JP2007218131A - Combustion condition estimating device for internal combustion engine - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To accurately estimate a combustion condition of an internal combustion engine. <P>SOLUTION: This device estimates the combustion condition of the internal combustion engine 10 and is provided with a storage means storing statistical distribution of crank angle acceleration at a time of engine start beforehand according to properties of fuel, an angular acceleration calculation means calculating crank angle acceleration at the time of engine start, and a combustion condition estimation means estimating fuel properties by applying crank angle acceleration calculated at the time of engine start to the statistical distribution. Since fuel properties are estimated by applying the crank angle acceleration calculated at the time of engine start to the pre-stored statistical distribution, fuel properties can be accurately estimated by taking variation of combustion due to disturbance or the like into consideration. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to a combustion state estimation device for an internal combustion engine, and is particularly suitable for application to a device that estimates a combustion state when the engine is started.

  Conventionally, for example, Japanese Patent Application Laid-Open No. 11-117787 discloses a method for determining the properties of fuel based on a combustion state detected by an ionic current. Japanese Patent Application Laid-Open No. 7-27010 discloses a method for determining the properties of fuel based on the change in the number of revolutions at the start of the engine and the charging efficiency into the cylinder.

Japanese Patent Laid-Open No. 11-117787 JP-A-7-27010

  However, it is difficult to accurately estimate the combustion state by a method based on ion current, engine speed, and the like. In particular, when the engine is started, the amount of fuel adhering to the intake pipe and the inner wall surface of the cylinder changes depending on environmental conditions such as temperature, so that the atomization of the fuel fluctuates and the amount of fuel supplied into the cylinder varies. . Further, the amount of intake air into the cylinder varies depending on environmental conditions such as temperature. Furthermore, engine friction also varies depending on environmental conditions. In particular, when the engine is started, it is easily affected by disturbances such as environmental conditions, and variations are likely to occur. Therefore, it is difficult for the conventional technology to accurately estimate the combustion state in consideration of these variations.

  For example, when determining the fuel properties based on the combustion state, if a misdetection of heavy fuel is made when light fuel is used, it is assumed that the fuel injection amount becomes excessive and the emission deteriorates. The In addition, if a misdetection of light fuel is made when heavy fuel is used, it is assumed that the fuel required for combustion will be insufficient, leading to misfire, engine stall, and worsening of emissions and drivability. Is done.

  The present invention has been made to solve the above-described problems, and an object thereof is to estimate the combustion state of an internal combustion engine with high accuracy.

  In order to achieve the above object, the first invention is an apparatus for estimating the combustion state of an internal combustion engine, and stores in advance a statistical distribution of crank angular acceleration at the time of engine start according to the properties of the fuel. And an angular acceleration calculating means for calculating the crank angular acceleration at the time of starting the engine, and a combustion state estimating means for estimating the properties of the fuel by applying the crank angular acceleration calculated at the time of starting the engine to the statistical distribution. It is characterized by that.

  According to a second invention, in the first invention, there is provided a detection value calculation means for calculating a detection value serving as an index representing a distribution state in the statistical distribution of the crank angular acceleration calculated at the time of engine start. The combustion state estimating means estimates the property of the fuel based on the detected value.

  In a third aspect based on the second aspect, the storage means stores the statistical distribution for each number of explosions since the start of the engine, and the detection value calculation means stores the statistical distribution for each number of explosions. Based on each of the crank angular accelerations calculated for each number of explosions, and includes a weighted average value calculating means for calculating a weighted average value of the detected values, and the combustion state estimating means Is characterized in that the fuel property is estimated based on the weighted average value.

  According to a fourth invention, in the third invention, a determination value calculating means for calculating a determination value for controlling a fuel injection amount based on the estimated fuel property and the weighted average value; and the determination value And a fuel injection amount control means for controlling the fuel injection amount based on the above.

  According to a fifth invention, in any one of the first to fourth inventions, the angular acceleration calculating means calculates the crank value as an average value in a crank angle section in which an average value of inertia torque by a reciprocating inertia mass is substantially zero. An angular acceleration is calculated.

  According to the first aspect of the invention, the crank angular acceleration calculated at the time of starting the engine is applied to a statistical distribution stored in advance to estimate the fuel properties. Therefore, the fuel properties are considered after taking into account combustion variations due to disturbances and the like. It is possible to estimate accurately.

  According to the second invention, by calculating the detection value, the distribution state in the statistical distribution of the crank angular acceleration calculated at the time of starting the engine is clarified, so that the fuel property can be estimated.

  According to the third aspect of the invention, the detection value is calculated from each of the crank angular accelerations calculated for each number of explosions, and the weighted average value of the detection values is calculated. Can be absorbed, and erroneous determination can be reliably prevented.

  According to the fourth aspect of the present invention, it is possible to calculate a determination value for controlling the fuel injection amount based on the estimated fuel property and the weighted average value. Therefore, it is possible to optimally control the fuel injection amount based on the determination value.

  According to the fifth aspect, since the crank angular acceleration is calculated as an average value in a section where the average value of the inertial torque due to the reciprocating inertial mass is almost zero, the influence of the inertial torque due to the reciprocating inertial mass on the angular acceleration is eliminated. can do. Accordingly, it is possible to accurately estimate the combustion state based on the angular acceleration.

  Several embodiments of the present invention will be described below with reference to the drawings. In addition, the same code | symbol is attached | subjected to the element which is common in each figure, and the overlapping description is abbreviate | omitted. The present invention is not limited to the following embodiments.

Embodiment 1 FIG.
FIG. 1 is a diagram for explaining a combustion state estimating device for an internal combustion engine and its surrounding structure according to each embodiment of the present invention. An intake passage 12 and an exhaust passage 14 communicate with the internal combustion engine 10. The intake passage 12 includes an air filter 16 at an upstream end. The air filter 16 is assembled with an intake air temperature sensor 18 for detecting the intake air temperature THA (that is, the outside air temperature). An exhaust purification catalyst 32 is disposed in the exhaust passage 14.

  An air flow meter 20 is disposed downstream of the air filter 16. A throttle valve 22 is provided downstream of the air flow meter 20. A throttle sensor 24 that detects the throttle opening degree TA and an idle switch 26 that is turned on when the throttle valve 22 is fully closed are disposed in the vicinity of the throttle valve 22.

  A surge tank 28 is provided downstream of the throttle valve 22. Further, a fuel injection valve 30 for injecting fuel into the intake port of the internal combustion engine 10 is disposed further downstream of the surge tank 28.

  Each cylinder of the internal combustion engine 10 includes a piston 34. A crankshaft 36 that is rotationally driven by the reciprocating motion is connected to the piston 34. The vehicle drive system and accessories (air conditioner compressor, alternator, torque converter, power steering pump, etc.) are driven by the rotational torque of the crankshaft 36. A crank angle sensor 38 for detecting the rotation angle of the crankshaft 36 is attached in the vicinity of the crankshaft 36. Further, a water temperature sensor 42 for detecting the cooling water temperature is attached to the cylinder block of the engine 10.

  As shown in FIG. 1, the combustion state estimation device of this embodiment includes an ECU (Electronic Control Unit) 40. In addition to the various sensors and the fuel injection valve 30 described above, a vehicle speed sensor 44 that detects the vehicle speed SPD and the like are connected to the ECU 40.

  Next, a method for estimating the combustion state of the internal combustion engine 10 using the system of FIG. 1 will be specifically described. First, mathematical formulas used for estimating the combustion state will be described. In the present embodiment, the combustion state is estimated using the following equations (1) and (2).

In the equations (1) and (2), the indicated torque _Ti is the torque generated in the crankshaft 36 by the combustion of the engine. Here, (2) the right side shows the torque generated indicated torque T _i, shows a torque consuming (1) is on the right side indicated torque T _i.

(1) On the right side of the formula, J is the moment of inertia of the driven member driven by the combustion or the like of the mixture, d [omega / dt is the angular acceleration of the crankshaft 36, T _f is the friction torque of the drive unit, T _l is the time of running The load torque received from the road surface is shown. Here, J × (dω / dt) is a dynamic loss torque (= T _ac ) caused by the angular acceleration of the crankshaft 36. The friction torque _Tf is a torque due to mechanical friction of each fitting portion such as friction between the piston 34 and the inner wall of the cylinder, and includes torque due to mechanical friction of accessories. The load torque _Tl is a torque due to a disturbance such as a road surface condition during traveling. In this embodiment, in order to estimate the combustion state with the shift gear in the neutral state, T _l = 0 in the following description.

Further, in (2) of the right _{side, T gas} is the torque due to cylinder gas pressure in the _{cylinder, T inertia} represents the inertial torque due to reciprocating inertia mass such as a piston 34. Torque T _gas due to in-cylinder gas pressure is torque generated by combustion of the air-fuel mixture in the cylinder. In order to accurately estimate the combustion state, it is necessary to obtain the torque T _gas by the cylinder gas pressure.

As shown in the equation (1), the indicated torque T _i can be obtained as the sum of dynamic loss torque J × (dω / dt) due to angular acceleration, friction torque T _f , and load torque T _l. it can. However, as shown in the equation (2), the indicated torque T _i and the torque T _gas due to the in-cylinder gas pressure do not coincide with each other, so that the combustion state cannot be accurately estimated from the indicated torque T _i .

FIG. 2 is a characteristic diagram showing the relationship between each torque and crank angle in equation (2). In FIG. 2, the vertical axis of the torque magnitude, the horizontal axis represents the crank angle, the one-dot chain line indicated torque T _i in FIG. 2, solid lines a torque T _gas by the in-cylinder gas pressure, the broken lines Indicates the inertia torque T _inertia due to the reciprocating inertia mass. Here, FIG. 2 shows the characteristics in the case of four cylinders. TDC and BDC in FIG. 2 are the top dead center (TDC) or the bottom dead center of the piston 34 of one of the four cylinders. The crank angle (0 °, 180 °) in the (BDC) position is shown. When the internal combustion engine 10 has four cylinders, every time the crankshaft 36 rotates 180 °, an explosion stroke is performed for each cylinder, and torque characteristics from TDC to BDC in FIG.

As shown by the solid line in FIG. 2, the torque T _gas due to the in-cylinder gas pressure rapidly increases and decreases between TDC and BDC. Here, the rapid increase in T _gas is due to the explosion of the air-fuel mixture in the combustion chamber in the explosion process. After the explosion, T _gas decreases and takes a negative value due to the influence of the cylinders in other compression strokes or exhaust strokes. When the crank angle reaches BDC, the change in the volume of the cylinder becomes zero, whereby T _gas takes a value of zero.

On the other hand, the inertia torque T _inertia due to the reciprocating inertia mass is an inertia torque generated by the inertia mass of the reciprocating member such as the piston 34, irrespective of the torque T _gas due to the in-cylinder gas pressure. The reciprocating member repeats acceleration / deceleration, and T _inertia always occurs as long as the crank rotates, even if the angular velocity is constant. As shown by the broken line in FIG. 2, the member that reciprocates is stopped at the position where the crank angle is TDC, and T _inertia = 0. When the crank angle advances from TDC toward BDC, the reciprocating member starts to move from the stopped state. At this time, T _inertia increases in a negative direction due to _inertia of these members. When the crank angle reaches around 90 °, the reciprocating members are moving at a predetermined speed, so that the crankshaft 36 is rotated by the inertia of these members. Therefore, T _interia changes from a negative value to a positive value between TDC and BDC. After that, when the crank angle reaches BDC, the reciprocating member stops and T _inertia = 0.

As shown in the equation (2), the indicated torque T _i is the sum of the torque T _gas caused by the in-cylinder gas pressure and the inertia torque T _inertia caused by the reciprocating inertia mass. For this reason, as shown by the one-dot chain line in FIG. 2, between TDC and BDC, the indicated torque T _i increases due to an increase in T _gas due to the explosion of the air-fuel mixture, and once decreases, then increases again due to T _inertia . It shows the complicated behavior.

However, paying attention to a section with a crank angle of 180 ° from TDC to BDC, the average value of the inertia torque T _inertia due to the reciprocating inertia mass in this section is zero. This is because the member having the reciprocating inertia mass moves in the opposite direction at a crank angle of about 0 ° to 90 ° and a crank angle of about 90 ° to 180 °. Therefore, when the torques in the equations (1) and (2) are calculated as average values from TDC to BDC, the inertia torque T _inertia due to the reciprocating inertia mass can be calculated as zero. Thus, the inertia torque T _inertia caused by the reciprocating inertia mass can eliminate the influence on the indicated torque T _i, it is possible to easily estimate an accurate combustion state.

When the average value of each torque in the interval from TDC to _{BDC, T} the average value of _inertia becomes zero, from equation (2), the torque T by the average value and the in-cylinder gas pressure indicated torque _{T i} The average value of _gas becomes equal. Therefore, it is possible to estimate accurately the combustion state based on the indicated torque T _i.

Further, when the average value of the angular acceleration of the crankshaft 36 is obtained in the section from TDC to BDC, the average value of T _{inertia in} this section is 0, so the influence of the reciprocating inertia mass on the angular acceleration is eliminated. Angular acceleration can be obtained. Therefore, it is possible to calculate the angular acceleration caused only by the combustion state, and it is possible to accurately estimate the combustion state based on the angular acceleration.

Next, a method for calculating each torque on the right side of the equation (1) will be described. First, a method of calculating dynamic loss torque T _ac = J × (dω / dt) resulting from angular acceleration will be described. FIG. 3 is a schematic diagram showing a method for obtaining the angular acceleration of the crankshaft 36. As shown in FIG. 3, in this embodiment, a crank angle signal is detected from the crank angle sensor 38 every 10 ° of rotation of the crankshaft 36.

The combustion state estimation apparatus according to the present embodiment calculates a dynamic loss torque _Tac caused by angular acceleration as an average value from TDC to BDC. For this purpose, the apparatus of the present embodiment obtains angular velocities ω ₀ (k) and ω ₀ (k + 1) at two crank angle positions of TDC and BDC, respectively, and at the same time, the crankshaft 36 rotates from TDC to BDC. Δt (k) is obtained.

When obtaining the angular velocity ω ₀ (k), for example, as shown in FIG. 3, the time Δt ₀ (k) and Δt ₁₀ (k) during which the crank angle is rotated 10 ° forward and backward from the TDC position are calculated. It is detected from the crank angle sensor 38. Since the crankshaft 36 rotates 20 ° during the time Δt ₀ (k) + Δt ₁₀ (k), ω ₀ (k) = (20 / (Δt ₀ (k) + Δt ₁₀ (k))) By calculating x (π / 180), ω ₀ (k) [rad / s] can be calculated. Similarly, when calculating ω ₀ (k + 1), times Δt ₀ (k + 1) and Δt ₁₀ (k + 1) are detected while the crank angle is rotating 10 ° forward and backward from the BDC position. _{Then, ω 0 (k + 1)} = (20 / (Δt 0 (k + 1) + Δt 10 (k + 1))) × (π / 180) ω by computing ₀ (k + 1) can be calculated [rad / s].

After obtaining the angular velocities ω ₀ (k), ω ₀ (k + 1), (ω ₀ (k + 1) −ω ₀ (k)) / Δt (k) is calculated, and the crankshaft 36 rotates from TDC to BDC. The average value of the angular acceleration is calculated.

  After the average value of angular acceleration is obtained, the average value of angular acceleration and the moment of inertia J are multiplied according to the right side of equation (1). As a result, an average value of dynamic loss torque J × (dω / dt) while the crankshaft 36 rotates from TDC to BDC can be calculated. The inertia moment J of the drive unit is obtained in advance from the inertia mass of the drive component.

Next, a method for calculating the friction torque _Tf will be described. FIG. 4 is a map showing the relationship between the friction torque _Tf , the engine speed (Ne) of the internal combustion engine 10 and the coolant temperature (thw). In FIG. 4, the friction torque T _f , the engine speed (Ne), and the cooling water temperature (thw) are average values when the crankshaft 36 rotates from TDC to BDC. Further, the cooling water temperature becomes higher in the order of thw1 → thw2 → thw3. As shown in FIG. 4, the friction torque _Tf tends to increase as the engine speed (Ne) increases and increase as the cooling water temperature (thw) decreases. The map of FIG. 4 measures the friction torque _Tf generated when the crankshaft 36 is rotated from TDC to BDC with the engine speed (Ne) and the cooling water temperature (thw) as parameters, and the average value thereof. It is created in advance by calculating. When estimating the combustion state, the average value of the cooling water temperature and the average value of the engine speed in the section from TDC to BDC are applied to the map of FIG. 4 to obtain the average value of the friction torque _Tf . At this time, the coolant temperature is detected from the water temperature sensor 42, and the engine speed is detected from the crank angle sensor 38.

The behavior of the friction torque _Tf accompanying the variation of the crank angle is very complicated and has a large variation. However, since the behavior of the friction torque _Tf mainly depends on the speed of the piston 34, the average value of the friction torque _Tf for each section in which the average value of the inertia torque _Tinertia due to the reciprocating inertial mass is 0 is substantially constant. ing. Therefore, by obtaining the average value of the friction torque _{T f} for each section in which the average value of the inertia torque _{T inertia} caused by the reciprocating inertial mass becomes 0 (TDC → BDC), accurately friction torque _{T f} showing the complex instantaneous behavior Can be sought. Moreover, the map shown in FIG. 4 can be created accurately by setting the friction torque _Tf to an average value for each section.

Further, as described above, the friction torque _Tf includes torque due to friction of auxiliary machinery. Here, the value of the torque due to the friction of the auxiliary machines varies depending on whether or not the auxiliary machines are operating. For example, the rotation of the engine is transmitted to a compressor of an air conditioner, which is one of the auxiliary machines, by a belt or the like, and torque due to friction is generated even when the air conditioner is not actually operating.

On the other hand, when the auxiliary machinery is operated, for example, when the air conditioner switch is turned on, the torque consumed by the compressor is larger than when the air conditioner is not operated. For this reason, the torque due to the friction of the auxiliary machinery increases, and the value of the friction torque _Tf also increases. Therefore, in order to accurately determine the friction torque _Tf , the operating state of the auxiliary machinery is detected, and when the auxiliary machinery is switched on (ON), the friction torque obtained from the map of FIG. It is desirable to correct the value of _Tf .

It is more preferable to correct the friction torque T _f in consideration of the difference between the temperature of the portion where the friction torque T _f is actually generated and the cooling water temperature at the time of extremely cold start. In this case, it is desirable to perform correction in consideration of the engine start time after the cold start, the in-cylinder inflow fuel amount, and the like.

In the present embodiment, as will be described in detail below, the property (combustion state) of the fuel is determined by a statistical method using the angular acceleration calculated by the above-described method. In the following description, a method of estimating the fuel property will be described with reference to angular acceleration, even when using the indicated torque T _i, is determined fuel property as in the case of using the angular acceleration be able to.

As a premise for determining the properties of the fuel, a determination index (probability density distribution difference) is acquired in advance. This determination index is statistical information on the distribution of angular acceleration. FIG. 5 is a diagram schematically illustrating a method for obtaining a determination index. First, as shown in FIG. 5 (A), the internal combustion engine 10 is started a plurality of times using a fuel having a predetermined fuel property (in this example, heavy fuel is used here), and the first explosion at the time of starting the engine is performed. Acquire angular acceleration multiple times in the stroke. Then, as shown in FIG. 5B, a histogram is created using the plurality of acquired angular accelerations. Thereafter, normalized histogram of FIG. 5 (B), as shown in FIG. 5 (C), to create a detection value G _F statistics to be determined index (statistic ^{N (μ, σ 2))} . Here, μ is an average value of angular accelerations acquired a plurality of times. Also, σ ² is the variance of the acquired angular acceleration.

According to the determination index shown in FIG. 5C, when heavy fuel is used, the level of angular acceleration in the first explosion stroke at the start of the engine is statistically clarified. Therefore, by determining the angular acceleration in the first explosion stroke at the time of starting the engine, it is possible to determine whether or not the fuel property is heavy based on the determination index of FIG. For example, if the value of the detection value G _F obtained from the judgment information shown in FIG. 5 (C) based on the angular acceleration is greater than the determination value T shown in FIG. 5 (C), heavy fuel angular acceleration obtained Therefore, it can be determined that the property of the fuel is heavy. Similarly, by obtaining a determination index for light fuel, it is possible to determine whether the fuel property is light or not based on the angular acceleration during the first explosion stroke at the time of engine start. It becomes.

  According to the above method, the fuel property can be determined based on the angular acceleration in the first explosion stroke at the time of starting the engine, and the fuel injection amount is increased or decreased according to the fuel property at an early stage immediately after the engine is started. can do. When the fuel injection amount is controlled, the heavier the fuel properties, the more the amount of fuel adhering to the intake passage 12 or the inner wall surface of the cylinder and the lower the degree of fuel atomization. In this case, control is performed to increase the fuel injection amount. As a result, the fuel injection amount from the fuel injection valve 30 can be optimally controlled based on the properties of the fuel.

  Further, in order to improve the accuracy of the determination, the fuel property may be determined using angular acceleration in a plurality of explosion strokes immediately after the engine is started. In this case, the determination index is acquired in advance by the method of FIG. 5 for the second, third and subsequent explosion strokes when the engine is started. Similarly, for the light fuel determination index, the determination index for the second, third and subsequent explosion strokes is acquired in advance. FIG. 6 is a characteristic diagram showing a determination index for obtaining the distribution of angular acceleration for each fuel property (heavy, light) and for each number of explosions from the start. When the fuel property is light, the amount of fuel adhering to the inner wall surface of the intake pipe and the inner wall surface of the cylinder is reduced as compared with the heavy fuel property, and atomization of the fuel is promoted. Accordingly, in FIG. 6, the angular acceleration when the fuel property is light is larger than the angular acceleration when the fuel property is heavy.

  When determining the fuel properties at the time of engine start using the determination index of FIG. 6, the angular accelerations in the first, second, third,... Apply to statistical information. Then, by determining whether the acquired angular acceleration belongs to the distribution of light fuel or heavy fuel in FIG. 6, the property of the fuel can be determined. In this case, since the fuel property is determined using the angular acceleration in a plurality of explosion strokes immediately after the engine is started, the determination accuracy can be improved. In addition, immediately after the engine is started, the engine is in a transient operation state in which the number of revolutions increases, and the influence of the fuel properties is likely to appear in the angular acceleration.Therefore, the fuel properties can be determined by making a determination based on the angular acceleration immediately after the engine starts. It becomes possible to determine with high accuracy.

  In addition, in the determination index of FIG. 6, two types of fuel property distributions, light fuel and heavy fuel, are created, but the fuel properties are further subdivided to create distributions for more fuel properties. May be. As a result, determinations corresponding to fuels having various properties can be made, and the continuously changing fuel property level can be obtained more accurately. For example, it is possible to determine whether the fuel properties are heavy fuel, light fuel, or intermediate fuel by creating a distribution for intermediate fuel that has fuel properties between light fuel and heavy fuel. Can do.

  In FIG. 5C, the angular acceleration distribution is standardized to a Gaussian distribution. However, as long as the statistical information can be effectively used and the combustion mode can be detected, other distributions can be used as shown in FIG. The shape may be standardized. Here, FIG. 7A shows an example in which the distribution of FIG. 5C is modified so that the frequency of the central value of the distribution of angular acceleration is increased.

FIG. 7B and FIG. 7C show an example in which the distribution is schematically illustrated in a staircase pattern. For example, when the distribution of heavy fuel is set to the shape shown in FIG. 7B, the fuel property is determined to be heavy only when the angular acceleration is u _{F1 to} u _F2 , and the heavy characteristic is determined otherwise. Can be avoided. Therefore, even when the accuracy of the angular acceleration data decreases due to some factor, the heavyness determination can be performed with high accuracy. Also, as shown in FIGS. 7B and 7C, by setting the distribution in a stepped shape, it is possible to simply configure the fuel property determination map.

  When the engine is started, variations in the amount of fuel supplied to the cylinder, the amount of intake air, etc. are likely to occur due to environmental conditions such as temperature, but according to the method of this embodiment, the variation is statistically and probabilistically. Since the analysis is performed and the determination process is performed based on the statistical information, it is possible to accurately estimate the fuel property and combustion state in consideration of variations.

  Next, a processing procedure in the system of the present embodiment will be described based on the flowchart of FIG. First, in step S1, the angular acceleration of the crankshaft 36 is calculated immediately after starting. Here, the average value of the angular acceleration from TDC to BDC is calculated by the method described above.

In the next step S2, fitting the angular acceleration calculated in step S1 to the determination indicator of FIG. 6, to calculate a detection value G _F. In the next step S3, it determines the property of fuel based on the detection value G _F. In the next step S4, the fuel injection amount is controlled based on the properties of the fuel. After step S4, the process ends (RETURN).

  As described above, according to the first embodiment, the fuel property (combustion state) can be accurately estimated by a statistical method based on the angular acceleration calculated at the time of engine start. Therefore, the internal combustion engine can be optimally controlled based on the fuel properties, and exhaust emission and drivability can be improved.

Embodiment 2. FIG.
Next, a second embodiment of the present invention will be described. In the second embodiment, a weighted addition average detection process is performed based on the angular acceleration acquired at the time of starting the engine, and the probability of determination is determined together with determination of fuel properties.

FIG. 9 is a characteristic diagram showing a weighted map representing the probability of fuel properties. Map of FIG. 9, in the determination indicator of Figure 5, which is normalized as described in the first embodiment (C), the average value mu = 0, the detected value g _F by integrating the detection value G _F as variance sigma ^{2 =} 1 Is calculated. When the angular acceleration data in FIG. 5A is heavy fuel data, the map in FIG. 9 obtained from the determination index in FIG. 5C is a map corresponding to the heavy fuel.

9, the horizontal axis represents the angular acceleration u _F, the vertical axis represents the detected value g _F corresponding to the value of the angular acceleration u _F. In FIG. 9, since the average value μ = 0 and the variance σ ² = 1, the detection value g _F is a value from 0 to 1. And when the map of FIG. 9 is a map corresponding to heavy fuel, the detection value g _F corresponding to the angular acceleration u _F with the highest frequency when heavy fuel is burned is 0.5.

Therefore, when the angular acceleration u _F is obtained at the time of engine start and applied to the map of FIG. 9 corresponding to heavy fuel, if the detected value g _F corresponding to the angular acceleration u _F is close to 0.5, the fuel is It can be judged that it is heavy. And it can be determined that the probability that the fuel is heavier is higher as the detected value is closer to 0.5.

Similarly, for light fuel, statistical information as a determination index is created using the method of FIG. 5, and a map similar to FIG. 9 is created. Thus, we determined angular acceleration u _F at the time of engine startup, when fit to the map of the light fuel, when the detected value g _F corresponding to the angular acceleration u _F is close to 0.5 can be determined that the fuel is light As the detection value g _F is closer to 0.5, it can be determined that the probability that the fuel is lighter is higher.

  Therefore, by preparing the map of FIG. 9 for both heavy fuel and light fuel, it is possible to determine the fuel properties based on the angular acceleration at the time of engine start, and to determine the likelihood of the fuel property determination. Is possible. Therefore, the fuel property can be determined with high accuracy.

Since the angular acceleration value obtained at the start belongs to one of the distributions of heavy fuel and light fuel in principle, it is calculated from the detected value g _F calculated from the map of light fuel and the map of heavy fuel. Both of the detected values g _F are not very close to 0.5. When one detected value g _F is a value in the vicinity of 0.5, the other detected value g _F is 0.5. The value is out of the range. Therefore, determining the property of fuel by one of the detection value g _F of the detection value g _F and heavy fuel for light fuel, either the detected value g _F to determine whether the value closer to 0.5 It is possible to determine the certainty.

When the statistical information in FIG. 5C represents the distribution of angular acceleration in the first explosion stroke at the time of starting the engine, according to the map in FIG. 9 corresponding thereto, the angle acquired in the first explosion stroke. The detection value g _F can be calculated based on the acceleration, and the fuel property and its likelihood can be determined based on the detection value g _F. Similarly, regarding the angular acceleration in the second and subsequent explosion strokes at the time of starting the engine, the map shown in FIG. 9 is created for each of the light fuel and the heavy fuel according to the number of explosions. The fuel property can be determined based on the angular acceleration acquired in the subsequent explosion stroke, and the likelihood of the fuel property can be determined.

In this case, first at engine starting, the second time, based on the respective angular acceleration u _F acquired in third · · · n-th power stroke, light fuel, depending on the number of explosions for each heavy fuel obtains a detection value g _F from the maps created Te, light fuel, for each of the heavy fuel, the following (3) obtaining a weighted average of n detected value g _F g _{F (n)} from the equation.

In this case, as the value of the average value g _F (n) approaches 0.5, the accuracy of the fuel property probability represented by the average value g _F (n) increases. For example, when the average value g _F (n) of the detected values g _F obtained from the heavy fuel map is close to 0.5, it is determined that the fuel is heavy and the probability thereof is very high. be able to. Therefore, by determining the fuel properties based on a plurality of angular accelerations, it is possible to absorb a small number of unique angular acceleration behaviors and suppress erroneous detection, and to determine the fuel properties and their accuracy with high accuracy. Is possible.

Next, a processing procedure in the second embodiment will be described based on the flowchart of FIG. First, in step S11, angular acceleration is acquired in each of the n explosion strokes immediately after starting. In the next step S12, weighted addition average detection value processing is performed using the map of FIG. Specifically, using the plurality of angular accelerations acquired in step S11, the detection value g _F is obtained from the light fuel and heavy fuel maps of FIG. 9 created according to the number of explosions, and the average value g _F (n) is calculated.

In the next step S13, the fuel property and its certainty are determined based on the average value g _F (n) calculated in step S12. At this time, it is determined that the probability of the fuel property is higher as the average value g _F (n) is closer to 0.5. In the next step S14, the fuel injection amount is controlled based on the fuel property obtained in step S13. After step S14, the process ends (RETURN).

  As described above, according to the second embodiment, the fuel property can be obtained based on the angular acceleration obtained at the time of starting the engine, and the probability of the obtained fuel property can be determined. Therefore, it is possible to further improve the reliability of determination of the fuel property, and it is possible to optimally control the internal combustion engine based on the fuel property.

Embodiment 3 FIG.
Next, a third embodiment of the present invention will be described. Embodiment 3 relates to a method for determining a fuel injection amount based on fuel properties. FIG. 11 is a characteristic diagram showing a map defining the relationship between the determination value determined based on the detection value g _F and the fuel injection amount in the third embodiment. As shown in FIG. 11, the map is defined so that the fuel injection amount decreases as the determination value increases.

In the third embodiment, the weighted average value g _F (n) of the detection value g _F is obtained by the same method as in the second embodiment, and the fuel property is determined. Then, as shown in FIG. 11, a determination value corresponding to the weighted average value g _F (n) is determined.

As shown in FIG. 11, when the average value g _F (n) calculated from the map of FIG. 9 corresponding to the heavy fuel is 0.2 to 0.8, the average value g _F (n) is 0. Since the value is in the vicinity of 5, the fuel is determined to be heavy. Similarly, when the average value g _F (n) calculated from the map of FIG. 9 corresponding to the light fuel is 0.2 to 0.8, the average value g _F (n) is a value in the vicinity of 0.5. Therefore, it is determined that the fuel is light.

Then, the value of the determination value is set according to the average value g _F (n) calculated from the heavy fuel and light fuel maps. In the example of FIG. 11, when the average value g _F (n) calculated from the map of FIG. 9 corresponding to the heavy fuel is 0.2 to 0.8, the determination value is set to 1 to 5, and the fuel The injection amount is set to a constant value TAU1. Further, when the average value g _F (n) calculated from the map of FIG. 9 corresponding to the light fuel is 0.2 to 0.8, the determination value is set to 7 to 11, and the fuel injection amount is a constant value. Set to TAU2. Further, when the determination value is in the range of 5 to 6 and the fuel property is determined to be intermediate between heavy and light, the fuel injection amount is set to change linearly between TAU1 and TAU2. .

In this way, the average value g _F (n) is given a range of a predetermined range, and within that range, the fuel injection amount is controlled to a constant value TAU1, TAU2 within a range that does not affect the emission, etc. The amount can be controlled simply, and the controllability of the fuel injection amount according to the fuel property can be improved.

  FIG. 12 is a schematic diagram showing another example of the map of FIG. The map in FIG. 11 can be changed to various types in consideration of the ratio of the properties of the fuel distributed in the market. Here, the maps of FIGS. 12A and 12B show that the fuel injection amount in the intermediate fuel is made more precise by changing the fuel injection amount nonlinearly between the heavy fuel and the light fuel. It is something to control.

Further, the map of FIG. 12C shows an example in which the map of FIG. 9 is set only for the intermediate fuel whose fuel properties are intermediate between heavy and light. That is, in the example of FIG. 12C, when the average value g _F (n) calculated from the map of FIG. 9 corresponding to the intermediate fuel is 0.2 to 0.8, the fuel properties are heavy and light. And the determination value is set to 4-7. In this case, the fuel injection amount is set to a constant value TAU3. When the average value g _F (n) is less than 0.2, the determination value is set to less than 4, and the value of the fuel injection amount is defined so as to change linearly with respect to the determination value. When the average value g _F (n) exceeds 0.8, the determination value is set to a value larger than 7, and the value of the fuel injection amount is defined so as to change linearly with respect to the determination value. . According to the map of FIG. 12C, the fuel injection amount is controlled based on the average value g _F (n) and the determination value by creating the map of FIG. 9 that performs weighted averaging processing only for the intermediate fuel. be able to.

Next, based on the flowchart of FIG. 13, the procedure of the process by the system of this embodiment will be described. First, in step S21, angular acceleration is acquired immediately after engine startup, and a weighted average value g _F (n) is calculated. In the next step S22, a fuel property determination value is calculated based on the weighted average value g _F (n). In the next step S23, the fuel injection amount is calculated from the map of FIG. 11 based on the fuel property determination value. In the next step S24, the fuel injection amount is controlled based on the fuel injection amount calculated in step S23. After step S24, the process ends (RETURN).

As described above, according to the third embodiment, since the determination value for determining the fuel property is calculated from the weighted average value g _F (n) calculated for each fuel property map, the determination value and the fuel The fuel injection amount can be optimally controlled based on the map that defines the relationship with the injection amount.

  In each of the embodiments described above, the properties of the fuel are determined based on the angular acceleration acquired at the time of starting the engine and the indicated torque. However, using the characteristic values representing the combustion state such as in-cylinder pressure and ion current, The properties (combustion state) of the fuel may be estimated by performing statistical processing.

It is a figure for demonstrating the combustion state estimation apparatus of the internal combustion engine concerning each embodiment of this invention, and its surrounding structure. It is a characteristic view showing the relationship between the indicated torque, the torque due to in-cylinder gas pressure, the inertia torque due to reciprocating inertia mass, and the crank angle. It is a schematic diagram which shows the method of calculating | requiring the angular acceleration of a crankshaft. It is a schematic diagram which shows the map showing the relationship between friction torque, engine speed, and cooling water temperature. It is a figure which shows typically the method of acquiring a determination parameter | index. FIG. 5 is a characteristic diagram showing a determination index obtained by obtaining the distribution of angular acceleration for each fuel property (heavy, light) and for each number of explosions from the start. It is a schematic diagram which shows the example which standardizes distribution of angular acceleration. 3 is a flowchart illustrating a processing procedure in the first embodiment. It is a characteristic view which shows the weighted map showing the certainty of a fuel property. 10 is a flowchart illustrating a processing procedure in the second embodiment. A determination value that is determined based on the detected value g _F, is a characteristic diagram showing a map defining the relationship between the fuel injection amount. It is a schematic diagram which shows the other example of the map of FIG. 12 is a flowchart illustrating a processing procedure in the third embodiment.

Explanation of symbols

10 Internal combustion engine 38 Crank angle sensor 40 ECU

Claims (5)

An apparatus for estimating the combustion state of an internal combustion engine,
Storage means for preliminarily storing the statistical distribution of crank angular acceleration at the time of engine start according to the properties of the fuel;
Angular acceleration calculating means for calculating crank angular acceleration at engine start;
Combustion state estimation means for estimating the properties of the fuel by applying the crank angular acceleration calculated at the time of engine start to the statistical distribution;
A combustion state estimation device for an internal combustion engine, comprising: Detection value calculation means for calculating a detection value serving as an index representing a distribution state in the statistical distribution of the crank angular acceleration calculated at the time of starting the engine;
The combustion state estimation device for an internal combustion engine according to claim 1, wherein the combustion state estimation means estimates a property of the fuel based on the detected value. The storage means stores the statistical distribution for each number of explosions since engine start,
The detection value calculation means calculates the detection value from each of the crank angular accelerations calculated for each number of explosions based on the statistical distribution for each number of explosions,
A weighted average value calculating means for calculating a weighted average value of the detected values;
The combustion state estimation device for an internal combustion engine according to claim 2, wherein the combustion state estimation means estimates a property of the fuel based on the weighted average value. Determination value calculating means for calculating a determination value for controlling the fuel injection amount based on the estimated fuel property and the weighted average value;
Fuel injection amount control means for controlling the fuel injection amount based on the determination value;
The combustion state estimating device for an internal combustion engine according to claim 3, further comprising:   5. The angular acceleration calculation means calculates the crank angular acceleration as an average value in a crank angle section where an average value of inertia torque due to a reciprocating inertia mass is substantially zero. The combustion state estimation apparatus of the internal combustion engine described in 1.

Images