JP4914732B2 - Rail displacement measuring method and rail displacement measuring device - Google Patents

Description

  The present invention relates to a method for correcting a detection value of a PSD element and a rail displacement amount measuring apparatus using the method, and more specifically, is mounted on a track inspection vehicle and uses a PSD element (position detection element) to measure the left and right displacement amounts of a rail. In this case, the present invention relates to a rail displacement amount measuring apparatus that can improve the measurement accuracy of the rail displacement amount by correcting the detected value of PSD corresponding to a PSD element (hereinafter referred to as PSD).

Since the generation timing of two detection signals having peaks obtained from both ends of the PSD shifts depending on the position at which the reflected light from the rail is received, the rail displacement measuring device determines the generation timing of the two detection signals as a voltage. The light receiving position is calculated as a value or a current value, and the ratio is used to calculate the amount of displacement of the rail.
As a device for measuring the left and right displacement amounts of the rail using this PSD, a rail displacement amount measuring device by the applicant is already known (Patent Document 1).
JP-A-6-42917

The PSD is an element using the surface resistance of a photodiode, and the light receiving position and the generation timing of two detection signals having peaks obtained from both ends do not necessarily correspond to 1: 1. Therefore, it is necessary to correct between the actual light receiving position and the position calculated from the detection signal.
As a simple correction, a correction value table is conventionally provided for the light receiving position of the PSD. However, if the resolution of the light receiving position is increased, the capacity of the correction value table is increased accordingly, and correction is performed. There is a problem that processing takes time. Therefore, an approximation method is adopted in which a certain range is corrected by linear approximation or broken line approximation. However, if the rail displacement amount measurement specification is ± 40 mm or less and the error is within ± 0.5 mm, it is difficult to measure the rail displacement amount while the track inspection vehicle is running at present.
On the other hand, if a laser length measuring device using a CCD other than PSD or the like is used in order to meet the requirements for the specifications described above, the device mounted on the track inspection vehicle must be large and expensive.
An object of the present invention is to solve such problems of the prior art, and a PSD detection value correction method capable of improving the position detection accuracy by correcting PSD detection values corresponding to PSD. It is to provide.
Another object of the present invention is to provide a rail displacement amount measuring apparatus capable of improving the accuracy of measuring a rail displacement amount by correcting a PSD detection value corresponding to PSD.

The present invention is a rail displacement amount measuring method and a rail displacement amount measuring apparatus that achieve the above object, characterized in that a logical PSD detected value y without error and the rail displacement amount x And a linear function expression y = kx (where k is a coefficient) corresponding to a straight line passing through the origin, and the detected value ya for at least three rail displacement amounts x and y = kx The coefficient A, B, C of this quadratic function is obtained from the relationship with the measured value ya with respect to the displacement amount x of the rail, which is known as a quadratic function ΔY = Ax ² + Bx + C. In search of
Further, the coefficients a and b of the linear function yc = ax + b established between the two different points of the detected value ya of the actually used PSD element and the rail displacement x are obtained, and the PSD element is detected at the time of measurement. Temporary rail displacement amount x ₁ according to the linear function equation is obtained from value ya , assuming ya = yc , and further, logical detection value y and detection value ya of the PSD element are obtained from temporary rail displacement amount x _1. rail near the detection value ya of the PSD element obtained by the error [Delta] Y to _{^{ΔY' = Ax 1 2 + Bx 1}} + C is corrected to yb = ya + ΔY ', the last yb = kx _d actual rail displacement x on the basis of the and requests the displacement _{x d.}

In the present invention, an error between a logical PSD detection value y having no error and an actual individual PSD detection value ya is calculated with respect to a rail displacement amount x calculated from the detection value ya in each PSD. This is because the inventors have found that a quadratic function can be approximated.
Therefore, in the present invention, the values of the coefficients A, B, and C of the quadratic function are obtained for the PSD actually used, and the rail (measurement object) is measured with the PSD actually used. The detected value ya is obtained, the provisional rail displacement amount x ₁ is obtained from the coefficients a and b of the linear equation, and ΔY ′ = Ax ₁ ² + Bx ₁ + C is obtained from this x ₁ to obtain the detected value ya as ΔY ′. The corrected detection value yb is obtained by correcting with.
As a result, a detection value yb close to the logical detection value ya without error can be obtained, and the rail displacement amount x as an actual measurement value can be calculated according to a logical linear function expression y = kx without error. .
By thus erroneous difference correcting the detection value ya in the rail displacement x ₁ of the provisional obtained from a linear function yc = ax + b quadratic function ^{_{ΔY = ΔY' = Ax 1 2 +}} Bx 1 + C, correcting the detected value ya The detected value yb obtained in this manner has a linear relationship with the light receiving position of the PSD and has a correspondence relationship close to 1: 1, so that the position detection accuracy can be improved.
As a result, the PSD detection accuracy can be improved, and the rail displacement measurement accuracy of the rail displacement measurement device can be improved.

FIG. 1 is an explanatory diagram of displacement measurement by a rail displacement measuring device that acquires coefficients of a linear function and a quadratic function in a rail displacement measuring device according to an embodiment to which the present invention is applied. FIG. FIG. 3 is a block diagram of the internal configuration of the detector, and FIG. 3 is an explanatory diagram of how to correct the PSD detection value in the rail displacement measuring device.
In FIG. 1, reference numeral 10 denotes a rail displacement measuring device mounted on the track inspection vehicle, which is installed in the measurement device 4 for each coefficient of the linear function and the quadratic function under the same conditions as the mounting state of the track inspection vehicle. Has been.
That is, the rail displacement measuring device 10 includes an optical measuring unit 7 and a rail displacement detecting unit 8 having a projector 5 and a light receiver 6 obliquely above the measuring rail 1 having a predetermined length, and a data processing / control device. 9 and is arranged in a relationship for measuring the measurement rail 1 of the measuring device 4 for each coefficient of the linear function and the quadratic function. The relationship between the measurement rail 1 and the rail displacement measuring device 10 coincides with the positional relationship with the running rail in the track inspection vehicle.
The optical measuring unit 7 irradiates a band-shaped laser pulse light L from the projector 5 in a direction perpendicular to the measurement rail 1, receives the reflected light R by the PSD 63 provided in the light receiver 6, and reflects the reflected light R on the PSD 63. Two detection signals corresponding to the light receiving position are obtained, and a detection value corresponding to the rail displacement amount (rail movement amount) x is generated.

A measuring rail 1 having a predetermined length is placed on a moving table 2 of a measuring device 4 for each coefficient, and the moving table 2 is moved about ± 40 mm (= 80 mm) at intervals of 5 mm by a ball / screw moving mechanism 3. . 4a is a stepping motor coupled to the screw shaft of the ball screw moving mechanism 3 and is driven by a drive circuit 4b. The drive circuit 4 b is controlled by the data processing / control device 9. As a result, the rail 1 moves in steps of 5 mm.
Reference numeral 5 denotes a projector. The laser light source 52 of the projector 5 uses a powerful LD that generates a laser pulse having a very short time width of several hundred to several hundreds of W, for example. Excitation is performed by a pulse drive circuit. This pulse drive circuit includes a pulse generator circuit 81 and a driver circuit 82 provided in the light projecting circuit section of the rail displacement detection section 8 shown in FIG.
The laser from the laser light source 52 is made into an optical band by the cylindrical lens 54 through the light projection lens 53, and is projected at a predetermined angle with respect to the measurement rail.
Here, it is assumed that the measurement rail is set at the initial position and moved forward and backward by ± 40 mm.

Reference numeral 6 denotes a light receiver. The light receiver 6 includes a mirror 61 that receives the wavelength of the laser, a focusing lens 62, and a PSD 63.
The laser reflected by the rail is reflected in the light receiving direction by the mirror 61, and is imaged by the focusing lens 62 at the light receiving position of the PSD 63 corresponding to the displacement of the measurement rail.
Detection signals from the two terminals A and B of the PSD 63 (see FIGS. 2A and 2B) are sent to the light receiving circuit of the rail displacement detector 8. As shown in FIG. 2A, the light receiving circuit outputs pre-amplifiers 64a and 64b, sample and hold circuits 65a and 65b connected to two terminals A and B of the PSD 63, and outputs of the sample and hold circuits 65a and 65b to A / It comprises a 16-bit A / D conversion circuit 66 for D conversion, and an optical fiber module transmission circuit 67 for transmitting the A / D converted data to the data processing / control device 9.
Note that a and b added to the reference numerals are circuits provided corresponding to the two terminals A and B of the PSD 63 shown in FIG. The A / D conversion circuit 67 converts the received light signals from the two terminals A and B of the PSD 63 alternately into digital values in a time division manner and sends them to the data processing / control device 9.
The light projecting circuit and the light receiving circuit of the rail displacement detector 8 are connected to and controlled by the data processing / control device 9, respectively.
The rail displacement detection unit 8 may be provided inside the optical measurement unit 7. In addition, a transmission filter for noise removal that transmits the wavelength of the laser may be provided in front of the mirror 61.

A data processing / control device 9 shown in FIG. 1 includes an MPU 91, a memory 92, an interface 93, a display 94, a keyboard 95, and a bus 96 for connecting them. A signal from the A / D conversion circuit 67 is received via the interface 93 and stored in a predetermined area of the memory 92. Reference numeral 97 denotes an external storage device such as an HDD.
The memory 92 stores a light receiving position calculation program 92a, a rail movement measurement program 92b, a PSD primary function calculation program 92c, a PSD secondary function calculation program 92d, a rail displacement calculation program 92e, a correction detection value calculation program 92f, and the like. A parameter area 92g and a work area 92h are provided.

The light reception position calculation program 92b is executed by the MPU 91 when it is called, and calculates a detection value corresponding to the light reception position on the PSD 63 from the following equation.
ya = (VA−VB) / (VA + VB)
However, ya is a distance from the terminal A, VA is a peak voltage value obtained from the terminal A, and VB is a peak voltage value obtained from the terminal B (see FIG. 2B). If current values are obtained from the two terminals A and B of the PSD 63, VA and VB are replaced with current values IA and IB, respectively. Hereinafter, the voltage value will be described.
Here, the initial position of the measurement rail on the moving table 2 is set so as to come to the center position of the PSD 63 where VA = VB at the position of ya = 0, but in reality, VA = VB Don't be. The light receiving position of the PSD 63 corresponding to the initial position of the measurement rail is defined as the reference position, the terminal B side direction is set to the-side (measurement rail is displaced inward), and the terminal A side direction from the center position is set to the + side (measurement rail is directed outward) As a displacement, the light receiving position ya on the PSD 63 is detected (calculated).
The MPU 91 further repeats the above measurement a plurality of times, calculates an average value of a plurality of measurement values at the position where the rail movement position is the same for each measurement position of the rail displacement stored sequentially, and corresponds to each position. After obtaining a large number of average values, for example, 16 and storing them in the work area 92h, the process returns to the caller main program.

  The rail movement measurement program 92b is called in response to a function key input for PSD characteristic measurement on the keyboard 95 and executed by the MPU 91. This is executed by the MPU 91, and the MPU 91 drives the pulse driving circuit (pulse generator circuit 81 and driver circuit 82) of the rail displacement detection unit 8, and the two terminals A and B of the PSD 63 from the light receiving circuit of the rail displacement detection unit 8. The detection signals VA and VB are obtained as digital values, the light reception position calculation program 92a is called, the measurement rail is moved to +40 mm at intervals of 5 mm for each light reception position calculation, and the detection value ya is measured and calculated. Returning to the initial position, the measurement rail is moved to −35 mm at −5 mm intervals, and the detected value ya is measured. A total of 16 calculated values are sequentially stored in the work area 92 h of the memory 92 so as to correspond to the rail moving position. To do. Then, the PSD primary function calculation program 92c is called.

The PSD primary function calculation program 92c is called and executed by the MPU 91. The MPU 91 executes this, and calculates a linear line yc = ax + b as a ride on the characteristic graph shown in FIG.
When there is no error, the light receiving position ya on the PSD 63 has a linear shape of y = k · x as shown in FIG. Should be listed. The vertical axis represents the light receiving position ya at 1000 magnification with the center position of PSD 63 where VA = VB as the origin, and 1000 magnifications.
The horizontal axis represents the movement amount X [mm] of the measurement rail. The slope k of y = k · x corresponds to a function established between the logical PSD detection value y with no error and the rail displacement amount x in a linear relationship passing through the origin. A coefficient obtained by adding 1000 times the above. This can be obtained by measuring the PSD of the same product actually used in a plurality or a plurality of rail displacement measuring devices and drawing an ideal straight line. K is stored in advance in the parameter area 92g.
However, the relationship between the light receiving position ya of the actual measurement data of the individual PSD 63 mounted on the rail displacement measuring device actually mounted on the inspection vehicle and the movement amount X of the measurement rail (corresponding to the rail displacement x) is Like the characteristic graph A shown in FIG. 3B, it is not ideal.
Therefore, the light receiving position ya and the measurement rail movement amount X are obtained from the PSD 63, and the slope a and the intercept b on the y-axis are obtained with the function yc = ax + b on the assumption that this is on a straight line.
The straight line can be calculated from, for example, two points: a value of ya when x = 30 mm and a value of ya when x = −30 mm.
For example, it is assumed that ya when x = 30 mm is ya = 4.5, and ya when x = −30 mm is ya = −2.8. Since ya is multiplied by 1000 as described above to obtain a functional expression,
4500 = a × 30 + b (1)
−2800 = a × −30 + b (2)
When the following two equations are solved, the slope a = 730/6 and the intercept b = 850 are obtained.
The MPU 91 stores the gradient a = 730/6 and the intercept b = 850 calculated by executing the PSD linear function calculation program 92c in the work area 92h, and calls the PSD quadratic function calculation program 92d.

The PSD quadratic function calculation program 92d is called and executed by the MPU 91. The MPU 91 executes this and calculates a quadratic curve shown as the characteristic graph B shown in FIG. This quadratic curve reads k from the parameter area 92g, and by using y = k · x shown in FIG. 3A, the value ya1 when x = 30 mm and the value ya2 when x = −30 mm are obtained. Further, when the slope a = 730/6 and the intercept b = 850 are read from the work area 92h and the function yc = ax + b, x = 30 mm, yc (= ya) value ya3 and x = −30 mm The value ya4 of yc (= ya) is calculated.
When the PSD linear function calculation program 92c is executed, if the value of ya at x = 30 mm and the value of ya at x = −30 mm, 1000 times the two times, are stored in the work area 92h, It is not necessary to calculate yc = ax + b. Therefore, if the value of the rail movement amount is the same two points, it is not necessary to calculate a linear function.
Next, the MPU 91 calculates a difference value ΔY = y−ya when x = 30 mm and x = −30 mm. This ΔY corresponds to the error between the logical PSD detection value y without error and the actual PSD detection value ya. In the present invention, this error is calculated by approximating a quadratic curve.
ΔY1 = ya3−ya1 (3)
ΔY2 = ya4−ya2 (4)
Assuming that the correction function is a quadratic curve function ΔY = Ax ² + Bx + C shown in FIG. 2C, a coefficient C when x = 0 is first obtained, and further, x = 30 mm, ya = ΔY 1, x = −30 mm, ya The coefficients A and B in this equation are obtained from .DELTA.Y2.
Next, the coefficients a and b of the linear function and the coefficients A, B and C of the quadratic function obtained as described above are stored in the parameter area 92g of the memory 92, respectively.

After obtaining the linear function coefficients a and b and the quadratic function coefficients A, B and C as described above, the rail displacement measuring device 10 is mounted on the track inspection vehicle. At this time, the measurement rail 1a is changed to a traveling rail for the track inspection vehicle. In this state, the rail displacement amount is measured by the rail displacement amount measuring apparatus 10.
In the running state of the track inspection vehicle equipped with the rail displacement measuring device 10, the rail displacement amount calculation program 92e performs the rotation of the wheels of the track inspection vehicle when a measurement start function key input is made on the keyboard 95. The MPU 91 is called by receiving a distance pulse generated in response to the distance pulse. This is executed by the MPU 91, and the MPU 91 calls and executes the light reception position calculation program 12b to obtain the calculated value of the light reception position as the detection signal ya. The detection signal ya at this time is a detection signal detected corresponding to the actual rail displacement in the running state of the track inspection vehicle.
Further, based coefficients of a linear function stored in the parameter area 92g of the memory 92 a, b from the detection value ya based read Te to the function yc = ax + b seek rail displacement x ₁ of the provisional. Next, the coefficients A, B, and C of the quadratic function stored in the parameter area 92g of the memory 92 are read, and based on this, an error ΔY ′ = Ax ₁ ² + Bx ₁ + C is calculated to detect the actual PSD detection value ya. Is corrected by adding an error ΔY ′ = Ax ₁ ² + Bx ₁ + C , and ya is added thereto to obtain a corrected detection value yb close to the logical PSD detection value y without error. Alternatively, the corrected detection value yb is obtained by yb = ya + Ax ₁ ² + Bx ₁ + C. Calculated from the calculated corrected detected value yb as last yb = k rail displacement near the rail displacement x to be actually measured based on x x _d.
As a result, the rail displacement amount x is calculated according to the distance pulse, and the rail displacement amount x is recorded in a predetermined area of the HDD 97 according to the current travel position.

  As described above, in the embodiment, the detected value of the PSD 63 is obtained by ya = (VA−VB) / (VA + VB). However, the detected value ya is obtained by obtaining each current value from both terminals of the PSD 63. It may be obtained as a value ratio.

FIG. 1 is an explanatory diagram of displacement amount measurement by a rail displacement amount measuring device that acquires coefficients of a linear function and a quadratic function in a rail displacement amount measuring device according to an embodiment to which the present invention is applied. FIG. 2 is a block diagram of the internal configuration of the rail displacement detector. FIG. 3 is an explanatory diagram of how to correct the PSD detection value in the rail displacement measurement device.

Explanation of symbols

1 ... measurement rail, 2 ... moving table,
3 ... Ball screw moving mechanism,
4 ... Measuring device for each coefficient of linear function and quadratic function,
4a ... stepping motor, 4b ... drive circuit,
5 ... projector, 52 ... laser light source,
53 ... Projection lens, 54 ... Cylindrical lens,
6 ... light receiver, 61 ... mirror, 62 ... focusing lens,
63 ... PSD, 64a, 64b ... preamplifier,
65a, 65b ... sample hold circuit,
66 ... A / D conversion circuit,
7: Optical measurement unit, 8 ... Rail displacement detection unit,
81 ... Pulse generator circuit, 82 ... Driver circuit,
9 ... Data processing / control device, 91 ... MPU,
92 ... Memory, 93 ... Interface,
94 ... Display, 95 ... Keyboard, 96 ... Bus,
97 ... an external storage device (HDD),
92a ... a light receiving position calculation program;
92b ... Rail movement measurement program,
92c ... PSD linear function calculation program,
92d ... PSD quadratic function calculation program,
92e ... Rail displacement calculation program,
92f... Correction detection value calculation program,
92g ... parameter area, 92h ... work area.

Claims (4)

Mounted on the track inspection and measurement vehicle, a projector and a light receiver are provided obliquely above the rail, and a belt-shaped laser pulse light is emitted from the light projector in a direction perpendicular to the rail, and the reflected light is provided to the light receiver. In a method for calculating a detection value of a PSD element, the detection value corresponding to the amount of rail displacement is calculated by obtaining two detection signals corresponding to the light receiving position of the reflected light on the PSD element received by the PSD element,
Corresponding to a straight line passing through the origin as a linear function expression y = kx (where k is a coefficient), a function established between the detected value y of the logical PSD element without error and the displacement amount x of the rail, From the detected value ya for at least three rail displacement amounts x and y = kx, the error ΔY (= y−ya) between them is known as a quadratic function ΔY = Ax ² + Bx + C with respect to the known rail displacement amount x. The coefficients A, B, and C of this quadratic function are obtained from the relationship with the actual measurement value ya,
Further, the coefficients a and b of the linear function yc = ax + b established between the two different points of the detected value ya of the actually used PSD element and the rail displacement x are obtained,
Determine the rail displacement x ₁ of the provisional according to formula of the linear function as if ya = yc from the detected value ya in the PSD during measurement, further wherein said logical detected value y from the rail displacement x ₁ of the temporary An error ΔY with respect to the detected value ya of the PSD element is obtained by ΔY ′ = Ax ₁ ² + Bx ₁ + C , and the detected value ya of the PSD element is corrected to yb = ya + ΔY ′ , and finally the actual rail based on yb = kx. A rail displacement amount measuring method for obtaining a rail displacement amount _xd close to the displacement amount x . Rail shift measuring method according to claim 1, wherein the light receiving position is calculated from the following equation on the PSD element.
ya = (VA−VB) / (VA + VB)
However, VA and VB are voltages for the two detection signals having peaks obtained from both ends of the PSD element. Mounted on the track inspection and measurement vehicle, a projector and a light receiver are provided obliquely above the rail, and a belt-shaped laser pulse light is emitted from the light projector in a direction perpendicular to the rail, and the reflected light is provided to the light receiver. In a rail displacement amount measuring apparatus for obtaining a detection value corresponding to a rail displacement amount by obtaining two detection signals corresponding to a light receiving position on the PSD element of the reflected light received by the PSD element,
Corresponding to a straight line passing through the origin as a linear function expression y = kx (where k is a coefficient), a function established between the detected value y of the logical PSD element without error and the displacement amount x of the rail, From the detected value ya for at least three rail displacement amounts x and y = kx, the error ΔY (= y−ya) between them is known as a quadratic function ΔY = Ax ² + Bx + C with respect to the known rail displacement amount x. Each coefficient A, B, C of this quadratic function is obtained from the relationship with the actual measurement value ya,
Further, coefficients a and b of a linear function yc = ax + b established between two different points of the detected value ya of the PSD element actually used and the displacement x of the rail are obtained, and the obtained a, b, A memory for storing A, B, and C;
From the detected value ya of the PSD element at the time of measurement, provisionally rail displacement amount x ₁ according to the linear function equation is obtained by assuming ya = yc , and yb is obtained by yb = Ax ₁ ² + Bx ₁ + C + ya , and finally yb = kx rail displacement measuring device for determining the rail displacement x _d close to the actual rail displacement x on the basis of. The rail displacement measuring device according to claim 3, wherein the light receiving position on the PSD element is calculated from the following equation.
ya = (VA−VB) / (VA + VB)
However, VA and VB are voltages for the two detection signals having peaks obtained from both ends of the PSD element.

Images