JPH03243808A - Relative displacement detecting method - Google Patents

Abstract

PURPOSE:To accurately detect the relative displacement of an automobile without a physical reference line by adding a constant correcting value to the acceleration signal of the automobile which is detected by an acceleration sensor fitted to the automobile and performing second order integration. CONSTITUTION:The acceleration sensor 2 is fitted to the automobile 1 to be measured. Then the constant correcting value with which the final value of the second order integration becomes zero is added to the acceleration signal from the sensor 2 first and the second order integration is carried out. Then a movement mean value as to such a sufficiently long selected necessary movement mean time that the movement average of displacement becomes zero is found from the second order integral value whose final value becomes zero and an error component based upon the offsets of an initial speed and the acceleration signal, etc., is found. Then the movement mean value which is found by the arithmetic is subtracted from the found second order integral value whose final value becomes zero to accurately detect the relative displacement of the automobile 1 without any physical reference line.

Description

DETAILED DESCRIPTION OF THE INVENTION [Object of the Invention] (Industrial Application Field) This invention relates to a relative displacement detection method used, for example, to detect height fluctuations of a running vehicle.

(Prior Art) As a conventional displacement detection method, for example, there is a method that measures unevenness of a road surface (Japanese Patent Laid-Open No. 63-95307). This conventional technology includes a laser beam transmitting/receiving section movably mounted on the main body frame, a signal processing circuit for processing detection information thereof, and the like. Then, a reference line is provided, and the distance to the target road surface is optically measured along the reference line using a laser beam, thereby measuring the unevenness of the road surface. From this, when considering an object that is displaced with the laser beam transmitting/receiving section fixed, it is possible to obtain the displacement of the object in time series.

(Problems to be Solved by the Invention) In the conventional displacement detection method, it is necessary to provide a physical reference line when detecting the displacement of an object. For this reason, it cannot be applied in cases where there is no means for providing a reference line, such as when detecting height fluctuations of a running vehicle.

Therefore, this invention enables even without a physical reference line,
It is an object of the present invention to provide a relative displacement detection method that can accurately detect the relative displacement of an object to be detected.

[Structure of the Invention] (Means for Solving the Problems) In order to solve the above problems, the present invention provides a relative displacement detection method for detecting the relative displacement of an object to be detected. 2 by adding a certain correction value to the acceleration signal of the object to be detected detected by the acceleration detection means
A first step of performing floor integration to obtain a second-order integral value that makes the closing price zero, and a second step of calculating a moving average value for a required moving average time from the second-order integral value that makes the closing price zero. and a third step of subtracting the moving average value from the second-order integral value whose final value becomes zero.

(Operation) An acceleration detection means is attached to the object to be detected, and in the first step, a constant correction value is applied to the acceleration signal so that the final value of the second-order integral of the acceleration signal from the acceleration detection means becomes zero. and second-order integration is performed. In the second step, a moving average value is calculated from the second-order integral value, which is the final value, for a required moving average time that is selected long enough so that the moving average for displacement becomes zero, and the moving average value is calculated for the initial velocity and An error component based on the offset of the acceleration signal, etc. is determined. Then, in the third step,
2 such that the closing price determined in the first step is zero
The moving average value obtained in the second step is subtracted from the floor integral value, and the relative displacement of the object to be detected can be detected with high accuracy even without a physical reference line.

(Example) Hereinafter, an example of the present invention will be described based on FIGS. 1 to 6. This embodiment is applied to the case where height fluctuations due to springs or the like of a running automobile are determined.

First, the equipment configuration etc. for executing the relative displacement detection method of this embodiment will be explained using FIGS. 1 to 3. In FIG. 1, 1 is a car as a detection target, and the car 1 detects the vertical acceleration of the running car 1 and outputs an acceleration signal (hereinafter also referred to as a G signal). An acceleration sensor (hereinafter also referred to as a G sensor) 2 is attached as acceleration detection means. The G signal from the G sensor is recorded via the A/D converter 3 on a digital data logger 4 as a recording means.

FIG. 2 shows a block configuration for determining height fluctuation (relative displacement) from the G signal recorded in the digital data logger 4, in which the digital data logger 4 is connected to the computer 6 via the I10 device 5. has been done. and,
As shown in the flowchart of FIG.
After reading the G signal data from the digital data logger 4 (step 11), performing correction processing by adding a predetermined value to the G signal, and performing second-order integration, the second-order integral value whose final value is zero is obtained. Determined bank processing (step 12.
13) and the process of calculating the moving average of the second-order integral value whose final value is zero, and subtracting the moving average from the second-order integral value whose final value is zero to obtain the relative displacement (step 14).
．． 15〉 will be carried out.

Next, a relative displacement detection method for detecting the relative displacement of the automobile 1 using the above-mentioned equipment configuration will be explained.

First, the principle of how accurate relative displacement can be determined by removing error components will be explained using FIGS. 4 and 5.

To describe what kind of error occurs when calculating the displacement by second-order integration of the G signal Q (t), generally speaking, the second-order integral value of the G signal after seconds is ff 0(t)dt = h(t)+C1・t+C2...(1) In the above equation (1), C1 and C2 are integral constants, C1 corresponds to the initial velocity Vo at the start of measurement, and C2 corresponds to the displacement at the start of measurement. , corresponds to . Also, in the G signal, there is a zero point offset (hereinafter referred to as G signal offset) Q that the G sensor has.
o is generally present. If the above equation (1) is rewritten based on these, it becomes as follows.

t ff Q(t)dt! = h (t) + < 1/2) CJ o
-t2+Vo-t+h.

-(1/2)qo -+2-vo, t...+2) Therefore, in order to find the relative displacement of h(t)-too, the two terms on the right side of equation (2) 1(1/2)go -+2-vo-t becomes an error. Now, assuming this error as ε, ε= (1/2) Qo-+2+Vo-t -f31
far. G signal offset g that is the basis of this error component.

It is necessary to find each of the initial velocity Vo and the initial velocity Vo separately. However, it is generally difficult to obtain an accurate initial speed for a moving object such as a car. In addition, since the G signal offset is minute, it is difficult to measure it accurately and with sufficient precision, and furthermore, it fluctuates depending on the temperature, so
It takes a lot of man-hours to measure each time. Therefore, it is often difficult to obtain the error component expressed by equation <3> in advance.

As mentioned above, the method of determining displacement by second-order integration of the G signal does not require determining a physical reference position, and it is possible to measure relative displacement simply by attaching the G sensor to the object to be detected, such as a car. It has a sexual nature. However, it is difficult to accurately determine error components due to two parameters, G signal offset and initial velocity.

In contrast, in the method of this embodiment, the error components due to the above two parameters are removed by signal processing as described below.

First, let's make one assumption. That is, it is assumed that the absolute heights at the start and end of the measurement are equal. Of course, instead of this assumption, the reference line may be a line that is actually measured by another means and directly connects the heights at the start and end of the measurement. However, since it is easier to explain if it is assumed that the heights at the start and end of the measurement are equal, such an assumption is made in this embodiment.

It is known that the error component ε is expressed by the above equation (3), and as is clear from the open equation, this error component ε is
This becomes a quadratic curve, and the displacement signal of the object to be actually measured is superimposed on the quadratic curve.

Now, consider the moving average signal of the error component, assuming that the measurement time is tl and the moving average time is t2.

First, since the heights at the start and end of the measurement are equal, the above equation (3) can be rewritten as follows.

0=(1/2)qo-t12+vo-t1■o=-(1
/2) go-tl Therefore, ε= (1/2)Qa-+2 - (1/2) go-tl・t...(4) Next, considering the moving average signal have (t), = ( 1/
2) Go −+2 − (1/2) go −tl ・t + (1/24) Qo−t 22=εten(1/2)
24) Go−j22・=(51 In equation (5), (1/
24) The offset of qo-j 22 is a constant. If the target displacement is oscillatory, select a sufficiently long moving average time 1Wlt2 so that the moving average value of displacement h(t) for +2 seconds is almost zero, then / t 2 ; ≧; ε + (1 /24)Q ・t 22
= ε'...(6) It can be seen that although an offset of (1/24)On-t2' occurs, this error component can be found.

Therefore, by subtracting this error component ε' from the second-order integral signal of the G signal, an accurate relative displacement can be determined.

That is, 7 If Q<t) dt2-ε' -ff C1(t) dt2-ε O (1/24) go -t 22 -ff C1(t) dt2 - (1/2) Qo -t 2 Va -i (1
/24〉 qo −t 22 =h (j) −ha −(1/24)Coo −
t 22 (a) Therefore, the above calculation can be performed by the computer shown in FIG. 2 above.

On the other hand, the error component ε is when the G signal offset is 0° and the initial velocity is V.
[], which becomes a very large number depending on the size of the moving average time t2, so when calculation processing is performed on a digital computer, the accuracy of the displacement signal often deteriorates due to loss of digits. For this, before taking the difference with the moving average,
Initial velocity V in advance. , it is necessary to reduce error components due to G signal offset errors.

This method will be described next. As mentioned above, the error component ε shown in the equation (3) is a quadratic equation, and can take various forms as shown in FIG. In cases such as those shown in FIGS. 3A and 3B, the absolute value of the error component ε becomes extremely large in a short period of time, and there is a high possibility that the digit will fall. Therefore, as described above, it is assumed that the absolute heights at the start and end of the measurement are equal. If the measurement time is tl, ε(t-tl)-(1/2) go t+2+vO・
tl -0 From the above equation, vO-(1/2)go t+, so ε-(1/2)go · t2- (1/2)g.

tl t = (1/2)go (t - (1/2)t+
) 2 (1/8) go ' ・t, 2
(8) From equation (8), the way in which error components occur is limited to the two series shown in FIG. Then, the maximum value of the error component is (1/8) go −t2, and if the G signal offset go is adjusted to be as small as possible (it is not necessary to make it completely zero, and it is difficult), then the maximum value of the error component is It can be seen that this can be suppressed below a certain level.

This will now be explained using a numerical example. Considering a 1% error and a 1% offset error go for a ±IG G sensor, the offset error is 10 cm/sec.

Now, if a pig for 30 seconds is acquired by sampling at 500 Hz and the moving average time is 4 seconds, the error component ε is ε - (1/2) go t2, so the measurement error after 30 seconds will be as much as 4500 cm. become. If moving average calculation is performed with an accuracy of 1 mm, the required calculation accuracy is (4500 (cm)
/1 (mm))X4X500-9X107, which requires an arithmetic precision of about 27 bits.

On the other hand, in the offset correction method of this embodiment, the maximum value of the error is (1/8) go-t2, so it falls within 20 cm. The calculation accuracy required at this time is (20 (cm)/1 (mm))X4X500-4
X 105 , and an arithmetic precision of about 19 bits is sufficient.

In this way, in the method of this embodiment, the incompleteness is also corrected in advance, and this correction can ensure the calculation accuracy of the moving average value calculation, making it possible to actually perform an error removal method using the moving average. becomes.

Next, the processing performed by the computer 6 based on the above-described relative displacement detection principle will be explained using the flowchart shown in FIG.

Read the G signal data from the digital data logger,
Record in internal memory (step 21).

Second-order integration is performed on the G signal data recorded in this memory to obtain a second-order integral final value. Under the assumption that the heights at the start and end of the measurement are equal, the final value of the second-order integral should be zero.

As mentioned above, the causes of error include the G signal offset go and the initial velocity vO, but ignoring the influence of the initial velocity vO,
The correction amount for the G signal offset is determined by regarding the error component only due to the error in the signal offset go. That is, when the measurement time is tl, go that satisfies (1/2) go ·t, 2 - fJ'g (t) dt2 is determined. Next, by calculating the second-order integral of the G signal with the go component corrected, a signal whose second-order integral final value becomes zero after 41 seconds of measurement time is obtained (
Step 22.23.24).

That is, the error components included at this time are as shown in FIG. 5, and the maximum value of the error components is (1/8) (go − go ) t+
2. The data of the second-order integral value whose final value is zero is recorded in the memory.

Next, calculate the moving average from the data of the second-order integral value whose closing price is zero to find the error component (step 25)
. Find the relative displacement by subtracting the moving average from the second-order integral value whose final value is zero, and output the result (
Step 26.27).

By using the method described above, it is possible to suppress the maximum value of error during calculation to a certain level or less, prevent deterioration of accuracy due to loss of digits in the computer, and as a result, reduce the initial velocity, G signal offset, etc. Error components based on the above are removed, and the relative displacement of the object to be detected can be detected with high accuracy even without a physical reference line.

Next, an application example of the above-described relative displacement detection method will be explained.

FIG. 7 shows a first application example, which is applied to measuring the longitudinal profile of a road surface. A G sensor 33 that detects the vertical acceleration of the vehicle 32 and a distance sensor 34 such as an ultrasonic sensor that measures the distance between the vehicle 32 and the road surface 31 are mounted on a bumper or the like of a vehicle 32 traveling on a road surface 31. is installed. The G sensor 33 has the G
A signal processing section 35 is connected to detect height fluctuations of the automobile 32 by performing second-order integration, moving average calculation, subtraction processing, etc. of the signal. The distance signal is then subtracted from the height variation of the vehicle 3.2 to determine the longitudinal shape of the road surface 31.

Further, as a means for obtaining partial displacement information of a moving object such as a car, a first G sensor is attached to the part where the displacement is to be detected so that a G signal in the direction of the displacement can be detected.
A second G sensor that detects displacement in the same direction as the detection direction of the first G sensor is attached to a reference part, and the G signal from the first G sensor and the G signal from the second G sensor are By processing the differential G signal, it is possible to obtain displacement information based on the location where the second G sensor is attached.

FIG. 8 shows a second application example using this detection principle, in which vibration displacement information of a hood 36 of a moving automobile 32 is obtained. In the same figure, 37 is the first G
Sensor 38 is a second G sensor.

As mentioned above, it is possible to incorporate a measurement device at low cost using an inexpensive G sensor without using an expensive optical displacement sensor, and it is possible to increase only the number of inexpensive G sensors and the memory for recording their G signals. This makes it possible to obtain not only one-point but two-dimensional vibration displacement information.

FIG. 9 shows a third application example, in which the first G sensor 37 and the second G sensor 38 are used to obtain stroke information of the automobile suspension, similar to the above example. . In the figure, 39 is the vehicle body, 40 is a spring, and 4
1 is a shock absorber.

Conventionally, suspension stroke information has been detected using a potentiometer, but this has a drawback in terms of durability. However, by applying this application example, reliability is significantly improved. Also, if you apply this application example,
Not only stroke information but also acceleration information can be used effectively, and it can also be applied to reliable damping force control of shock absorbers to improve riding comfort.

[Effects of the Invention] As explained above, according to the present invention, a certain correction value is added to the acceleration signal of the detected object detected by the acceleration detection means attached to the detected object, and the A first step of performing integration to obtain a second-order integral value whose final value is zero, and a second step of calculating a moving average value for a required moving average time from the second-order integral value whose final value is zero. , a third step of subtracting the moving average value from the second-order integral value at which the closing price becomes zero;
In the second step, an error component based on the initial velocity and the offset of the acceleration signal is determined, and in the third step, this error component is removed and the detected target is This method has the advantage of being able to detect relative displacements with high accuracy.

[Brief explanation of drawings]

1 to 6 are for explaining an embodiment of the relative displacement detection method according to the present invention. is a flowchart showing an example of the processing content of the computer in the above configuration example, FIG. 4 is a characteristic diagram showing the change over time of a general error component that occurs during second-order integration of an acceleration signal, and FIG. 6 is a flowchart for explaining the relative displacement detection method, and FIG. 7 is a configuration diagram showing a first application example of the relative displacement detection method according to the present invention. FIG. 8 is a block diagram showing the second application example, and FIG. 9 is a block diagram showing the third application example. 2: Acceleration sensor (acceleration detection means); 4: Digital data logger (recording means); 6: Computer that executes second-order integration, moving average calculation, subtraction processing, etc. Acting Patent Attorney Hidekazu San Yoshikazu Figure 1 Figure 2 Figure 6 Figure 3

Claims (1)

[Claims] A relative displacement detection method for detecting the relative displacement of an object to be detected, wherein the acceleration signal of the object to be detected detected by an acceleration detection means attached to the object to be detected is constant. A first step in which a second-order integral is performed by adding a correction value to obtain a second-order integral value whose closing price is zero, and a moving average for the required moving average time from the second-order integral value whose closing price is zero. A relative displacement detection method comprising: a second step of calculating a value; and a third step of subtracting the moving average value from the second-order integral value whose final value is zero.