TWI769695B - Calibration method of magnetic linear position sensor - Google Patents

Abstract

A method for calibrating a magnetic linear position sensor, which causes a magnet to move along a straight line while making a processing unit of the magnetic linear position sensor request a magnetic angle sensing element of the magnetic linear position sensor to periodically return a sinusoidal signal generated at the moment and the cosine signals until the magnet completes the linear travel; the signal processing device receives the digitized sine signals and the cosine signals from the processing unit, and subtracts a first preset value from the sine signal values and the subtraction. A second-order derivative (atan2) operation of the arctangent function is performed on the cosine signal values from a second preset value to obtain a curvature curve, and a polynomial equation fitting an effective line segment of the curvature curve is obtained, and the The polynomial equation is written in the processing unit as the mathematical model of the magnetic angle-type sensing element.

Description

Calibration method of magnetic linear position sensor

The present invention relates to a calibration method of a position sensor, in particular to a calibration method of a magnetic linear position sensor.

Referring to FIG. 1 , one way to obtain the detection distance of a magnetic angle sensing element S (such as AMR, Hall, TMR....) is to make the magnetic angle sensing element S sense a magnet moving along a linear stroke P1 The magnetic field of M causes a sine wave signal and a cosine wave signal related to a moving distance of the magnet M to be generated and output to a signal processor 10. The signal processor 10 digitizes the sine wave signal and the cosine wave signal into The digitized sine wave signal sin formed by the complex sine signal value Usin as shown in Fig. 1 and the digitized cosine wave signal cos formed by the complex cosine signal value Ucos shown in Fig. 1, and the sine signal values Usin and These cosine signal values Ucos are subjected to the second derivative (atan2) operation of the arc tangent function (ARCTAN), namely atan2(Usin/Ucos), to obtain complex curvature values, and these curvature values form a curvature as shown in Figure 1 Curve C1, the curvature value of each point on an effective line segment L1 in the curvature curve C1 corresponds to a detection distance, and because the curvature value of the effective line segment L1 only ranges from 0.2 to 1.2, it represents the magnetic angle sensor element S. An effective detection distance is not For example, as shown in Figure 1, it is only 6mm. Therefore, if the length to be detected is 6 times of 6mm, at least 6 magnetic angle sensing elements S need to be connected in parallel to detect at the same time.

Therefore, the purpose of the present invention is to provide a calibration method for a magnetic linear position sensor, which can increase the detection distance of the magnetic angle sensing element in the magnetic linear position sensor, thereby reducing the use of the magnetic angle sensing element quantity.

的m值以及β₀~β_m的值，其中i=1,2,3...n，y_i代表該磁性角度型感應元件與該磁鐵的一相對距離，xi代表該曲率值，β₀~β_m是該磁性角度型感應元件的係數；及(G)該訊號處理裝置將該多項式方程式寫入該處理單元中做為該磁性角度型感應元件的一數學模型。 Therefore, the present invention provides a method for calibrating a magnetic linear position sensor. The magnetic linear position sensor is arranged on one side of a linear travel to detect the position of a magnet that reciprocates along the linear travel. The magnetic linear position sensor includes a Magnetic angle type induction element and a processing unit; the calibration method includes: (A) making the magnet move along the linear stroke, and at the same time, a signal processing device makes the processing unit request the magnetic angle type induction element to return with a sampling frequency A sine signal and a cosine signal are generated at the moment until the magnet completes the linear travel; (B) the processing unit digitizes the sine signals and the cosine signals continuously transmitted from the magnetic angle sensing element and outputs them to the signal processing device; (C) the signal processing device obtains the digitized sine signal values and the cosines of the position of each sampling point of the magnet in the linear stroke according to the length of the linear stroke and the sampling frequency The corresponding relationship between the signal values, and the signal processing device subtracts a first predetermined value from the sine signal values, and subtracts a second predetermined value from the cosine signal values; (D) the signal processing device pairs The sine signal values minus the first preset value and the cosine signal values minus the second preset value are subjected to the second derivative (atan2) operation of the arc tangent function, that is, atan2 (minus the first preset value) setting the sine signal value of the value/subtracting the cosine signal value of the second preset value) to obtain a corresponding complex curvature value, and the curvature values form a curvature curve; (E) the signal processing device uses An effective line segment in the curvature curve is used as a characteristic curve of the magnetic angle sensing element, and an effective detection distance and a corresponding curvature value range of the magnetic angle sensing element are determined according to the characteristic curve; (F ) The signal processing device determines a polynomial equation for fitting the characteristic curve by curve fitting and linear regression analysis according to the effective detection distance and the curvature value range

The m value and the value of β ₀ ~β _m , where i=1,2,3...n, _yi represents a relative distance between the magnetic angle-type induction element and the magnet, xi represents the curvature value, β ₀ ~ _βm is a coefficient of the magnetic angle sensing element; and (G) the signal processing device writes the polynomial equation into the processing unit as a mathematical model of the magnetic angle sensing element.

In some embodiments of the present invention, the first preset value is one of the sum of the maximum value and the minimum value among the sine signal values; the second preset value is one of the cosine signal values One of two of the sum of the maximum and minimum values.

In some embodiments of the present invention, m is 6.

In some embodiments of the present invention, the magnetic angle sensing element has two magnetoresistive bridges with a difference of 45°, wherein one magnetoresistive bridge induces the magnetic field of the magnet and generates a sine wave signal, and the other magnetoresistive bridge induces a sine wave signal. The resistive bridge senses the magnetic field of the magnet and generates a cosine wave signal with a phase difference of 45° from the sine wave signal.

The effect of the present invention is: in the calibration procedure, the sine signal values and the cosine signals obtained by the induction of the magnetic angle-type sensing element are appropriately shifted and then the second derivative of the arc tangent function (ARCTAN) is performed on them ( atan2) operation can increase the effective detection distance of the magnetic angle sensing element, thereby reducing the number of magnetic angle sensing elements used and making the detection distance of the magnetic linear position sensor 2 longer. .

1: Cylinder

2: Magnetic Linear Position Sensor

21: Magnetic angle sensing element

211, 212: Magnetoresistive bridge

22: Processing unit

3: Signal processing device

M: Magnet (Piston)

P2: Linear stroke

SIN: Sine wave signal

COS: cosine wave signal

sin: digitized sine wave signal

cos: digitized cosine wave signal

sin': digitized sine wave signal after translation

cos': digitized cosine wave signal after translation

C2: Curvature Curve

L2: valid line segment

L2': Effective line segment after replacement

S1~S7: Steps

Other features and effects of the present invention will be clearly shown in the embodiments with reference to the drawings, wherein: FIG. 1 illustrates a method for obtaining the detection distance of the magnetic angle sensing element; FIG. 2 is the magnetic linear position of the present invention. A flow chart of an embodiment of a method for calibrating a sensor; FIG. 3 illustrates the components and their arrangement of the magnetic linear position sensor to be calibrated in this embodiment; FIG. 4 is a magnetic linear position sensor included in this embodiment A schematic diagram of a detailed circuit of an angle-type sensing element; FIG. 5 is a schematic diagram of waveforms of the digitized sine wave signal sin and the digitized cosine wave signal cos after being digitized by the processing unit of the present embodiment; FIG. 6 illustrates the digitized sine wave signal of FIG. 5 wave signal sin and digitized cosine wave signal The number cos is shifted downward by a first preset value and a second preset value, respectively; FIG. 7 illustrates the second derivative of the arc tangent function (ARCTAN) performed on the sine wave signal sin' and the cosine wave signal cos' shown in FIG. 6 (atan2) operation to obtain corresponding complex curvature values, which constitute a curvature curve C2; and FIG. 8 illustrates the replacement of the X-axis data and the Y-axis data of the effective line segment L2 of the curvature curve C2 shown in FIG. 7, Instead, it becomes the effective line segment L2' after replacement.

Before the present invention is described in detail, it should be noted that in the following description, similar elements are designated by the same reference numerals.

Referring to FIG. 2, it is the main flow of an embodiment of the calibration method of the magnetic linear position sensor of the present invention, and as shown in FIG. side, to detect the position of a magnet M reciprocating along the linear stroke P2, for example, the magnet M is a piston set in a cylinder 1, and the linear stroke P2 is the reciprocating motion of the piston in the cylinder 1 Piston stroke. The magnetic linear position sensor 2 mainly includes a magnetic angle sensing element 21 and a processing unit 22, such as a microprocessor or a microcontroller (MCU). The magnetic angle sensing element 21 is arranged on one side of the linear stroke P2, such as the outer wall of the cylinder 1, to sense the magnetic field of the magnet M and output an analogous sine signal and a cosine signal; the processing unit 22 and the The magnetic angle type induction element 21 is electrically connected to connect receive the sine signal and the cosine signal.

Since the characteristics of the magnetic angle-type sensing element 21 in each magnetic linear position sensor 2 are different, the magnetic linear position sensor 2 needs to undergo a calibration procedure before the magnetic linear position sensor 2 leaves the factory. To find the mathematical model that conforms to the characteristics of the magnetic angle sensing element 21; and as shown in FIG. 4, the magnetic angle sensing element 21 (eg AMR, Hall, TMR... Therefore, in the calibration method of this embodiment, as shown in step S1 of FIG. 2 , first, the magnet M is moved away from the linear stroke P2 One end of the magnetic angle sensing element 21 moves toward the magnetic angle sensing element 21 , and then moves away from the magnetic angle sensing element 21 to the other end of the linear stroke P2 after passing through the magnetic angle sensing element 21 . During this process, the two magnetoresistive bridges 211 and 212 will continue to induce the magnetic field of the magnet M to generate a sine wave signal SIN and a cosine wave signal COS with a phase difference of 45°.

At the same time, a signal processing device 3, such as but not limited to a personal computer, causes the processing unit 22 to request the magnetic angle sensing element 21 to return the currently generated sinusoidal signal (ie the The value of a certain point of the sine wave signal SIN, the analog voltage value) and the cosine signal (that is, the value of a certain point of the cosine wave signal COS, the analog voltage value), until the magnet M completes the linear stroke P2, thereby, The magnetic angle sensing element 21 will return 10 sine signals and 10 cosine signals to the processing unit 22 every second. Then, as shown in step S2 in FIG. 2 , the processing unit 22 digitizes the received sine signals and the cosine signals and outputs them to the signal processing device 3 . The digitized values of the sine signal values Usin and the cosine signal values Ucos can be compared with the values indicated by the vertical axis on the left side of FIG. 5 .

Therefore, the signal processing device 3 can know and digitize the position and digitization of each sampling point of the magnet M in the linear stroke P2 according to a moving distance of the magnet M (ie the length of the linear stroke P2 ) and the sampling frequency The corresponding relationship between the sine signal values Usin and the cosine signal values Ucos, such as the digitized sine wave signal sin composed of the digitized sine signal values Usin shown in FIG. A digitized cosine wave signal cos composed of the digitized cosine signal values Ucos.

Then, in step S3 of FIG. 2 , the signal processing device 3 subtracts a first preset value from the sinusoidal signal values Usin to obtain Usin′, as shown in FIG. 6 , which is equivalent to the digitized sinusoidal signal sin to the The first preset value is shifted downward to become a shifted digitized sine wave signal sin', and the signal processing device 3 subtracts a second preset value from the cosine signal values Ucos to become Ucos', such as As shown in FIG. 6, it is equivalent to shifting the digitized cosine wave signal cos downward (offset) the second preset value to become a shifted digitized cosine wave signal cos'; and in this embodiment, the first The default value is one of two, but not limited to, the sum of the maximum value and the minimum value among the sine signal values Usin; the second default value is the maximum value and the minimum value among the cosine signal values Ucos One of two, but not limited to, the sum of the values.

Next, as shown in step S4 in FIG. 2 , the signal processing device 3 subtracts the sine signal values Usin' (ie, the digitized sine wave signal sin' after the translation) from the first preset value and subtracts the second preset value. These cosine signal values Ucos' (that is, the digitized cosine wave signal cos' after translation) are set to perform the second derivative (atan2) operation of the arc tangent function (ARCTAN), namely atan2(Usin'/Ucos'), and obtain the phase Corresponding complex curvature values, these curvature values can be compared with the values indicated by the vertical axis on the right side of FIG. 7, and these curvature values constitute a curvature curve C2 as shown in FIG. 7, and an effective line segment L2 of the curvature curve C2 represents A characteristic curve of the magnetic angle sensing element 21 determines an effective detection distance of the magnetic angle sensing element 21 and a corresponding curvature value range. For example, when the curvature value is -1, it can be determined from the effective line segment L2 (the characteristic curve) shows that the distance value is -2mm. Compared to FIG. 3 , it can represent that the magnet M is located on the left side of the magnetic angle type induction element 21 and is 2mm away from the magnetic angle type induction element 21 ; Similarly, when the curvature value is 2, it can be known that the distance value is 2.6mm, which means that the magnet M is on the right side of the magnetic angle type induction element 21 and is about 2.6mm away from the magnetic angle type induction element 21. position; and as shown in FIG. 7, in this embodiment, since the curvature value range can reach -π~π, the corresponding effective detection distance can be increased to 14mm, it can be seen that this embodiment through After the signal value sensed by the magnetic angle sensing element 21 is shifted, and the second derivative (atan2) of the arc tangent function (ARCTAN) is calculated, the detection distance (14 mm) of the magnetic angle sensing element 21 can be increased. to more than twice the conventional technology (6mm), therefore, Assuming that the length to be detected is 6 times as long as 6 mm, only three magnetic angle sensing elements 21 need to be connected in parallel for detection at the same time, thereby reducing the number of magnetic angle sensing elements 21 used.

的m值以及β₀~β_m的值，其中 i=1,2,3...n，y_i代表該磁性角度型感應元件21與該磁鐵M的一相對距離，xi代表該曲率值，β₀~β_m是該磁性角度型感應元件21的係數。 Then, as shown in step S6 of FIG. 2 , the signal processing device 3 uses the effective line segment L2 (ie, the effective detection distance and the corresponding curvature value range) as the characteristic curve of the magnetic angle sensing element 21 , and Determine the effective detection distance of the magnetic angle type sensing element 21 and the corresponding curvature value range, and the signal processing device 3 performs curve fitting and linear regression analysis according to the effective detection distance and the curvature value range determine a polynomial equation that fits the characteristic curve

The m value and the value of β ₀ ~β _m , where i=1,2,3...n, _yi represents a relative distance between the magnetic angle type induction element 21 and the magnet M, xi represents the curvature value, β ₀ to β _m are coefficients of the magnetic angle type induction element 21 .

，例如採用多項式擬合(Polynomial Fitting)，以一元m次多項式回歸方程式

來擬合該置換後有效線段L2’，並以 如下所示的矩陣解多項式回歸方程式：

Specifically, as shown in FIG. 8(A) , it is assumed that the effective line segment L2 (the characteristic curve) is defined by the distance values corresponding to 100 positions (that is, the distances between 100 magnets M and the magnetic angle type induction element 21 , that is, the The effective detection distance is composed of 100 curvature values (the curvature value range), in this embodiment, the signal processing device 3 will first replace the data of the X axis and the Y axis of the effective line segment L2, so that the X axis Instead, the curvature values are displayed and the Y-axis is replaced with the corresponding distance values, so that the effective line segment L2 becomes the replaced effective line segment L2', as shown in FIG. 8(B). Then, the signal processing device 3 determines the most suitable polynomial equation for the permuted effective line segment L2'

, for example, using polynomial fitting (Polynomial Fitting), a univariate m-degree polynomial regression equation

to fit the permuted effective line segment L2' and solve the polynomial regression equation with the matrix shown below:

Wherein y ₁ ~y _n represent the 100 distance values, x ₁ ~x _n represent the 100 curvature values, β ₀ ~β _m are coefficients, and since the characteristics of each of the magnetic angle sensing elements 21 are different, the The coefficients β ₀ to β _m are also different, so through the above matrix operation, the values of the coefficients β ₀ to β _m of the magnetic angle type sensing element 21 can be found out, and the fitting to be used is determined to be valid after the replacement. Polynomial equation for line segment L2'. For example, in this embodiment, the polyfit command provided by METLAB can be used to find the optimal coefficients (parameters) of the above-mentioned one-variable m-th degree polynomial equation and the polynomial equation that best matches the permuted effective line segment L2'.

For example, use the polyfit command to substitute the 100 known curvature values (x ₁ ~x _n ) and the corresponding 100 distance values (y ₁ ~y _n ) into 1 yuan 1 ~1 yuan respectively Among the 8 polynomial equations of the 8th degree, the respective optimal coefficients of these 8 polynomial equations and their fitting results with the effective line segment L2' after replacement will be obtained, and the univariate 6th degree polynomial will be found from these 8 polynomial equations. When the equation fitting the effective line segment L2' after replacement has the best result (that is, the unary 6th-degree polynomial equation is closest and best represents the effective line segment L2' after replacement), the signal processing device 3 uses the best fitting result. The 6th degree polynomial equation of one variable, for example, is y=2.4431x ⁶ -8.2418x ⁵ +15.967x ⁴ -10.349x ³ +7.6091x ² +1.567x+1.0009, where y represents the magnet M and the magnetic angle sensing element A relative distance of 21 (that is, the above-mentioned distance value), and x represents the curvature value. Then, as shown in step S7 in FIG. 1 , the signal processing device 3 writes the unary sixth-order polynomial equation and the curvature value range into the processing unit 22 as a mathematical model of the characteristic curve of the magnetic angle sensing element 21 , that is, the calibration procedure is completed.

Therefore, when the magnetic linear position sensor 2 is actually applied to the cylinder 1 shown in FIG. 3 to detect the position of the piston (ie the magnet M) in the cylinder 1, the processing unit 22 receives the magnetic angle type sensing element When the analogous sine signal and the cosine signal are transmitted from 21, the processing unit 22 digitizes the sine signal and the cosine signal into a sine signal value and a cosine signal value, and subtracts the first sine signal value from the sine signal value. a preset value and the cosine signal value minus the second preset value, the sine signal value minus the first preset value and the cosine signal value minus the second preset value are processed The second derivative (atan2) operation of the arc tangent function (ARCTAN), namely atan2 (subtract the sine signal value of the first preset value/subtract the cosine signal value of the second preset value value) to obtain a curvature value; then, when the processing unit 22 judges that the curvature value is within the range of the curvature value, and substitutes the curvature value into the mathematical model to obtain the univariate sixth degree polynomial equation A distance value, which represents the relative distance between the magnetic angle-type sensing element 21 and the magnet M; then, the processing unit 22 can directly output the distance value for back-end application, or convert the distance value into phase The corresponding analog signal, such as analog voltage (such as a voltage value of 1~5V) or analog current (such as a current value of 4~20mA), is output to the back-end application.

To sum up, in the calibration procedure of the above embodiment, the sine signal values and the cosine signals obtained by the induction of the magnetic angle sensing element 21 are appropriately shifted and then the arc tangent function (ARCTAN) is performed on them. The second-order derivative (atan2) operation can increase the effective detection distance of the magnetic angle sensing element 21, thereby reducing the number of magnetic angle sensing elements 21 used, and making the detection distance of the magnetic linear position sensor 2 change. long-term effect and purpose; and, through the calibration procedure, the mathematical model in the processing unit 22 of the magnetic linear position sensor 2 is fitted to the characteristic curve of the magnetic angle type sensing element 21, and the magnetic angle type is determined. An effective detection distance of the sensing element 21 and a corresponding range of curvature values; thus, the processing unit 22 only needs to digitize the sine signal and the cosine signal transmitted from the magnetic angle sensing element 21, and then Subtract the sine signal value of the first preset value and subtract the cosine signal value of the second preset value to perform atan2 (subtract the sine signal value of the first preset value/subtract the second preset value the cosine signal value) operation, after obtaining a corresponding curvature value, the curvature By inputting the value into the mathematical model for calculating distance, a relative distance between the magnetic angle-type sensing element 21 and the magnet M can be quickly obtained through an easy-to-calculate one-dimensional m-th degree polynomial equation.

However, the above are only examples of the present invention, and should not limit the scope of the present invention. Any simple equivalent changes and modifications made according to the scope of the application for patent of the present invention and the content of the patent specification are still within the scope of the present invention. within the scope of the invention patent.

S1~S7: Steps

Claims (5)

A method for calibrating a magnetic linear position sensor, the magnetic linear position sensor is arranged on one side of a linear stroke to detect the position of a magnet that reciprocates along the linear stroke, the magnetic linear position sensor comprises a magnetic angle sensor Element and a processing unit; the calibration method includes: (A) making the magnet move along the linear stroke, and at the same time, a signal processing device makes the processing unit request the magnetic angle-type sensing element to return a currently generated one at a sampling frequency The sine signal and a cosine signal until the magnet completes the linear travel; (B) the processing unit digitizes the sine signal and the cosine signal continuously transmitted from the magnetic angle type induction element and outputs it to the signal processing device (C) The signal processing device obtains the correspondence between the position of each sampling point of the magnet in the straight line travel and the digitized sine signal values and the cosine signal values according to the length of the straight line travel and the sampling frequency relationship, and the signal processing device subtracts a first predetermined value from the sine signal values, and subtracts a second predetermined value from the cosine signal values; (D) the signal processing device subtracts the first predetermined value from the The sine signal values of a preset value and the cosine signal values minus the second preset value are subjected to the second derivative (atan2) operation of the arc tangent function, namely atan2 (subtracting the first preset value of the second derivative (atan2) sine signal value/subtracting the cosine signal value of the second preset value) to obtain a corresponding complex curvature value, and the curvature values constitute a curvature curve; (E) the signal processing device uses the curvature curve in the An effective line segment of the magnetic angle sensing element is used as a characteristic curve of the magnetic angle sensing element, and an effective detection distance and a corresponding curvature value range of the magnetic angle sensing element are determined according to the characteristic curve; (F) The signal processing The device determines a polynomial equation for fitting the characteristic curve through curve fitting and linear regression analysis according to the effective detection distance and the curvature value range

The m value and the value of β ₀ ~β _m , where i=1,2,3...n, y _i represents a relative distance between the magnetic angle sensing element and the magnet, xi represents the curvature value, β ₀ ~ _βm is the coefficient of the magnetic angle sensing element; and (G) the signal processing device writes the polynomial equation and the curvature value range into the processing unit as a characteristic curve of the magnetic angle sensing element mathematical model. The calibration method for a magnetic linear position sensor as claimed in claim 1, wherein the first preset value is one of the sum of the maximum value and the minimum value among the sinusoidal signal values; the second preset value is One of two of the sum of the maximum value and the minimum value of the cosine signal values. The calibration method of a magnetic linear position sensor as claimed in claim 1 or 2, wherein m is 6. The method for calibrating a magnetic linear position sensor according to claim 1 or 2, wherein the magnetic angle sensing element has two magnetoresistive bridges with a difference of 45°, wherein one magnetoresistive bridge induces the magnetic field of the magnet and generates For a sine wave signal, the other magnetoresistive bridge induces the magnetic field of the magnet and generates a cosine wave signal with a phase difference of 45° from the sine wave signal. The calibration method for a magnetic linear position sensor according to claim 1 or 2, wherein the curvature value range is -π~π.

Images