JPS58144755A - Detection of rotating speed - Google Patents

Abstract

PURPOSE:To improve the quick response of the detection by determining the sum of (n) sets of data which are memorized in a sequential time series for a specified time at an interval of 1/n rotation while each set of data memorized is updated in a sequential shift at an interval of 1/n rotation. CONSTITUTION:An example of n=8 is mentioned. A reference pulse P of a fixed frequency is counted with a counter 10 and the counts pm are sent to a register 11 with the falling of an output signal of a rotary sensor. Then, the counter is cleared, and the pm is time required of 1/n=1/8 rotation. Registers 11-17 are arranged in the number n-1=8-1=7 and include pm-1, pm-2...pm-7 which are shifted simultaneously at each falling of an output signal of the rotary sensor and transferred to the sybsequent stage. Simultaneously with the shift, the data pm, pm-1,...pm-7 are sent to an adder 19 to determine the sum Pm of the data. When the constant K is divided by the sum, the rotating speed N'=K/Pm is determined. Thus, the calculation of the speed can be done stably.

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a rotational speed detection method for detecting the speed based on the time required for a rotating body to rotate by a certain angle.

For example, the rotational speed of the output shaft of an internal combustion engine changes before and after ignition, resulting in pulsations within one rotation.

In order to eliminate the influence of this pulsation, it is necessary to increase the rotation angle at which time is measured. Figure 1 shows that this angle is 3
600, that is, one rotation, and the speed is determined by measuring the time required for one rotation. A magnetic disk 1 for a rotation sensor in the form of a gear with one tooth is attached to a rotating shaft (not shown), and the pickup 2 responds to the rotation of the disk 1 and outputs an output signal S as shown in FIG. Configure it to do so. As shown in the figure, for example, a falling edge of 9 is used for this output signal S. If the time between falling a and falling b, that is, one period, is T, then the rotation speed is T.
It can be obtained as the reciprocal of 3/T. In addition to the magnetic type shown in the figure, the rotation sensor may be of the photoelectric type, as long as it produces pulse rises and falls at every predetermined rotation angle (one rotation in Figure 1), and may be of a specific type. There are no limitations. When the output signal S of the pickup 2 falls (or rises), the divider 5 shown in FIG. 3 is started and the counter 4 is cleared. The counter 4 counts the constant frequency reference pulse l generated by the signal generator 3, and immediately before it is cleared at the falling edge of the signal S, the counter 4 calculates the count value, that is, the count value during the time from the previous falling edge to the current falling edge. The value, Pn f is sent to the divider 5 as a signal corresponding to the time required for one rotation.The divider 5 performs an operation to divide the constant K by the transmitted data, that is, the count value Pn, and the quotient is /Pn. is output as a signal representing the rotational speed N.

In this method, the speed information is not updated unless it rotates by a large rotation angle (one rotation in the illustrated example) corresponding to the period of the signal S, and division is performed every time corresponding to a constant rotation in the past. Therefore, from a microscopic point of view, there is a tendency to measure the past speed, and there is a drawback that it is not possible to follow sudden changes in speed without delay.The present invention eliminates the above-mentioned drawbacks, It is an object of the present invention to provide a method for detecting speed that is responsive and can track speed well.

In order to achieve this objective, the present invention further divides the reference angle used for speed calculation, for example 360°, into a plurality of parts, provides a register for each division, stores the required time for each division, and stores the required time for each division. By sequentially shifting and determining the time required to rotate the above reference angle, which is determined as the sum of the time required for each division, for each division, the speed obtained using the rotation time of a relatively large angle can be finely calculated. This allows calculations to be made for each divided angle.

The present invention also provides a speed signal for flexible and fine-grained speed control by modifying the speed for each segment.

FIG. 4 shows an example of the configuration of an apparatus for implementing the present invention. First, the rotation sensor used in the apparatus shown in FIG. 4 has a structure in which n pulses are generated per rotation of the rotating shaft, that is, the number of falls and falls of n pulses is 9.

- As an example, the sensor output in the case of n=8 is shown in Fig. 5,
The following explanation will be based on this. The device shown in FIG. 4 operates in synchronization with the fall of the sensor output shown in FIG. The counter 10 counts the fixed frequency reference pulse P, and is cleared after transmitting the count value pm to the first register 11 at the fall of the rotation sensor output signal. In other words, the count value pm corresponds to the time required for 14 rotations minus IA rotation. Registers 11-17 are registers in which n-1=8-1=7 are arranged. Registers 11-17 contain p-1, p
m-2, . . . pm-7 are stored, and these are shifted all at once every time the rotation sensor output signal falls and are transferred to the next stage register. Therefore, the contents of registers 11 to 17 after shifting are pm, pm -1, pm
-2,”'.

It becomes pm-6. Here, pm-i is 1 as current fm
This is the count value transmitted from the counter 10 to the registers 11 to 17 at the previous fall of the rotation sensor output signal.

At the same time as the shift, data pm, pm-1,...

pm-7 is also transmitted to the adder 19~, where the total sum Pm of each data is determined.

Pm-22, %-1 (1) 5 SO In this (11 formula, in the example of Figure 4, (n-1) = 7, OPm corresponds to the time required for one past revolution at time m''& The divider addition divides the constant K by this Pm to obtain a temporary rotational speed N' = K/Pm.This rotational speed N'
is characterized in that it is updated n=8 times per rotation. Incidentally, in the circuit of FIG. 3, even if the number of output pulses per rotation of the rotation sensor is increased, as long as the time required for one rotation is measured, the speed information is updated only once per rotation.

The provisional rotational speed N' obtained in the above manner is written to the register 21, and at the same time, the
1. The contents written in 22 are register n.

It will be transferred to Party B. The contents of register n, that is, the rotational speed N'(m-3) determined three times before, are input to the subtraction terminal of the subtractor. In Fig. 4, the past rotational speed N'
The registers used for storing are registers 21 and 22.2.
3 is illustrated, but this is not limited to three, and may be more or less than this. The rotational speed N' output from the register B is expressed as N'(m-3), meaning that it was calculated three times before the current time m. Similarly, the rotational speed calculated at the current time m is N
'6nl.

The circuit consisting of circuit components 24 to 30 divides the rotation speed N' into three speed stages: 1) N' width) ≧ out 2) N, ≦ H' (output 3) N' (%) (output 3) N' (%) (output) 2. The subtracter 5 divides the rotation speed N'
and the constant N set in the constant setting circuit, and the output B1 is N'(N.

It outputs a '1' signal when it is N2H, and it outputs a '0' signal when it is N2H.The subtracter γ compares the magnitude between the rotational speed N' and the constant N2 set by the constant setting circuit 26. As the output B2, a "1" signal is output when N' (N2), and a °'0" signal is output when N/≧N2.

Signal B1 is inverted by inverter A, B,=C
It becomes 1. Therefore, the signal CI becomes "pul" only when N'≧N1. C2-B1.B2 is formed by the gate 4, and this signal C2 is set to N' (N and N'≧
When N2, that is, when N2≦N'<N, C2-“
1”.

Gate I outputs C3=B2, and C3=''1'' only when N' (N2). In this way, the signals C, , C2, C
, only one of them becomes '1' depending on the magnitude of the rotational speed N'.

The sum of the rotational speeds N' (e) and N' (m-3) is added to the adder 32.
The sum is multiplied by the multiplier 33. In the end, the output of the multiplier 33 is the average of the rotational speeds N' (e) and N'(m-3)'(N'tn) + N' (m
−3))=Nh is obtained. The subtracter 34 calculates the difference N'(rrj-N
'(m −3) is calculated, and the difference is multiplied by a constant α set by the constant setting circuit by the multiplier 36, and α(N'
)-N'(m-3)) is calculated.

The sum of the rotational speed N' (c) output from the adder circuit and divider adder and the output of the multiplier 36 is calculated, and only when the enable input supplied from the gate 30 is 1', the calculation result (=
N'6nl+α(N'fm)+N'(m-3))=N
t) is output.

The output N'- of the divider addition is further set as the rotational speed N(e), and the gate 3 opens only when the output C2 of the gate 29 is "1".
The output Nh of the 0 multiplier 33, which is output as the final output N through the inverter 1, is output as the final output N through the gate 37, which opens only when the output 6 of the inverter is '1'.

The output Nt of the adder directly forms the final output N.

In the above process, as an element of speed N, N^20fN'r>> dimension 1, 1' (outside - 3)
N'(-nF4L t N'(z)+((fN'(%)+N?x-
31. Although it was mentioned that s >'j are calculated, the circuit in FIG. 4 calculates one of these three. :J7
and the gate part of the adder 38, and the circuit that selects using the output signals Ct, C2, and Ca of the gate 29.30. More specifically, 1) N'(%)≧ 11 prayers N=N table 2') Nz,'
i; N' (%) < Nl 4fj'f N ''14
%3) i, J″c%>’<Jh’s 1 N=’NL.Since the signals C1, C2, and Ca are never turned on at the same time, the gate 31°37 The outputs between the adders and the adders are directly coupled.

The circuit configuration shown in FIG. 4 is merely an example, and the portions indicated by circuit elements 24 to 30 are as follows.

Appropriate modifications and changes can be made, such as increasing the number of speed divisions and changing the functional forms of the signals Nh and Nt for the circuit elements 32 to 38. Furthermore, the above series of operations can also be executed using a microcomputer.

Next, the speed signal Nh described in the circuit operation of the device shown in FIG.
The functional forms of Nm and Nt are intended to be used as a speed signal for speed control of a diesel engine as an example, and the technical significance of each will be explained below.

First, Nh = 7 (N' @ + N' (m 3 )
) takes into account the phenomenon that occurs when a diesel engine is suddenly accelerated from low speed to high speed. As shown in FIG. 6, if the accelerator is suddenly depressed at time to, the controlled variable f1, for example, the rack position of the fuel injection pump, is actually 1
It is desirable that the curve changes with a polygonal characteristic as shown in 1161.

However, control using only the speed signal N' results in curve characteristics with rounded corners as shown by the broken line 62.

Therefore, if the speed is calculated according to Nh in the above equation, that is, if the calculation is performed slightly later than the actual speed change, a sharp characteristic closer to a polygonal line characteristic can be obtained.

Next, Nt=N'(rrj+α(N'(m)+N'(m-
3)) takes into consideration the phenomenon that occurs when a diesel engine is suddenly decelerated from high speed to low speed. As shown in FIG. 7, when the accelerator pedal depression is reduced to zero at time to, the engine rotational speed suddenly decreases and enters a low speed region. In such a low speed region, there is generally a control action that tries to fix the rotational speed at a target speed, such as the idle target speed N. However, in the case of sudden deceleration, an undershoot ΔN occurs with respect to the target speed No, and this have a negative impact. As already mentioned, the form of the N10 formula corrects the fact that the speed signal N' (e) is calculated too late in the first place, and also strengthens the feedback function of the entire control system by predicting the future speed. The aim is to eliminate undershoot ΔN. The effectiveness of this was confirmed through experiments.

As described above, according to the present invention, n sets of the time required for each rotation are stored sequentially in time series, and the data is updated by sequentially shifting each rotation, and based on the sum of the n sets of stored data. Since the speed is determined by using the engine, stable speed calculations that do not respond to, for example, speed fluctuations before and after ignition of the internal combustion engine can be repeated at high speed.

Although the speed detection method of the present invention can be applied to detecting the speed of any rotating body, it is particularly useful for detecting the speed of an internal combustion engine, which inherently generates pulsations within one revolution, in a coordinated manner while alleviating the effects of pulsation. suitable for

[Brief explanation of the drawing]

Fig. 1 is a layout diagram of a known rotation sensor, Fig. 2 is an output signal waveform diagram of the rotation sensor shown in Fig. 1, Fig. 3 is a block diagram of a known rotation speed detection device, and Fig. 4 is an implementation of the present invention. Fig. 5 is a block diagram of an example of a device that performs the engine control, Fig. 5 is an output waveform diagram of the rotation sensor used in the device shown in Fig. 4, and Figs. 6 and 7 are for explaining the characteristics of diesel engines during sudden acceleration and deceleration. FIG. ]O...Counter, 11-17...Register, 19...
...Adder, Title...Divider, S...Rotation sensor output signal, P...Reference amount Fuji Electric Manufacturing Co., Ltd.

Claims (1)

[Scope of Claims] 1) In a rotational speed detection method that detects the speed based on the time required for a rotating body to rotate by a certain angle, n sets of required times are sequentially stored in chronological order, and A rotational speed detection method characterized in that data is updated by sequentially shifting each stored data, and the speed is detected each time based on the sum of the 0 sets of stored data. 2. In the method described in claim 1, the time required for each rotation is n
Obtaining by the number of fixed frequency reference pulses counted during n: f: Characteristic rotational speed detection method.