JPS632449B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a signal processing device for a Karman vortex flow sensor that detects the intake air amount of an internal combustion engine.

Karman vortices are vortices that occur regularly and alternately in proportion to the flow velocity downstream of an object placed in a flow. This is the Karman vortex flow sensor.

The vortex frequency is measured, for example, by detecting alternating pressure changes due to separation of fluid when a vortex is generated from a rod-shaped vortex generator using a means such as a hot wire or an ultrasonic wave.

Figure 1 shows an example of an electronically controlled fuel injection engine that uses such a Karman vortex flow rate sensor (hereinafter referred to as "Karman sensor") as an intake air amount detection means.
Karman sensor 2 installed in the middle of engine intake passage 1
The control circuit 3 controls the opening time of the electromagnetic fuel injection valve 4 based on the intake air amount signal and other signals representative of the engine operating state (mainly the engine speed signal), and controls the amount of fuel supplied to the internal combustion engine 5. The amount is controlled to an appropriate value depending on the operating condition.

By the way, when the Kalman sensor is applied to detect the intake air amount of an internal combustion engine, it is difficult to detect the intake pulsation that occurs in high-load operating ranges and the turbulence and noise superimposed on this pulsation, which is especially noticeable in engines with fewer cylinders. There is a problem in that this causes a large detection error. This is because, as shown in Figure 2, when the intake pulsation is intense near the throttle valve fully open, the flow velocity rises in the intake pulsation region (the region where the vortex form generated in the wake of the vortex generator transitions from twin vortices to Karman vortices). This is because the sensor output becomes very small, and when the original waveform is shaped, a square wave is missing, which causes a so-called missing tooth phenomenon.In this case, the detection cycle becomes abnormally long, and the intake air volume appears to decrease. Become. Also. If the intake air flows backwards in the full-open operation range, this backward flow will also be detected as the actual intake flow rate.
A positive error appears and the apparent intake air amount increases (see Figure 3).

Figure 4 shows the experimental results regarding the above-mentioned phenomenon, and shows the relationship between the load (intake air flow rate) and sensor frequency when a 4-cylinder engine for a passenger car is operated at a constant speed of 1200 RPM. A negative error occurs from around -40mmHg when the engine suction negative pressure is set.
At -20mmHg, the error rate reaches a maximum of 20%.
In addition, the correct error due to the influence of intake backflow in the fully open range is 40
%.

For this reason, conventionally, as shown in Fig. 1, a vacuum switch 6 that responds to the suction negative pressure is provided to detect the load condition, and for example, the suction negative pressure is set to 0 to -50 mm.
In the Hg high-load operating range, countermeasures have been taken such as switching to fuel injection amount control that mainly depends only on the engine speed without using the output signal of the Kalman sensor 2 (for example, JP-A-55-46033).

However, in control dependent on engine speed, the fuel injection amount remains constant as long as the engine speed does not change, so the air-fuel ratio (A/F) in this control range changes depending on the load condition, as shown in Figure 5. Therefore, fluctuations were unavoidable, and there were still problems such as large fluctuations in the air-heat ratio at the control switching point (the operating point of the vacuum switch 6), resulting in poor drivability.

In addition, a cycle variation rate (σ/Ti) is determined from the average value Ti and standard deviation σ of a predetermined number of cycles (e.g., 8 cycles or 16 cycles) of the Kalman sensor output signal, and this is used as the cycle variation rate to obtain the average value. The present applicant proposed a signal processing device for a Kalman sensor configured to take into consideration the correction of
(No.) There are problems such as an increase in calculation processing time and the fact that the above-mentioned missing phenomenon cannot be precisely corrected from the standard deviation alone.

The present invention has been made in consideration of these conventional problems, and focuses on the fact that the period of the output pulse of the Kalman sensor changes significantly during intake pulsation (see Figure 2). By correcting the period value that suddenly changes to an appropriate period value individually in relation to the normal period values before and after it, and using this as a signal for flow rate detection, control errors can be reduced as much as possible. , the purpose is to shorten the flow rate calculation time.

To this end, the present invention uses a period detection circuit that detects the period of the pulse signal from the Kalman sensor, a storage circuit that sequentially stores the detected period values, and a comparison of the period values stored in succession. A comparison circuit for detecting excessive period fluctuations during intake pulsation and a correction circuit for correcting the stored value of the detection period at a predetermined ratio are provided, and the correction circuit stores the period storage value that has been corrected (including the stored value that has not been corrected). ) is used as the flow rate signal. In other words, even if an error occurs in the sensor output due to intake pulsation, the flow rate signal supplied to the control system such as the fuel injection control device can be corrected to a value that is extremely close to the value corresponding to the actual intake flow rate. Therefore, the error appearing in the final control target value can be significantly reduced.

Hereinafter, the present invention will be explained based on embodiments shown in the drawings.

FIG. 6 shows an embodiment in which the present invention is applied to fuel injection control of an internal combustion engine, where 10 is a signal processing circuit according to the present invention, 20 is a rotation sensor 7 that detects the engine speed, and signals from the signal processing circuit 10. This is a control circuit that controls the basic fuel injection amount, that is, the valve opening time (basic pulse width) of the fuel injection valve 4 based on the following.

The signal processing circuit 10 includes the period detection circuit 1 described above.
1. Comparison circuit 12, correction circuit 13, memory circuit 14
Basically, the detection circuit 11 detects the cycle of the pulse signal from the Kalman sensor 2 and stores it in the memory circuit 14, which is called the Kalman cycle IRQ. On the other hand, the control circuit 20 uses the cycle value stored in the storage circuit 14 as the final information for flow rate detection, and performs fuel injection control.

The respective circuits and operations forming the signal processing circuit 10 will be explained along the flowchart shown in FIG. It consists of a counter that measures the period by capturing the rising edge, and detects the period of the pulse wave shaped into a rectangular wave every period as shown in FIG.
Here, the subsequent processing will be explained focusing on the period (or period value) Ta detected at point A (shaped wave rising point) in FIG. 8.

The comparison circuit 12 compares the period value Ta multiplied by a constant C with a preceding period value already stored in the storage circuit 14 (for example, Ta-1 in FIG. 8). The constant C is determined by the dimensions of the Karman sensor 2 used (the projected width of the vortex generator in the flow direction). When using a relatively large engine, it is desirable to set C≒1/2.On the other hand, when applying it to an engine with a larger intake air amount than the fully open intake air volume (for example, about 2800 c.c. for 6 cylinders), a small Karman engine is recommended. Sensor 2 is required. For example, when using a relatively small Kalman sensor 2 with dimensions of approximately 3 mm, C
It has been experimentally confirmed that it is desirable to set it according to the dimensions of the Kalman sensor 2, such as approximately 1/4.

In other words, as shown in Fig. 13, the relationship between adjacent period values, the ratio Ta/Ta-1 of Ta and Ta-1, and the tooth extraction rate (ratio (%) of pulse output zero to the total measurement period) Based on the experimental results obtained using the dimensions of the Kalman sensor 2 as a parameter, Ta/Ta-1, which exhibits a gap phenomenon based on the dimensions of the Kalman sensor and the displacement of the target engine.
It is possible to set the value of . For example, in the case of a large Kalman sensor, if the tooth loss rate is approximately 40% (this value can be determined experimentally, and favorable results have been obtained if it is approximately 40% or less), then the periodicity value should be corrected. If necessary, the ratio of adjacent period values
When Ta/Ta-1 becomes 2 (that is, C=1/2), the period value Ta is set to a constant d (for example, d=
1/2) for correction.

In addition, in the same figure, the dashed-dotted line shows the case where the period value of the small Kalman sensor is read every two shaping pulses, and similarly when using the small Kalman sensor, the period value is read every pulse. Although the detectable maximum flow rate is suppressed compared to the solid line characteristic, the frequency of reading periodic values is reduced, resulting in the advantage that the time required for flow rate calculation is shortened.

The correction circuit 13 cooperates with the comparison circuit 12 and corrects the period value Ta according to the comparison result. That is,
If we are using a large Kalman sensor and the condition [Ta-1<1/2Ta] is satisfied, it will be determined that Ta has fluctuated too much with respect to Ta-1, and Ta will change. Correct it in the shortening direction by multiplying it by a constant d that is less than 1 (for example, 1/2, ...), and use this as a new
When it is sent to the storage circuit 14 as Ta (Ta-1≧1/2 Ta), the period value Ta from the detection circuit 11 is sent to the storage circuit 14 as it is without being corrected.

The storage circuit 14 includes a RAM (readable and writable storage element) that stores the cycle values from the correction circuit 13, and an address circuit that sequentially addresses and stores the cycle values that are inputted one after another in the cycle storage area of the RAM. The control circuit 20 sequentially rewrites old data that has already been used with new cycle values, so that a large amount of new data can be read out at any time.

In this way, the signal processing circuit 10 judges whether the output of each pulse signal from the Kalman sensor 2 is appropriate or not, and if the periodic fluctuation is excessive, the periodic value related to this excessive fluctuation is set as the preceding periodic value. According to this relationship, the period value is corrected at a predetermined rate so as to approach a normal period value, and this is stored in the storage circuit 14. Therefore, by referring to the cycle value stored in the memory circuit 14 as a flow rate signal, it is possible to calculate the intake flow rate with less error from the actual intake flow rate even when intake pulsation occurs, and the intake flow rate can be calculated based on this intake flow rate calculation value. This makes it possible to increase the accuracy of fuel injection amount control.

In the above embodiment, the correction process is performed before the period value is stored, but the period values that have been stored may be compared with each other and the period value to be corrected may be rewritten.

Next, the control circuit 20 will be explained along the flowchart shown in FIG.

The control circuit 20 uses the flow rate calculation circuit 21 based on the corrected cycle value stored in the storage circuit 14.
The intake flow rate Qa is synchronized with time or engine rotation.
is calculated (table lookup or calculation, the same applies hereafter). In this case, as shown in Fig. 9, a large number of cycle values are read depending on the engine load condition, and
It is preferable to take the average value for accuracy, and the appropriate number to read is 1 to 4 revolutions of the crankshaft, for example, 32 in the high load region and 10 in the low load region (in the low load region Since the operating conditions change relatively rapidly, it is preferable to reduce the number of readings from the viewpoint of responsiveness).

Next, in the injection amount calculation circuit 22, based on the flow rate Qa and the engine rotation speed Nr from the rotation sensor 7,
Basic fuel injection amount (injector drive signal pulse width)
Tp is calculated. In addition, in the case of rotation-synchronous injection, when K is a constant determined according to the characteristics of the injection valve, etc., Tp=K·Qa/Nr.

On the other hand, the maximum injection amount calculation circuit 23 determines the maximum injection amount Tp· _MAX according to the rotational speed Nr at that time.

The above two injection amount values Tp and Tp・_MAX are compared via the comparison circuit 24, and when [TP<TP・_MAX ], Tp remains unchanged, and when [Tp≧Tp・_MAX ], Tp・_MAX is new. It is stored in the storage circuit 25 as Tp.

The control circuit 20 sequentially reads out the injection amounts Tp stored one after another in the storage circuit 25 in this way, and drives the fuel injection valve 4 based on this. Furthermore, since Tp accurately reflects the engine operating state, it is used for exhaust gas recirculation control, ignition timing control, etc. as necessary.

Note that the signal processing circuit 10 and the control circuit 20 are
Of course, it is also possible to configure any of them with a microprocessor and perform calculations.

In addition, in this embodiment, the maximum injection amount is regulated by setting Tp・_MAX as described above, but this is due to excessive error caused by intake backflow in the full load operating range (see Figure 4). This measure is intended to deal with the occurrence of air reflux, so if the intake air reflux phenomenon is not significant, there is no need to take such measures. Incidentally, in order to regulate the maximum injection amount, for example, in the case of a rotation-synchronized injection method, the minimum value of the average period Ta or the maximum value of the flow rate signal Qa is set depending on the engine speed, and the injection amount is synchronized with the Karman period. For an injection method (in this case, the injection valve pulse width is approximately constant), a minimum value of the injection period may be set.

FIG. 10 shows the effect of the present invention when the fuel injection device is configured in the manner of the above embodiment, and shows the effect of the present invention.
As is clear from a comparison with the figure or FIG. 5, the correction processing of the Kalman sensor signal significantly reduces the error that ultimately appears in the injector pulse width.
Especially when looking at the air-fuel ratio in the high-load operating range, the error is 0 to 3% on the rich side, which is less than half of the conventional ratio. Furthermore, since no sudden changes in the air-fuel ratio occur, it can be seen that good drivability can be maintained.

Incidentally, in the above embodiment, the Kalman period Ta detected at the rising point of each shaped pulse is corrected based on comparison with the preceding period, but as shown in FIGS. 11 and 12, Further, it may be corrected in comparison with the subsequent period Ta+1, etc. That is, for example, when [1/2Ta>Ta-1], Ta is determined to be an excessive period, and this is corrected to a short period.

By doing this, when the intake flow rate of the engine rapidly decreases and the Karman period rapidly decreases, such as during sudden acceleration, the period preceding this period (in the example above)
As a result of correcting Ta) to a short period, the fuel supply amount changes to the increasing side, which has the advantage of improving the acceleration response of the engine.

As described above, according to the present invention, individual shaped pulses (periodic values) are corrected for errors in the output signal period of the Karman vortex flow rate sensor caused by pulsation phenomena, etc., to obtain periodic values corresponding to the actual flow rate. Since this is corrected, it is possible to significantly improve the accuracy of a control system such as a fuel injection device that operates based on the corrected cycle value or flow rate signal.

[Brief explanation of the drawing]

FIG. 1 is a schematic diagram of a conventional example of a fuel injection type internal heat engine in which the intake air flow rate is detected via a Karman vortex flow rate sensor. FIGS. 2 and 3 are explanatory diagrams showing the influence of intake pulsation and intake backflow on flow rate detection, respectively, and FIG. 4 is a characteristic line diagram of flow rate and error rate that similarly represents the influence error. Fifth
The figure is a characteristic diagram showing the error in the fuel injection amount in the high-load operating range in the apparatus of FIG. 1. FIG. 6 is a block diagram of one embodiment of the present invention, and FIG. 7 is a flowchart showing the operation contents of the signal processing circuit. FIG. 8 is a waveform diagram for explaining the detection of the Karman vortex generation period. FIG. 9 is a flowchart showing the operation of the control circuit of the embodiment. FIG. 10 is a characteristic diagram showing the effect of the embodiment with respect to the amount of heat injection. FIG. 11 is a flowchart showing the signal processing operation of another embodiment of the present invention, and FIG. 12 is a waveform diagram for explaining the detection of the vortex generation period.
FIG. 13 is a diagram showing the relationship between the ratio of adjacent cycles and the tooth extraction rate. 2...Karman vortex flow rate sensor, 10...signal processing circuit, 11...period detection circuit, 12...comparison circuit, 13...correction circuit, 14...memory circuit.

Claims (1)

[Claims] 1 A period detection circuit that detects the period of the pulse signal from the Karman vortex flow sensor, a storage circuit that sequentially stores the period values detected via the period detection circuit, and a period value mutually stored in the storage circuit one after the other. A comparator circuit that detects an excessive cycle fluctuation based on a comparison of A signal processing device for a Karman vortex flow rate sensor, characterized in that the average value of the stored values stored in the storage circuit is used as a flow rate signal.