JP2006312945A - Throttle valve control device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a throttle valve control device further improving its safety and reliability and reducing manufacturing costs by revising a value even if a full-close position is erroneously learned since the full-close position is learned on the basis of results of monitoring of output signals of accelerator sensors of multiple system. <P>SOLUTION: This throttle valve control device of an engine has: the accelerator sensors of multiple system detecting an opening of an accelerator pedal; and a throttle valve driven to be opened and closed through a motor on the basis of the output signals of the accelerator sensors. The device further comprises: a means for learning a full-close voltage of the accelerator sensors; and a means for determining each idle switch to the accelerator sensors of the multiple system. The means for learning the full-close voltage of the accelerator sensors upwardly revises a learned value of the full-close point voltage of the accelerator sensor on the side on which the idle-switch-off is determined, and learns the value, when a mismatching is found in on/off determinations of the idle switches by the means for determining the idle switch. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to a throttle valve control device, and more particularly to a throttle valve control device that corrects a learned value of a full-close point voltage of an accelerator sensor so that reliability and safety can be further improved.

  Conventionally, there is known an electronic throttle valve control device that detects an accelerator pedal depression amount with an accelerator sensor, sets a target throttle opening according to the detected amount, and controls driving of the throttle valve with a motor. An electronic throttle valve control device controls the opening of a throttle valve that adjusts the engine output to improve engine operating performance.

  Further, in the control of the opening degree of the throttle valve, the reliability of the accelerator operation is improved by constituting a so-called multiplex system in which a plurality of accelerator sensors and throttle sensors are provided.

  Here, the electronic throttle valve control device learns a full-close point voltage signal or a reference voltage signal of the accelerator sensor, which is a basis of an opening degree command from the throttle valve control device, to accurately control the throttle valve. Various techniques for achieving the above have been proposed (see, for example, JP-A-10-297314, JP-A-11-36896, JP-A-10-9036, JP-A-10-176582).

  In the technique described in Japanese Patent Laid-Open No. 10-297314, the fully closed point voltage of the accelerator sensor is stored, and when the subsequent read value is smaller than the stored value, the stored value is used as the read value. An accelerator opening detection system that is updated each time, and the technology described in Japanese Patent Laid-Open No. 11-36896 drives a throttle valve in consideration of detection errors, mechanical and electrical errors in the sensor. When the motor is locked, the throttle valve control device sets the opening of the throttle valve at this time to the valve fully closed position. The technology described in Japanese Patent Laid-Open No. 10-9036 discloses a plurality of accelerator sensors. Are connected to a common ground, and offsets are provided on the maximum value side and the minimum value side of the output characteristics of the plurality of sensors to diagnose sensor abnormalities and adjust the offset. A sensor abnormality diagnosing apparatus that corrects all the reference points of a plurality of sensors. The technique described in Japanese Patent Laid-Open No. 10-176582 is disclosed in order to improve the detection accuracy of sensor abnormality. The throttle valve control device learns a voltage value, takes the deviation, and compares the deviation with a threshold value.

  Various technologies have been proposed for the electronic throttle valve control device that have a fail-safe function against abnormalities in the multiple accelerator sensors and throttle sensors (for example, JP 2000-54868, JP 2000-97087, JP-A 2000-97087, etc.).

  The technique described in Japanese Patent Laid-Open No. 2000-54868 is capable of running with the minimum output by balancing the springs provided between the sensors when one of the multiple sensors fails. The technology described in Japanese Patent Application Laid-Open No. 10-238389 discloses a fail-safe property and a sensor abnormality in the case where a signal having the same level as that at the time of abnormality is generated in the throttle sensor due to noise, instantaneous interruption, or the like. In order to prevent erroneous detection, a determination delay period is provided. The technique described in Japanese Patent Laid-Open No. 2000-97087 learns the fully closed position of the multi-system sensor, and the throttle sensor and the accelerator sensor fail. In this case, the fuel cut condition is set according to the failure state.

  By the way, the signal of the accelerator sensor input to the throttle control system is a starting point of the opening command of the throttle valve control device, and in the engine control calculation unit of the throttle valve control device, By further improving the reliability of the learning value of the closing point voltage (full closing learning value), the safety of the throttle valve control device can be improved. In addition, the safety of the throttle valve control device has been studied from all aspects, and the above-described fail safe has been established.

  FIG. 18 shows an input circuit of a multiple accelerator sensor in the engine control arithmetic unit ECM. In the input circuit, each accelerator sensor APS1 is used as a fail safe when the harness between the connectors is disconnected. , APS2 are connected to pull-down resistors R1, R2, respectively.

  As a result, when the harness is disconnected, the input voltage of the accelerator sensors APS1 and APS2 can be set to 0 (V), that is, the accelerator pedal is not depressed, and danger can be avoided in the event of an abnormality. Yes. Therefore, the pull-down resistors R1 and R2 are essential from the viewpoint of safety of the throttle valve control device.

  On the other hand, the pull-down resistors R1 and R2 are affected by contact resistance due to deterioration over time between terminals used for connection to the engine control arithmetic unit ECM, accelerator sensors APS1 and APS2, and the harness. Due to the fluctuation, there is a problem that the signals of the accelerator sensors APS1 and APS2 fluctuate.

In addition, although it is impossible in a normal connection state, when the contact state of the terminal becomes extremely unstable, or when an oxide film is formed on the contact surface between the terminals, the contact resistance between the terminals increases or fluctuates. To do. Since the contact resistance between the terminals can be further varied due to the influence of vehicle vibration or the like, the output signals of the accelerator sensors APS1 and APS2 are in a so-called indeterminate state that varies independently of the driver's intention. There is a problem.

  FIG. 19 illustrates a high idle state due to false learning of the full-close point voltage when an indefinite state occurs in the output signal of the accelerator sensor.

  When the contact resistance between the terminals increases, the outputs of the accelerator sensors APS1 and APS2 recognized by the engine control calculation unit ECM are further reduced even though the depression amount of the accelerator pedal is zero. When the learning update delay period elapses stably while this output remains low, the learning condition is satisfied, and the learning value is updated with the decreased output as the full-close point voltage of the accelerator sensors APS1 and APS2. This learning value update operation is erroneous learning because it learns an abnormal signal.

  Thereafter, when the contact state between the terminals recovers and the contact resistance decreases, the outputs of the accelerator sensors APS1 and APS2 rise and return despite the depression amount of the accelerator pedal being zero, The engine control calculation unit ECM recognizes the deviation between the voltage level of the restored signal and the erroneously learned fully closed voltage level as the depression amount of the accelerator pedal, and controls the throttle valve to the open side. It can be seen that a high idle state occurs in which the engine speed increases even though the driver does not step on the accelerator pedal.

  Note that if the acceleration sensor signal drop is so great that it exceeds the lower limit value of the voltage abnormality determination, or if a deviation occurs that exceeds the determination value of the mutual deviation diagnosis of the multiplex system sensor, the diagnosis is regarded as abnormal. Then, the operation shifts to the fail-safe operation. However, even within the determination threshold based on the diagnosis, when the non-idle determination region changes, the high idle state still occurs.

  20 and 21 show the measurement results of the accelerator sensor output when the fully closed point voltage is erroneously learned.

  As will be described later, the fully closed learning in the conventional engine control arithmetic unit ECM provided with the multiple accelerator sensors APS1 and APS2 does not monitor the output signals of the accelerator sensors APS1 and APS2 with each other. When the output signal of the accelerator sensor fluctuates, the fully closed learning is performed. As shown in FIG. 20, when the output of only the accelerator sensor APS1 fluctuates due to a poor contact state. While the signal is fluctuating, the learning condition is not satisfied and the learning value is not updated. However, when the output of only the accelerator sensor APS1 is stable while being abnormal, the fully closed learning value of the accelerator sensor APS1 is It can be seen that the fully closed learning value of the accelerator sensor APS2 is updated even though it is not updated.

  Further, when the accelerator sensor output is measured over a longer period of time, it can be seen that the fully closed learning value of the accelerator sensor APS1 may be updated one after another as shown in FIG.

  On the other hand, FIG. 22 shows the measurement result of the fluctuation of the accelerator sensor output, and the time when the output signal drops and is stable may last for a long time (about 10 seconds in this figure). Therefore, it is not possible to solve the problem simply by providing a learning update delay period. In this case, it can also be prevented by writing each fully closed point voltage value of each accelerator sensor in the ROM, but there is a problem in terms of manufacturing cost.

  That is, the inventor monitors each output signal of the accelerator sensor and sets a learning condition based on the monitoring result, that is, by providing a guard for all accelerator sensors based on the monitoring result of each other, Although new knowledge that further improvement of the safety and reliability of the valve control device can be achieved has been obtained, none of the above-mentioned conventional techniques give special consideration to the above points.

  The present invention has been made in view of such problems, and the object of the present invention is to monitor the output signals of multiple accelerator sensors from each other and learn the fully closed position based on the monitoring results. An object of the present invention is to provide a throttle valve control device that can achieve further improvement in safety and reliability and reduction in manufacturing cost by correcting the learning value even when erroneous learning is performed.

  In order to achieve the above object, a throttle valve control device according to the present invention includes an accelerator pedal, a multiple accelerator sensor that detects the opening degree of the accelerator pedal, and a motor that opens and closes based on an output signal of the accelerator sensor. In the throttle valve control device for an engine having a driven throttle valve, the control device learns a fully closed voltage of the accelerator sensor, and determines a respective idle switch for the accelerator sensor of the multiplex system. And the means for learning the fully closed voltage of the accelerator sensor determines that the idle switch is off when the means for determining the idle switch has a mismatch in the on / off determination of each idle switch. Correct the learning value of the full-close point voltage of the accelerator sensor on the other side upward, and learn that value. It is a symptom.

  The throttle valve control device of the present invention configured as described above raises the learning value of the full-close point voltage of the accelerator sensor on the side on which the idle switch is determined to be off when an inconsistency occurs in the idle determination. Therefore, even if the learning value of the full-closed point voltage is learned by mistake, it can be corrected to a normal value and the occurrence of a high idle state can be prevented.

  Further, in a specific aspect of the throttle valve control device according to the present invention, the means for learning the fully closed voltage of the accelerator sensor corrects the learned value of the fully closed point voltage of the accelerator sensor upward after a predetermined delay time. In addition, the means for learning the value or learning the fully closed voltage of the accelerator sensor includes a detection signal of the accelerator sensor in which the idle switch is determined to be off and a signal of the accelerator sensor in which the idle switch is determined to be on. The difference from the detection signal is added to the learned value of the full-closed point voltage of the accelerator sensor in which the idle switch is determined to be off, and the learned value of the full-closed point voltage of the accelerator sensor is corrected upward, and the value It is characterized by learning.

  Further, in another specific aspect of the throttle valve control device according to the present invention, the control device includes means for determining each idle switch for the multiplex accelerator sensor, and the means for determining the idle switch includes: Comprehensive idle determination is performed based on the output signal of each idle switch, and when all the idle switches are determined to be off, the comprehensive idle determination is turned off, or the control device includes: When the comprehensive idle determination is turned on by the means for determining the idle switch, the target opening for the throttle valve is set to zero.

  In addition, the control device has a pull-down resistor in the input circuit of the multiple accelerator sensor, or the control device drives the motor and a multiple throttle sensor that detects the opening of the throttle valve. An engine throttle valve control device having a default mechanism that keeps the throttle valve at an initial opening when not, the control device based on an output signal from means for learning a fully closed voltage of the accelerator sensor It is characterized by having means for setting the target opening of the throttle valve and means for driving the motor based on the output signal of the means for setting the target opening.

  As can be understood from the above description, the throttle valve control device of the present invention monitors each output signal of the multi-system sensor, sets the learning condition based on the monitoring result, determines the idle state, or sets the learning value. Since upward correction is performed and learning and determination are performed when all of the multiple sensors are applicable, the occurrence of a high idle state due to fluctuations in the sensor signals of the multiple systems is prevented, and the reliability of the throttle valve control device is reduced. The safety and safety can be further improved.

  Hereinafter, an embodiment of a throttle valve control device of the present invention will be described in detail with reference to the drawings.

  FIG. 1 shows an overall configuration of an engine control device 1 including a throttle valve control device of the present embodiment. The engine control device 1 including the throttle valve control device includes an input I / F (A / D converter) 9, an engine control calculation unit (main engine control means) 2, and a throttle control calculation unit (throttle valve control means) 3. The vehicle speed signal and the engine speed signal are input to the A / D converter 9 from the vehicle speed detector 22 and the engine speed detector 23.

  An accelerator pedal 24 and two accelerator sensors 25, 26 are arranged in the accelerator unit 6, and the driver's intention is indicated by the amount of depression of the accelerator pedal 24 (accelerator depression amount). The angle is detected by the accelerator sensors 25 and 26 and input to the A / D converter 9. The accelerator sensors 25 and 26 are configured in a multiple system in order to improve the reliability of the accelerator operation, and in the present embodiment, are in a dual system as described above. The A / D converter 9 receives an accelerator depression signal from the accelerator unit 6 and a throttle opening degree detection signal from the throttle unit 7.

  The throttle unit 7 includes a throttle valve 18, a motor 14 and a gear 15 for rotationally driving the throttle valve 18, and a default mechanism 22, and throttle sensors 16 and 17 configured in a multiplex system are disposed. In this embodiment, a double system is used.

  The main engine control means 2 performs various control of the engine based on the signal from the A / D converter 9, and also performs failure diagnosis of the two accelerator sensors 25, 26, etc. The detection signals 25 and 26 are taken into the main engine control means 2 to diagnose the presence or absence of abnormality of the accelerator sensors 25 and 26. If the abnormality is detected, it is determined that the accelerator sensors 25 and 26 are faulty. On the other hand, if it is normal, after learning the fully closed point voltage of the accelerator sensors 25 and 26 based on the signals of the accelerator sensors 25 and 26, the accelerator depression amount is determined.

  That is, the main engine control means 2 recognizes the increase in the output voltage of the current accelerator sensors 25 and 26 with respect to the accelerator fully closed point voltage learned in advance as the accelerator depression amount, and the output voltage is the same as the fully closed point voltage. In this case, it is determined that the accelerator pedal 24 is not depressed by the driver. When the accelerator pedal 24 is depressed, the output voltage becomes a value larger than the full closing voltage.

  Then, after determining the opening degree of the accelerator pedal 24 based on the accelerator depression amount, the main engine control means 2 performs an idle SW (switch) determination to determine whether or not the driver depresses the accelerator pedal 24. And whether the engine is in an idle state or a non-idle state is determined. Then, the target opening of the throttle valve 18 corresponding to the accelerator opening is stored, the target opening (required throttle opening) of the throttle valve 18 is set, and the required throttle opening is set to the throttle valve. Output to the control means 3.

  The throttle valve control means 3 includes a throttle control calculation unit 4 and a motor drive circuit unit 5. The throttle control calculation unit 4 controls the drive of the motor 14 based on a signal from the A / D converter 9. And the failure diagnosis of the two throttle sensors 16 and 17 are also performed. The detection signals of the throttle sensors 16 and 17 are taken into the throttle control calculation unit 4 and the abnormality of the throttle sensors 16 and 17 is detected. The presence / absence is diagnosed, and if it is abnormal, it is determined that the throttle sensors 16, 17 are out of order. On the other hand, if it is normal, it is used as a parameter for engine fuel injection control or ignition control not shown in FIG. 1, and the actual throttle opening is determined based on the signals of the throttle sensors 16 and 17. Then, based on the deviation between the required throttle opening and the actual opening of the throttle valve 18, feedback (F / B) control is performed so that the required opening and the actual opening coincide with each other, and the throttle valve 18 is rotated. The driving amount of the motor 14 is calculated, the driving amount is output to the motor 14 of the throttle unit 7 through the motor driving circuit unit 5, and the throttle valve 18 is opened and closed through the motor 14 and the gear 15.

  In this series of operations, the main engine control means 2 constantly monitors whether the throttle control calculation unit 4 is functioning normally by comparing and verifying the required opening instruction and the actual throttle opening, On the other hand, the throttle valve control means 3 also constantly monitors whether or not the main engine control means 2 is functioning normally by comparing and verifying the required opening degree instruction and the accelerator opening degree. As a matter of course, the components such as the motor 14 and the gear 15 are also diagnosed, and if an abnormality is detected, the failure is determined according to a predetermined condition.

  FIG. 2 shows a control block diagram of the engine control means 2 including accelerator sensor diagnosis and control thereof.

  In FIG. 2, the accelerator depression amount detection signals output from the accelerator sensors 25 and 26 are converted by the A / D converter 9, and then the accelerator sensors 25 and 26 are disconnected or short-circuited by the accelerator first diagnostic means 31. If the diagnosis means 31 determines that the sensor position is normal, the accelerator sensor full-close learning means 32 learns and reads the sensor output value of the accelerator position when fully closed, and the accelerator sensor depression amount calculation means In step 33, the accelerator depression amount is calculated from the current output signal.

  Next, in the accelerator sensor second diagnostic means 34, correlation diagnosis such as comparison of output signals between the two accelerator sensors 25 and 26 is performed, and either one or both of the two accelerator sensors 25 and 26 are performed. If the diagnosis means 34 determines that the vehicle is malfunctioning and the like is normal, the accelerator pedal opening calculation means 35 calculates the accelerator based on the accelerator depression amount. Calculate the opening.

  Then, based on the accelerator opening, the accelerator depression amount opening calculation means 38 calculates the depression opening of the accelerator pedal 24 depressed by the driver's intention, and the idle switch determination means 36 determines whether the engine is in an idle state. The idle speed control amount opening calculating means 39 calculates the accelerator opening during idling. The accelerator opening at the time of idling includes outputs of an air conditioner, a generator, and a power steering oil pump that obtain power from the engine. Further, the change amount calculation means 37 calculates the change amount of the accelerator depression from the difference in the depression amount between the previous time and the current time, and the acceleration feeling correction opening degree calculation means 40 calculates the accelerator opening correction value according to the change amount. Is calculated.

  The requested throttle opening setting means 41 multiplies the values calculated by the accelerator depression opening calculation means 38, the idle speed control opening calculation means 39, and the acceleration sense correction opening calculation means 40, thereby multiplying the required throttle opening. A degree value is set, and the set value is output to the throttle valve control means 3.

  When the accelerator section 6 is diagnosed as abnormal by the accelerator first diagnosis means 31 and the accelerator sensor second diagnosis means 34, the abnormality control means 41 diagnoses various components, and as a result When the normal control of the throttle valve 18 is impossible and a failure is determined, the motor power supply relay 19 is shut off to stop the operation of the motor drive circuit 5 and the driver is abnormally turned on by turning on the diagnostic lamp 20. Is further output to the fuel injection / ignition timing control means 42 to cut the fuel injection (decrease the fuel amount), change the ignition timing, and the like.

  FIG. 3 shows a control block diagram of the throttle valve control means 3 including diagnosis and control of the throttle sensors 16 and 17.

  In FIG. 3, the detection signal of the opening degree of the throttle valve 18 output from the throttle sensors 16 and 17 is converted by the A / D conversion means 9, and then the throttle sensor first diagnosis means 51 of the throttle control calculation unit 4. When the diagnosis of the throttle sensors 16 and 17 such as disconnection / short circuit is made and the diagnosis means 51 determines that the sensor is normal, the diagnosis is output to the actual opening degree calculation means 53. The fully closed learning means 52 of the throttle sensors 16 and 17 reads and stores the sensor output value of the position of the throttle valve 18 when fully closed.

  In the actual opening calculation means 53, when the throttle sensor first diagnosis means 51 diagnoses that the throttle sensors 16, 17 are normal, the throttle value is calculated from the output value when fully closed and the current detection signal. The actual opening of each of the sensors 16 and 17 is calculated.

  Next, the second throttle diagnosis means 54 performs correlation diagnosis such as comparison of the output signals of the two throttle sensors 16 and 17, and one or both of the two throttle sensors 16 and 17 fail. Diagnose whether or not When it is determined that the diagnosis means 54 is normal, the actual opening degree of the throttle valve 18 is calculated based on the detection values of the two throttle sensors 16 and 17 by the actual throttle opening degree calculation means 55. The calculated actual opening of the throttle valve 18 is output to the main engine control means 2 and is also output to the return spring diagnosis means 59 and the actuator diagnosis means 60 for use in the diagnosis.

  The PID control means 56 calculates the correction value by PID control of the deviation between the actual opening of the throttle valve 18 and the required throttle opening calculated by the engine control means 2, and the motor DUTY calculation means 57 The DUTY value for driving the motor is calculated, and the motor 14 is driven via the motor driving circuit means 5 to rotate the throttle valve 18. The motor 14 itself is monitored by the monitoring circuit means 61, and the motor overcurrent diagnosis means 62 diagnoses whether or not an overcurrent flows through the motor 14.

  When the throttle first diagnosis means 51 and the throttle second diagnosis means 54 diagnose that the throttle unit 7 is abnormal, the abnormality control means 58 outputs to the motor drive circuit means 5 to stop the motor drive. At the same time, the abnormal state is output to the main engine control means 2.

  FIG. 4 is a characteristic diagram of the two accelerator sensors 25 and 26, and shows the output voltage values of the accelerator sensors 25 and 26 with respect to the operating angle of the accelerator pedal 24. FIG. 5 shows the two throttle sensors 16 and 17. In the characteristic diagram, the output voltage values of the throttle sensors 16 and 17 with respect to the operating angle of the throttle valve 18 are shown. FIG. 6 is a characteristic diagram of the motor 14 and shows the relationship between the current value with respect to the torque of the motor 14 and the rotational speed.

  In the present embodiment, when the abnormal state is diagnosed by the main engine control means 2 or the throttle valve control means 3 and the motor drive is instructed to stop, the protection function is activated. That is, the throttle unit 7 includes a default mechanism 22 (see FIG. 1), and the default mechanism 22 holds the throttle valve 18 at the default opening that is initially set when an abnormality occurs. The default opening is set such that the engine intake air amount can be secured within a range that allows the vehicle to travel while being parked and can be safely stopped by the brake of the vehicle.

  By the way, since the signals from the accelerator sensors 25 and 26 become the starting point of the opening degree command of the throttle valve control device of the engine control device 1, the main engine control means 2 fully closes the accelerator sensors 25 and 26. By further improving the reliability of the learning value of the point voltage (full closing learning value), the safety of the throttle valve control device can be further improved. Here, as described above, the input circuit of the multiple accelerator sensors 25 and 26 of the main engine control means 2 has a pull-down resistor that is indispensable as a fail-safe when the connector is disconnected or the harness between the connectors is disconnected. R1 and R2 are connected to each other. However, when the contact resistance between the terminals fluctuates, the signals of the accelerator sensors 25 and 26 fluctuate and become indefinite. , 26 is erroneously learned, and as a result, a high idle state occurs even though the driver does not actually step on the accelerator pedal 24. Therefore, in the main engine control means 2 of the engine control device 1 including the throttle valve control device of the present embodiment, as shown below, the accelerator sensor 25 is used to prevent erroneous learning of the full-close point voltage of the accelerator sensors 25 and 26. , 26 are mutually monitored, and the accelerator sensor full-close learning means 32 provides learning conditions based on the monitoring results.

  FIGS. 7 to 10 are diagrams for explaining the learning of the fully closed point voltage of the accelerator sensor in the main engine control means 2, FIG. 7 is a block diagram of a conventional fully closed learning, and FIG. 8 is a fully closed accelerator sensor. FIG. 9 is a block diagram of the depression amount calculation in the accelerator depression amount calculation means 33, and FIG. 10 shows the confirmation result of the accelerator sensor output voltage full closure learning in the main engine control means 2. FIG. ing.

  First, the accelerator sensor fully-closed learning means 32 performs learning in advance when the engine control apparatus 1 is turned on for the first time or immediately after replacement of the in-vehicle battery as an initial setting. In these cases, the value is still set. Since there is no opportunity to store, the fully closed learning values MLPPS1b and MLPPS2b of the accelerator sensors 25 and 26 are replaced with preset data values KDMPPS1 # and KDMPPS2 # in block 32A, respectively.

  When the ignition switch is turned on from off, the predetermined value APSFI # is added to the fully closed learning values MLPPS1b and MLPPS2b of the accelerator sensors 25 and 26 currently stored (backed up) in the block 32B. Every time the ignition switch is turned on, the fully closed learning operation is performed again. The fully closed learning value is provided with the upper and lower limits shown below.

[Equation 1]
MLPMIN1 # ≦ MLPPS1b ≦ MLBMAX1 #
MLPMIN2 # ≦ MLPPS2b ≦ MLBMAX2 #
Here, MLBMAX1 # indicates the upper limit value of the accelerator sensor (APS1) 25, MLBMIN1 # indicates the lower limit value, MLBMAX2 # indicates the upper limit value of the accelerator sensor (APS2) 26, and MLBMIN2 # indicates the lower limit value. .

  The accelerator sensor full-close learning means 32 monitors each output signal of the accelerator sensors 25 and 26 and provides a learning condition based on the monitoring result.

  That is, the conventional fully closed learning establishment condition is as shown in FIG. 7 when either one of the two types of accelerator sensors satisfies the following conditions (a) to (f). Whereas the fully closed learning condition is satisfied, the accelerator sensor fully closed learning means 32 of the present embodiment has an accelerator sensor (APS1) 25 of the following (a) as shown in FIG. In addition to (f) to (f), the conditions (k) and (l) are all satisfied, and the accelerator sensor (APS2) 26 satisfies the following conditions (a) and (b), and (g) to (l ), The fully closed learning condition is satisfied. When not established, learning value update references MLPREF1 and MLPREF2 described later are set to 0.

[Equation 2]
Condition (a): The starter switch is turned off. Condition (b): The battery voltage is a predetermined value or more (for example, 10 V or more).
Condition (c): APS1 voltage diagnosis result is normal Condition (d): MLBPPS1 <MLPPS1b
Here, MLBPPS1 is a buffer value of the input voltage of APS1, and MLPPS1b is a fully closed learning value of APS1. The current input voltage is lower than the stored fully closed learning value.
Condition (e): MLBPPS1 ≧ MLBMIN1 #
Condition (f): MLBPPS1 ≦ MLBMAX1 #
Condition (g): APS2 voltage diagnosis result is normal Condition (h): MLBPPS2 <MLPPS2b
Here, MLBPPS2 is a buffer value of the input voltage of APS2, and MLPPS2b is a fully closed learning value of APS2. The current input voltage is lower than the stored fully closed learning value.
Condition (i): MLBPPS2 ≧ MLBMIN2 #
Condition (j): MLBPPS2 ≦ MLBMAX2 #
Condition (k): MLBPPS1 ≦ MLPPS1b−MLPOFS #
Condition (l): MLBPPS2 ≦ MLPPS2b−MLPOFS #
Here, MLPPS1b is the current fully closed learning value of the accelerator sensor 25, MLPOFS # is the learning update offset value, MLPPS2b is the current fully closed learning value of the accelerator sensor 26, and MLPOFS # is the learning update offset value.

  Then, the fully closed learning after the learning condition is satisfied by the accelerator sensor fully closed learning means 32 is performed in block 32C. In block 32CA, (1) the initial value of the learning update reference of the fully closed learning value is set and (2) After the learning update reference is cleared, (3) fully closed learning values MLPPS1b and MLPPS2b are updated in block 32CB.

First, as shown in FIG.
[Equation 3]
A. Fully closed learning value update reference MLPREF1
(1) Setting an initial value When the previously obtained update reference is 0, the update reference MLPREF1 is set as the buffer value MLBPPS1 of the APS1 input voltage.
(2) Clear learning update reference | MLBPPS1−MLPREF1 | ≧ 10mV When MLPREF1 = 0.
B. Update of fully closed learning value When (2) non-establishment has passed for a predetermined time KMLTHR #, MLPPS1b = MLBPPS1 and MLPREF1 = MLBPPS1. However, at this time, the lowest value of MLPPS1b and MLPREF1 is MLBMIN1 #.

On the other hand, the accelerator sensor full-close learning means 32 of the present embodiment, as shown in FIG.
[Equation 4]
A. Fully closed learning value update reference MLPREF1
(1) Setting an initial value When the previously obtained update reference is 0, the update reference MLPREF1 is set as the buffer value MLBPPS1 of the APS1 input voltage.
(2) Clear learning update reference When | MLBPPS1−MLPREF1 | ≧ 10 mV, MLPREF1 = 0 and MLPREF2 = 0.
B. Update of fully closed learning value When (2) non-establishment has passed for a predetermined time KMLTHR #, MLPPS1b = MLBPPS1 and MLPREF1 = MLBPPS1. However, at this time, the lowest value of MLPPS1b and MLPREF1 is MLBMIN1 #.

  The accelerator sensor 26 is also performed in the block 32D, and after the initial value is set in the block 32DA in the same manner as described above, the absolute value of the difference between the MLBPPS2 and the MLPREF2 becomes a predetermined voltage value (10 mv) or more. Sometimes MLPREF1 is set to 0 and MLPREF2 is also set to 0. The MLPPS 2b is updated in the block 32DB in the same manner as described above.

  That is, the accelerator sensor fully closed learning means 32 adds the following learning conditions (1) to (4) to the conventional fully closed learning condition, and in addition to the conventional learning condition, the (1) to (4) When all the learning conditions are satisfied, the fully closed learning of the accelerator sensors 25 and 26 is performed.

Condition (1) is that the A / D value as the buffer value of the input voltage of the accelerator sensor 25 is equal to or less than the difference between the fully closed learning value of the accelerator sensor 25 and a predetermined offset amount (20 mV).
Condition (2) is that the A / D value as the buffer value of the input voltage of the accelerator sensor 26 is equal to or less than the difference between the fully closed learning value of the accelerator sensor 26 and a predetermined offset amount (20 mV).
Condition (3) is that the difference between the buffer value, which is the voltage change amount of the accelerator sensor 25, and the update reference is a predetermined amount (10 mV) or less.
Condition (4) is that the difference between the buffer value, which is the voltage change amount of the accelerator sensor 26, and the update reference is a predetermined amount (10 mV) or less.

  In other words, the accelerator sensor full-close learning unit 32 is prohibited from lowering the learning value of any one of the accelerator sensors by the conditions (1) and (2), and either one of the accelerator sensors by (3) and (4). Learning is prohibited when the signal of the accelerator sensor is fluctuating. The signals of the two accelerator sensors 25 and 26 are monitored with each other, and all of the learning conditions for the accelerator sensor 25 and the learning conditions for the accelerator sensor 26 are monitored. Only when the condition is established, fully closed learning is performed, and learning is prohibited by assuming that the learning value on only one side is lowered or the signal on only one side is fluctuating and regarded as abnormal.

FIG. 9 is a block diagram for calculating the accelerator depression amount by the accelerator depression amount calculating means 33.
The accelerator depression amount calculation means 33 calculates the accelerator depression amount in the block 33A as follows based on the fully closed learning value by the accelerator sensor full closure learning means 32.

[Equation 5]
DISPPPS1 = NORMPPS1-MLPPS1b
DISPPPS2 = NORMMPPS2-MLPPS2b
However, DISPPPS1 ≧ 0, DISPPPS2 ≧ 0, DISPPPS1 is the accelerator depression amount detected by the accelerator sensor 25, NORMPPS1 is its A / D value, MLPPS1b is its fully closed learning value, and DISPPPS2 is the accelerator sensor 26 The detected accelerator depression amount, NORMPPS2 is its A / D value, and MLPPS2b is its fully closed learning value.

  FIG. 10 shows the result of calculating the fully closed learning value by the accelerator sensor fully closing learning means 32 and the accelerator depression amount by the accelerator depression amount calculating means 33 and confirming it with an actual vehicle. The first accelerator sensor Even if the output voltage of 25 is reduced, the output voltage of the second accelerator sensor 26 is not reduced, so that all conditions are not satisfied, and the fully closed learning is prohibited based on the monitoring of each other's signals, It can be seen that the output voltage of the lowered accelerator sensor 25 is not learned.

  11 and 12 show the idle state determination of the idle switch determination means 36 in the main engine control means 2, FIG. 11 is a block diagram of a conventional idle determination, and FIG. 12 is an idle determination of the idle switch determination means 36. A determination block diagram is shown, and the idle switch determination means 36 is determined based on an idle determination result for each of the accelerator sensors 25 and 26.

As shown in FIG. 11, the conventional idle determination is used for throttle valve control with the signal of one accelerator sensor APS1 as a main signal at block 36A 'and at normal time at block 36AA'. On the basis of the value of the accelerator depression amount DISPPPS1 obtained by the side accelerator sensor APS1, a fixed value ACCV of the accelerator depression amount is calculated. When the main accelerator sensor APS1 is abnormal, the sub-side accelerator sensor APS2 is determined in block 36AB '. When the sub-side accelerator sensor APS2 is abnormal, the determined value ACCV of the accelerator depression amount is calculated based on the signal of the main-side accelerator sensor APS1 in block 36AC ′. The idle switch is a soft switch and not a hard switch having mechanical contacts. Then, in block 36B ′, the determination is performed as follows.

[Equation 6]
When the idle switch is ON
DISPPPS1 ≦ IDTVO # (with hysteresis)
Here, IDTVO # is an idle determination voltage. In other cases, the idle switch is turned off.

  Then, the accelerator depression amount is obtained from the difference between the accelerator sensor output voltage and the accelerator fully closed learning value, and it is determined from the depression amount whether or not the engine is in the idle state. In the idle state, idle control is performed. If not, the instruction opening corresponding to the depression amount is given to the throttle valve control means to control the throttle opening.

  In contrast, as shown in FIG. 12, the idle switch determination means 36 of the present embodiment is configured to block 36A for each system based on the accelerator depression amount detected by the two accelerator sensors 25 and 26. Judgment is performed as follows.

[Equation 7]
Accelerator sensor 25
DISPPPS1 ≦ IDTVO1 # (with hysteresis)
Accelerator sensor 26
DISPPPS2 ≦ IDTVO2 # (with hysteresis)
Here, IDTVO1 # is an idle determination voltage of the accelerator sensor 25, and IDTVO2 # is an idle determination voltage of the accelerator sensor 26.

  Then, based on the respective determinations in the two systems and the diagnosis results of the accelerator sensors 25 and 26, a comprehensive idle determination is performed in block 36B according to the following (a) to (d), and the throttle valve control means 3 Output.

[Equation 8]
(A) When the diagnosis results of the accelerator sensors 25 and 26 are normal
IDLE SW1 = ON or IDLE SW2 = ON
In other words, the overall determination is turned on by any idle-on determination in block 36BA, and the idle state is determined. When both systems are idle-off determined, the overall determination is turned off and it is determined that the system is not in the idle state.
(B) When abnormality of the accelerator sensor 25 is confirmed
IDLE SW2 = ON
That is, in block 36BB, ON / OFF determination is performed only with the accelerator depression amount DISPPPS2 on the accelerator sensor 26 side.
(C) When abnormality of accelerator sensor 26 is confirmed
IDLE SW1 ＝ ON
That is, in block 36BC, the ON / OFF determination is performed only with the accelerator depression amount DISPPPS1 on the accelerator sensor 25 side.
(D) When an abnormality is determined for both the accelerator sensors 25 and 26, or when an abnormality is determined by the sensor correlation diagnosis means 34, the comprehensive determination is turned off at block 36BD.

  13 to 17 show the upward correction of the fully closed learning value of the accelerator sensor full closing learning means 32 in the engine control means 2, and FIG. 13 is an explanatory view of the characteristic error of the accelerator sensor. FIG. 15 is a diagram showing a measurement result of the fully closed learning upward correction when there is a characteristic error, FIG. 15 is an upper learning block diagram by the accelerator sensor fully closed learning unit 32, and FIGS. 16 and 17 are accelerator sensor fully closed learning units. It is a figure which shows the confirmation result of the fully closed learning upward correction of the accelerator sensor output voltage by 32.

  As described above, it is possible to prevent the accelerator sensor full-close learning unit 32 from erroneously learning the voltage value of the full-close point with respect to the fluctuation of the accelerator sensor signal due to the change in the contact resistance between the terminals. In the case where a phenomenon such as a decrease in the accelerator sensor signal occurs at the same time as the ignition switch is turned on, for example, the depressed accelerator sensor signal may be learned. Therefore, when the signal returns to normal after such learning, the fully closed learning value needs to be corrected upward.

  Further, the accelerator sensor full-close learning unit 32 according to the present embodiment determines that the side that has determined that it is not in the idle state is abnormal when there is a mismatch in the signals of each idle switch, that is, when the idle determination results of the two system signals are different. Therefore, the upward correction of the fully closed learning value is performed even when the learning is normally performed. For example, as shown in FIG. 13, the accelerator sensor 25 and the accelerator sensor 26 may have different inclination characteristics for each sensor, and even if the depression amount of the accelerator pedal 24 is the same, there is inconsistency in the idle determination. It can be seen that it can occur.

  Therefore, in this case, since the idle switch is turned off, that is, the side that is determined not to be in the idle state (accelerator sensor 26 side in FIG. 13) is erroneously learning, the fully closed learning value of the accelerator sensor 26 is It can be considered that the idle switch is turned on by lifting it upward so that the accelerator pedal 24 is not depressed.

  This is because the idle switch can be turned on by raising the fully closed learning value by the amount corresponding to the current accelerator depression amount. As shown below, the amount of depression of the accelerator pedal 24 is added to make the total amount of depression 0, thereby determining that the vehicle is in an idle state.

[Equation 9]
MLPPS1 = mlpps1 + disppps1
Here, MLPPS1 is the fully closed learning value of the current accelerator sensor 25 after upward correction, mlpps1 is the previous fully closed learning value of the accelerator sensor 25 before correction, and disppps1 is the previous accelerator before correction. This is the amount of depression of the sensor 25.

  However, in this upward correction, as shown in FIG. 14, when the accelerator pedal 24 is slowly depressed, the fully closed learning values of the accelerator sensor (APS1) 25 and the accelerator sensor (APS2) 26 are both depressed. Since it is erroneously learned upward one after another by the amount, it can be seen that it is not determined that the engine is not in the idle state, that is, the idle switch is off.

  In this case, since the desired upward learning cannot be obtained, the accelerator sensor fully closed learning means 32 of the present embodiment performs upward correction of the fully closed learning values shown in the following (a) and (b). .

[Equation 10]
(A) When IDLE SW1 = OFF and IDLE SW2 = ON, and this state continues for a predetermined time (TMCID #),
APS1 learning value after correction = current APS1 learning value + (APS1 side depression amount-APS2 side depression amount)
(B) When IDLE SW1 = ON and IDLE SW2 = OFF, and this state continues for a predetermined time (TMCID #),
APS2 learning value after correction = current APS2 learning value + (APS2 side depression amount-APS1 side depression amount)

  That is, the accelerator sensor full-close learning means 32 reads the full-close learning value in block 32E, and after the predetermined idle switch mismatch determination time continues in block 32F, the accelerator sensor 25 detects the accelerator detected on the accelerator sensor 25 side. The difference between the depression amount and the accelerator depression amount detected on the accelerator sensor 26 side is used as a correction amount for upward learning, and the fully closed learning value is corrected upward in block 32G. The correction amount is converged and the accelerator pedal 24 is converged. The influence by how to step in is lost.

  FIGS. 16 and 17 show the results of confirming the upward correction of the fully closed learning value by the accelerator sensor full closing learning means 32.

  As shown in FIG. 16, the correction amount for upward learning, which is the difference in the accelerator depression amount, converges, and the amount of mislearning is about 15 mV at the maximum, which is within the error range.

  As shown in FIG. 17, when the output of the first accelerator sensor 25 returns to a normal value, it is recognized that the accelerator pedal 24 has been depressed and is determined to be in an idle state, so that the accelerator pedal 24 is not depressed. 2 is determined to be not in the idle state on the accelerator sensor 26 side, but after the determination delay time for monitoring the idle switch mismatch, the current fully closed learning value on the accelerator sensor 25 side It can be seen that the fully closed learning value on the accelerator sensor 25 side can be corrected upward to a normal value by adding the difference in the depression amount.

  Further, after this upward correction, the idle switch on the accelerator sensor 25 side is turned on, and the amount of depression recognized as the accelerator pedal 24 is depressed due to the return of the signal is erased, thus preventing the occurrence of a high idle state. In addition, even when inconsistency occurs in the idle switch, since it is generally controlled as an idle state, it is possible to prevent the occurrence of a high idle state during this period.

  As described above, the embodiment of the present invention achieves the following functions by adopting the above configuration.

  That is, the throttle valve control device in the engine control device 1 of the present embodiment is composed of a main engine control means 2 and a throttle valve control means 3, and the main engine control means 2 includes a double accelerator sensor 25, 26, etc., based on each output signal, accelerator sensor fully closed voltage learning means 32, accelerator pedal depression amount calculating means 33, idle switch determining means 36, required opening setting means 41 of throttle valve 18, fuel injection / Ignition timing control means 42, the throttle valve control means 3 is provided with motor drive circuit means 5 based on the required opening degree, and the signals of the accelerator sensors 25, 26 are used as opening degree commands of the throttle valve control device. And, for example, pull-down resistors R1 and R2 are provided as fail safes in the input circuits of the accelerator sensors 25 and 26. In the case where the two sensors are connected, the output signals of the accelerator sensors 25 and 26 are monitored from each other in view of the high idle state caused by the fluctuation of the signals of the accelerator sensors 25 and 26, and learning based on the monitoring results. Since the setting of the condition, the determination of the idle state, or the upward correction of the learning value is performed, and learning and determination are performed when both of the accelerator sensors 25 and 26 are applicable, the fully closed point voltage of the accelerator sensors 25 and 26 The reliability of the learning value (full-learning learning value) and the like can be improved more than ever, and the safety of the throttle valve control device can be further improved.

  The accelerator sensor full-close learning unit 32 sets a learning condition for monitoring the output voltages of the two accelerator sensors 25 and 26 to each other in the conventional learning condition, and the learning condition for all the accelerator sensors 25 and 26 is set. Is established, the full-closed point voltage is learned.If either one of the full-closed learning values decreases or fluctuates, it is regarded as abnormal and learning is prohibited. The output voltage of the accelerator sensors 25 and 26 is equal to or less than the difference between the fully closed learning value of each accelerator sensor 25 and 26 and the offset amount of 20 mV, and the change amount of the voltage of each accelerator sensor 25 and 26 is a predetermined value of 10 mV. The following conditions are set, and, as in the former, the difference between the current learning value of each full-close point voltage of the accelerator sensors 25 and 26 and the output voltage of the accelerator sensors 25 and 26 is predetermined. If the offset amount is greater than or equal to the offset amount, it is prohibited to decrease one of the fully closed learning values, and as in the latter case, the difference between each output voltage of the accelerator sensors 25 and 26 and each learning update reference is different. When the amount of change is equal to or less than the predetermined amount of change, learning when any one of the signals is fluctuating is prohibited. Therefore, the accelerator sensors 25 and 26 are determined based on the correlation between the signals of the accelerator sensors 25 and 26. The occurrence of a high idle state due to the fluctuation of the signal can be prevented.

  Further, the idle switch determination means 36 performs an idle determination for each of the accelerator sensor 25 and the accelerator sensor 26 and makes a determination based on the correlation between the signals of the accelerator sensors 25 and 26 without being influenced only by the main sensor output. Therefore, even if mis-learning occurs in one of the two and the judgment does not match, comprehensive idle determination can be performed, the occurrence of a high idle state is prevented, and an appropriate required throttle opening is set to the throttle valve control means. 3 can be output.

  Furthermore, since the accelerator sensor full-close learning unit 32 performs the upward correction of the full-close learning value after the delay time for determining the predetermined idle switch mismatch, the accelerator sensor full-close learning unit 32 can perform stable correction. The upward correction amount converges the correction amount as a difference between the stepping amount detected by the accelerator sensor 25 and the stepping amount detected by the accelerator sensor 26. Even in the case of erroneous learning, it can be corrected to a normal value, so that malfunction due to the depression of the accelerator pedal 24 can be eliminated, and the occurrence of a high idle state can also be prevented.

  Although one embodiment of the present invention has been described in detail above, the present invention is not limited to the above-described embodiment, and various changes in design can be made without departing from the spirit of the invention described in the claims. It can be done.

  For example, the engine control device 1 including the throttle valve control device of the above embodiment takes in the output signals of the two accelerator sensors 25 and 26. In this case, the same effect can be obtained.

1 is an overall configuration diagram of an engine control device including a throttle valve control device according to an embodiment of the present invention. The control block diagram of the main engine control means of the throttle valve control apparatus of FIG. FIG. 2 is a control block diagram of throttle valve control means of the throttle valve control device of FIG. 1. FIG. 3 is an accelerator sensor characteristic diagram showing an output voltage value with respect to an operating angle of the accelerator of FIG. The throttle sensor characteristic figure which shows the output voltage value with respect to the operating angle of the throttle of FIG. The motor characteristic figure which shows the relationship between the electric current value with respect to the torque of FIG. 1, and rotation speed. The conventional fully-closed learning block diagram. FIG. 3 is a fully closed learning block diagram of accelerator sensor fully closed learning means in the main engine control means of FIG. 2. FIG. 3 is a block diagram for calculating a depression amount of an accelerator depression amount calculation unit in the main engine control unit of FIG. 2. The figure which shows the confirmation result of the fully closed learning of the accelerator sensor output voltage in the main engine control means of FIG. The conventional idle determination block diagram. The idle determination block diagram of the idle switch determination means in the main engine control means of FIG. Explanatory drawing of the characteristic error of an accelerator sensor. The figure which shows the measurement result of the fully closed learning upward correction example with respect to the characteristic error of an accelerator sensor. The upper learning block diagram of the accelerator sensor full-close learning means in the main engine control means of FIG. The figure which shows the confirmation result of the fully closed learning upward correction of the accelerator sensor output voltage in the accelerator sensor fully closed learning means of the main engine control means of FIG. The figure which shows the confirmation result of the fully closed learning upward correction of the accelerator sensor output voltage in the accelerator sensor fully closed learning means of the main engine control means of FIG. The input circuit figure of the accelerator sensor in a main engine control means. The figure explaining the high idle state by the fully-closed false learning when the indefinite state of an accelerator sensor signal arises. The figure which shows the measurement result of the accelerator sensor output by the fully closed error learning of FIG. The figure which shows the measurement result of the accelerator sensor output by the fully closed error learning of FIG. The figure which shows the measurement result of the fluctuation of an accelerator sensor output.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Engine control apparatus containing a throttle valve control apparatus 5 Motor drive means 14 Motor 16 Throttle sensor 17 Throttle sensor 18 Throttle valve 22 Default mechanism 24 Accelerator pedal 25 Accelerator sensor 26 Accelerator sensor 32 Means to learn fully closed voltage of accelerator sensor 36 Means for determining idle switch 41 Means for setting target opening of throttle valve R1 Pull-down resistor R2 Pull-down resistor

Claims (7)

In a throttle valve control device for an engine having an accelerator pedal, a multiple accelerator sensor that detects an opening degree of the accelerator pedal, and a throttle valve that is driven to open and close via a motor based on an output signal of the accelerator sensor,
The control device comprises means for learning a fully closed voltage of the accelerator sensor and means for determining each idle switch for the accelerator sensor of the multiple system, and means for learning the fully closed voltage of the accelerator sensor, If there is a mismatch in the on / off determination of each idle switch by the means for determining the idle switch, the learning value of the full-close point voltage of the accelerator sensor on the side where the idle switch is determined to be off is increased upward. A throttle valve control device characterized by correcting and learning the value.   The means for learning the fully closed voltage of the accelerator sensor corrects the learning value of the fully closed point voltage of the accelerator sensor upward after a predetermined delay time and learns the value. Throttle valve control device.   The means for learning the fully closed voltage of the accelerator sensor is configured to calculate a difference between a detection signal of the accelerator sensor in which the idle switch is determined to be off and a detection signal of the accelerator sensor in which the idle switch is determined to be on. The learning value of the full-closed point voltage of the accelerator sensor determined to be OFF is added to the learned value of the full-closed point voltage of the accelerator sensor, and the learned value is learned while being corrected upward. The throttle valve control device according to 1 or 2.   The control device comprises means for determining each idle switch for the accelerator sensor of the multiplex system, and the means for determining the idle switch performs comprehensive idle determination based on an output signal of each idle switch, The throttle valve control apparatus according to any one of claims 1 to 3, wherein when the idle switches are all determined to be off, the comprehensive idle determination is turned off.   5. The control device according to claim 4, wherein when the comprehensive idle determination is turned on by the means for determining the idle switch, the control device sets the target opening degree to the throttle valve to 0. 6. Throttle valve control device.   6. The throttle valve control device according to claim 1, wherein the control device has a pull-down resistor in an input circuit of the multiple accelerator sensor. The control device is a throttle valve control device for an engine having a multiple throttle sensor that detects the opening of the throttle valve, and a default mechanism that maintains the throttle valve at an initial opening when the motor is not driven. And
The control device is configured to set a target opening of the throttle valve based on an output signal from a means for learning a fully closed voltage of the accelerator sensor, and based on an output signal of the means for setting the target opening. The throttle valve control device according to any one of claims 1 to 6, further comprising means for driving a motor.

Images