Engine combustion control device

Abstract

To suppress an occurrence of abnormal combustion in advance while improving thermal efficiency.SOLUTION: An engine combustion control device: applies bias voltage between plug electrodes of an ignition plug in synchronization with an initiation of a low-temperature oxidation reaction of a mixed gas; calculates an ion current characteristic value (maximum charge amount) correlated with an in-cylinder gas state, which represents a state of the mixed gas inside a cylinder, based on ion current detected during a detection period, from the time the bias voltage starts being applied until a predetermined time prior to an ignition timing; determines the ignition timing in an identical cycle with the cycle in which the ion current characteristic value is calculated based on the ion current characteristic amount; and ignites the mixed gas at the determined ignition timing.SELECTED DRAWING: Figure 12

Description

Technical Field

[0001] The present invention relates to a combustion control device applied to an engine including a cylinder and a spark plug that ignites an air-fuel mixture in the cylinder.

Background Art

[0002] A combustion control device for an engine disclosed in Patent Document 1 below is known. The combustion control device of this Patent Document 1 includes an ignition coil including a primary coil and a secondary coil, a switching element that controls energization to the primary coil, a spark plug that discharges in response to an induced voltage of the secondary coil, an ion signal detection circuit that detects an ion current flowing between plug electrodes of the spark plug, and an ECU electrically connected to the switching element and the ion signal detection circuit. When the ignition timing arrives and the energization period to the primary coil ends, the ECU calculates a cumulative value of the ion current from the start of energization to the primary coil, and determines whether pre-ignition (premature ignition) has occurred based on the calculated cumulative value.

Prior Art Document

Patent Document

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] In the above Patent Document 1, when the occurrence of pre-ignition is determined, appropriate control for suppressing pre-ignition is to be executed. However, in Patent Document 1, pre-ignition is determined based on the cumulative value of the ion current detected during the period from the start of energization to the end of energization (in other words, the ignition timing) of the primary coil. Therefore, at the time when pre-ignition is determined, ignition has already ended, and it is considered that combustion has progressed considerably. For this reason, there has been a problem that substantial pre-ignition suppression control cannot be executed unless at least the next cycle is reached. Also, in order to avoid the surfacing of such problems, it is conceivable to design the engine on the safe side so that abnormal combustion such as pre-ignition does not occur as much as possible. However, if this is done, there is a problem that the thermal efficiency of the engine decreases.

[0005] The present invention has been made in view of the above circumstances, and an object thereof is to provide an engine combustion control device capable of suppressing the occurrence of abnormal combustion while improving the thermal efficiency.

Means for Solving the Problems

[0006] As a solution to the above problems, the present invention is a combustion control device applied to an engine including a cylinder and a spark plug that ignites an air-fuel mixture in the cylinder, the combustion control device including: an ignition coil including a primary coil and a secondary coil; an igniter that induces a high voltage in the secondary coil through energization and de-energization of the primary coil, and causes a discharge to occur between the plug electrodes of the spark plug by the induced high voltage; a bias voltage generation unit that applies a bias voltage for detecting an ion current generated between the plug electrodes due to ions in the cylinder between the plug electrodes; an ion current detection unit that detects the ion current; an ignition control unit that controls the bias voltage generation unit so that the bias voltage is applied between the plug electrodes in accordance with the start of a low-temperature oxidation reaction occurring in the cylinder as the air-fuel mixture is compressed, and controls the igniter so that a discharge occurs between the plug electrodes at an ignition timing set at a time later than the start of the low-temperature oxidation reaction; and an arithmetic unit that calculates an ion current feature amount correlated with the in-cylinder gas state, which is the state of the air-fuel mixture in the cylinder, based on the ion current detected by the ion current detection unit during a detection period that is a period from the time when the application of the bias voltage starts to a predetermined time earlier than the ignition timing. The ignition control unit determines the ignition timing in the same cycle as the cycle in which the ion current feature amount is calculated based on the ion current feature amount, and controls the igniter so that the discharge is performed at the determined ignition timing (Claim 1).

[0007] According to the present invention, an ion current feature amount correlated with the in-cylinder gas state is calculated based on the ion current detected before the ignition timing, and the ignition timing is determined based on the calculated ion current feature amount. Therefore, the ignition timing can be adjusted to an appropriate timing considering the in-cylinder gas state. That is, the ion current feature amount correlated with the in-cylinder gas state can be used as an index indicating the likelihood of knocking that changes depending on the in-cylinder gas state. Therefore, according to the present invention in which the ignition timing is determined based on the ion current feature amount, the air-fuel mixture can be ignited at an ignition timing with as high a thermal efficiency as possible within a range where knocking does not occur. Moreover, since the ion current feature amount is calculated before the ignition timing, the ignition timing in the same cycle as the cycle in which the calculation is performed can be adjusted in advance to an appropriate timing considering knocking. Thereby, it is possible to suppress the occurrence of knocking while improving the thermal efficiency.

[0008] Preferably, the ion current feature amount is the maximum charge amount which is the maximum value of the charge amount between the plug electrodes during the detection period, and the ignition control unit sets the ignition timing more advanced as the maximum charge amount is smaller (Claim 2).

[0009] In this aspect, by utilizing the knowledge that knocking is less likely to occur when the maximum charge amount is small, an ignition timing that can achieve both suppression of knocking and improvement of thermal efficiency can be appropriately obtained from the maximum charge amount.

[0010] The ion current feature amount may be the maximum ion current which is the maximum value of the ion current during the detection period. In this case, it is preferable that the ignition control unit sets the ignition timing more advanced as the maximum ion current is smaller (Claim 3).

[0011] Similarly, with this aspect, an ignition timing that can achieve both suppression of knocking and improvement of thermal efficiency can be appropriately obtained.

[0012] Preferably, the ignition control unit controls the bias voltage generation unit such that the bias voltage gradually increases from the start of the low-temperature oxidation reaction and then gradually decreases (Claim 4).

[0013] In this way, when the bias voltage is changed in a mountain shape, the amount of charge between the plug electrodes can be changed with a similar mountain-shaped tendency. This facilitates the calculation of the above-mentioned maximum charge amount, which is the maximum value of the charge amount, or the above-mentioned maximum ion current, which is the maximum value of the ion current. Further, since the ion current that increases with the start of the low-temperature oxidation reaction is further amplified by the gradual increase of the bias voltage, the sensitivity of the maximum charge amount or the maximum ion current that changes depending on the difference in the in-cylinder gas state can be improved. Thereby, the ignition timing suitable for the in-cylinder gas state can be appropriately obtained.

[0014] Preferably, the bias voltage generation unit is a capacitor device connected to the secondary coil (Claim 5).

[0015] In this aspect, the bias voltage can be appropriately controlled by adjusting the charge amount of the capacitor device.

[0016] Preferably, the ignition control unit controls the bias voltage generation unit such that the application start of the bias voltage is advanced as the engine speed increases (Claim 6).

[0017] Alternatively, the ignition control unit controls the bias voltage generation unit such that the application start of the bias voltage is advanced as the engine load increases (Claim 7).

[0018] In these aspects, the bias voltage can be applied in accordance with the start of the low-temperature oxidation reaction, which advances as the engine load or engine speed increases, and the ion current that increases with the start of the low-temperature oxidation reaction can be appropriately detected.

[0019] Preferably, the combustion control device further includes an injector that injects fuel into the cylinder before the start of the low-temperature oxidation reaction, an abnormal combustion determination unit that determines the occurrence of pre-ignition based on the ion current detected by the ion current detection unit, and an injection control unit that causes the injector to inject additional fuel when the occurrence of pre-ignition is determined (Claim 8).

[0020] In this aspect, since the latent heat of vaporization of the additionally injected fuel decreases the internal temperature of the cylinder, the progress of the combustion of the air-fuel mixture can be suppressed, and the influence caused by pre-ignition can be reduced.

Advantages of the Invention

[0021] As described above, according to the combustion control device for an engine of the present invention, it is possible to suppress the occurrence of abnormal combustion while improving the thermal efficiency.

Brief Description of the Drawings

[0022]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Figure 8

Figure 9

Figure 10

Figure 11

Figure 12

Figure 13

Figure 14A

Figure 14B

Figure 15

Mode for Carrying Out the Invention

[0023] (1) Overall configuration of the engine FIG. 1 is a system diagram showing the overall configuration of an engine to which a combustion control device according to an embodiment of the present invention is applied. The engine shown in this figure is a four-cycle spark ignition engine mounted on a vehicle as a power source for traveling. The engine includes an engine body 1, an intake passage 20 and an exhaust passage 30 connected to the engine body 1, and an ECU 40 that controls each part of the engine. In this embodiment, the upper and lower directions are defined with reference to FIG. 1, but this is for convenience of explanation and is not intended to limit the installation posture of the engine.

[0024] The engine body 1 includes a cylinder block 3 and a cylinder head 4 that define a cylinder 2 inside, and a piston 5 that is reciprocally accommodated in the cylinder 2. A combustion chamber C is formed above the piston 5. Although only one cylinder 2 is shown in FIG. 1, the engine body 1 can be a multi-cylinder type having a plurality of cylinders 2.

[0025] The cylinder head 4 is equipped with an ignition circuit 10, an ignition plug 11, and an injector 15. The injector 15 is an injection valve that injects fuel (e.g., gasoline fuel) supplied from a fuel tank (not shown) through a fuel supply pipe 15a toward the combustion chamber C. The ignition plug 11 is a plug that ignites the air-fuel mixture generated by the fuel injected from the injector 15 into the combustion chamber C mixing with air. The ignition circuit 10 is a circuit that applies a high voltage for generating an ignition spark to the ignition plug 11. When the air-fuel mixture in the combustion chamber C burns triggered by the ignition by the ignition plug 11, the piston 5 reciprocates under the expansion force of the combustion.

[0026] Figure 2 is a diagram showing an enlarged view of the tip of the ignition plug 11. As shown in this figure, the ignition plug 11 includes a plug body 12, a center electrode 13, and a ground electrode 14. The plug body 12 has a cylindrical shape extending along the cylinder axis X1 which is the central axis of the cylinder 2, and is attached to the cylinder head 4 with its tip exposed to the combustion chamber C. The center electrode 13 is formed to protrude downward from the center of the tip of the plug body 12. The ground electrode 14 is formed to extend downward while bending in an L shape from the peripheral edge of the tip of the plug body 12. The tip of the ground electrode 14 faces the center electrode 13 with a predetermined gap G therebetween.

[0027] The center electrode 13 is connected to a secondary coil 103 (to be described later) of the ignition circuit 10. The ground electrode 14 is connected to the ground through the plug body 12 and the cylinder head 4. During ignition, a high voltage is applied from the ignition circuit 10 to the center electrode 13, causing a discharge between the center electrode 13 and the ground electrode 14. Then, the air-fuel mixture is ignited by the spark generated along with this discharge. Incidentally, hereinafter, the center electrode 13 and the ground electrode 14 may be collectively referred to as the plug electrodes 13, 14.

[0028] As shown in FIG. 1, below the piston 5, a crankshaft 9 which is the output shaft of the engine is disposed. The crankshaft 9 is rotatably supported by the cylinder block 3. The reciprocating motion of the piston 5 described above is transmitted to the crankshaft 9 via a crank mechanism including a connecting rod 7 and the like, and the crankshaft 9 is rotated.

[0029] A crank angle sensor SN1 is attached to the cylinder block 3. The crank angle sensor SN1 is a sensor that detects the crank angle which is the rotation angle of the crankshaft 9 and the engine rotation speed which is the rotation speed of the crankshaft 9.

[0030] An intake port 16 and an exhaust port 17 are formed in the cylinder head 4. The intake port 16 is a port that communicates the combustion chamber C and the intake passage 20. The exhaust port 17 is a port that communicates the combustion chamber C and the exhaust passage 30. An intake valve 18 and an exhaust valve 19 that open and close the intake port 16 and the exhaust port 17 in conjunction with the rotation of the crankshaft 9 are attached to the cylinder head 4.

[0031] The intake passage 20 is connected to one side surface of the cylinder head 4 so as to communicate with the intake port 16. The intake passage 20 is provided with an openable and closable throttle valve 21 for adjusting the intake air flow rate. An air flow sensor SN2 for detecting the intake air flow rate is provided at a position upstream of the throttle valve 21 in the intake passage 20.

[0032] The exhaust passage 30 is connected to the other side surface of the cylinder head 4 so as to communicate with the exhaust port 17. The exhaust passage 30 is provided with a catalyst (not shown) and the like for purifying harmful components in the exhaust gas.

[0033] The ECU 40 is a controller mainly composed of a microcomputer including a processor (CPU) that performs various operations, memories such as ROM and RAM, and various input / output buses. The ECU 40 receives input information from sensors provided in the engine. For example, the ECU 40 is electrically connected to the above-described crank angle sensor SN1 and air flow sensor SN2. Detection information from these respective sensors SN1, SN2, that is, information such as crank angle, engine speed, and intake air flow rate, is sequentially input to the ECU 40.

[0034] The ECU 40 also receives input information from sensors provided in the vehicle. For example, the ECU 40 is electrically connected to a vehicle speed sensor SN3 and an accelerator sensor SN4 provided in the vehicle. The vehicle speed sensor SN3 is a sensor that detects the traveling speed of the vehicle, that is, the vehicle speed, and the accelerator sensor SN4 is a sensor that detects the opening degree of the accelerator pedal 50 operated by a driver who drives the vehicle, that is, the accelerator opening degree. Detection information (vehicle speed and accelerator opening degree) from these respective sensors SN3, SN4 is sequentially input to the ECU 40.

[0035] The ECU 40 controls each part of the engine while executing various operations and determinations based on the input information from the above-described respective sensors (such as SN1 to SN4). For example, the ECU 40 is electrically connected to the ignition circuit 10 and the injector 15, and appropriately outputs control signals to these respective elements.

[0036] As functional elements related to the above control, the ECU 40 includes an ignition control unit 41, an injection control unit 42, an arithmetic unit 43, an ion current detection unit 44, and a storage unit 45. The ignition control unit 41 is a module that controls the ignition operation of the spark plug 11 through energization control of the ignition circuit 10. The injection control unit 42 is a module that controls the injection operation of the injector 15. The arithmetic unit 43 is a control module that performs various calculations and determinations necessary to determine the content of control by the ignition control unit 41 and the injection control unit 42. The ion current detection unit 44 is a module that detects the ion current generated between the plug electrodes 13 and 14 of the spark plug 11 based on a signal from the ignition circuit 10. The storage unit 45 is a module that stores various data necessary for control or calculation.

[0037] (2) Configuration of the Ignition Circuit FIG. 3 is a circuit diagram showing the configuration of the ignition circuit 10. As shown in this figure, the ignition circuit 10 includes an ignition coil 101 that applies a high voltage for discharge to the spark plug 11, and a coil driver 105 that realizes operations such as energization and de-energization of the ignition coil 101.

[0038] The ignition coil 101 includes a primary coil 102 connected to a battery (not shown) and a secondary coil 103 connected to the center electrode 13 of the spark plug 11. The primary coil 102 and the secondary coil 103 are wound around a common core (iron core).

[0039] The coil driver 105 includes an igniter 106, a capacitor device 107, a current amplification circuit 110, and a Zener diode 111. The igniter 106 is a transistor interposed between the primary coil 102 and the ground (Gnd), with its collector connected to the primary coil 102 and its emitter connected to the ground. The capacitor device 107 is interposed between the secondary coil 103 and the ground. The Zener diode 111 is connected in parallel with the capacitor device 107. The current amplification circuit 110 is interposed between a terminal 121 connected to the ion current detection unit 44 of the ECU 40 and a terminal 125 on the negative electrode side of the capacitor device 107.

[0040] A resistor 115 and a diode 116 are provided in parallel between the secondary coil 103 and the coil driver 105. That is, the secondary coil 103 is connected to the terminal 122 of the coil driver 105 via the resistor 115 or the diode 116.

[0041] The igniter 106 induces a high voltage in the secondary coil 103 through energization and de-energization of the primary coil 102, and causes a discharge to occur between the plug electrodes 13 and 14 of the spark plug 11 due to the induced high voltage.

[0042] Specifically, the igniter 106 is switched to the ON state before the ignition timing arrives. As a result, the flow of current between the collector and the emitter is allowed, and the primary coil 102 is energized. That is, a current flows from the battery through the primary coil 102 and the igniter 106 to the ground. Then, when the ignition timing arrives, the igniter 106 is switched from the ON state to the OFF state. As a result, the flow of current between the collector and the emitter is prohibited, and the energization of the primary coil 102 is stopped. Then, due to electromagnetic induction accompanying the stop of the energization, a high voltage corresponding to the winding ratio of the primary coil 102 and the secondary coil 103 is induced in the secondary coil 103.

[0043] When a high voltage is induced in the secondary coil 103 as described above, a discharge due to the high voltage occurs between the plug electrodes 13 and 14 of the spark plug 11, and a spark is generated. The air-fuel mixture in the combustion chamber C is ignited by this spark. FIG. 4 is a diagram showing the flow of current (discharge current) during this discharge. As shown in this figure, the discharge current flows from the center electrode 13 of the spark plug 11 to the ground via the secondary coil 103, the diode 116, the Zener diode 111, etc.

[0044] The capacitor device 107 applies a bias voltage for detecting an ion current to the ignition plug 11. That is, ions are generated inside the cylinder 2, namely, in the combustion chamber C, along with the reaction of the air-fuel mixture. During the generation of such ions, when a bias voltage is applied to the ignition plug 11, an ion current as shown in FIG. 5 is generated according to the potential difference between the plug electrodes 13 and 14 due to the bias voltage. The ion current flows toward the center electrode 13 of the ignition plug 11 via the current amplification circuit 110, the capacitor device 107, the resistor 115, the secondary coil 103, etc. The capacitor device 107 functions as a voltage supply source (bias voltage generation unit) that applies a bias voltage leading to the generation of such an ion current to the ignition plug 11.

[0045] The capacitor device 107 includes a capacitor 108 and a voltage control circuit 109 arranged in parallel. The capacitor 108 has a pair of positive and negative chargeable electrode plates. The voltage control circuit 109 supplies charge to the capacitor 108 for generating the above-described bias voltage. That is, when charge is supplied from the voltage control circuit 109 and the capacitor 108 is charged, a potential difference is generated between the electrode plates of the capacitor 108, and this potential difference also generates a potential difference between the plug electrodes 13 and 14 of the ignition plug 11, which functions as a bias voltage. The charge amount of the capacitor 108, in other words, the magnitude of the bias voltage, is increased or decreased according to the charge supply amount from the voltage control circuit 109. In other words, the voltage control circuit 109 can control the bias voltage through the charge supply amount to the capacitor 108.

[0046] The current amplification circuit 110 is a circuit that amplifies the ion current. The current amplified by this current amplification circuit 110 is detected by the ion current detection unit 44.

[0047] The coil driver 105 has a terminal 123 that receives a control signal from the ignition control unit 41 of the ECU 40. The above-described igniter 106 and capacitor device 107 (voltage control circuit 109) operate in response to the control signal input through the terminal 123. That is, the ignition control unit 41 controls energization and de-energization of the primary coil 102 through ON / OFF control of the igniter 106, and controls the bias voltage of the spark plug 11 through adjustment of the charge amount of the capacitor device 107 (capacitor 108).

[0048] (3) Determination of ignition timing based on ion current Here, in a spark ignition engine as in this embodiment, there is a need to ignite the air-fuel mixture at a timing with as high a thermal efficiency as possible. However, depending on the state of the air-fuel mixture in the cylinder 2 (combustion chamber C), that is, the in-cylinder gas state, knocking, which is abnormal combustion in which end gas (unburned gas) auto-ignites during combustion of the air-fuel mixture, may occur unless the ignition timing is retarded compared to the timing at which the thermal efficiency is best. For this reason, in a spark ignition engine, it is desired to achieve both improvement in thermal efficiency and suppression of knocking, in other words, to ignite the air-fuel mixture at a timing with as high a thermal efficiency as possible within a range where knocking does not occur.

[0049] Therefore, the inventors of the present application focused on the correlation between the in-cylinder gas state, that is, the state of the air-fuel mixture in the combustion chamber C, and the above-described ion current, and came up with the idea of predicting the in-cylinder gas state based on the ion current and then determining the ignition timing. That is, the ion current is detected before ignition by the spark plug 11, and a predetermined feature amount (maximum charge amount qmax, described later) that represents the in-cylinder gas state and thus the likelihood of knocking is calculated from the detected ion current, and further, the ignition timing in the current cycle is determined based on the calculated feature amount. This will be described in detail below.

[0050] (Principle of the correlation between ion current and in-cylinder gas state) FIG. 6 is a schematic diagram for explaining the relationship between the ion current and the in-cylinder gas state. As shown in this figure, the ion current depends on the inter-electrode resistance R, which is the resistance between the plug electrodes 13 and 14 of the ignition plug 11. That is, under the condition that the potential difference between the plug electrodes 13 and 14, i.e., the bias voltage, is the same, the larger the inter-electrode resistance R, the smaller the ion current.

[0051] The inter-electrode resistance R depends on the path length L of the charged particles (ions or electrons) moving between the plug electrodes 13 and 14, the gas temperature T in the cylinder 2, and the gas density ρ in the cylinder 2. That is, the inter-electrode resistance R increases as the path length L becomes longer, decreases as the gas temperature T becomes higher, and increases as the gas density ρ becomes larger.

[0052] The gas density ρ depends on the gas temperature T, the gas pressure P, and the gas composition represented by the air excess ratio λ of the air-fuel mixture. That is, the gas density ρ decreases as the gas temperature T becomes higher, increases as the gas pressure P becomes higher, and decreases as the air excess ratio λ becomes larger.

[0053] The path length L depends on the gas flow velocity u. FIG. 7 is a schematic diagram for explaining the relationship between the path length L and the gas flow velocity u. As shown in this figure, the larger the gas flow velocity u, the longer the path length L. That is, under the condition that the gas flow velocity u is large, due to the influence of the gas flow velocity u, the movement path of the charged particles curves, and as a result, the path length L becomes longer. Note that the path length L under the condition that the gas flow velocity u is very small is equal to the shortest distance between the plug electrodes 13 and 14. As the gas flow velocity u increases from this state, the path length L becomes longer.

[0054] From the above, it is understood that the ion current depends on the gas flow velocity u, the gas temperature T, the gas pressure P, and the air excess ratio λ in the cylinder 2. This suggests that the in-cylinder gas state can be estimated by examining the ion current.

[0055] (Verification experiment) As a verification experiment based on the above considerations, a discharge simulation was conducted to confirm the correlation between the ion current and the in-cylinder gas state. That is, using a model that modeled the behavior of charged particles (ions or electrons) between the plug electrodes 13 and 14 under various conditions with different in-cylinder gas states, a virtual discharge was caused between the plug electrodes 13 and 14, and the ion current generated during the period until the discharge was investigated. Here, as a value related to the ion current, the charge quantity q of the space charge between the plug electrodes 13 and 14 was identified, and how the charge quantity q changes with the gas flow velocity u, gas temperature T, gas pressure P, and air excess ratio λ was investigated. Note that as the charge quantity q, the absolute value of the space charge (negative) of electrons was used.

[0056] Figure 8 is a graph showing the change of the bias voltage applied between the plug electrodes 13 and 14 with the crank angle for generating an ion current. In this graph, the change of the bias voltage is shown together with the charge signal which is the energization signal to the primary coil 102. Also, the first crank angle CA1 and the third crank angle CA3 on the horizontal axis of this graph are the crank angles when the energization (charging) to the primary coil 102 starts and stops, respectively. That is, the period from the first crank angle CA1 to the third crank angle CA3 is the charging period during which the energization to the primary coil 102 is performed. As described above, since the stop of the energization (charging) to the primary coil 102 causes a discharge between the plug electrodes 13 and 14, the third crank angle CA3 which is the end point of the charging period is equivalent to the timing of the discharge, that is, the ignition timing.

[0057] As shown in FIG. 8, in this verification experiment, a bias voltage was applied from the start to the end of the charging period. That is, the application of the bias voltage was started at the first crank angle CA1 at which the charging period started, and the application of the bias voltage was stopped at the third crank angle CA3 at which the charging period ended. Further, in this verification experiment, the bias voltage was changed in a mountain shape. That is, from the first crank angle CA1 to the second crank angle CA2 between the first crank angle CA1 and the third crank angle CA3, the bias voltage was gradually increased, and from the second crank angle CA2 to the third crank angle CA3, the bias voltage was gradually decreased. In other words, in this verification experiment, the bias voltage was applied so that the voltage changed along a mountain-shaped waveform having a peak near the center of the charging period.

[0058] Next, the change in the charge amount q that occurred when the bias voltage was applied as described above was examined, and the results shown in FIG. 9 were obtained. Specifically, under various conditions where the gas flow rate u, the gas temperature T, the gas pressure P, and the air excess ratio λ were different, a bias voltage that changed in a mountain shape as shown in FIG. 8 was applied to the spark plug 11, and the change in the charge amount q at that time was examined. That is, graph (a) in FIG. 9 is a graph showing the change in the charge amount q under a plurality of conditions where only the gas flow rate u is different (T, P, and λ are constant), graph (b) is a graph showing the change in the charge amount q under a plurality of conditions where only the gas temperature T is different (u, P, and λ are constant), graph (c) is a graph showing the change in the charge amount q under a plurality of conditions where only the gas pressure P is different (u, T, and λ are constant), and graph (d) is a graph showing the change in the charge amount q under a plurality of conditions where only the air excess ratio λ is different (u, T, and P are constant).

[0059] As shown in Fig. 9, the charge quantity q varies along a mountain-shaped waveform with a peak in the middle of the charging period, similar to the bias voltage. Also, the value of the peak, i.e., the maximum charge quantity qmax which is the maximum value of the charge quantity q, varies depending on the respective parameters of u, T, P, and λ described above. For example, as shown in graph (b) of Fig. 9, the maximum charge quantity qmax increases as the gas temperature T increases. Similarly, as shown in graph (d), the maximum charge quantity qmax increases as the air excess ratio λ increases. On the other hand, the relationship between the gas flow velocity u and the maximum charge quantity qmax is the opposite. That is, as shown in graph (a), the maximum charge quantity qmax decreases as the gas flow velocity u increases. Regarding the gas pressure P, as shown in graph (c), it has almost no influence on the maximum charge quantity qmax.

[0060] Fig. 10 is a graph summarizing the above results and shows the relationship between the gas state quantity and the maximum charge quantity qmax. The gas state quantity on the horizontal axis is the value obtained by dividing the product of the gas density ρ and the gas flow velocity u by the gas temperature T (ρ·u / T), which is a parameter that varies depending on any of u, T, P, and λ described above. In other words, the graph in Fig. 10 is obtained by reducing the relationship between U, T, P, λ and the charge quantity q shown in Fig. 9 to the relationship between the gas state quantity (ρ·u / T) and the maximum charge quantity qmax. In this graph, the difference in the plot shapes represents the difference in the parameters to be varied. For example, the plurality of plots ● representing the relationship between u and qmax show the respective values of qmax when only the gas flow velocity u is varied while the other parameters (P, T, λ) are constant. The same applies to the other plots representing the relationship between the respective parameters of T, P, λ and qmax. Note that this graph also shows the results obtained under conditions different from those in Fig. 9. From the degree of distribution of the plots in the horizontal axis direction, it can be seen that the maximum charge quantity qmax is most affected by the gas flow velocity u. This is consistent with the fact that the change in the waveform of the charge quantity q due to the gas flow velocity u shown in graph (a) of Fig. 9 is larger than that in the other graphs (b) to (c).

[0061] As shown in Fig. 10, the maximum charge amount qmax decreases as the gas state quantity increases. Here, from the definition of the gas state quantity (ρ·u / T), an increase in the gas state quantity means that the gas flow velocity u is high, the gas temperature T is low, or the gas density ρ is high. However, according to the results of Fig. 9, the influence of the gas flow velocity u and the gas temperature T on the maximum charge amount qmax is considerably greater than the influence of the gas pressure P and the air excess ratio λ (especially the gas pressure P), which are the fluctuation factors of the gas density ρ, on the maximum charge amount qmax. From this, it is considered that the gas state quantity mainly changes according to the gas flow velocity u and the gas temperature T, which brings about a change in the maximum charge amount qmax. And both the high gas flow velocity u and the low gas temperature T act in the direction of suppressing knocking. Therefore, the maximum charge amount qmax mainly determined by such parameters (u, T) can be used as an index representing the ease of occurrence of knocking. That is, it can be determined that the smaller the maximum charge amount qmax, the larger the gas state quantity and the less likely knocking is to occur, and the larger the maximum charge amount qmax, the smaller the gas state quantity and the more likely knocking is to occur.

[0062] The above findings mean that the ignition timing that can achieve the highest possible thermal efficiency within the range where knocking does not occur can be determined based on the maximum charge amount qmax. Fig. 11 is a graph showing the relationship between the maximum charge amount qmax determined from such a perspective and the model ignition timing. The model ignition timing here refers to the advance limit of the ignition timing determined by performing simulations for each in-cylinder gas state corresponding to each value of the maximum charge amount qmax. That is, if the ignition timing is at the model ignition timing or a retarded timing than this, the occurrence of knocking is highly likely to be prevented. On the other hand, if the ignition timing is advanced more than the model ignition timing, the possibility of knocking occurring increases.

[0063] As shown in Fig. 11, the model ignition timing is set on the retard side as the maximum charge amount qmax increases. That is, since the maximum charge amount qmax increases in the in-cylinder gas state where knocking is likely to occur, it is necessary to retard the ignition timing to suppress knocking under the condition that the maximum charge amount qmax is large. From such a viewpoint, the model ignition timing is set on the retard side as the maximum charge amount qmax increases.

[0064] (4) Actual control Next, the details of the engine combustion control performed based on the above findings will be described. As shown below, in the present embodiment, the standard ignition timing Igr determined from the operating state of the engine is compared with the model ignition timing Igm determined from the maximum charge amount qmax which is a characteristic amount of the ion current, and the ignition timing is determined from the result. That is, if the model ignition timing Igm is on the advanced side of the standard ignition timing Igr, the ignition timing is advanced more than the standard ignition timing Igr, and if the model ignition timing Igm is on the retard side of the standard ignition timing Igr, the ignition timing is retarded more than the standard ignition timing Igr. Moreover, since such an operation for determining the ignition timing is performed during the energization of the primary coil 102 (during the charge period), it is possible to correct the ignition timing within the same cycle as the cycle in which the operation was performed. This will be described in detail below.

[0065] Figs. 12 and 13 are flowcharts showing the details of the combustion control performed by the ECU 40 during the operation of the engine. When this combustion control starts, the arithmetic unit 43 of the ECU 40 acquires various information regarding the engine (step S1). Specifically, the arithmetic unit 43 acquires information such as the crank angle, engine speed, intake air flow rate, vehicle speed, and accelerator opening from the detection values of the crank angle sensor SN1, air flow sensor SN2, vehicle speed sensor SN3, and accelerator sensor SN4. Also, based on the acquired accelerator opening and vehicle speed, the load (required torque) of the engine is specified.

[0066] Next, the arithmetic unit 43 determines the fuel injection amount and injection timing (step S2). That is, the arithmetic unit 43 determines the injection amount, which is the amount of fuel to be injected from the injector 15, based on the intake air flow rate and engine load acquired in step S1 above. Further, the arithmetic unit 43 determines the injection timing, which is the timing at which fuel injection from the injector 15 should start, based on the determined injection amount and the engine speed acquired in step S1 above.

[0067] Next, the arithmetic unit 43 determines the standard ignition timing Igr based on the operating conditions including the engine load and engine speed acquired in step S1 above (step S3). The standard ignition timing Igr is determined in advance based on prior experiments or the like for each operating condition of the engine as the ignition timing that maximizes the thermal efficiency as much as possible within the range where knocking does not occur. For example, the storage unit 45 of the ECU 40 stores in advance map data that defines the relationship between the operating conditions of the engine (load, speed, etc.) and the standard ignition timing Igr so that the standard ignition timing Igr can be derived from the operating conditions of the engine. In this case, the arithmetic unit 43 determines the standard ignition timing Igr from the operating conditions of the engine by referring to the stored map data. Note that the data stored in the storage unit 45 for deriving the standard ignition timing Igr is not limited to map data and may be, for example, an arithmetic expression.

[0068] Next, the arithmetic unit 43 predicts the start timing of the low-temperature oxidation reaction that occurs as the air-fuel mixture is compressed (step S4). The low-temperature oxidation reaction is a slow oxidation reaction that occurs before the high-temperature oxidation reaction (substantive combustion reaction) that generates high thermal energy while accompanied by a flame, and can occur in the latter half of the compression stroke in which the combustion chamber C becomes hot. The start timing of the low-temperature oxidation reaction can be predicted from the operating conditions of the engine based on prior experiments or the like.

[0069] FIG. 14A is a graph showing the relationship between the engine load and the start timing of the low-temperature oxidation reaction, and FIG. 14B is a graph showing the relationship between the engine speed and the start timing of the low-temperature oxidation reaction. In each graph, it is assumed that the conditions other than the parameter on the horizontal axis (engine load or engine speed) are the same. As shown in FIG. 14A, the low-temperature oxidation reaction starts at an advanced timing as the engine load increases. Also, as shown in FIG. 14B, the low-temperature oxidation reaction starts at an advanced timing as the engine speed increases. In the storage unit 45, map data or arithmetic expressions corresponding to these FIGS. 14A and 14B are stored in advance. The arithmetic unit 43 predicts the start timing of the low-temperature oxidation reaction from the operating conditions (load, speed, etc.) of the engine using the stored map data or arithmetic expression.

[0070] Next, the arithmetic unit 43 determines the charge start timing, which is the timing to start energizing, that is, charging, the primary coil 102 of the ignition circuit 10 (step S5). In the present embodiment, the charge start timing is set to be substantially simultaneous with the start of the low-temperature oxidation reaction. The arithmetic unit 43 determines the charge start timing based on the start timing of the low-temperature oxidation reaction predicted in step S4 above so that charging starts at such a timing.

[0071] Next, the injection control unit 42 of the ECU 40 injects fuel into the injector 15 at the timing when the injection timing determined in step S2 above arrives (step S6). The fuel injection continues until the amount of fuel corresponding to the injection amount determined in step S2 above is completely injected.

[0072] Next, the ignition control unit 41 of the ECU 40 starts energization of the primary coil 102, that is, starts charging, at the timing when the charge start timing determined in step S5 arrives, and also starts applying a bias voltage to the spark plug 11 (step S7). That is, the ignition control unit 41 controls the igniter 106 so that energization of the primary coil 102 is started at the timing when the charge start timing substantially simultaneous with the low-temperature oxidation reaction arrives, and controls the capacitor device 107 (voltage control circuit 109) so that a bias voltage is applied between the plug electrodes 13 and 14 of the spark plug 11 at that time. Thus, in the present embodiment, the energization of the primary coil 102 and the application of the bias voltage are started at the same timing in accordance with the start of the low-temperature oxidation reaction. Applying the bias voltage in accordance with the start of the low-temperature oxidation reaction is to capture the increasing tendency of ions in the cylinder 2 that increases with the start of the low-temperature oxidation reaction.

[0073] Here, as shown in FIGS. 14A and 14B, the start timing of the low-temperature oxidation reaction is more advanced as the engine load or the engine speed is higher. From this, the timing at which the application of the bias voltage is started in step S7 is set more advanced as the engine load or the engine speed is higher.

[0074] Also, in the above step S7, similar to the verification experiment described above, a bias voltage is applied so that the voltage changes in a mountain shape (see FIG. 8). That is, in the present embodiment, the ignition control unit 41 applies a bias voltage such that the voltage gradually increases and then gradually decreases from the start to the end of charging. More specifically, the ignition control unit 41 starts applying the bias voltage at the first crank angle CA1 corresponding to the charge start timing determined in the above step S5, and gradually increases the bias voltage therefrom. Further, the ignition control unit 41 adjusts the bias voltage so that the voltage reaches a peak at the second crank angle CA2 near the center of the charge period and gradually decreases toward the third crank angle CA3 at which the charge period ends. Note that the charge end timing corresponding to the third crank angle CA3, in other words, the ignition timing, may be changed (corrected) from the standard ignition timing Igr determined in the above step S3 by the processing of each step described later. Here, however, the bias voltage is adjusted on the assumption that the standard ignition timing Igr is the charge end timing.

[0075] Next, the arithmetic unit 43 determines whether or not the detection end timing CAx, which is slightly delayed from the peak timing (the second crank angle CA2) of the bias voltage, has arrived (step S8). The detection end timing CAx is the timing at which it is expected that the data of the ion current necessary for calculating the maximum charge amount qmax, which is performed in the next step S9, will be complete, and is set at a timing that is retarded by a predetermined small crank angle from the second crank angle CA2 at which the bias voltage reaches a peak. In other words, the period from the first crank angle CA1 at which the bias voltage starts to be applied to the detection end timing CAx is the detection period of the ion current necessary for calculating the maximum charge amount qmax.

[0076] When it is determined as NO in step S8 above and it is confirmed that the detection end time CAx has not yet arrived, the arithmetic unit 43 determines whether or not the ignition of the air-fuel mixture has been detected based on the ion current detected by the ion current detector 44 of the ECU 40 (step S12). That the air-fuel mixture has ignited at this stage means that abnormal combustion, that is, pre-ignition, has occurred in which the air-fuel mixture self-ignites before ignition by the ignition plug 11. The occurrence of pre-ignition, that is, the self-ignition of the air-fuel mixture during the charge period, appears as a phenomenon in which the ion current rises at an abnormal rate of increase. In step S12, when such an abnormal increase in the ion current is confirmed, the arithmetic unit 43 determines that pre-ignition has occurred. Note that the arithmetic unit 43 that determines the occurrence of pre-ignition corresponds to the "abnormal combustion determination unit" in the present invention.

[0077] When it is determined as YES in step S12 above and the occurrence of pre-ignition is confirmed, the injection control unit 42 causes the injector 15 to inject additional fuel (step S13). That is, for the injector 15 that has once finished injecting a specified amount of fuel in step S6 above, an additional amount of fuel is further injected. The injected additional fuel causes a temperature drop due to the latent heat of vaporization and plays a role in suppressing the progress of combustion.

[0078] On the other hand, when it is determined as NO in step S12 above and it is confirmed that pre-ignition has not occurred, the arithmetic unit 43 returns to step S8 above and waits for the arrival of the detection end time CAx.

[0079] When it is determined as YES in the above step S8 and the arrival of the detection end time CAx is confirmed at the end of detection, the calculation unit 43 calculates the maximum charge amount qmax, which is the maximum value of the charge amount q between the plug electrodes 13 and 14, based on the ion current detected by the ion current detector 44 so far (step S9). Specifically, the calculation unit 43 calculates the maximum charge amount qmax using a predetermined calculation formula based on the data of the ion current detected by the ion current detector 44 during the detection period (FIG. 8) from the first crank angle CA1, which is the start time of applying the bias voltage, to the detection end time CAx, and the data of the bias voltage during the same period. The maximum charge amount qmax calculated in this way can be said to be a value characterizing the ion current during the above detection period, that is, one of the ion current characteristic amounts.

[0080] Here, as described above, the detection end time CAx is a time somewhat later than the second crank angle CA2 (FIG. 8) at which the bias voltage reaches its peak. On the other hand, as understood from FIGS. 8 and 9, the charge amount q (space charge) between the plug electrodes 13 and 14 changes along a mountain-shaped waveform similar to the trend of the bias voltage. Therefore, at the timing of step S9 immediately after reaching the detection end time CAx, the time when the charge amount q between the plug electrodes 13 and 14 becomes maximum should already have passed. That is, in the present embodiment, the detection end time CAx is set at a timing slightly later than the time when the charge amount q is predicted to be maximum. For this reason, even at the timing of step S9 during the application of the bias voltage, the maximum charge amount qmax, which is the maximum value of the charge amount q, can be calculated.

[0081] Next, the arithmetic unit 43 determines the model ignition timing Igm based on the maximum charge amount qmax calculated in the above step S9 (step S10). As described above, the model ignition timing Igm is the advance limit of the ignition timing determined in consideration of the current in-cylinder gas state from the viewpoint of knocking avoidance, and is determined with a tendency as shown in FIG. 11 in relation to the maximum charge amount qmax. That is, the model ignition timing Igm is set on the retard side as the maximum charge amount qmax is larger. Map data or arithmetic expressions corresponding to FIG. 11 are stored in advance in the storage unit 45. The arithmetic unit 43 determines the model ignition timing Igm from the maximum charge amount qmax using the map data or arithmetic expressions.

[0082] Next, the arithmetic unit 43 determines whether or not the model ignition timing Igm determined in the above step S10 is on the advance side of the standard ignition timing Igr determined in the above step S3 (step S15).

[0083] When it is determined as YES in the above step S15 and it is confirmed that the model ignition timing Igm is on the advance side of the standard ignition timing Igr, the arithmetic unit 43 determines the advance correction amount ΔIg (step S16). The advance correction amount ΔIg is the correction amount for correcting the standard ignition timing Igr to the advance side.

[0084] In the present embodiment, the ignition advance correction amount ΔIg is variably set according to the ignition deviation amount, which is the deviation amount in the ignition advance direction from the standard ignition timing Igr to the model ignition timing Igm. FIG. 15 is a graph showing the relationship between the ignition deviation amount and the ignition advance correction amount ΔIg. As shown in this figure, the ignition advance correction amount ΔIg is increased at a rate of change of 1 as the ignition deviation amount increases so that it becomes the same value as the ignition deviation amount within the range where the ignition deviation amount is equal to or less than a predetermined value ΔX. On the other hand, when the ignition deviation amount exceeds the predetermined value ΔX, the ignition advance correction amount ΔIg is fixed at the predetermined value ΔX. In other words, the predetermined value ΔX is a value that functions as the upper limit of the ignition advance correction amount ΔIg. In this way, the upper limit is provided for the ignition advance correction amount ΔIg in order to prevent the ignition timing from being inadvertently advanced significantly. That is, due to the upper limit of the ignition advance correction amount ΔIg, even in a situation where the deviation amount in the ignition advance direction (ignition deviation amount) from the model ignition timing Igm to the standard ignition timing Igr is large, the ignition timing is not advanced all at once but is advanced step by step according to the repetition of the cycle.

[0085] Next, the ignition control unit 41 determines the timing obtained by advancing the standard ignition timing Igr by ΔIg determined in step S16 above as the ignition timing by the ignition plug 11 (step S17).

[0086] Next, the ignition control unit 41 causes the ignition plug 11 to ignite at the timing when the ignition timing determined in step S17 above, that is, the timing obtained by advancing the standard ignition timing Igr by ΔIg, arrives (step S18). That is, the ignition control unit 41 controls the igniter 106 so that the energization (charging) of the primary coil 102 is stopped as the ignition timing arrives. As a result, a high voltage is induced in the secondary coil 103, and a discharge occurs between the plug electrodes 13 and 14 of the ignition plug 11 in response to the induced high voltage.

[0087] Thus, in this embodiment, the maximum charge amount qmax is calculated from the ion current detected during the detection period (FIG. 8) before the ignition timing, and based on the calculated maximum charge amount qmax, the ignition timing in the current cycle, that is, the ignition timing in the same cycle as the cycle in which the maximum charge amount qmax was calculated, is determined.

[0088] Next, the ignition control unit 41 updates the standard ignition timing Igr (step S19). That is, the ignition control unit 41 sets, as the new standard ignition timing, the timing at which ignition was performed in step S18 above, that is, the timing obtained by advancing the standard ignition timing Igr by ΔIg. The new standard ignition timing set in this way is used as the standard ignition timing Igr in the next cycle.

[0089] Next, the control when the determination in step S15 above is NO, that is, when the model ignition timing Igm is not on the advanced side of the standard ignition timing Igr, will be described. In this case, the calculation unit 43 determines whether the model ignition timing Igm is on the retarded side of the standard ignition timing Igr (step S21).

[0090] When it is determined YES in step S21 above and it is confirmed that the model ignition timing Igm is on the retarded side of the standard ignition timing Igr, the ignition control unit 41 determines the model ignition timing Igm as the ignition timing as it is (step S22). As a result, the ignition timing will be retarded from the standard ignition timing Igr.

[0091] Next, the ignition control unit 41 causes the ignition plug 11 to be ignited at the timing when the ignition timing determined in step S22 above, that is, the model ignition timing Igm, arrives (step S23).

[0092] Next, the ignition control unit 41 updates the standard ignition timing Igr (step S24). That is, the ignition control unit 41 sets the model ignition timing Igm, which is the timing at which ignition was performed in step S23 above, as the new standard ignition timing Igr to be used in the next cycle.

[0093] Next, the control when the determination in step S21 is NO will be described. The fact that the determination is NO means that the model ignition timing Igm is neither on the advanced side nor on the retarded side with respect to the standard ignition timing Igr, in other words, the model ignition timing Igm is substantially the same as the standard ignition timing Igr. In this case, the ignition control unit 41 determines the standard ignition timing Igr as the ignition timing as it is (step S26).

[0094] Next, the ignition control unit 41 causes the ignition plug 11 to ignite at the timing when the ignition timing determined in step S26, that is, the standard ignition timing Igr, arrives (step S27).

[0095] Next, the arithmetic unit 43 estimates the combustion center of gravity of the combustion generated by the ignition in step S27 (step S28). The combustion center of gravity is the time when 50% of the mass of the injected fuel has burned. Such a combustion center of gravity can be estimated, for example, through calculation of the heat generation amount accompanying combustion. The method for calculating the heat generation amount is not particularly limited. For example, in the case of an engine equipped with a cylinder pressure sensor that detects the in-cylinder pressure, which is the pressure in the combustion chamber C, the heat generation amount can be calculated from the detection value of the cylinder pressure sensor, and the combustion center of gravity can be estimated from the change in the calculated heat generation amount.

[0096] Next, the arithmetic unit 43 determines whether or not the combustion center of gravity estimated in step S28 deviates from the target combustion center of gravity (step S29). The target combustion center of gravity is determined in advance for each operating condition including the engine load and the engine speed and stored in the storage unit 45. The arithmetic unit 43 compares the stored target combustion center of gravity with the combustion center of gravity estimated in step S28 and determines whether there is a deviation between the two.

[0097] When it is determined as YES in the above step S29 and it is confirmed that there is a deviation in the combustion center of gravity, the ignition control unit 41 corrects the standard ignition timing Igr according to the deviation (step S30). For example, when the estimated combustion center of gravity is deviated to the retard side with respect to the target combustion center of gravity, the ignition control unit 41 corrects the standard ignition timing Igr to the advance side. Conversely, when the estimated combustion center of gravity is deviated to the advance side with respect to the target combustion center of gravity, the ignition control unit 41 corrects the standard ignition timing Igr to the retard side.

[0098] On the other hand, when it is determined as NO in the above step S29 and it is confirmed that there is no deviation in the combustion center of gravity, the standard ignition timing Igr is not corrected, and the flow returns to the above step S1.

[0099] (5) Operational effects As described above, in the present embodiment, a bias voltage is applied between the plug electrodes 13 and 14 of the ignition plug 11 in accordance with the start of the low-temperature oxidation reaction, and based on the ion current detected during the detection period from the time (CA1) when the application of the bias voltage starts to the detection end time CAx before the ignition time (CA3), the maximum charge amount qmax, which is an ion current feature amount correlated with the in-cylinder gas state (gas state amount ρ·u / T), is calculated. Then, the ignition time in the same cycle in which the calculation is performed is determined based on the maximum charge amount qmax, and ignition is performed by the ignition plug 11 at the determined ignition time. According to such a configuration, there is an advantage that the occurrence of knocking can be suppressed while improving the thermal efficiency.

[0100] That is, in the present embodiment, based on the ion current detected before the ignition timing, the maximum charge amount qmax correlated with the in-cylinder gas state is calculated, and the ignition timing is determined based on the calculated maximum charge amount qmax. Therefore, the ignition timing can be adjusted to an appropriate timing considering the in-cylinder gas state. Specifically, the maximum charge amount qmax correlated with the in-cylinder gas state can be used as an index indicating the ease of occurrence of knocking that changes depending on the in-cylinder gas state. Therefore, according to the present embodiment in which the ignition timing is determined based on the maximum charge amount qmax, the air-fuel mixture can be ignited at an ignition timing with as good a thermal efficiency as possible within a range where knocking does not occur. Moreover, since the maximum charge amount qmax is calculated before the ignition timing, the ignition timing in the same cycle as the cycle in which the calculation is performed can be adjusted in advance to an appropriate timing considering knocking. Thereby, it is possible to suppress the occurrence of knocking while improving the thermal efficiency.

[0101] More specifically, in the present embodiment, as an ion current feature amount, the maximum charge amount qmax, which is the maximum value of the charge amount q (space charge) between the plug electrodes 13 and 14 during the detection period, is calculated, and the ignition timing is determined based on the model ignition timing Igm (FIG. 11) that is advanced more as the calculated maximum charge amount qmax is smaller. According to such a configuration, when the maximum charge amount qmax is small, it is possible to appropriately obtain an ignition timing that can achieve both suppression of knocking and improvement of thermal efficiency by using the knowledge that the in-cylinder gas state is in a state where knocking is unlikely to occur (the in-cylinder gas state quantity ρ·u / T is large).

[0102] Further, in the present embodiment, during the application of the bias voltage, the bias voltage is adjusted so that the voltage gradually increases and then gradually decreases (see FIG. 8). In this way, when the bias voltage is changed in a mountain shape, the charge amount q between the plug electrodes 13 and 14 can be changed with a similar mountain-shaped tendency. This facilitates the calculation of the above-described maximum charge amount qmax, which is the maximum value of the charge amount q. In addition, since the ion current that increases with the start of the low-temperature oxidation reaction is further amplified by the gradual increase of the bias voltage, the sensitivity of the maximum charge amount qmax that changes depending on the difference in the gas state inside the cylinder can be improved. As a result, the ignition timing suitable for the gas state inside the cylinder can be appropriately obtained from the maximum charge amount qmax.

[0103] Further, in the present embodiment, the capacitor device 107 (FIG. 3) is connected to the secondary coil 103, and a bias voltage is applied between the plug electrodes 13 and 14 by the charging of the capacitor device 107. According to such a configuration, the bias voltage can be appropriately controlled by adjusting the charge amount of the capacitor device 107.

[0104] Further, in the present embodiment, the start timing of the application of the bias voltage is adjusted so that it becomes earlier as the engine load or the engine speed is higher. According to such a configuration, the bias voltage can be applied in accordance with the start of the low-temperature oxidation reaction that becomes earlier as the engine load or the engine speed is higher (see FIGS. 14A and 14B), and the ion current that increases with the start of the low-temperature oxidation reaction can be appropriately detected.

[0105] Further, in the present embodiment, the occurrence of pre-ignition is determined based on the detected ion current, and when the occurrence of pre-ignition is determined, additional fuel is injected. According to such a configuration, since the latent heat of vaporization of the additionally injected fuel decreases the internal temperature of the cylinder 2, the progress of the combustion of the air-fuel mixture can be suppressed, and the influence caused by pre-ignition can be reduced.

[0106] (6) Modification The preferred embodiments of the present invention have been described above. However, the present invention is not limited to the above-described embodiments, and various modifications are possible without departing from the spirit of the present invention.

[0107] For example, in the above embodiment, as the ion current feature amount correlated with the in-cylinder gas state, the maximum charge amount qmax, which is the maximum value of the charge amount q (space charge) between the plug electrodes 13 and 14, was calculated. However, the ion current feature amount can be calculated from the ion current detected by the ion current detection unit 44 and may be a value correlated with the in-cylinder gas state, and is not limited to the maximum charge amount qmax. For example, the maximum ion current, which is the maximum value of the ion current detected by the ion current detection unit 44, may be calculated as the ion current feature amount. In this case, after setting the model ignition timing at which the ignition timing is advanced as the maximum ion current becomes smaller, the ignition timing may be determined based on this model ignition timing. Thereby, similarly to the above embodiment, it is possible to ignite the air-fuel mixture at a timing on the advanced angle side as much as possible while suppressing knocking.

[0108] In the above embodiment, the energization (charging) of the primary coil 102 and the application of the bias voltage between the plug electrodes 13 and 14 were started simultaneously, but these may be started at different timings. That is, a bias voltage may be applied between the plug electrodes 13 and 14 in accordance with the start of the low-temperature oxidation reaction, and the energization of the primary coil 102 may be started at a time slightly earlier or later than the start of the application of the bias voltage.

Explanation of Reference Numerals

[0109] 2 cylinders 11 Ignition plug 15 Injector 41 Ignition control unit 42 Injection control unit 43 Arithmetic unit (abnormal combustion determination unit) 44 Ion current detection unit 101 Ignition coil 102 Primary coil 103 Secondary coil 106 Igniter 107 Capacitor device (bias voltage generation unit) qmax Maximum charge amount

Claims

1. A combustion control device applied to an engine including a cylinder and a spark plug for igniting an air-fuel mixture in the cylinder, an ignition coil including a primary coil and a secondary coil, an igniter that induces a high voltage in the secondary coil through energization and de-energization of the primary coil, and causes a discharge between the plug electrodes of the spark plug by the induced high voltage, a bias voltage generation unit that applies a bias voltage for detecting an ion current generated between the plug electrodes due to ions in the cylinder between the plug electrodes, an ion current detection unit that detects the ion current, an ignition control unit that controls the bias voltage generation unit so that the bias voltage is applied between the plug electrodes in accordance with the start of a low-temperature oxidation reaction occurring in the cylinder as the air-fuel mixture is compressed, and controls the igniter so that a discharge occurs between the plug electrodes at an ignition timing set at a time later than the start of the low-temperature oxidation reaction, an arithmetic unit that calculates an ion current feature amount correlated with the in-cylinder gas state, which is the state of the air-fuel mixture in the cylinder, based on the ion current detected by the ion current detection unit during a detection period that is a period from the time when the application of the bias voltage starts to a predetermined time earlier than the ignition timing, wherein the ignition control unit determines the ignition timing in the same cycle in which the ion current feature amount is calculated based on the ion current feature amount, and controls the igniter so that the discharge is performed at the determined ignition timing, the combustion control device for an engine.

2. In the combustion control device for an engine according to Claim 1, the ion current feature amount is a maximum charge amount that is the maximum value of the charge amount between the plug electrodes during the detection period, and the ignition control unit sets the ignition timing more advanced as the maximum charge amount is smaller, the combustion control device for an engine.

3. In the combustion control device for an engine according to Claim 1, the ion current feature amount is a maximum ion current that is the maximum value of the ion current during the detection period, and the ignition control unit sets the ignition timing more advanced as the maximum ion current is smaller, the combustion control device for an engine.

4. In the combustion control device for an engine according to Claim 2 or 3, The ignition control unit controls the bias voltage generation unit so that the bias voltage gradually increases from the start of the low-temperature oxidation reaction and then gradually decreases, in an engine combustion control device.

5. In the engine combustion control device according to any one of claims 1 to 3, The bias voltage generation unit is a capacitor device connected to the secondary coil, in an engine combustion control device.

6. In the engine combustion control device according to any one of claims 1 to 3, The ignition control unit controls the bias voltage generation unit so that the application start of the bias voltage is earlier as the engine speed is higher, in an engine combustion control device.

7. In the engine combustion control device according to any one of claims 1 to 3, The ignition control unit controls the bias voltage generation unit so that the application start of the bias voltage is earlier as the engine load is higher, in an engine combustion control device.

8. In the engine combustion control device according to any one of claims 1 to 3, An injector that injects fuel into the cylinder before the start of the low-temperature oxidation reaction; An abnormal combustion determination unit that determines the occurrence of pre-ignition based on the ion current detected by the ion current detection unit; And an injection control unit that causes the injector to inject additional fuel when the occurrence of the pre-ignition is determined, in an engine combustion control device.

Images