Ceramic piston rod absolute stroke measurement self-learning method

Abstract

The invention provides a ceramic piston rod absolute stroke measurement self-learning method, and the method comprises the steps: reading a code and a signal of a piston rod through a sensor array, and forming a cycle direction period through a voltage wave crest and a wave trough of the signal; according to the period number of the sensor array, the absolute stroke information of the piston rod is obtained through a self-learning method in combination with the direction area array in the current circulation direction period; the sensor array comprises a first single-row sensor and a second single-row sensor which are arranged in a front-back staggered mode, the sensor array and the piston rod are arranged in parallel, grooves are formed in the piston rod, each groove corresponds to a binary number, and a plurality of binary numbers form a code. According to the method, errors caused by environmental factors or electronic elements are corrected through a self-learning method, so that the absolute stroke value of the piston rod is accurately obtained, and the long-term stability and environmental adaptability of absolute stroke measurement of the piston rod are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of piston rod stroke measurement, and particularly to a self-learning method for absolute stroke measurement of a ceramic piston rod. Background Art

[0002] In the fields of industrial automation, aerospace, heavy machinery, and new energy, the piston rod, as the core moving part of actuators such as hydraulic cylinders and air cylinders, its stroke directly determines the output displacement, movement speed, and control accuracy of the actuator. High-precision stroke measurement is not only the basis for realizing closed-loop control but also the key requirement for ensuring the safe operation of equipment and extending its service life. Traditional piston rods are mostly made of metal materials, but they are prone to deformation or failure under extreme working conditions such as high temperature, strong corrosion, and high wear. In recent years, due to its excellent properties such as high hardness, high temperature resistance, corrosion resistance, and low thermal expansion coefficient, ceramic materials have gradually become an ideal choice for piston rods in special environments. However, the non-conductivity and complex surface characteristics of ceramic materials pose new challenges to the design of stroke measurement sensors and signal acquisition.

[0003] With the continuous development of electronic technology, the stroke of the piston rod is measured by using the changes in the electrical signals of components such as resistors, inductors, and capacitors. This method further improves the measurement accuracy and has strong anti-interference ability. However, since it works based on the electromagnetic field principle, it will inevitably be affected by the external electromagnetic field, thus interfering with the measurement accuracy. Further, optoelectronic technology has been applied in stroke measurement, such as using optical devices such as grating rulers or laser interferometers for measurement. Such methods can achieve high-precision measurements at the micron or even nanometer level and have good measurement stability, but they have high requirements for the use environment and the equipment price is relatively expensive, which limits their wide application to a certain extent. In recent years, magnetic encoders have gradually been applied to piston rod stroke measurement, such as Hall effect encoders, magnetoresistive encoders, etc., which use the changes in magnetic signals to determine the position of the piston rod. Such methods have no wear during the measurement process, have a long service life, and can achieve high-precision measurements, but are easily interfered by magnetic fields.

[0004] Chinese Patent CN101280796B Absolute Stroke Detection Method and Device for Use with Ceramic Piston Rods, which encodes by setting area markers and area number markers on the detection section of the piston rod. The area number markers are encoded in binary order and are achieved through a specific arrangement of grooves and ridges. However, when using multiple multi-pack sensors, the adaptability to cylinders with relatively small piston rod diameters is relatively poor. Chinese Patent CN112431820A Absolute Position Displacement Measurement System and Method for Ceramic Piston Rod with Grooves discloses that upper and lower groove groups are provided on the piston rod, and the structures of the upper and lower groove groups are different. The upper and lower magnetic Hall sensors respectively measure the upper and lower groove groups to obtain the absolute position of the piston rod. However, two sets of grooves with different structures need to be machined on the piston rod, namely the equally spaced first grooves of the upper groove group and the non-equally spaced grooved sections and non-grooved sections of the lower groove group. The overall process is relatively complex, and the problem of reading errors is not considered. Summary of the Invention

[0005] The main object of the present invention is to provide a self-learning method for absolute stroke measurement of a ceramic piston rod, which solves the problem of incorrect piston rod stroke measurement caused by deviations due to inaccurate detection caused by environmental interference or electronic component failures.

[0006] To solve the above technical problems, the technical solution adopted by the present invention is: a self-learning method for absolute stroke measurement of a ceramic piston rod, which reads the encoding and signals of the piston rod through a sensor array, and forms a cycle direction period from the voltage peaks and valleys of the signals; according to the number of cycles of the sensor array, combined with the direction area array in the current cycle direction period, the absolute stroke information of the piston rod is obtained through a self-learning method; The sensor array includes a first single-row sensor and a second single-row sensor arranged in a staggered front and back manner. The sensor array is arranged parallel to the piston rod. Grooves are provided on the piston rod, and each groove corresponds to a binary number. Multiple binary numbers form an encoding.

[0007] In a preferred embodiment, the absolute stroke information of the piston rod is S = X×L + Y×(L / 4) + Z; where S is the absolute stroke value of the piston rod, i.e., the cylinder stroke value; X is the number of cycles. When the sensor array is in the i-th cycle, X = i - 1; Y is the pulse count corresponding to the direction area array in the current cycle direction period. When the direction area array is XX, Y = 1; when the direction area array is 0X, Y = 2; when the direction area array is 00, Y = 3; when the direction area array is X0, Y = 4; L is the length of a cycle direction period; Z is the deviation value obtained by the self-learning method; The self-learning method includes combining the current stroke information and historical stroke information of the piston rod, and automatically correcting using a lightweight learning method to obtain the deviation value Z.

[0008] In the preferred solution, four consecutively acquired codes form a cycle direction period. Within each cycle direction period, based on the direction area array of the first single-row sensor, the direction area array is used to obtain the moving direction of the piston rod and pulse counting; whenever the pulse signal of the piston rod changes from a high potential to a low potential, or from a low potential to a high potential once, the pulse counting is incremented by 1. The pulse signal includes low potentials and high potentials with equal widths and equal spacings. When the voltages of the first single-row sensor and the second single-row sensor are in the same state, that is, both are at the peak value or both are at the valley value, a low potential is generated; when the voltages of the first single-row sensor and the second single-row sensor are in different states, that is, one is at the peak value and the other is at the valley value, a high potential is generated.

[0009] In the preferred solution, the sensor array outputs signals according to the change in magnetic field strength, and respectively obtains the voltage signals of the first single-row sensor and the second single-row sensor. The voltage signals include the voltage valley value in the weak magnetic field region and the voltage peak value in the strong magnetic field region; according to the positional relationship between the voltage valley value and the peak value, a cycle direction period is formed by the wave peaks and wave valleys, and a cycle direction period includes four direction area arrays.

[0010] In the preferred solution, each direction area array is composed of the voltage values obtained by the first single-row sensor and the second single-row sensor at the same position. Denote the voltage valley value in the weak magnetic field region as X and the voltage peak value in the strong magnetic field region as 0. Then the four direction area arrays included in the cycle direction period are respectively denoted as XX, X0, 00, and 0X.

[0011] In the preferred solution, according to the cycle direction period sorting method, the moving direction of the piston rod is obtained; if the cycle direction period sorting of the first single-row sensor and the second single-row sensor is XX, 0X, 00, and X0, then the piston rod moves from left to right; if the cycle direction period sorting of the first single-row sensor and the second single-row sensor is XX, X0, 00, and 0X, then the piston rod moves from right to left.

[0012] In the preferred solution, the strong magnetic field region corresponds to the groove region of the piston rod, and the weak magnetic field region corresponds to the rack region between adjacent grooves.

[0013] In the preferred solution, the first single-row sensor and the second single-row sensor respectively include a plurality of Hall sensors with the same quantity, and each Hall sensor is used to obtain the binary number corresponding to one groove; The code includes a column of code values obtained by the first single-row sensor and two columns of code values obtained by the second single-row sensor. The column of code values and the two columns of code values are respectively composed of a plurality of binary numbers.

[0014] In the preferred solution, the grooves are distributed on the piston rod at the same interval distance. The grooves include shallow grooves and deep grooves. The depth of the shallow groove is less than the depth of the deep groove, and the shallow groove and the deep groove respectively correspond to different binary numbers.

[0015] In the preferred solution, the width of the groove is 2 mm, the depth of the shallow groove is 1 mm, and the depth of the deep groove is 2 mm; the first single-row sensor and the second single-row sensor are arranged with a 1-mm offset front and back, and the second single-row sensor is located in the right-rear direction of the first single-row sensor.

[0016] The present invention provides a self-learning method for measuring the absolute stroke of a ceramic piston rod. This method reads the code and signal of the piston rod through a sensor array, and forms a cycle direction period from the voltage peaks and valleys of the signal; according to the number of periods of the sensor array, combined with the direction area array in the current cycle direction period, the absolute stroke information of the piston rod is obtained through the self-learning method; the sensor array includes a first single-row sensor and a second single-row sensor arranged with an offset front and back, the sensor array is arranged parallel to the piston rod, grooves are provided on the piston rod, each groove corresponds to a binary number, and multiple binary numbers form a code. This method, through the collaborative design of binary coding, sensor array and self-learning method, is applicable to oil cylinders with a relatively small diameter of the piston rod, processes the signal by combining pulse difference, and uses the self-learning method to correct the errors caused by environmental factors or electronic components, so as to accurately obtain the absolute stroke value of the piston rod and improve the long-term stability and environmental adaptability of the absolute stroke measurement of the piston rod. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] The present invention will be further described below in conjunction with the drawings and embodiments: Figure 1 is the flowchart of the method of the present invention; Figure 2 is the schematic diagram of the sensor array of the present invention; Figure 3 is the schematic diagram of the first single-row sensor reading when the piston rod of the present invention moves; Figure 4 is the schematic diagram of the groove in Embodiment 3 of the present invention; Figure 5 is the schematic diagram of the output signal of the sensor array in Embodiment 3 of the present invention; In the figure: sensor array 1; first single-row sensor 2; second single-row sensor 3; Hall sensor 4; groove 5; shallow groove 6; deep groove 7; rack 8. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0018] Embodiment 1 As Figures 1 to 5 shown, a self-learning method for measuring the absolute stroke of a ceramic piston rod. This method reads the code and signal of the piston rod through the sensor array 1, and forms a cycle direction period from the voltage peaks and valleys of the signal; according to the number of periods of the sensor array 1, combined with the direction area array in the current cycle direction period, the absolute stroke information of the piston rod is obtained through the self-learning method; The sensor array 1 includes a first single-row sensor 2 and a second single-row sensor 3 arranged in a front-and-back offset manner. The sensor array 1 is arranged parallel to the piston rod. A groove 5 is provided on the piston rod, and each groove 5 corresponds to a binary number. Multiple binary numbers form a code.

[0019] With this solution, in this embodiment, the piston rod is a ceramic piston rod, and a magnet or magnetic material is provided on the piston rod. Using laser etching or machining, grooves 5 with equal width and equal spacing are formed on the surface of the piston rod, and each groove 5 corresponds to a binary number. The sensor array 1 is arranged on the oil cylinder near the piston rod. The binary numbers of the continuous grooves 5 are obtained through the sensor array 1 to form a code. The first single-row sensor 2 is located below the second single-row sensor 3 and is arranged in a front-and-back offset manner along the axial direction of the piston rod, covering different phases of the same groove 5 to realize phase difference measurement. The first single-row sensor 2 and the second single-row sensor 3 obtain the voltage values of the output signals, obtaining a complete cycle direction period composed of wave peaks and wave valleys. According to the cycle direction period sorting method, the moving direction of the piston rod can also be obtained. Preferably, the offset distance between the first single-row sensor 2 and the second single-row sensor 3 is half of the width of the groove 5, eliminating the error codes caused by magnetic field interference or installation errors of the single-row sensor. The code is obtained by the sensor array 1, the number of periods and the direction area array within the current cycle direction period are determined for obtaining pulse counting, and the absolute stroke value of the piston rod is obtained according to the self-learning method. This method through the collaborative design of binary coding, the sensor array 1 and the self-learning method makes it applicable to oil cylinders with a smaller piston rod diameter, processes the signal by combining pulse difference, and corrects the errors caused by environmental factors or electronic components using the self-learning method, so as to accurately obtain the absolute stroke value of the piston rod and improve the long-term stability and environmental adaptability of the absolute stroke measurement of the piston rod.

[0020] Embodiment 2 Further described in combination with Embodiment 1, as Figures 1 to 5 shown, in the preferred solution, the absolute stroke information of the piston rod is S = X×L + Y×(L / 4) + Z; where S is the absolute stroke value of the piston rod, that is, the oil cylinder stroke value; X is the number of periods. When the sensor array 1 is in the i-th period, then X = i - 1; Y is the pulse counting corresponding to the direction area array within the current cycle direction period. When the direction area array is XX, Y = 1; when the direction area array is 0X, Y = 2; when the direction area array is 00, Y = 3; when the direction area array is X0, Y = 4; L is the length of a cycle direction period; Z is the deviation value obtained by the self-learning method; The self-learning method includes combining the current stroke information and historical stroke information of the piston rod and automatically correcting using a lightweight learning method to obtain the deviation value Z.

[0021] In a preferred embodiment, four consecutively acquired codes form a cycle direction period. Within each cycle direction period, based on the direction area array of the first single-row sensor 2, the direction area array is used to obtain the moving direction of the piston rod and pulse counting; whenever the pulse signal of the piston rod undergoes a high-to-low potential jump or a low-to-high potential jump, the pulse count is incremented by 1. The pulse signal includes low and high potentials with equal width and equal spacing. When the voltages of the first single-row sensor 2 and the second single-row sensor 3 are in the same state, that is, both are at the peak or both are at the valley, a low potential is generated; when the voltages of the first single-row sensor 2 and the second single-row sensor 3 are in different states, that is, one is at the peak and the other is at the valley, a high potential is generated.

[0022] With this solution, the voltage signals of the first single-row sensor 2 and the second single-row sensor 3 are differentially processed to obtain a pulse signal; whenever the pulse signal of the piston rod undergoes a high-to-low potential jump or a low-to-high potential jump, the pulse count is cumulatively incremented by 1; when the direction area array within the cycle direction period is XX, the corresponding pulse count Y = 1; when the direction area array within the cycle direction period is 0X, the corresponding pulse count Y = 2; when the direction area array within the cycle direction period is 00, the corresponding pulse count Y = 3; when the direction area array within the cycle direction period is X0, the corresponding pulse count Y = 4. Whenever a column of code values read by the first single-row sensor 2 changes, a new cycle direction period is entered, and four consecutively acquired codes form a cycle direction period; according to the column of code values read by the first single-row sensor 2, it is used to obtain the period in which the first single-row sensor 2 is located, that is, to determine the period in which the sensor array 1 is located. Preferably, the length L of the cycle direction period is 4 mm.

[0023] Preferably, the method for obtaining the deviation value Z using a self-learning method includes moving the piston rod to the starting position, ensuring the accuracy of the position through mechanical limit or external tools, and measuring the actual stroke value using other high-precision measuring devices; reading the current stroke information using the sensor array 1 of this method, calculating the theoretical stroke value without including the deviation value Z, and preliminarily calculating the estimated difference between the actual stroke value and the theoretical stroke value; moving the piston rod to other preset calibration points, calculating multiple obtained estimated differences, and establishing a deviation model using a lightweight learning method; by combining historical stroke information, including historical stroke information, deviation values, and model parameters, updating the deviation model to obtain the deviation value Z. Preferably, the lightweight learning method adopts one or any combination of incremental linear regression, Kalman filtering, or online clustering. Incremental linear regression is used to obtain the deviation value linearly related to time or cycle; Kalman filtering is used to obtain the deviation value caused by pulse loss resulting in abrupt errors; online clustering is used to obtain the deviation value caused by multi-condition switching in different load or temperature ranges. The deviation value Z is obtained through the self-learning method to correct the absolute stroke information of the piston rod, improving the long-term stability and environmental adaptability of the absolute stroke measurement of the piston rod.

[0024] Embodiment 3 Combined with Embodiments 1 and 2 for further illustration, as Figures 2 to 5 shown, in the preferred solution, the sensor array 1 outputs signals according to the change in magnetic field strength, respectively obtaining the voltage signals of the first single-row sensor 2 and the second single-row sensor 3. The voltage signals include the voltage valley value in the weak magnetic field region and the voltage peak value in the strong magnetic field region; according to the positional relationship between the voltage valley value and the peak value, a cycle direction period is formed by the wave peak and the wave valley. One cycle direction period includes four direction region arrays.

[0025] With this solution, when the groove 5 on the piston rod passes through the sensor array 1, the voltage signal output by the sensor array 1 changes with the change in magnetic field strength. When the sensor array 1 is in the weak magnetic field region, the magnetic field strength is low, and the output voltage is low, forming a voltage valley value; when the sensor array 1 is in the strong magnetic field region, the magnetic field strength is high, and the output voltage is high, forming a voltage peak valley value. The signal processing unit filters, amplifies, and shapes the signal output by the sensor array 1 to clearly reflect the wave peak and the wave valley. The voltage values of the first single-row sensor 2 and the second single-row sensor 3 at different positions are collected, and the corresponding relationship between the voltage valley value and the peak value and the position of the piston rod is established. Four consecutive positions correspond to four direction region arrays, and the four direction region arrays change cyclically, forming a cycle direction period.

[0026] In the preferred solution, as Figures 2 to 5As shown, each array of direction regions is composed of the voltage values obtained by the first single-row sensor 2 and the second single-row sensor 3 at the same position. Denote the voltage valley value in the weak magnetic field region as X, and the voltage peak value in the strong magnetic field region as 0. Then the four arrays of direction regions included in the cyclic direction period are respectively denoted as XX, X0, 00, and 0X.

[0027] According to this solution, when in the strong magnetic field region, when the read voltage peak value is about 3V or 5V, it is all denoted as 0; when in the weak magnetic field region, when the read voltage valley value is about 1V, it is denoted as X. The cyclic direction period includes that both the first single-row sensor 2 and the second single-row sensor 3 are at the voltage valley value X at the same position, and the array of direction regions is denoted as XX; both the first single-row sensor 2 and the second single-row sensor 3 are at the voltage peak value 0 at the same position, and the array of direction regions is denoted as 00; at the same position, the first single-row sensor 2 is at the voltage peak value 0, and the second single-row sensor 3 is at the voltage valley value X, and the array of direction regions is denoted as 0X; and at the same position, the first single-row sensor 2 is at the voltage valley value X, and the second single-row sensor 3 is at the voltage peak value 0, and the array of direction regions is denoted as X0. Since the first single-row sensor 2 and the second single-row sensor 3 are arranged with a dislocation front and back, there is a phase difference in the voltage values obtained at the same position. By the sorting method of the four arrays of direction regions within the cyclic direction period, the moving direction of the piston rod is judged. Preferably, the distances between the four arrays of direction regions are the same and depend on the dislocation distance between the first single-row sensor 2 and the second single-row sensor 3.

[0028] In the preferred solution, as Figure 4 and 5 shown, according to the sorting method of the cyclic direction period, the moving direction of the piston rod is obtained; if the cyclic direction period sorting of the first single-row sensor 2 and the second single-row sensor 3 is XX, 0X, 00, and X0, then the piston rod moves from left to right; if the cyclic direction period sorting of the first single-row sensor 2 and the second single-row sensor 3 is XX, X0, 00, and 0X, then the piston rod moves from right to left.

[0029] According to this solution, if the binary number corresponding to the groove 5 is 0011111010, the peak voltage in the region with a stronger magnetic field of the shallow groove 6 is about 3V, denoted as 0, the peak voltage in the region with a stronger magnetic field of the deep groove 7 is about 5V, denoted as 1, and the valley voltage in the region with the weakest magnetic field of the rack 8 is about 1V, denoted as X; for the convenience of calculation and program writing, both 0 and 1 are regarded as 0, that is, for the peak voltage of about 3V or 5V, they are both denoted as 0. Then, the direction region array and position relationship of the above binary number 0011111010 are shown in Table 1; according to the cyclic direction period sorting, the moving direction of the piston rod can be judged. If the cyclic direction period sorting of the obtained direction region array is XX, 0X, 00, and X0, the current piston rod moves to the right; if the cyclic direction period sorting of the obtained direction region array is XX, X0, 00, and 0X, the current piston rod moves to the left.

[0030] Table 1 Position Relationship Table

[0031] In the preferred solution, as Figure 4 and 5 shown, the region with a strong magnetic field corresponds to the groove 5 region of the piston rod, and the region with a weak magnetic field corresponds to the rack 8 region between adjacent grooves 5. According to this solution, the groove 5 region of the piston rod is concave, and the rack 8 region of the piston rod is convex, so that the magnetic field intensities of the groove 5 and the rack 8 are different. By measuring the change of the magnetic field intensity with a sensor and converting it into a corresponding signal output. The magnetic field intensity increases in the groove 5 region to form a region with a strong magnetic field for obtaining the peak voltage; the magnetic field intensity is weak in the rack 8 region to form a region with a weak magnetic field for obtaining the valley voltage. The length of one groove 5 and one rack 8 forms the length of a cyclic direction period. The peak voltage of the region with a strong magnetic field corresponds to the bottom position of the groove 5, and the valley voltage of the region with a weak magnetic field corresponds to the top position of the rack 8.

[0032] Example 4 Combined with Examples 1 to 3 for further illustration, as Figure 2 and 3 shown, in the preferred solution, the first single-row sensor 2 and the second single-row sensor 3 each include a plurality of Hall sensors 4 with the same quantity. Each Hall sensor 4 is used to obtain a binary number corresponding to one groove 5; the encoding includes a column of code values obtained by the first single-row sensor 2 and two columns of code values obtained by the second single-row sensor 3. The column of code values and the two columns of code values are each composed of a plurality of binary numbers.

[0033] With this solution, the sensor array 1 includes a plurality of Hall sensors 4. The Hall sensors 4 are small in size and light in weight, enabling a compact measurement system. Only two rows of Hall sensors 4 are provided, making it suitable for oil cylinders with a smaller piston rod diameter. There is a gap between the sensor array 1 and the piston rod. The first single-row sensor 2 and the second single-row sensor 3 are arranged axially misaligned along the piston rod, covering the same groove 5 but with a phase shift, for forming a phase difference to identify the moving direction of the piston rod. A column code value and a two-column code value are composed of a plurality of consecutive binary numbers. A column code value and a two-column code value form a group of codes. The column code value and the two-column code value are respectively composed of 8-bit, 12-bit or 16-bit binary numbers.

[0034] In a preferred solution, as Figures 2 to 4 shown, the grooves 5 are distributed on the piston rod at the same interval distance. The groove 5 includes a shallow groove 6 and a deep groove 7. The depth of the shallow groove 6 is less than that of the deep groove 7. The shallow groove 6 and the deep groove 7 respectively correspond to different binary numbers. With this solution, due to the greater depth of the deep groove 7, the change in magnetic field intensity generated is more significant, and an obvious voltage difference is formed when detected by the Hall sensor 4. The design of the shallow groove 6 and the deep groove 7 is used to enhance the magnetic field contrast and reduce the risk of signal confusion. The Hall sensor 4 outputs the corresponding binary number by detecting the magnetic field intensity of the groove 5. Preferably, a single shallow groove 6 is the binary number 0, and a single deep groove 7 is the binary number 1.

[0035] In a preferred solution, as Figures 2 to 4 shown, the width of the groove 5 is 2 mm, the depth of the shallow groove 6 is 1 mm, and the depth of the deep groove 7 is 2 mm. The first single-row sensor 2 and the second single-row sensor 3 are arranged with a 1-mm front-back misalignment. The second single-row sensor 3 is located in the right-rear direction of the first single-row sensor 2. With this solution, the width of the rack 8 is also 2 mm. The code composed of binary numbers and the grooves 5 with equal width and equal spacing are used to simplify the decoding logic. Each Hall sensor 4 corresponds to a groove 5 with a width of 2 mm. The first single-row sensor 2 and the second single-row sensor 3 are misaligned 1 mm front and back, and decoding with an accuracy of 1 mm can be achieved.

[0036] The above embodiments are only the preferred technical solutions of the present invention and should not be regarded as limitations on the present invention. The protection scope of the present invention should be the technical solutions recorded in the claims, including equivalent replacement solutions of the technical features in the technical solutions recorded in the claims. That is, equivalent replacement improvements within this scope are also within the protection scope of the present invention.

Claims

1. A self-learning method for measuring the absolute stroke of a ceramic piston rod, characterized in that: This method reads the encoding and signals of the piston rod through the sensor array (1), and forms a cycle direction period from the voltage peaks and valleys of the signals; according to the number of periods of the sensor array (1), combined with the direction area array within the current cycle direction period, the absolute stroke information of the piston rod is obtained through a self-learning method; The sensor array (1) includes a first single-row sensor (2) and a second single-row sensor (3) arranged in a front-back staggered manner. The sensor array (1) is arranged parallel to the piston rod. A groove (5) is provided on the piston rod, and each groove (5) corresponds to a binary number, and multiple binary numbers form an encoding.

2. The self-learning method for measuring the absolute stroke of the ceramic piston rod according to claim 1, characterized in that: The absolute stroke information of the piston rod is S = X×L + Y×(L / 4) + Z; where S is the absolute stroke value of the piston rod, that is, the cylinder stroke value; X is the number of periods. When the sensor array (1) is in the i-th period, then X = i - 1; Y is the pulse count corresponding to the direction area array within the current cycle direction period. When the direction area array is XX, Y = 1; when the direction area array is 0X, Y = 2; when the direction area array is 00, Y = 3; when the direction area array is X0, Y = 4; L is the length of a cycle direction period; Z is the deviation value obtained by the self-learning method; The self-learning method includes combining the current stroke information and historical stroke information of the piston rod, and automatically correcting it using a lightweight learning method to obtain the deviation value Z.

3. The ceramic piston rod absolute stroke measurement self-learning method according to claim 2, characterized in that: Four continuously obtained encodings form a cycle direction period. Within each cycle direction period, based on the direction area array of the first single-row sensor (2), the direction area array is used to obtain the moving direction and pulse count of the piston rod; whenever the pulse signal of the piston rod has a high-to-low potential jump or a low-to-high potential jump, the pulse count is incremented by 1; The pulse signal includes low potential and high potential with equal width and equal spacing. When the voltages of the first single-row sensor (2) and the second single-row sensor (3) are in the same state, that is, both are peaks or both are valleys, a low potential is generated; when the voltages of the first single-row sensor (2) and the second single-row sensor (3) are in different states, that is, one is a peak and the other is a valley, a high potential is generated.

4. The absolute stroke measurement self-learning method for a ceramic piston rod according to any one of claims 1 to 3, characterized in that: The sensor array (1) outputs signals according to the change of magnetic field strength, and obtains the voltage signals of the first single-row sensor (2) and the second single-row sensor (3) respectively. The voltage signals include the voltage valley value in the weak magnetic field area and the voltage peak value in the strong magnetic field area; according to the position relationship between the voltage valley value and the peak value, a cycle direction period is formed by the peaks and valleys. A cycle direction period includes four direction area arrays.

5. The absolute stroke measurement self-learning method of the ceramic piston rod according to claim 4, characterized in that: Each direction area array is composed of the voltage values obtained by the first single-row sensor (2) and the second single-row sensor (3) at the same position. The voltage valley value in the weak magnetic field area is denoted as X, and the voltage peak value in the strong magnetic field area is denoted as 0. Then the four direction area arrays included in the cycle direction period are denoted as XX, X0, 00, and 0X respectively.

6. The absolute stroke measurement self-learning method for a ceramic piston rod according to claim 5, characterized in that: Obtain the moving direction of the piston rod according to the cyclic direction periodic sorting method; if the cyclic direction periodic sorting of the first single-row sensor (2) and the second single-row sensor (3) is XX, 0X, 00, and X0, the piston rod moves from left to right; if the cyclic direction periodic sorting of the first single-row sensor (2) and the second single-row sensor (3) is XX, X0, 00, and 0X, the piston rod moves from right to left.

7. The absolute stroke measurement self-learning method for the ceramic piston rod according to claim 4, characterized in that: The strong magnetic field region corresponds to the groove (5) region of the piston rod, and the weak magnetic field region corresponds to the rack (8) region between adjacent grooves (5).

8. The self-learning method for measuring the absolute stroke of the ceramic piston rod according to claim 1, characterized in that: The first single-row sensor (2) and the second single-row sensor (3) each include a plurality of Hall sensors (4) with the same quantity, and each Hall sensor (4) is used to obtain the binary number corresponding to one groove (5); The encoding includes a column of code values obtained by the first single-row sensor (2) and two columns of code values obtained by the second single-row sensor (3), and the column of code values and the two columns of code values are each composed of a plurality of binary numbers.

9. The self-learning method for measuring the absolute stroke of a ceramic piston rod according to claim 8, characterized in that: The grooves (5) are distributed on the piston rod at the same interval distance. The grooves (5) include shallow grooves (6) and deep grooves (7). The depth of the shallow grooves (6) is less than the depth of the deep grooves (7), and the shallow grooves (6) and the deep grooves (7) respectively correspond to different binary numbers.

10. The method for self-learning the absolute stroke measurement of the ceramic piston rod according to claim 9, characterized in that: The width of the groove (5) is 2 mm, the depth of the shallow groove (6) is 1 mm, and the depth of the deep groove (7) is 2 mm; the first single-row sensor (2) and the second single-row sensor (3) are arranged with a 1-mm offset front and back, and the second single-row sensor (3) is located in the right-rear direction of the first single-row sensor (2).

Images