JP5171738B2 - Electric throttle characteristic learning control device and method - Google Patents

Description

  The present invention relates to an engine control apparatus having an idle stop function for automatically stopping / starting an engine, and to an electric throttle characteristic learning control apparatus and method for learning electric throttle characteristics.

  The driving force (engine torque) generated in the engine is mainly determined by the amount of air taken into the engine. The intake air amount, that is, the engine generated torque is controlled by adjusting the valve opening degree of an electrically controlled throttle valve (hereinafter referred to as ETC) arranged in the intake pipe of the engine according to the target engine torque. At this time, the engine control device previously includes the relationship between the ETC opening degree and the intake air amount as an ETC characteristic in the device, calculates the intake air amount for realizing the target engine torque, and then uses the ETC opening as a reference. And the ETC opening is adjusted by an electrical signal. Since this ETC characteristic changes due to machine differences and deterioration over time, it is common to improve the engine torque realization accuracy by correcting the deviation from the characteristic provided in advance by learning using the idle operation state. Known in the art. Further, as the ETC characteristic learning described above, the deviation of the ETC characteristic is learned from the correlation of the intake air amount with respect to the ETC opening degree by using the intake air amount detected by the air flow sensor provided in the engine. The intake air amount detected by the airflow sensor in a transient state where the engine speed and the load change has a phase lag due to the intake pipe volume, so that a stable operating state, that is, an idle operating state from the viewpoint of securing a correlation with the ETC opening degree. It is common to be executed in

JP 2007-92711 A

  In a vehicle with an idle stop function including a hybrid vehicle in recent years, the engine is stopped and the idle stop is performed in a scene where the conventional idle operation is continued, and therefore the idle operation period is shortened.

  However, in these vehicles with the idle stop function, it is difficult to provide a learning opportunity for the conventional ETC characteristics because there is no or short period for idling operation. Due to the decrease in learning opportunities, there is a concern that engine torque realization accuracy may deteriorate due to characteristic deviation due to deterioration over time.

  Also, it is not desirable to prolong idle operation for ETC characteristic learning because it results in fuel consumption and exhaust deterioration.

  An object of the present invention is made in view of deterioration in torque realization accuracy due to a decrease in an opportunity to learn ETC characteristics in a vehicle with an idle stop function, and ETC learns even in a vehicle that does not have an idle state or has a shorter period of time than conventional ones. An object of the present invention is to provide an ETC characteristic learning control device and a learning method that secure an opportunity.

  The ETC characteristic learning control apparatus of the present invention has an electric throttle characteristic of an internal combustion engine having an electric throttle valve that adjusts the amount of air flowing into the intake pipe and a pressure sensor that directly or indirectly detects the pressure in the intake pipe. The learning control device includes an intake pipe negative pressure predicting unit that predicts the pressure in the intake pipe, and uses the intake pipe negative pressure detected by the pressure sensor and the estimated intake pipe negative pressure to control the electric throttle. An electric throttle characteristic learning control device that learns a characteristic of an air flow rate with respect to a valve opening degree.

  More specifically, learning permission determining means for determining whether or not to learn ETC characteristics based on signals such as engine speed, ETC opening, and water temperature, and intake air predicted in the operation scene / region. A target negative pressure calculating means for calculating the negative pressure in the pipe is provided, and the actual intake pipe negative pressure detected by a pressure sensor attached to the engine is compared with the target negative pressure to deviate from the ETC characteristic provided in advance. It comprises ETC characteristic learning means for calculating and correcting.

  The ETC characteristic learning value calculated by the ETC characteristic learning means is reflected in the target ETC opening degree calculating means for calculating the ETC opening degree according to the target torque value required for the engine, and the desired ETC opening degree, i.e., suction. Control air volume.

  In a transient state where the engine shifts from partial operation range to idle operation, such as when the vehicle is decelerating, the intake air amount calculated from the air flow sensor is relative to the amount of air that is truly inhaled due to phase lag by the intake pipe volume or engine speed change. Due to the deviation, the conventional ETC characteristic learning has been executed after sufficiently shifting to idle operation. On the other hand, in the present invention in which ETC characteristic learning is performed in the dimension of the intake pipe negative pressure, the ETC opening degree and engine rotation in the operating state are in a state other than the idle operation such as a transient state during vehicle deceleration. The target negative pressure predicted from the number is calculated and the ETC characteristic deviation is learned by comparison with the actual intake pipe negative pressure detected by the pressure sensor. It is possible to secure an opportunity for ETC characteristic learning.

  By performing ETC characteristic learning even in the operating state other than the steady state such as the idle state using the intake pipe negative pressure, the deviation of the ETC characteristic is reduced, and the output torque required for the engine is improved. can do.

1 shows an embodiment of an ETC characteristic learning control device according to the present invention. The whole control system composition of cylinder injection engine 213. The main part of the engine control unit 216. The time chart example of ETC learning on the basis of the amount of intake air. Example 1 of time chart of ETC learning using intake pipe negative pressure. Intake pipe negative pressure waveform during engine deceleration. Example 2 of time chart of ETC learning using intake pipe negative pressure. An example of the ETC characteristic learning value (deviation correction allowance) for the difference between the target negative pressure and the actual negative pressure. Correlation between air flow rate and ETC opening. Example of learning value calculation time chart based on multiple learning experiences. Time chart when the engine speed during deceleration is kept constant. Correlation between load value and ETC opening at the time of predetermined rotation.

  Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.

  FIG. 1 shows an embodiment of an ETC characteristic learning control apparatus according to the present invention. The engine control unit 216 includes a learning permission determination unit a1 that determines whether or not to learn ETC characteristics based on signals such as the engine speed, the ETC opening degree, and the water temperature, and is predicted in the operation state. The target negative pressure calculating means a2 for calculating the negative pressure in the intake pipe and the actual intake pipe negative pressure detected by the pressure sensor attached to the engine are compared with the target negative pressure to obtain a deviation from the ETC characteristic provided in advance. It comprises ETC characteristic learning means a3 for calculating and correcting. The ETC characteristic learning value calculated by the ETC characteristic learning means a3 is reflected in the target ETC opening degree calculation means a4 that calculates the ETC opening degree in accordance with the target torque value required for the engine, and the desired ETC opening degree, that is, Control the amount of intake air.

  The target ETC opening calculation means a4 is provided with an ETC characteristic in advance, and after calculating the necessary intake air amount according to the target torque value, the ETC opening is calculated using the ETC characteristic. An example of the ETC characteristic is shown in FIG. 9 as a correlation between the air flow rate and the ETC opening. The ETC characteristic indicated by the solid line is provided in advance in the engine control unit 216, and the ETC opening degree can be uniquely determined from the required air flow rate in a predetermined rotational speed state. Here, since the actual ETC characteristic has a machine difference variation X as shown by a broken line, it is necessary to learn the ETC characteristic. In addition, the characteristics generally change in the flow rate decreasing direction with respect to the ETC opening indicated by the arrow when the deposit is deteriorated over time. In order to maintain the engine torque realization accuracy, the ETC characteristics are continuously learned. Correction needs to be performed.

  The learning permission determination unit a1 determines whether or not the current operation state is a state in which the intake pipe negative pressure can be predicted based on various engine control parameters (engine speed, ETC opening, water temperature, etc.). Further, when the ETC opening is determined in advance when the vehicle is decelerated and the intake pipe negative pressure can be uniquely predicted by the engine speed or the like, it is determined that learning is permitted. The target negative pressure calculating means a2 calculates the intake pipe negative pressure predicted or expected in this state as the target negative pressure. The ETC characteristic learning means a3 compares the actual intake pipe negative pressure actually detected by a pressure sensor or the like with the target negative pressure and calculates a learning value.

  FIG. 8 shows an example of an ETC characteristic learning value (deviation correction allowance) for the difference between the target negative pressure and the actual negative pressure (differential pressure ΔP). The dimension of the pressure is defined as 0 kPa at -1 atm and 103 kPa at the atmospheric pressure, and the state in which the intake pipe negative pressure is generated is described as 103 kPa or less. When the target negative pressure minus the actual negative pressure is the differential pressure ΔP, the differential pressure is positive, indicating that the actual negative pressure is smaller than the target, that is, the negative pressure is generated more than expected. Indicates that the actual characteristic has a smaller flow rate than the ETC characteristic value (the relationship between the ETC opening degree and the air flow rate) prepared in advance. At this time, the characteristic deviation can be absorbed by correcting / learning the learning value in the opening direction of the ETC opening as shown in FIG. On the contrary, when the differential pressure ΔP is on the negative side, the flow rate is larger than the ETC opening, and the negative pressure does not develop. The learning value is corrected and learned in the closing direction. Do. In this example, the table shows the relationship between the differential pressure ΔP and the learning value, but the learning value may be derived by calculation using the comparison result between the target negative pressure and the actual negative pressure.

  By performing ETC characteristic learning in the pressure dimension in an operation state in which the intake pipe negative pressure can be predicted in advance even in an idle operation state in this way, the vehicle with an idle stop function with few idle operation opportunities can be used. The opportunity and frequency of ETC characteristic learning can be secured.

  FIG. 2 shows an overall configuration of the present embodiment, taking a control system for the in-cylinder injection engine 213 as an example in the present embodiment. This system includes a high-pressure fuel pump 201. The intake air introduced into the cylinder 229 is taken in from the inlet 219 of the air cleaner 220, passes through an air flow sensor 218, which is one of the engine operating state measuring means, and an electric throttle valve (ETC 224) that controls the intake flow rate. It enters the collector 223 through the accommodated throttle body 221. The air flow sensor 218 outputs a signal representing the intake air flow rate to an engine control unit 216 that is an engine control device.

  The throttle body 221 includes a throttle sensor 217 that is one of engine operating state measuring means for detecting the opening of the ETC 224. The throttle sensor 217 outputs the throttle opening to the engine control unit 216. The engine control unit 216 outputs a drive signal to the motor 222 attached to the throttle body 221 based on the target torque required for the engine, opens and closes the ETC 224, and controls the intake air amount.

  The air sucked into the collector 223 is distributed to the intake pipe 225 connected to the cylinder 229 of the in-cylinder injection engine 213 and then guided to the combustion chamber 228 in the cylinder 229. Although FIG. 2 shows only one cylinder for simplification, even a plurality of cylinders does not depart from the spirit of the present invention.

  A negative pressure sensor 226 attached to the collector 223 outputs a signal indicating the negative pressure in the collector 223 (intake pipe) to the engine control unit 216.

  On the other hand, fuel such as gasoline is primarily pressurized from the fuel tank 205 by the fuel pump 204 and regulated to a constant pressure (eg, 0.3 MPa) by the fuel pressure regulator 203. Thereafter, the secondary pressure is increased to a higher pressure (for example, 5 MPa) by the high-pressure fuel pump 201 and is pumped to a fuel pressure accumulation chamber (not shown).

  The high-pressure fuel stored in the fuel pressure storage chamber is injected from the injector 214 provided in the cylinder 229 into the combustion chamber 228 at a predetermined amount and at a predetermined timing based on the drive signal of the engine control unit 216. The fuel injected into the combustion chamber 228 is ignited by the spark plug 215 by an ignition signal that has been increased in voltage by the ignition coil 211 at a timing similarly controlled by the engine control unit 216.

  A cam angle sensor 207 attached to the cam shaft of the exhaust valve outputs a signal for detecting the phase of the cam shaft to the engine control unit 216. Here, the cam angle sensor may be attached to the camshaft on the intake valve side. Further, a crank angle sensor 230 is provided on the crankshaft shaft to detect the rotation and phase of the crankshaft of the engine, and the output is output to the engine control unit 216.

  An A / F sensor 208 provided upstream of the catalyst 210 in the exhaust pipe 209 detects exhaust gas and outputs a detection signal to the engine control unit 216.

  As shown in FIG. 3, the main part of the engine control unit 216 includes an MPU 302, a ROM 301, a RAM 303, an I / O LSI 304 including an A / D converter, and the like. The engine control unit 216 includes a crank angle sensor 230, a cam angle sensor 207, a fuel pressure sensor 206, a throttle sensor 217, an accelerator sensor 231, an A / F sensor 208, which are one of means for measuring (detecting) the operating state of the engine. Predetermined arithmetic processing for the purpose of fuel injection amount control, ignition timing control, intake air amount control, and fuel pressure control by a high-pressure pump is performed by taking in signals from various sensors such as a negative pressure sensor 226 and an air flow sensor 218 as input. This is executed, and various control signals calculated as the calculation results are output. Injector 214, ignition coil 211, ETC motor 222, and high-pressure fuel pump solenoid 201 operate in response to a control signal from engine control unit 216.

  As described above, there are various types of analog sensors in engine control, and each of them is used after being processed by signal processing such as a filter and a mask in accordance with an object to be controlled.

  Further, a shift range information 500 by a driver operation detected by a control unit other than the engine control unit 216, an idle stop execution request (I) from a control unit that supervises engine control, motor control, and battery control in a hybrid vehicle or the like. / S request 501), idle stop prohibition request (I / S prohibition request 502) indicating a state in which idle stop control is prohibited as engine control, etc., is transmitted and received by inter-unit communication such as CAN communication and Flex Ray communication. Done.

  FIG. 4 shows a time chart example of ETC characteristic learning based on the intake air amount by the conventional airflow sensor. When the required torque to the engine becomes unnecessary, such as when the vehicle enters a deceleration state (time T0), the engine control unit 216 controls the ETC opening to the fully closed position. Although not shown, fuel cut is also performed as necessary, and the engine speed starts to decrease. Since the ETC opening is closed to the fully closed position, the intake air amount Q decreases, the load value (actual intake air amount) calculated by Q / N moves as shown in the figure, and the engine speed is idle. After reaching several ranges (time T1), the load value is in a stable state (after time T2). Conventionally, an ETC characteristic learning region is set after time T2 when the load value is stabilized.

  As an example of an actual learning method, FIG. 12 shows a correlation (ETC characteristic) between a load value (intake air amount) at a predetermined rotation and an ETC opening degree. When the actual load value calculated from the actual intake air amount is La when the ETC opening degree Et is controlled using the ETC characteristic with respect to the required load (air amount) Lt in the idle operation, If the ETC opening degree is estimated with the load value La, the opening degree Ee is obtained, and it is determined that the ETC characteristic is shifted in the opening direction with respect to the target Et. Using the difference between the demand and the actual load value, the apparent ETC characteristic is read as a characteristic indicated by a broken line, and the target ETC opening relative to the demand load value is Et1. As a result, the actual load value La approximates the required value Lt, and the difference between Et and Et1 becomes the ETC characteristic learning allowance. Since the intake air amount detection indicating the actual load is shifted in the transient state, the ETC characteristic learning is executed using the idle state. However, learning cannot be performed because the engine stops in an idle stop compatible vehicle. On the other hand, the intake pipe negative pressure (intake pressure) detected by the pressure sensor is gradually approaching a desired pressure value in the vicinity of time Ta even in a transient state where the engine speed changes. It is characterized by performing learning using.

  FIG. 5 shows a time chart of ETC characteristic learning using the intake pipe negative pressure of this embodiment. In the conventional learning using the load value, the time T2 and later are set as the learning region. However, in the idle stop compatible vehicle, the engine is stopped before the time T2 and learning is impossible. On the other hand, in the present embodiment, a period from the time Ta after the intake pipe negative pressure responds (stable) after the ETC opening degree is controlled to the fully closed position until the engine stops is defined as a learning region. Therefore, an ETC characteristic learning opportunity can be secured even in a vehicle having a short idle operation time or not present.

  The example in which the ETC characteristic learning is performed using the intake pipe negative pressure in a transient state such as when the vehicle is decelerated has been described so far. However, when the engine is decelerated, the intake pipe negative pressure may be set or controlled to be a predetermined intake pipe negative pressure in advance. It is general, and ETC characteristic learning can be performed by comparing the setting or control target value as a target negative pressure with the actual negative pressure.

  FIG. 6 shows a specific intake pipe negative pressure waveform during engine deceleration. When the required torque for the engine disappears, the ETC opening is controlled in the closing direction in order to reduce the amount of intake air. Here, when the opening shown by the broken line, ie, the fully closed position is closed, the intake pipe negative pressure may develop excessively (negative pressure increases) because the engine speed is still high at the early stage of engine deceleration. If the negative pressure develops too much, so-called oil rise / fall phenomenon that sucks engine oil from the gap between the intake / exhaust valve and the piston ring occurs, and there is a concern that exhaust deterioration and engine damage may occur. It is necessary to manage or control the negative pressure during engine deceleration. Specifically, the ETC opening that realizes the required torque and the ETC opening for the negative pressure management during deceleration are calculated separately, and the larger of the two is selected as the final target ETC opening. It is common to achieve this. The broken line shown in FIG. 6 corresponds to the ETC opening degree that achieves the required torque, and the solid line corresponds to the ETC opening degree for the negative pressure management. Even when there is no torque request, even if the ETC opening for torque realization requires a fully closed position, by selecting the ETC opening in the opening direction so that the negative pressure does not develop too much by the negative pressure management part, The intake pipe negative pressure is maintained at a predetermined negative pressure value. The ETC characteristic learning can be executed if the region is maintained and managed at such a desired negative pressure value.

  Moreover, it is known that the ETC opening corresponding to the negative pressure management during deceleration is set for each engine speed, and an ETC opening capable of holding a desired negative pressure is set according to the engine speed. For this reason, as shown in FIG. 7, the ETC opening also changes in the process of decelerating the engine speed. If learning is performed at the times Ta, Tb, and Tc shown in FIG. It is possible to detect deviations in ETC characteristics, and learning accuracy / reliability can be improved by setting values such as average characteristic deviation or maximum / minimum values at each learning point as final learning values. . Furthermore, it is possible to calculate the final learning value from the learning value obtained in each deceleration transition after experiencing a plurality of learning opportunities instead of only the learning value in one deceleration transient state. . FIG. 10 shows an example of a learning value calculation time chart based on a plurality of learning experiences. Each learning value in learning opportunities A, B, and C in the figure is stored, and a final learning value is calculated based on the stored learning value when a predetermined opportunity learning is experienced or a condition is satisfied. Thus, the accuracy and reliability of the learning value itself can be ensured.

  Although it has been explained that idle operation is not required by updating the learning value in the transient state using the intake pipe negative pressure, provisional learning using the intake pipe negative pressure is executed in the transient state, and the ETC characteristic deviation It may be used with the intention of only determining the necessity of learning. When it is determined that the characteristic deviation occurs in the transient state, the idle stop operation is prohibited, the idle operation is ensured as before, and the ETC characteristic learning is executed in that state. ETC characteristic deviation due to intake pipe negative pressure is determined between times Ta and T1 in FIG. 4, and when the target differential and actual pressure difference due to intake pipe negative pressure is greater than or equal to a predetermined value, the idle operation range after time T2 is continued. Control to ensure. Although there are idle driving points, it is only when it is determined that there is a need for learning in a transient state, and the frequency / opportunity of idle driving itself can be reduced, and fuel economy rebound can be minimized. Is possible.

  In this case, the determination of the necessity of learning in the transient state does not limit the determination parameter as in the determination using the intake pipe negative pressure, but the ETC characteristic deviation, that is, the air amount, like the air-fuel ratio sensor value in the deceleration transient state. Any parameter can be used as long as the deviation and the correlation can be determined.

  FIG. 11 shows a time chart when the engine speed during deceleration is kept constant in order to secure an ETC characteristic learning opportunity. In a system having a driving force source other than an engine such as a hybrid vehicle, it is possible to control the deceleration force simultaneously with the driving force with a motor or the like. In FIG. 11, when the vehicle is decelerating, the vehicle deceleration is conventionally secured by so-called engine braking using engine friction, but the target acceleration (deceleration) is set and the engine speed is kept constant. An example is shown in which a possible section is provided and the vehicle deceleration in that section is realized by controlling the driving force of the motor. In this way, even if there is no idle speed range, a stable speed range is provided for a certain section while the vehicle is decelerating, and ETC characteristic learning is performed in that section, thereby ensuring learning frequency and opportunity and fuel consumption performance Improvements can be realized.

201 High Pressure Fuel Pump 203 Fuel Pressure Regulator 204 Fuel Pump 205 Fuel Tank 206 Fuel Pressure Sensor 207 Cam Angle Sensor 208 A / F Sensor 209 Exhaust Pipe 210 Catalyst 211 Ignition Coil 213 In-Cylinder Injection Engine 214 Injector 215 Spark Plug 216 Engine Control Unit 217 Throttle Sensor 218 Air flow sensor 219 Inlet portion 220 Air cleaner 221 Throttle body 222 Motor 223 Collector 224 ETC
226 Negative pressure sensor 228 Combustion chamber 229 Cylinder 230 Crank angle sensor 231 Accelerator sensor 301 ROM
302 MPU
303 RAM
304 I / O LSI
500 Shift range information

Claims (3)

Used in vehicles with an idle stop function and control, an electric throttle valve that adjusts the amount of air flowing into the intake pipe, a pressure sensor that directly or indirectly detects the pressure in the intake pipe, and an intake air amount in the intake pipe In an electric throttle characteristic learning control device for an internal combustion engine having a flow meter for detecting,
An intake pipe negative pressure predicting means for predicting the pressure in the intake pipe,
Using the intake pipe negative pressure detected by the pressure sensor and the predicted intake pipe negative pressure, the characteristic of the air flow rate with respect to the valve opening of the electric throttle is learned in the deceleration transient state, and a characteristic deviation occurs. An electric throttle characteristic learning control device that prohibits idle stop control and executes learning of electric throttle characteristics in an idle operation state when it is determined that Used in vehicles with an idle stop function and control, an electric throttle valve that adjusts the amount of air flowing into the intake pipe, a pressure sensor that directly or indirectly detects the pressure in the intake pipe, and an intake air amount in the intake pipe In an electric throttle characteristic learning control device for an internal combustion engine having a flow meter for detecting,
An intake pipe negative pressure predicting means for predicting the pressure in the intake pipe;
Using the intake pipe negative pressure detected by the pressure sensor and the predicted intake pipe negative pressure to learn the characteristics of the air flow rate with respect to the valve opening of the electric throttle,
The learning using the intake pipe negative pressure is based on the control throttle characteristic compared with the actual negative pressure based on the target negative pressure value provided to prevent excessive development of the intake pipe negative pressure during the control of the throttle control. An electric throttle characteristic learning control device characterized by learning a deviation. The learning value using the intake pipe negative pressure according to any one of claims 1 and 2 is a result obtained by calculating a plurality of times in a continuous learnable state and a result obtained by calculating in each state in a discontinuous learnable state. A learning throttle characteristic learning control apparatus and method, wherein a learning value is determined / updated by one or a combination of both.

Images