JPH01153909A - Correcting method for pulse/distance conversion coefficient of wheel speed sensor - Google Patents

Abstract

PURPOSE:To calculate travel-directional variation quantity data even at the time of tire replacement, variation in tire air pressure, tire wear, etc., by inputting pulse signal trains outputted by right and left wheel speed sensors selectively and calculating the relation between conversion coefficients set for the respective sensors. CONSTITUTION:The turning angle of a moving body is detected according to the output pulse signals of the wheel speed sensors 1L and 1R mounted on the left and right wheels of the moving body. Here, the output pulse signal trains of the sensors 1L and 1R in a specific travel state are inputted as independent output pulse signal trains. Those are inputted selectively to a conversion coefficient mean value calculation part 9 through a data decision part 8 according to said output pulse signal trains and distance data in a digital map memory 7. Then the pulse interval-distance conversion coefficients and the ratio of the coefficient are calculated and the pulse interval-distance conversion coefficients of the sensors 1L and 1R are corrected 13 according to them. Consequently, a state wherein the travel-directional variation quantity data is accurately calculated is secured.

Description

[Detailed Description of the Invention] <Industrial Application Field> The present invention relates to a pulse-to-distance conversion coefficient correction method for a wheel speed sensor, and more specifically, the present invention relates to a pulse-distance conversion coefficient correction method for a wheel speed sensor. The present invention relates to a pulse distance conversion coefficient correction method applied to a wheel speed sensor that outputs a pulse signal for calculating a rotation angle of a moving object.

<Prior art and problems to be solved by the invention> Navigation systems and location systems have been studied as systems for detecting the position of vehicles, which are representative of moving objects.
Since it is necessary to recognize the current position of the vehicle by detecting the vehicle's travel distance and rotation angle and sequentially integrating the detected data into the initial data, a travel distance sensor and a rotation angle sensor are required. It is.

As the mileage sensor, a wheel speed sensor mounted on the left and right wheels is used in the membrane, and as the rotation angle sensor, a geomagnetic direction sensor and a wheel speed sensor are used in the membrane. ing.

To explain the wheel speed sensor in detail, since pulse signals corresponding to the rotation angle are generated for each of the left and right wheels, the number of pulses p1 output from one wheel speed sensor and the moving distance M d l is di -a
t pi (however, al is the conversion coefficient),
The number of pulses p2 output from the other wheel speed sensor and the moving distance d2 are d2-a2p2 (however, a2 is a conversion coefficient)
They have the following relationship.

Therefore, the traveling distance #, L of the vehicle is the above-mentioned travel distance d
1. The simple average of d2 (i.e. L-(dl +d2)/
2+, and the rotation angle θ of the vehicle can be calculated as θ-(di-d2)/(where J is the distance M between the left and right wheels).

As a result, distance data can be obtained by multiplying the number of output pulses from the wheel speed sensors by respective conversion coefficients, and based on both distance data, it is possible to obtain the travel distance of the vehicle and the rotation angle of the vehicle. Regarding the rotation angle of the vehicle,
- It is calculated based on the output signal from the geomagnetic direction sensor in the membrane, but if the output of the geomagnetic direction sensor deviates significantly from the true direction due to reasons such as body magnetization, the wheel speed sensor The output pulse number data and the rotation angle of the vehicle obtained based on the conversion coefficient are very important data. That is, the conversion coefficient a1. when calculating the travel distance based on the output pulse number data from the wheel speed sensor. Since a2 is determined based on the mounting condition of the wheel speed sensor, the diameter of the wheel, the air pressure, etc., if the initial settings are made accurately for each vehicle, the output pulse from the wheel speed sensor will be It is possible to calculate fairly accurate rotation angle data based on several data, and by using it under the above circumstances, it is possible to continue detecting a fairly high current vehicle position.

However, since the tire air pressure, degree of wear, etc. change depending on the driving duration, tire type, etc., the initially set conversion coefficient a1. Even if a2 is accurate, there is no guarantee that it will be accurate throughout the journey, and there is a high possibility that the conversion coefficient will become inaccurate.If we look at the rotation angle detection operation, even though the vehicle is actually traveling straight, there is a high possibility that the conversion coefficient will be inaccurate. However, if the current vehicle position detection operation is continued based on this misrecognition, the cumulative error will become significantly large. There's a problem.

To explain in more detail, although most of the vehicle is running in a straight line, the above conversion coefficient al.

Since a2 inevitably contains some error,
Errors in detecting running direction due to running will be superimposed.

As a result, there is a problem in that a fairly large running direction detection error occurs when the vehicle is running in a straight line.

Further, the conversion coefficient a1. a2, the actual situation of the vehicle 4
When making corrections corresponding to this, a preset driving situation, for example, a figure-8 driving situation appears, and a new conversion coefficient is calculated based on the output pulse number data from the wheel speed sensor obtained in this situation. Since there are very few geographical conditions that allow such driving conditions, in reality, the conversion coefficient a
t.

Without correcting a2, leave it at the initial setting value, and based on the output pulse number data from the wheel speed sensor <--
It performs line short distance detection and rotation angle detection.
Most of the above problems have not been resolved.

<Object of the invention> This invention was made in view of the above problems,
Wheel speed sensors installed on the left and right wheels are based on the output pulse number data from the wheel speed sensors in driving conditions that frequently appear during normal driving, and the map data installed on the moving object. It is an object of the present invention to provide a method for correcting a conversion coefficient for output pulse number data from.

Means for Solving the Problems> In order to achieve the above object, the wheel speed sensor pulse-distance conversion coefficient correction method of the present invention is based on the pulse signal train output from each wheel speed sensor in a predetermined running state. is selectively taken in, and the magnitude of the pulse interval-distance conversion coefficient is calculated based on each pulse signal train and the distance data of the map memory mounted on the mobile object, and the pulse interval-distance conversion coefficient of each pulse signal is calculated. Calculate the ratio of the coefficients and calculate the above pulse interval -
This method corrects the pulse interval-distance conversion coefficient of each wheel speed sensor based on the magnitude of the distance conversion coefficient and the ratio of the pulse interval-distance conversion coefficients.

However, the above-mentioned predetermined running state is preferably a state in which the moving body is running straight, and it is also preferable that the second differential value of the rotation angle of the moving body with respect to the running distance is a running state in which the moving body is in an atmosphere. .

With the wheel speed sensor pulse-distance conversion coefficient correction method described above, the rotation angle of the moving object can also be detected based on the output pulse signal train from the wheel speed sensors attached to the left and right wheels of the moving object. In this case, the output pulse signal train from each wheel speed sensor is selectively taken in as a mutually independent output pulse signal train in a predetermined running state, and the travel distance is calculated based on the distance data of the map memory mounted on the mobile object. The magnitude of the pulse interval-distance conversion coefficient in each pulse signal train is calculated based on the calculated moving distance and the pulse signal train.

On the other hand, the ratio between the pulse conversion coefficients is calculated based on the pulse signal train, and the ratio between the calculated pulse conversion coefficients and
The pulse interval-distance conversion coefficient of each wheel speed sensor is calculated based on the magnitude of the above-mentioned pulse conversion coefficient, and the previous pulse-distance conversion coefficient is calculated.
By correcting the distance conversion coefficient, the relative relationship between the distance data converted based on the pulse signal trains output from the left and right wheel speed sensors can be approximated to the actual driving state.

If the above-mentioned predetermined traveling state is a state in which the moving object is traveling straight, the distance data after conversion will be equal, so each coefficient is calculated based on the ratio of both coefficients so that the distance data are equal. It is only necessary to correct the pulse-distance conversion coefficient of the wheel speed sensor, and the process for correction can be simplified.

Further, if the above-mentioned predetermined running state is a running state in which the second derivative value of the rotation angle of the moving body with respect to the running distance is zero,
Since the straight running state can be detected based on the state where the acceleration of the rotation angle of the moving body is zero, under this condition, the pulse-distance conversion coefficient of each wheel speed sensor is adjusted so that the distance data after conversion is equal. can be corrected, the detection of the straight-ahead traveling state can be ensured, and the processing for correction can be simplified.

To explain in more detail, let the pulse-distance conversion coefficients of the wheel speed sensors attached to each wheel of the moving body be al and a2, and the number of pulses output from each wheel speed sensor be pl and p2.
Then, each wheel is di -al pi S
It will move a distance of d2-a2 p2. Then, if the moving distance calculated based on the map memory is dO, then the new corrected pulse-distance conversion coefficient X1. X
2 becomes Xi-al do /di X2-a2 do /d2, respectively. Then, the moving distance of the moving object thereafter is L −
(pi XI + p2 X2 ) /2 (I
].

If the distance between each wheel is J, then the rotation angle θ of the moving body when traveling for a specified period is θ−(dl
-d2) /ノ= (Xi pi -X2 p2) /J
(I) becomes.

From the above formula (1), the moving distance can be found by the average of the moving distances of both + lj wheels, so as long as the average of both pulse-distance conversion m coefficients (i.e., the magnitude of the pulse-distance conversion coefficient) is accurate. , the distance can be calculated accurately. As long as the distance between points in the map memory is accurate, which is the basis for calculating the +2 average of the pulse-distance conversion coefficient,
The average of the pulse-distance conversion coefficients is an accurate value.

However, from the above equation [1[), since the rotation angle θ is calculated based on the difference in the moving distance of both wheels, the pulse conversion coefficients of both wheels need to be accurate. For example, if there is an error of about 5 x 10-' in the ratio of the conversion coefficients, a rotation error of 20 to 30 degrees will occur after traveling 1 km.

Furthermore, the pulse-distance conversion coefficient X1. Since X2 has a slight error due to the way the lane is set when curved on a wide road, a large error will occur if the rotation angle is calculated based on the above-corrected pulse-distance conversion coefficients Xi and X2. It ends up.

Therefore, the ratio of both pulse-distance conversion coefficients is separately calculated by the method shown below, and the pulse-distance conversion coefficient is corrected based on this ratio.

That is, in the above equation (I[), if the rotation angle θ is known, the interval J is known and the number of pulses p1. p2
Since is output from the wheel speed sensor, it becomes Xl-(θJ+X2 p2 )/pl, and the relationship between Xl and X2 is obtained.

In this case, if θ is very small compared to the traveling distance, θ can be ignored, and XI ′, X2 p
2/pi. Therefore, if the initial conversion coefficient does not satisfy the above relationship, by correcting the conversion coefficient based on the ratio Xi /X2 obtained in this way, a state that satisfies the above relationship can be obtained. Based on the pulse train data output from each wheel speed sensor, it is possible to accurately detect whether the wheel is traveling straight or rotating, and in the rotating state, it is possible to accurately determine the rotation angle. can be detected. It should be noted that either the calculation of the magnitude of the pulse-distance conversion coefficient or the calculation of the ratio of the pulse-distance conversion coefficient may be performed first.

In addition, when the conversion coefficient is corrected based on the pulse train data output from each wheel speed sensor while the moving object is moving straight, the above θ becomes zero, so the above ratio becomes very large. This becomes an accurate value, and the ratio between the conversion coefficients can be corrected with high precision.

Furthermore, when correcting the conversion coefficient based on the pulse train data output from each wheel speed sensor in a running state where the acceleration of the rotation angle of the moving object is zero, it is easy to detect the state in which the moving object is moving straight. In this state, the ratio between the conversion coefficients can be corrected with high precision in the same way as above.

<Example> Hereinafter, embodiments will be described in detail with reference to the accompanying drawings showing examples.

FIG. 1 is a block diagram showing an embodiment of a device for carrying out the present invention, which includes wheel speed sensors (LL) and (IR) respectively installed in the left and right wheels of a vehicle, and from each wheel speed sensor. Pulse number counter (2L) that counts the number of pulses pL and pR by inputting the output pulse signal train
(2R), a conversion coefficient memory (3!,) (3R) storing conversion coefficients aL and aR for pulse-to-distance conversion corresponding to each wheel speed sensor (IL) (IR), A distance calculation unit (4L) (4R) generates distance data dL, dr by inputting the pulse number count data and the conversion coefficient and multiplies both, and averages both distance data dL, dR calculated above. a mileage calculation unit (5) that generates vehicle mileage data d;
An angle calculation unit (6) that generates rotation angle data θ by calculating (dL - dR) /it based on the calculated distance data and preset wheel spacing data J; Digital map memory (7
), and a data discrimination unit ( 8), and both distance data dL calculated on the condition that calculation is possible by the data discriminator (8),
Using dR and the distance dO between the points in the map memory (7) as input, XL -aL dO/dL, XR-aRdO/
dRaL' -aL (XL +XR)
/ (aR+aL)aR'-aR(XL
+XR) / (aR+aL) The average magnitude of the pulse-distance conversion coefficient aL', aR
A conversion coefficient average value calculation unit (9) that calculates the conversion coefficient average value calculation unit (9) that calculates the second-order differential value calculation unit (10) that generates the second-order differential value d2θ/dd2 regarding the distance of the rotation angle data θ,
A second differential value determination unit (1
1) and the discrimination result signal from the second-order differential value discriminator (■1) indicating that the second-order differential d2θ/dd2 is zero is supplied.
Output data dL, dl from (4R)? a conversion coefficient ratio calculation unit (12) that calculates the ratio aL'/aR'-pR/pL between the conversion coefficients based on
Based on L1 and the conversion coefficients aR and aL stored in the conversion coefficient memory (3R), (aR+aL)·(pR/(pR+pL)).

It is composed of a conversion coefficient correction unit (13) that calculates a new conversion coefficient by performing the calculation of (aR+aL)・l pL / (pR+pL)l and supplies it to the conversion coefficient memory (3L) (3R). .

To explain in more detail, each of the wheel speed sensors described above has a magnetic belt, in which N and S poles are alternately located, attached in a ring shape to the inside of the rim, and is placed close to the magnetic belt at a predetermined position on the vehicle body. It has a configuration with a sensor coil installed at the
A pulse signal is output every time the wheel rotates by a predetermined angle.

Further, the conversion coefficients aL and aR are initially set based on the live pipe state of the magnetic belt, the outer diameter of the wheel, etc.
Specifically, it is calculated by performing a predetermined calculation based on the number of pulses obtained by driving a predetermined distance or driving in which the driving direction changes 360', and the calculation result is converted. These are stored as initial values in the coefficient memories (3L) and (3R), respectively.

FIG. 2 is a diagram illustrating a case in which the average magnitude of the pulse-distance conversion coefficient can be corrected. The solid line is the trajectory of the vehicle, the thick solid line is the road pattern in the map memory (7), and the point A, B is stored in the map memory (7), and the distance between points with the center of the intersection as a reference is stored as additional data.

The data discriminator (8) detects the case where both of the two points A and B stored in the map memory (7) turn right based on the trajectory of the vehicle, and only in the above case, the average of the pulse-distance conversion coefficients is determined. In other cases, the average magnitude of the pulse-distance conversion coefficient cannot be corrected, and the starting point A' of the curve is recognized as the digital map memory (force point A). At the same time, the starting point B' of the next curve is recognized as point B in the digital map memory (7).

That is, when turning right, the vehicle turns close to the left side of the center of the intersection, so the distance between the starting points of this curve matches the distance between points A and B in the map memory (7).

The operation of the device having the above configuration is as follows.

As the vehicle travels, the left and right wheels rotate, and pulse signal trains are output from the wheel speed sensors (IL) and (IR), respectively. Then, each pulse signal train is supplied to a pulse number counter (2L) <2R) to count the number of pulses, and each count data pL, pR and conversion coefficient memory (3L) (3R) are stored. Based on the conversion coefficients aL and aR, the distance calculation unit (4L
)<4R) Distance data dL-aL for each wheel
Calculate pL, dR=aRpR.

Then, by supplying the calculated distance data dL, dR to the mileage calculation unit (5), the mileage data d-(dL+dR)/2 of the vehicle is calculated, and the distance data dL, d R to angle calculation unit (6)
By supplying vehicle rotation angle data θ−(d
l - dR)/1 and supplying the travel distance data d1 and rotation angle data θ to an integration section (not shown), it is possible to detect the current position of the vehicle.

In addition, while the current position detection operation is being performed as described above, the data discriminator (8) determines the trajectory of the vehicle generated from the travel distance data d and rotation angle data θ of the heavy vehicle, and the map memory ( 7) Based on the road pattern stored in 7), detect the case where both points A and B turn right, and determine that the average magnitude of the pulse-distance conversion coefficient can be calculated only in the above case; In other cases, calculation of the average magnitude of the pulse-distance conversion coefficients is not performed. Then, the starting point A' of the curve is recognized as the point A of the digital map memory (7), and the starting point B' of the next curve is recognized as the point B of the digital map memory (7).

Distance data dL, dR from the distance calculation sections (4L) (4R) on the condition that the average magnitude of the pulse-distance conversion coefficients can be calculated by the data discrimination section (8).

By inputting the point-to-point distance data dO from the map memory into the conversion coefficient average value calculation unit (9),
Calculation of aL d(1/dL, XR-aRd[l/dR, and aL'-aL(XL+XR)/(aR+
aL )aR' -aR(XL +XR) / (aR
+aL) to correct the average size of the conversion coefficients.

Further, the rotation angle data θ is calculated by the quadratic differential gi calculation unit 00.
), a second-order differential dθ/dd2 with respect to the distance of the rotation angle data θ is generated, and a second-order differential value determining unit (11) determines whether the second-order differential value is 0 or not, that is, whether it is a straight line. It is determined whether the vehicle is traveling or not (see FIG. 3). If it is determined that the vehicle is traveling in a straight line, the distance data for each wheel is dL −aL pL,
Although dR-aRpR should be equal to each other,
dL-dl? The following relationship holds true, and based on this relational expression, the conversion coefficient ratio calculation unit (12) can calculate the ratio aL'/aR'-pg/pL of both conversion coefficients. After that, the obtained ratio a L' /a
A conversion coefficient correction unit (I3) based on R' and conversion coefficients aR and aL stored in the conversion coefficient memory (3R).
at (aR+aL)-1pR/(pR+pL)l.

A new conversion coefficient is calculated by performing the calculation of (at?+aL)・l pL / (pR+pL)l, and is supplied to the conversion coefficient memory (3L), and both conversion coefficient memories (3L
) (3R) is the ratio of the conversion coefficients stored in pR/pL
can be made to be equal to

Therefore, the distance data dL and dR for each wheel calculated in the course of subsequent driving are equal to each other in a straight-line driving state, and relatively accurate rotation angle data can be obtained even in a curved driving state. become. In other words, when the geomagnetic orientation sensor is not installed, or when the geomagnetic orientation sensor is affected by the magnetic field of the vehicle body and cannot accurately detect the orientation, the wheel speed sensor (LL)
Accurate travel direction change amount data can be obtained based on the pulse signal trains output from the (IR), and by integrating the data with the travel distance, the current position can be accurately detected.

FIG. 4 shows a flowchart for carrying out the present invention. In step (3), by driving between predetermined points, a pulse signal train pL, pR is output from a wheel speed sensor (IR).
, and point-to-point distance data dO from the map memory. In step ■, it is determined whether the data is sufficient to calculate the magnitude of the pulse-distance conversion coefficient, and if it is determined that the data is sufficient, in step ■, the currently stored pulse - distance conversion coefficient a L,
Based on aR, calculate dL = aL pL, dR - aRpR, and in step (2), XL - aL dO/dL, XR - aRdO/dR
Then, in step (■), aL
' -aL (XL +XR) / (aR+aL
)aR' -aR(XL +XR) / (aR+aL
) to correct the average size of the conversion coefficients. If it is determined in step (2) that the data is not sufficient to calculate the magnitude of the pulse-distance conversion coefficient, step (2) is performed.

Next, in step (2), it is determined whether the data is sufficient to calculate the ratio of the left and right pulse-distance conversion coefficients, and if it is determined that the data is sufficient, in step (2), aL' −(aR+aL)・l pR/(pR+
pL)1aR'-(aR+aL)・(pL/(pR
+ pL) 1 and use this value as the corrected conversion coefficient (however, for aR and aL, if aR' and aL' have already been calculated in step ① above, use these values. do).

In addition, in the data discriminator (8), the map memory (7)
The pulse-distance conversion coefficient is corrected only when turning right at both points A and B stored in
This is because if the starting points of a curve for a left turn on a wide road are recognized as points A and B on the map memory, the distance will be shorter than the actual travel distance, making it impossible to obtain accurate distance data.

However, if road width data at the intersection Z point is stored, accurate distance data can be obtained by taking this road width data into account.

In addition, in the above embodiment, the conversion coefficient is corrected only in a straight-line driving state, but it is possible to correct the conversion coefficient even in a driving state that includes curved driving as follows. be.

That is, since the rotation angle data θ is given by (dL - dR) /J, when the rotation angle data θ is obtained from a direction sensor, a road map, etc., θJ + dR - dL
becomes. Therefore, θ10aRpRwaL pL, which becomes aL−(θ)+aRpR)/pL.

As a result, by correcting the conversion coefficients based on the formula for aL, both conversion coefficients can be brought into a state in which travel direction change amount data can be accurately calculated. In particular, if the rotation angle data θ is extremely small, θ can be ignored, and in this case, the conversion coefficient can be corrected based on the ratio between the two. In other words, when a geomagnetic azimuth sensor is not installed, or when the geomagnetic azimuth sensor is affected by the magnetization of the vehicle body and cannot detect accurate azimuth, each wheel speed sensor (LL>(IR) Accurate travel direction change amount data can be obtained based on the output pulse signal train, and by integrating it together with the travel distance, accurate current position detection can be performed.

Note that the present invention is not limited to the above-described embodiments; for example, when the second-order differential value of the rotation angle data θ with respect to the distance is O, instead of correcting the conversion coefficient,
When the driver recognizes that the driver is driving in a straight line and operates a switch, or when the driver recognizes that the driver is driving in a straight line due to matching with the road traffic network, the conversion coefficient can be corrected. In addition, it is possible to appropriately set the side on which the conversion coefficients are corrected, and it is also possible to make various design changes without changing the gist of the present invention.

<Effects of the Invention> As described above, the present invention selectively takes in the pulse signal train output from the left and right wheel speed sensors in a predetermined running state, and converts the conversion coefficient to be set for each wheel speed sensor. Since the relationship between the two is calculated and the conversion coefficient is corrected based on the calculated relationship, even when tires are replaced, tire air pressure changes, tires are worn out, etc. It is possible to ensure a state in which the azimuth change amount data can be calculated accurately, which has the unique effect of increasing the accuracy of detecting the current position of the vehicle.

[Brief explanation of the drawing]

FIG. 1 is a block diagram showing an embodiment of an apparatus for carrying out the present invention, FIG. 2 is a diagram illustrating a case where the average magnitude of the pulse-distance conversion coefficient can be corrected, and FIG. FIG. 3 is a diagram illustrating a relationship between an actual running state and a second-order differential d2θ/dd2. Figure 4 is a flowchart. (IL) (IR)...Wheel speed sensor, (2L) (2
R)...Pulse number counter, (3L) (3R)...
・Conversion coefficient memory, (7)...Digital map memory,
(8)...Data discrimination unit, (9)...Conversion coefficient average value calculation unit, (12)...Conversion coefficient ratio calculation unit, (13)...Conversion coefficient correction unit Patent applicant Sumitomo Electric Industries, Ltd. Co., Ltd. (2 others) 1-U11 Figure 3 Figure 4

Claims (3)

[Claims] 1. In a position detection device that also detects the rotation angle of a moving object based on the output pulse signal train from wheel speed sensors attached to the left and right wheels of the moving object, the output pulse signal from each wheel speed sensor in a predetermined running state is The pulse interval-distance conversion coefficient is calculated based on each pulse signal train and the distance data of the map memory mounted on the mobile object by selectively importing the pulse interval-distance of each pulse signal. A ratio between the conversion coefficients is calculated, and the pulse interval-distance conversion coefficient of each wheel speed sensor is corrected based on the magnitude of the pulse interval-distance conversion coefficient and the ratio of the pulse interval-distance conversion coefficients. A pulse-distance conversion coefficient correction method for a wheel speed sensor. 2. The pulse-distance conversion coefficient correction method for a wheel speed sensor according to claim 1, wherein the predetermined running state is a state in which the moving object is running straight. 3. The pulse-distance conversion of the wheel speed sensor according to claim 1 or 2, wherein the predetermined running state is a running state in which the second derivative of the rotation angle of the moving body with respect to the running distance is zero. Coefficient correction method.