU 10 estimates θci by each of the two methods, the first method and the second method. Even if the first ignition timing estimating unit 101o cannot estimate θci1 by the first method, the second ignition timing estimating unit 101p can estimate θci2 by the second method.

(Method of suppressing erroneous estimation (part 2))
The functional configuration of the ECU 10 is the same as that of FIG. The adjustment unit 101u adjusts the first threshold value Cj1 stored in the threshold value storage unit 1024.

  The lower diagram of FIG. 16 illustrates the adjustment map 162 of the first threshold value Cj1. The adjustment map 162 for the first threshold value Cj1 defines the relationship between the θmfb10-50 period and the first threshold value Cj1. The vertical axis of the adjustment map 162 indicates the absolute value of the first threshold value Cj1. The adjustment map 162 of the first threshold value Cj1 is stored in the threshold value storage unit 1024.

  The adjusting unit 101u sets the first threshold value Cj1 based on the value of the θmfb10-50 period calculated by the combustion parameter calculating unit 101t and the adjustment map 162 of the first threshold value Cj1. Specifically, the adjustment unit 101u increases the absolute value of the first threshold value Cj1 when the θmfb10-50 period is short, and decreases the absolute value of the first threshold value Cj1 when the θmfb10-50 period is long. .

  As can be seen from the heat generation amount waveform 93 in FIG. 9, when the rising angle of the heat generation rate of SI combustion becomes steep, the θmfb10-50 period becomes short. When the absolute value of the first threshold value Cj1 is increased, the minimum value Cmin of the first specific frequency output value at a timing advanced from the actual CI combustion start timing θci becomes the same as the method of “Part 1”. It can be suppressed that the value falls below one threshold value Cj1. That is, the first ignition timing estimation unit 101o erroneously estimates the CI combustion start timing θci based on the minimum value Cmin of the first specific frequency output value advanced from the actual CI combustion start timing θci. Is suppressed.

(Method of suppressing erroneous estimation (part 3))
The functional configuration of the ECU 10 is the same as that of FIG. The adjustment unit 101u adjusts the first threshold value Cj1 stored in the threshold value storage unit 1024.

  FIG. 17 illustrates an adjustment map 171 of the first threshold value Cj1. The adjustment map 171 of the first threshold value Cj1 defines the relationship between the θmfb10-50 period, θmfb10, and the absolute value of the first threshold value Cj1. The vertical axis of the adjustment map 171 is θmfb10, where θmfb10 is a retard angle as going upward and θmfb10 is an advanced angle as going downward. The horizontal axis of the adjustment map 171 is the θmfb10-50 period, where the period is longer as going to the right and shorter as going to the left. The adjustment map 171 sets the first threshold value Cj1 to three, large, medium, and small, in accordance with θmfb10-50 period and θmfb10. Although not shown, the adjustment map 171 may set the value of the first threshold value Cj1 more finely. The adjustment map 171 for the first threshold value Cj1 is also stored in the threshold value storage unit 1024.

  The adjustment unit 101u sets the first threshold value Cj1 based on the value of the θmfb10-50 period calculated by the combustion parameter calculation unit 101t, θmfb10, and the adjustment map 171 of the first threshold value Cj1. Specifically, if the θmfb10-50 period is short and θmfb10 is on the advance side, the absolute value of the first threshold value Cj1 is increased, and if the θmfb10-50 period is long and θmfb10 is on the retard side, The absolute value of the first threshold value Cj1 is reduced, and the intermediate value between the absolute values of the first threshold value Cj1 is set to be medium.

  When the rising angle of the heat release rate of the SI combustion becomes steep, the θmfb10-50 period is shortened and θmfb10 is advanced. By increasing the absolute value of the first threshold value Cj1, the minimum value Cmin of the first specific frequency output value at a timing advanced from the actual CI combustion start timing θci falls below the first threshold value Cj1. Can be suppressed. The first ignition timing estimating unit 101o may erroneously estimate the CI combustion start timing θci based on the minimum value Cmin of the first specific frequency output value advanced from the actual CI combustion start timing θci. Be suppressed. In “part 3”, since both the θmfb10-50 period and θmfb10 are considered, the combustion speed of SI combustion in SPCCI combustion can be grasped more accurately, and the first threshold value Cj1 is more appropriately set. Can be set to As a result, the first ignition timing estimating unit 101o can suppress erroneous estimation, and can accurately estimate the CI combustion start timing θci based on the minimum value Cmin_ci of the first specific frequency output value.

  Here, the adjustment map 171 for the first threshold value Cj1 increases the absolute value of the first threshold value Cj1 when the θmfb10-50 period is longer than a predetermined period and θmfb10 is on the retard side than a predetermined period. (The upper right area of the adjustment map 171). If the θmfb10-50 period is longer than the predetermined period and θmfb10 is on the more retarded side than the predetermined period, CI combustion in the SPCCI combustion does not occur in the first place. Therefore, by increasing the absolute value of the first threshold value Cj1, it is possible to prevent the first ignition timing estimation unit 101o from erroneously estimating θci.

(Method of suppressing erroneous estimation (part 4))
In the methods of “Part 1” to “Part 3”, erroneous estimation is suppressed by adjusting the first threshold value Cj1. The method of “Part 4” does not perform adjustment of the first threshold value Cj1, and when there are a plurality of minimum values of the first specific frequency output value exceeding the first threshold value Cj1, the minimum values thereof are used. The correct CI combustion start timing θci is selected from θmin corresponding to the value according to a predetermined rule.

  Specifically, the ECU 10 (the first ignition timing estimating unit 101o) extracts a plurality of local minimum values in steps S3 and S4 of the flow of FIG. 11, and then in step S5, among the extracted local minimum values, A local minimum value smaller than one threshold value Cj1 is specified. When there is only one minimum value less than the first threshold value Cj1, the ECU 10 (first ignition timing estimation unit 101o) determines the crank angle θmin corresponding to the minimum value as the CI combustion start timing θci. .

  On the other hand, when there are a plurality of minimum values less than the first threshold value Cj1, the ECU 10 (first ignition timing estimating unit 101o) determines the most retarded timing θmin as the CI combustion start timing θci. As described above, since the erroneous estimation of θci is caused by an increase in the heat release rate during SI combustion, the first ignition timing estimating unit 101o erroneously estimates θci to a more advanced angle than the actual θci. Estimate. Therefore, the most retarded timing of the plurality of candidates is determined as the CI combustion start timing θci. In the example of FIG. 9, the first ignition timing estimating unit 101o selects Cmin_ci from the minimum value Cmin of the first specific frequency output value and Cmin_ci. The first ignition timing estimating unit 101o can suppress erroneous estimation.

(Method for suppressing erroneous estimation (part 5))
FIG. 18 illustrates functional blocks of the ECU 10 according to the “part 5” method. As described above, the first ignition timing estimating unit 101o of the ECU 10 determines the value of the component of the first specific frequency band S1 included in the in-cylinder pressure waveform 1801 measured by the in-cylinder pressure sensor SW6 (that is, the first specific frequency output). Based on the value 1802), the minimum values of the plurality of first specific frequency output values and the crank angles θmin corresponding to those minimum values are extracted (steps S3 and S4 in the flow of FIG. 11 and the waveform 1802). See white circle).

  The first ignition timing estimating unit 101o also predicts the CI combustion start timing θci_pre based on the operating state of the engine 1 obtained from the measurement signals of the sensors SW1 to SW17. Specifically, the first ignition timing estimating unit 101o calculates θmfb10 and θmfb50 calculated according to the flow of FIG. 13, the crank angle at which the heat release rate dQ / dθ is maximum, the maximum value of dQ / dθ, and the in-cylinder state quantity. Based on the above, a CI combustion start timing is predicted using a model. Further, the first ignition timing estimating unit 101o corrects the predicted CI combustion start timing based on the rotation speed ne and the charging efficiency ce of the engine 1 and a preset correction map, and calculates the predicted CI combustion start timing θci_pre. Determine (see dashed line in waveform 1802).

  The first ignition timing estimating unit 101o compares the crank angle θmin corresponding to each of the plurality of local minimum values with the predicted CI combustion start timing θci_pre, and is on the more retarded side than the predicted CI combustion start timing θci_pre and performs the predicted CI combustion. The minimum crank angle θmin closest to the start timing θci_pre is determined as the CI combustion start timing θci. In the example of the waveform 1802, the crank angle θmin of the minimum value of the most retarded angle is θci.

  As described above, the first ignition timing estimating unit 101o erroneously estimates θci to be more advanced than the actual θci. By determining the CI combustion start timing θci from the plurality of candidates according to the above rule, the first ignition timing estimating unit 101o can suppress erroneous estimation.

(Second method for estimating θci)
The present inventors have found that the frequency of a pressure wave generated during CI combustion in SPCCI combustion is close to the frequency of a pressure wave (standing wave) generated when knocking occurs in SI combustion. If the occurrence of the pressure wave is detected, the timing at which the CI combustion of the SPCCI combustion starts can be estimated. The second method for estimating θci estimates CI combustion start timing θci by detecting the occurrence of pressure waves generated during CI combustion.

  The frequency of the pressure wave generated at the time of the CI combustion of the SPCCI combustion corresponds to a second specific frequency band S2 having a higher frequency than the first specific frequency band S1, as indicated by a dashed line in FIG. The second specific frequency band S2 is a frequency band equal to or higher than the third frequency f3 and equal to or lower than the fourth frequency f4. The third frequency and the fourth frequency are set in a range from 5.5 kHz to 8.0 kHz.

  The second ignition timing estimating unit 101p estimates θci by the second technique. Specifically, the second method is to pass the measured value of the in-cylinder pressure sensor SW6 through a second band-pass filter 101m whose second specific frequency band S2 is set as a pass band, The crank angle at which the second specific frequency output value exceeding the second threshold value Cj2 first exceeds the second threshold value Cj2 is estimated as the CI combustion start timing θci.

  Here, FIG. 19 shows the absolute value of the waveform 1901 of the in-cylinder pressure and the absolute value of the second specific frequency output value obtained by passing the signal of the in-cylinder pressure sensor SW6 through the second band-pass filter 101m when the rotation speed of the engine 1 is relatively high. A value waveform 1902 is illustrated. The dashed line in the figure indicates the second threshold value Cj2. The absolute value of the second specific frequency output value exceeds the second threshold value Cj2 at the timing θci at which CI combustion starts.

  However, on the advanced side of θci, the absolute value of the second specific frequency output value exceeds the second threshold value Cj2. The second ignition timing estimation unit 101p erroneously estimates that the timing is θci. Such an erroneous estimation occurs because the spectrum of the in-cylinder pressure increases in the second specific frequency band S2 when the rotation speed of the engine 1 increases, as shown in FIG. FIG. 20 shows the result of simulating the in-cylinder pressure spectrum when the same amount of heat is generated in the combustion chamber 17 when the rotation speed of the engine 1 is 1000, 2000, 4000, and 6000 rpm, respectively. ing. According to this, when the number of revolutions of the engine 1 increases, the progress of the time with respect to the progress of the crank angle decreases, so that the spectrum is extended toward higher frequencies. As a result, when the rotational speed of the engine 1 increases, the spectrum of the in-cylinder pressure increases in the second specific frequency band S2.

  Therefore, the second ignition timing estimation unit 101p is configured to suppress erroneous estimation in the second estimation method. Specifically, as shown in FIG. 15, the second ignition timing estimating unit 101p has an adjusting unit 101v. The adjustment unit 101v changes the second threshold value Cj2 stored in the threshold value storage unit 1024 according to the operating state of the engine 1.

  FIG. 21 illustrates an adjustment map 211 of the second threshold value Cj2. The adjustment map 211 of the second threshold value Cj2 defines the relationship between the rotation speed of the engine 1 and the second threshold value Cj2. The vertical axis of the adjustment map 211 indicates the absolute value of the second threshold value Cj2. The adjustment map 211 of the second threshold value Cj2 is stored in the threshold value storage unit 1024.

  The adjusting unit 101v sets the second threshold value Cj2 based on the rotation speed of the engine 1 and the adjustment map 211 of the second threshold value Cj2. Specifically, when the rotation speed of the engine 1 is less than the predetermined rotation speed N3, the adjustment unit 101v keeps the absolute value of the second threshold value Cj2 constant at a predetermined value.

  When the rotation speed of the engine 1 is equal to or higher than the predetermined rotation speed N3, the adjustment unit 101v increases the absolute value of the second threshold value Cj2 as the rotation speed of the engine 1 increases. By doing so, as the rotation speed of the engine 1 increases, the second threshold value Cj2 in FIG. 19 moves upward in the drawing, so that the absolute value of the second specific frequency output value becomes equal to the second threshold value Cj2. , And the second ignition timing estimation unit 101p can suppress erroneous estimation.

  Here, the predetermined rotation speed N3 in the adjustment map 211 is lower than the lower limit rotation speed N2 of the high rotation region A5 where SI combustion is performed, and higher than the rotation speed N1.

  The adjustment map 211 is the same regardless of the level of the load of the engine 1. When the rotation speed of the engine 1 is the same, the adjustment unit 101v sets the absolute value of the second threshold value Cj2 to be the same with respect to the level of the load of the engine 1.

  FIG. 22 is a flowchart illustrating a procedure in which the ECU 10 estimates θci by the second technique. In step S21, the ECU 10 reads the signal of the in-cylinder pressure sensor SW6 stored in the sensor signal storage unit 1023.

  Next, in step S22, the ECU 10 passes the signal of the in-cylinder pressure sensor SW6 read in step S21 through the second band-pass filter 101m. The second bandpass filter 101m passes only the signal of the second specific frequency band S2. Thereby, in step S22, the second specific frequency output value is extracted from the measurement signal of the in-cylinder pressure sensor SW6. The second specific frequency output value output by the second bandpass filter 101m is sent to the second ignition timing estimating unit 101p.

  Next, in step S23, the ECU 10 changes the second threshold value Cj2 according to the rotation speed of the engine 1. As shown in FIG. 21, the ECU 10 sets the absolute value of the second threshold value Cj2 to a predetermined value when the rotation speed of the engine 1 is less than the predetermined rotation speed N3, and when the rotation speed exceeds the predetermined rotation speed N3, the engine 1 Is set so that the absolute value of the second threshold value Cj2 becomes larger as the rotation speed becomes higher.

  Next, in step S24, the ECU 10 (the second ignition timing estimation unit 101p) determines whether or not the extracted second specific frequency output value is equal to or greater than the set second threshold value Cj2. If the determination in step S24 is YES, the process proceeds to step S25, in which the ECU 10 (second ignition timing estimating unit 101p) determines the time when the second specific frequency output value first becomes equal to or more than the second threshold value Cj2. The CI combustion start timing θci2 is determined, and θci2 is output to the selection unit 101q.

  On the other hand, if the determination in step S24 is NO, the process ends without proceeding to step S25. It is assumed that the second ignition timing estimating unit 101p cannot estimate θci2.

  The second estimation method can be used not only for estimating the CI combustion start timing θci in SPCCI combustion but also for detecting occurrence of knocking in SI combustion.

(Configuration of the selection unit)
As described above, θci1 estimated by the first ignition timing estimation unit 101o and θci2 estimated by the second ignition timing estimation unit 101p are each input to the selection unit 101q. The selection unit 101q selects θci according to the following conditions.
1) When only one of θci1 and θci2 is estimated, the CI combustion start timing is selected as θci.
2) When both θci1 and θci2 are estimated, the CI combustion start timing on the advance side is selected as θci.

  The θci selected by the selection unit 101q is stored in the memory 102. As described above, θci is used for suppressing combustion noise. If θci is on the advance side of target θci, target in-cylinder state quantity setting unit 101b corrects the model so that the ignition timing is retarded. . By selecting the timing on the advance side when both θci1 and θci2 are estimated, θci is more likely to be shifted toward the advance side than the target θci, so that the generation of combustion noise is more effectively suppressed. It becomes possible.

(Correction of estimated CI combustion start timing)
In the configuration described above, the ECU 10 suppresses erroneous estimation of the CI combustion start timing θci. As a second embodiment different from this configuration, the ECU 10 may correct the estimated CI combustion start timing so that the estimated CI combustion start timing is a correct CI combustion start timing.

  FIG. 23 illustrates a functional configuration of the ECU 10 according to the second embodiment. The functional configuration according to the second embodiment is a configuration that replaces the functional configuration of FIG. Therefore, in the configuration according to the second embodiment, the ECU 10 does not change the first threshold value Cj1. The first threshold value Cj1 is a predetermined fixed value.

  The first ignition timing estimation unit 101o has a correction unit 101w. The correction unit 101w corrects the estimated θci according to the operating state of the engine 1.

  The upper diagram of FIG. 24 illustrates the correction map 241. The correction map 241 defines the relationship between the heat release rate dQ / dθ and the correction value θcor. Specifically, if the heat release rate dQ / dθ during the predetermined period after ignition is high, the correction value θcor is increased in the retard direction, and if the heat release rate dQ / dθ during the predetermined period is low, the correction value θcor is reduced. . The correction map 241 is stored in the memory 102.

  The correction unit 101w calculates the value of the heat release rate dQ / dθ calculated by the combustion parameter calculation unit 101t based on the value of the heat release rate dQ / dθ during a predetermined period after ignition of the spark plug 25 and the correction map 241. Thus, the correction value θcor is determined.

  Then, the first ignition timing estimating unit 101o adds the determined correction value θcor to the timing θmin1 (see the waveform 92 in FIG. 9) at which the minimum value of the first specific frequency output value exceeds the first threshold value Cj1. Thus, the CI combustion start timing θci1 is determined. Specifically, θci1 = θmin1 + θcor.

  As described above, the erroneous estimation in the first method is based on the assumption that the CI combustion start timing θci is advanced from the actual CI combustion start timing θci when the heat generation rate during SI combustion increases. It depends. After the ignition of the ignition plug 25, when the heat release rate dQ / dθ during a predetermined period is high, θci1 temporarily corrected based on the minimum value of the first specific frequency output value is corrected in the direction of retarding, whereby θci1 Comes closer to the actual CI combustion start timing θci. That is, the accuracy of estimating θci of the first ignition timing estimating unit 101o increases.

  Here, the first ignition timing estimating unit 101o may estimate the CI combustion start timing θci1 using the correction map 242 shown in the lower diagram of FIG. 24 instead of the correction map 241 shown in the upper diagram of FIG. .

  The correction map 242 defines the relationship between the θmfb10-50 period and the correction value θcor. Specifically, if the period θmfb10-50 is long, the correction value θcor is increased in the retard direction, and if the period θmfb10-50 is short, the correction value θcor is reduced. The correction map 242 is stored in the memory 102.

  By using the correction map 242, the first ignition timing estimating unit 101o delays θci temporarily set based on the minimum value of the first specific frequency output value in a direction in which the θmfb10-50 period is short. Can be corrected. θci1 approaches the actual CI combustion start timing θci, and the accuracy of the first ignition timing estimation unit 101o in estimating θci increases.

  FIG. 25 illustrates still another correction map 251. The first ignition timing estimation unit 101o may estimate the CI combustion start timing θci1 using the correction map 251 shown in FIG.

  The correction map 251 defines the relationship between the θmfb10-50 period, θmfb10, and the correction amount θcor. The vertical axis of the correction map 251 is θmfb10, where θmfb10 is a retard angle as going upward and θmfb10 is an advanced angle as going downward. The horizontal axis of the correction map 251 is the θmfb10-50 period, where the period is longer as going to the right and shorter as going to the left. The correction map 251 sets the correction value θcor to three, that is, a large retardation amount, a medium retardation amount, and a small retardation amount, according to the θmfb10-50 period and θmfb10. Although not shown, the correction map 251 may set the correction value θcor more finely. The correction map 251 is also stored in the memory 102.

  The correction unit 101w corrects the temporarily set θci based on the value of the θmfb10-50 period calculated by the combustion parameter calculation unit 101t, θmfb10, and the correction map 251. Specifically, if the θmfb10-50 period is short and θmfb10 is on the advance side, the correction value θcor is set to a large retardation amount, and if the θmfb10-50 period is long and θmfb10 is on the retard side, the correction is made. The value θcor is set such that the retard amount is small, and in the middle between them, the correction value θcor is set so that the retard amount is medium. As a result, θci1 comes closer to the actual CI combustion start timing θci, and the accuracy of the first ignition timing estimator 101o in estimating θci increases.

  If the θmfb10-50 period is longer than a predetermined period and θmfb10 is on the retard side than a predetermined period, as described above, CI combustion in the SPCCI combustion does not occur in the first place. Do not estimate θci.

(Other embodiments)
Note that the technology disclosed herein is not limited to being applied to the engine 1 having the above-described configuration. Various configurations can be adopted for the configuration of the engine 1.

1 engine 10 ECU (control unit)
11 Cylinder 13 Cylinder head 17 Combustion chamber 101i Ignition controller 101l First bandpass filter 101m Second bandpass filter 101o First ignition timing estimator 101p Second ignition timing estimator 101q Selector 101u Adjuster 101v Adjuster 101w Corrector 1022 Target θci storage unit (target timing storage unit)
1024 threshold value storage unit 25 spark plug (ignition unit)
3 piston 6 injector (fuel injection unit)
SW1 air flow sensor (measurement unit)
SW2 1st intake air temperature sensor (measuring unit)
SW3 1st pressure sensor (measurement part)
SW4 Second intake air temperature sensor (measuring unit)
SW5 Intake pressure sensor (measuring unit)
SW6 In-cylinder pressure sensor (measuring unit)
SW7 Exhaust gas temperature sensor (measuring unit)
SW8 linear _{O 2} sensor (measuring section)
SW9 lambda _{O 2} sensor (measuring section)
SW10 Water temperature sensor (measuring unit)
SW11 Crank angle sensor (measuring unit)
SW12 accelerator opening sensor (measuring unit)
SW13 Intake cam angle sensor (measuring unit)
SW14 Exhaust cam angle sensor (measuring unit)
SW15 EGR differential pressure sensor (measuring unit)
SW16 Fuel pressure sensor (measuring unit)
SW17 Third intake air temperature sensor (measuring unit)

Claims (7)

A combustion chamber of an engine formed by a cylinder, a piston reciprocating in the cylinder, and a cylinder head closing one end of the cylinder;
A fuel injection unit attached to the cylinder head and injecting fuel to be supplied into the combustion chamber;
An ignition unit disposed facing the combustion chamber and igniting an air-fuel mixture in the combustion chamber,
A measurement unit that measures parameters related to the operation of the engine,
Each of the ignition unit, the fuel injection unit, and the measurement unit is connected, receives a measurement signal from the measurement unit, performs an operation, and outputs a signal to the ignition unit and the fuel injection unit. And a control unit,
The air-fuel mixture, after a part of the air-fuel mixture starts combustion with flame propagation by the forced ignition of the ignition unit, the remaining unburned air-fuel mixture burns by self-ignition,
The control unit outputs an ignition signal to the ignition unit before the target timing so that the unburned mixture self-ignites at the target timing, and the combustion unit measures the combustion measured by the measurement unit. Based on the in-cylinder pressure parameter related to the pressure in the room, the in-cylinder pressure parameter has a timing exceeding a threshold, an ignition timing estimating unit that estimates the timing at which the unburned mixture self-ignites,
The control device for a compression ignition engine, further comprising: a correction unit that retards the estimated timing when the speed of the combustion accompanied by the flame propagation is high. The control device for a compression ignition engine according to claim 1,
The control device for a compression ignition type engine, wherein the measuring unit is provided facing the combustion chamber and includes an in-cylinder pressure sensor for measuring a pressure in the combustion chamber. The control device for a compression ignition type engine according to claim 2,
The control unit has a first bandpass filter that passes a signal having a frequency equal to or higher than a first frequency and equal to or lower than a second frequency among measurement signals of the in-cylinder pressure sensor,
The ignition timing estimation unit estimates a timing at which the measurement signal passing through the first band-pass filter exceeds a first threshold value as a timing at which the unburned air-fuel mixture self-ignites,
The control device for a compression ignition engine, wherein the first frequency and the second frequency are set in a range of 0.5 kHz or more and 4.0 kHz or less. The control device for a compression ignition engine according to claim 3,
The control unit calculates a heat generation rate for a predetermined period from the forced ignition based on the parameter measured by the measurement unit,
The ignition timing estimating unit is a control device for a compression ignition engine that increases a correction amount in a retard direction when the heat generation rate is high as compared with when the heat generation rate is low. The control device for a compression ignition engine according to claim 3,
The control unit calculates a crank angle period until the mass combustion ratio reaches a predetermined value based on the parameter measured by the measurement unit,
The ignition timing estimating unit is a control device for a compression ignition engine that increases a correction amount in a retard direction when the crank angle period is short compared to when the crank angle period is long. The control device for a compression ignition type engine according to claim 2,
The control unit has a first bandpass filter that passes a signal having a frequency equal to or higher than a first frequency and equal to or lower than a second frequency from a measurement signal of the in-cylinder pressure sensor,
The ignition timing estimating unit has a first ignition timing estimating unit that estimates a timing at which the measurement signal passing through the first bandpass filter exceeds a first threshold value as a timing at which the unburned air-fuel mixture self-ignites,
The first frequency and the second frequency are set in a range of 0.5 kHz or more and 4.0 kHz or less,
The control unit further includes a second bandpass filter that passes a component having a frequency equal to or higher than a third frequency and equal to or lower than a fourth frequency from the measurement signal of the in-cylinder pressure sensor,
The ignition timing estimating unit has a second ignition timing estimating unit that estimates a timing at which the measurement signal passing through the second bandpass filter exceeds a second threshold value as a timing at which the unburned air-fuel mixture self-ignites,
The third frequency and the fourth frequency are set in a range of 5.5 kHz or more and 8.0 kHz or less,
The control unit further includes a selection unit that selects one of the timings estimated by the first ignition timing estimation unit and the second ignition timing estimation unit,
When only one of the first ignition timing estimation unit and the second ignition timing estimation unit estimates the timing, the selection unit estimates the timing as the timing at which the unburned air-fuel mixture self-ignites,
The selecting unit also sets the timing on the relatively advanced side to the timing at which the unburned mixture self-ignites when both the first ignition timing estimation unit and the second ignition timing estimation unit estimate the timing. The control device for the compression ignition engine to be estimated. A combustion chamber of an engine formed by a cylinder, a piston reciprocating in the cylinder, and a cylinder head closing one end of the cylinder;
A fuel injection unit attached to the cylinder head and injecting fuel to be supplied into the combustion chamber;
An ignition unit disposed facing the combustion chamber and igniting an air-fuel mixture in the combustion chamber,
A measurement unit disposed facing the combustion chamber and including at least a cylinder pressure sensor for measuring a pressure in the combustion chamber, and measuring a parameter related to operation of the engine,
Each of the ignition unit, the fuel injection unit, and the measurement unit is connected, receives a measurement signal from the measurement unit, performs an operation, and outputs a signal to the ignition unit and the fuel injection unit. And a control unit,
The air-fuel mixture, after a part of the air-fuel mixture starts combustion with flame propagation by the forced ignition of the ignition unit, the remaining unburned air-fuel mixture burns by self-ignition,
The control unit includes:
A target timing storage unit that stores a target timing at which the unburned air-fuel mixture self-ignites,
An ignition control unit that outputs an ignition signal to the ignition unit before the target timing so that the unburned mixture self-ignites at the target timing based on the target timing of the target timing storage unit. Have
The control unit also includes
Among the measurement signals of the in-cylinder pressure sensor, a band-pass filter that passes a signal in a specific frequency band,
A threshold storage unit for storing a threshold,
An ignition timing estimating unit that estimates the timing at which the value of the measurement signal passing through the bandpass filter exceeds the threshold value as the timing at which the unburned air-fuel mixture self-ignites,
A correction unit that corrects the timing estimated by the ignition timing estimation unit based on a measurement signal from the measurement unit,
The control unit for a compression ignition engine that retards the estimated timing when the speed of the combustion accompanied by the flame propagation is high.