A rotor shaft speed measurement method using non-uniform reflective zebra strips

Abstract

The present invention provides a rotor shaft speed measurement method using non-uniform reflective zebra strips. The non-uniformly distributed reflective zebra strips are used as a measurement reference, and combined with the signal acquisition and processing method of the pulse timing method, a more accurate rotor shaft speed measurement is achieved. Compared with traditional methods, this method can effectively reduce the assembly error of the reflective zebra strips, especially the distortion of the zebra strips caused by factors such as rotor structural deformation, or the processing error of the zebra strip printing and drawing. The measurement method first requires the design and processing of non-uniform reflective zebra strips; then, based on the assembled zebra strips, a reflective fiber optic sensor is used to collect signals; secondly, the time difference sequence of the pulse signal is extracted based on the collected data to obtain a real-time rough speed measurement result; then, the rough speed measurement result is Fourier transform filtered to remove the interference frequency term, and finally, an accurate real-time measurement result of the rotor shaft speed is obtained.

Description

Technical Field

[0001] The present invention relates to the technical field of speed measurement of a rotor shaft system of a power machinery, and in particular to a method for measuring the speed of a rotor shaft system by using a non-uniform reflective zebra strip. Background Art

[0002] Today, mechanical rotors are widely used in industries such as shipping, vehicles, aviation, and power generation. With the continuous development of modern industry, various mechanical structures have become increasingly complex, posing increasingly daunting challenges to the normal operation of rotors. Rotor operation is often accompanied by various vibrations, which lead to rotor fatigue accumulation, shaft damage, and ultimately various mechanical failures and even disasters, causing immeasurable losses. To nip mechanical failures and disasters in the bud, it is necessary to install sensors on mechanical rotors to monitor and evaluate their operating status, preemptively addressing various faults and minimizing losses. Currently, a common monitoring method is to collect vibration signals during operation and perform time-frequency domain analysis on these signals to determine whether there is a mechanical fault. Accurately measuring the rotating equipment's speed is a crucial step in performing time-frequency analysis of vibration signals.

[0003] Current speed measurement methods are divided into two categories based on whether the sensor is in contact with the object being measured: contact measurement and non-contact measurement [Wang Siwen. Research on Torsional Vibration Measurement Technology Based on Photoelectric Pulse Sequence Method [D]. Dalian University of Technology, 2019]. Contact measurement typically uses a resistance strain gauge or accelerometer attached to the rotor surface to measure speed. Non-contact measurement mainly includes the laser Doppler method, which relies on the Doppler effect, and the pulse sequence method, which uses optical or electrical sensors. Non-contact measurement methods offer relatively simple and practical signal transmission systems, a relatively high measurement range and accuracy, and a broader application prospect than contact measurement methods. The pulse sequence method, a non-contact measurement method, is widely used due to its many advantages, including a wide measurement range, high measurement accuracy, and long equipment life. Optical sensors, compared to electrical sensors, are immune to electromagnetic interference and have a faster response speed, offering broader application prospects.

[0004] When measuring the rotational speed using the pulse timing method based on a reflective fiber optic sensor, a black and white zebra strip as a dividing device needs to be pasted on the shaft. The fiber optic sensor generates a pulse signal by receiving the reflected light, which is then converted into an electrical signal through a photoelectric converter. The signal is then converted into a digital signal through an acquisition card and transmitted to the host computer for signal processing. Since the tiny mass of the striped strip does not affect the vibration characteristics of the rotor, and the fiber optic sensor is not subject to electromagnetic interference, this method is simple to operate and easy to implement real-time monitoring. Therefore, this method has developed rapidly in recent years. However, the zebra strip as a dividing device is generally printed on paper, so printing errors, stretching effects, and division errors such as difficulty in precise connection at the interface will all lead to large rotational speed measurement errors. One way to reduce the error is to measure the angular velocity ω i A Fourier transform is performed to eliminate the peak frequency domain points of the speed multiplication error signal, and then an inverse Fourier transform is performed to obtain a more accurate speed measurement result. However, in traditional zebra strip designs, the width of each cycle is equal, resulting in a wide frequency band of the speed multiplication error signal. After eliminating the peak frequency domain points of the speed multiplication error signal, the speed multiplication error signal is not completely eliminated. As a result, although the accuracy of the speed result obtained after the inverse Fourier transform is improved, it still has a large error [Resor BR, Groover CL, Trethewey MW, et al. Natural frequency identification in torsional vibration with high-level order content [C] / / 22nd International Modal Analysis Conference, Dearborn, Michigan, USA. 2004.]. Therefore, a method is needed to completely remove the speed multiplication error signal caused by the indexing error, thereby improving the accuracy of the speed measurement. Summary of the Invention

[0005] The purpose of the present invention is to provide a method for measuring the rotational speed of a rotor shaft system using non-uniform reflective zebra strips.

[0006] The object of the present invention is achieved in that the steps are as follows:

[0007] Step 1: Design the width L of each cycle according to the size of the mechanical rotor to be tested. i The unequal black and white non-uniform reflective zebra strips are fixed on the mechanical rotor to be measured to ensure that the reflective fiber optic probe can be aligned with the reflective zebra strips during the rotation of the mechanical rotor.

[0008] Step 2: The reflective fiber optic sensor has an emitting hole at the center, and six receiving holes are distributed around the emitting hole. The light emitted by the light source is irradiated on the zebra strip through the emitting hole of the fiber optic sensor. The fiber optic sensor receives the reflected light from the zebra strip through the receiving hole, and then transmits it to the host computer through the photoelectric converter and the acquisition card. Since the white stripes have a high reflectivity, the reflected light is a high-level signal after photoelectric conversion, while the black stripes are a low-level signal. As the rotor rotates, the reflected light appears as a pulse signal similar to a square wave in the host computer. Set a suitable threshold voltage according to the high and low voltage values ​​of the pulse signal. Start counting when the rising edge of the i-th pulse breaks through the threshold voltage, and end counting when the rising edge of the i+1-th pulse breaks through the threshold voltage. The time interval between the i-th pulse and the i+1-th pulse is T i =n / F s , where F s is the sampling frequency of the acquisition card, and n is the number of points counted in one cycle.

[0009] Step 3: Based on the time interval sequence T of adjacent pulses in step 2 i , and combined with the non-uniform reflective zebra strip period width L designed in step 1 i Changes, according to ω i =L i / (T i R) Calculate the speed and obtain real-time speed rough measurement data ω i , where R is the radius of the axis of rotation.

[0010] Step 4: Roughly measure the real-time speed of the mechanical rotor i Perform fast Fourier transform to obtain the frequency domain result X of the speed change n , X n The peak value (X k 、X 2k 、X 3k …), k is the fundamental frequency of the rotation speed.

[0011] Step 5: further process the frequency domain result of the speed variation obtained in step 4, and find the peak frequency domain point (X k 、X 2k 、X 3k ...), replace the intensity amplitude of the error peak frequency domain point with the average of the intensity amplitudes of the two points next to the error peak frequency domain point, or replace the intensity amplitude of the error peak frequency domain point with the point with the smaller intensity amplitude next to the error peak frequency domain point, and obtain the speed change frequency domain result X after removing the error N .

[0012] Step 6: Convert the speed change frequency domain result X obtained in step 5 N Perform inverse Fourier transform to obtain the accurate result of the real-time speed of the mechanical rotor.

[0013] Furthermore, in step 1, the width L of each period is designed. i The unequal black and white non-uniform reflective zebra strips should meet the conditions of periodic gradual change, which can be linearly increased or decreased, or increased or decreased in a higher power, or in the shape of a trigonometric function, etc., so that it can be conveniently combined with the time interval sequence T i corresponding;

[0014] In step 3, the corresponding method for obtaining the rough speed measurement result is as follows: the optical fiber probe receives the reflected signal and converts it into an electrical signal through the photoelectric detector, which is collected in real time by the acquisition card and transmitted to the host computer. The time interval sequence T of the pulse signal generated by the non-uniform reflective zebra strip triggering is converted into an electrical signal by the acquisition card. i The period width L of the non-uniform zebra strip i Correspondingly, according to ω i =L i / (T i R) Calculate the speed and get the speed data ω i , where R is the radius of the axis of rotation;

[0015] Furthermore, in step 5, the peak frequency domain points of the speed multiplication error signal can be eliminated in two ways:

[0016] Method 1: Replace the intensity amplitude of the error peak frequency domain point with the average of the intensity amplitudes of the two points next to the error peak frequency domain point

[0017]

[0018] Method 2: Replace the intensity amplitude of the error peak frequency domain point with the smaller point of the intensity amplitude next to the error peak frequency domain point

[0019]

[0020] Compared with the prior art, the beneficial effect of the present invention is: the present invention provides a rotor shaft speed measurement method using non-uniform reflective zebra strips. Compared with traditional uniform reflective zebra strips, its significant advantage is that the speed frequency multiplication signal caused by the division error is easier to remove. Therefore, in the speed measurement, the influence of the division error can be effectively eliminated, thereby improving the accuracy of the speed measurement. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 Schematic diagram of the experimental setup;

[0022] Figure 2 Schematic diagram for obtaining the time interval between adjacent pulses

[0023] Figure 3 The speed diagram directly measured for the uniform zebra strip;

[0024] Figure 4 This is a rotation speed diagram directly measured by the non-uniform zebra strip in an embodiment of the present invention;

[0025] Figure 5 Spectrum diagram of the rotational speed measured for a uniform zebra strip;

[0026] Figure 6 The frequency spectrum of the rotation speed measured by the non-uniform zebra band in the example of the present invention;

[0027] Figure 7 This is the final speed diagram of the uniform zebra strip;

[0028] Figure 8 The final speed diagram of the non-uniform zebra strip in the example of the present invention;

[0029] Figure 9 This is a comparison chart of the relative error of speed measurement between uniform zebra strips and non-uniform zebra strips; DETAILED DESCRIPTION

[0030] The present invention will be described in further detail below with reference to the accompanying drawings and specific embodiments.

[0031] The present invention provides a rotor shaft speed measurement method using non-uniform reflective zebra strips. Non-uniform zebra strips with different period widths are used to replace traditional uniform zebra strips, so that the speed multiplication error signal caused by the indexing error can be completely removed, thereby reducing the influence of the indexing error in the speed measurement and effectively improving the accuracy of the speed measurement.

[0032] The embodiment of the present invention uses a period width L i Take the linearly increasing non-uniform reflective zebra strip as an example. Figure 1 As shown in the figure, a non-uniform zebra strip and a uniform zebra strip are attached to the shaft, and a fiber optic sensor is placed on each zebra strip, aligned with the strip. In the figure, ① is the shaft being measured, ② is the fiber optic probe, ③ is the ASE light source, ④ is the photoelectric converter, ⑤ is the acquisition card, ⑥ is the host computer, and ⑦ is the 1×6 beam combiner. Because the motor used in the experiment is an AC servo motor, the output speed value is very accurate. The shaft used is relatively short, so torsional vibration and bending vibration are not generated. Therefore, the value on the motor tachometer can be approximately equal to the actual shaft speed.

[0033] Adjust the speed value on the motor tachometer to 500rpm. When the shaft rotates, the reflected light received by the fiber optic sensor appears on the host computer as a pulse timing signal similar to a square wave. Set the appropriate threshold value according to the high and low voltage values ​​of the pulse signal. Figure 2 As shown, counting starts when the rising edge of the i-th pulse breaks through the threshold voltage, and ends when the rising edge of the i+1-th pulse breaks through the threshold voltage. The i-th pulse period is T i =n / Fs , where F s is the sampling frequency of the acquisition card, n is the number of points counted in one cycle, and ① in the figure is the time interval T between adjacent pulses i , ② is the threshold voltage, ③ is the time interval between adjacent sampling points of the acquisition card 1 / F s , ④ is the pulse signal.

[0034] Then by ω i =L i / (T i R) calculates the rough angular velocity of the axis rotation, Figure 3 and Figure 4 The speed diagrams are measured for uniform zebra strips and non-uniform zebra strips respectively. i Perform fast Fourier transform to get X n =fft(ω i ), where the peak value should be (X k 、X 2k 、X 3k …)k is the fundamental frequency of the rotation speed, Figure 5 and Figure 6 The frequency spectra obtained by fast Fourier transform of the speed measured for uniform zebra bands and non-uniform zebra bands show that Figure 6 The single frequency of the speed multiplication error signal is better than Figure 5 Much better. Use the peak-finding function to find the peak frequency domain points in the speed multiplication error signal. In this example, we will use the method described above to eliminate these peak frequency domain points and use the sum of the intensity amplitudes of the two points next to the peak frequency domain point to replace the intensity amplitude of the peak frequency domain point:

[0035]

[0036] Then perform inverse Fourier transform to get the final speed map. Figure 7 and Figure 8 The following are the final speed diagrams obtained using uniform zebra bands and non-uniform zebra bands. Compared with the uniform zebra band, the non-uniform zebra band eliminates the peak frequency domain points in the speed multiplication error signal, and the speed obtained is very close to 500rpm, which effectively reduces the speed measurement error caused by the zebra band division error.

[0037] In order to verify the universal applicability of the non-uniform zebra strip, the speed was changed and multiple experiments were carried out to obtain the comparison chart of the relative error of the speed measurement of the uniform zebra strip and the non-uniform zebra strip, as shown in the figure below. Figure 9 As shown in the figure, the non-uniform reflective zebra strips combined with the method of eliminating the peak frequency domain points in the speed multiplication error signal can effectively eliminate the influence of the division error and improve the accuracy of the speed measurement.

[0038] In summary, the present invention belongs to the technical field of speed measurement of rotor shaft systems of power machinery, and proposes a rotor shaft speed measurement method using non-uniform reflective zebra strips, which belongs to the field of optical measurement in non-contact measurement. Based on the existing reflective fiber optic sensor, the innovative use of non-uniformly distributed reflective zebra strips as the measurement reference, combined with the signal acquisition and processing method of the pulse timing method, can achieve more accurate rotor shaft speed measurement. Compared with traditional methods, this method can effectively reduce the assembly error of the reflective zebra strip, especially the distortion of the zebra strip caused by factors such as rotor structural deformation, or the processing error of the zebra strip printing and drawing. The measurement method first requires the design and processing of non-uniform reflective zebra strips; then, based on the assembled zebra strips, a reflective fiber optic sensor is used to collect signals; secondly, the time difference sequence of the pulse signal is extracted based on the collected data to obtain a real-time rough speed measurement result; then, the rough speed measurement result is Fourier transform filtered to remove the interference frequency term, and finally, an accurate real-time measurement result of the rotor shaft speed is obtained.

Claims

1. A method for measuring the rotor shaft speed using non-uniform reflective zebra strips, characterized in that: Here are the steps: Step 1: Design the width L of each cycle according to the size of the mechanical rotor to be tested i Unequal black and white non-uniform reflective zebra strips are fixed on the mechanical rotor to be tested to ensure that the reflective fiber optic sensor can align with the reflective zebra strips during the rotation of the mechanical rotor; Step 2: Use a reflective fiber optic sensor to receive the reflected light from the black and white zebra strips to generate a pulse signal. The pulse signal is transmitted to the host computer through a photoelectric converter and an acquisition card, and the time difference sequence T of adjacent pulses is obtained in the host computer. i ; Step 3: According to the time interval sequence T of adjacent pulses in step 2 i , and combined with the non-uniform reflective zebra strip period width L i Changes, get the real-time speed rough measurement result ω during the rotation of the mechanical rotor i ; Step 4: Roughly measure the real-time speed of the mechanical rotor ω i Perform fast Fourier transform to obtain the frequency domain result X of the speed change n , X n The peak frequency domain point of the speed multiplication error signal (X k 、X 2k 、X 3k …), k is the fundamental frequency of the rotation speed; Step 5: Further process the frequency domain results of the speed variation obtained in step 4, and find the peak frequency domain point (X k 、X 2k 、X 3k ...), replace the intensity amplitude of the error peak frequency domain point with the average of the intensity amplitudes of the two points next to the error peak frequency domain point, or replace the intensity amplitude of the error peak frequency domain point with the point with the smaller intensity amplitude next to the error peak frequency domain point, and obtain the speed change frequency domain result X after removing the error N ; Step 6: Convert the speed change frequency domain result X obtained in step 5 N Perform inverse Fourier transform to obtain the accurate result of the real-time speed of the mechanical rotor.

2. The method for measuring rotor shaft speed using non-uniform reflective zebra strips according to claim 1, characterized in that: In the step 1, the width L of each period is designed. i The unequal black and white non-uniform reflective zebra strips should meet the conditions of periodic gradient, increasing or decreasing linearly or increasing or decreasing with a higher power or in the shape of a trigonometric function, and the time interval sequence T i Corresponding.

3. The method for measuring rotor shaft speed using non-uniform reflective zebra strips according to claim 1, characterized in that: In step 3, the corresponding method for obtaining the rough speed measurement result is as follows: the optical fiber probe receives the reflected signal and converts it into an electrical signal through the photoelectric detector, which is collected in real time by the acquisition card and transmitted to the host computer. The time interval sequence T of the pulse signal generated by the non-uniform reflective zebra strip triggering is converted into an electrical signal by the acquisition card. i The period width L of the non-uniform zebra strip i Correspondingly, according to ω i =L i / (T i R) Calculate the speed and get the speed data ω i , where R is the radius of the axis of rotation.

4. The method for measuring rotor shaft speed using non-uniform reflective zebra strips according to claim 1, characterized in that: In step 5, there are two ways to eliminate the peak frequency domain points of the speed multiplication error signal: Method 1: Replace the intensity amplitude of the error peak frequency domain point with the average of the intensity amplitudes of the two points next to the error peak frequency domain point: Or method 2: Use the smaller point of the intensity amplitude of the two points next to the error peak frequency domain point to replace the intensity amplitude of the error peak frequency domain point:

Images