An Adaptive Extraction Method for the Initial Phase Point of the Axis Orbit of a Rotating Machine

Abstract

An adaptive method for extracting the initial phase point of the shaft center orbit of a rotating machine. First, the vibration displacement signal and the coaxial key phase signal of the measurement section are obtained through discrete sampling. The maximum value, minimum value, and mean value of the key phase signal are calculated, and the pulse type is automatically judged using the distance relationship among the three. Then, a first-order backward difference operation is performed on the key phase signal to construct a new sequence, and all local maximum points are searched. The pulse detection threshold and the pulse trigger position are adaptively determined according to the characteristics of the key phase signal itself. Finally, using the vibration waveform of the measurement section and the pulse trigger position index, the coordinates of the initial phase point of the shaft center orbit are obtained, and the corresponding shaft center orbit diagram is drawn and the position of the initial phase point is marked. The present invention realizes the adaptive extraction of the initial phase point of the shaft center orbit under different pulse trigger directions, pulse amplitudes, and pulse widths, provides key information for the quantitative evaluation of the stability and complexity of the shaft center orbit, and helps to achieve the accurate diagnosis of rotating equipment faults.

Description

Technical Field

[0001] The present invention belongs to the technical field of mechanical equipment fault diagnosis, and particularly relates to a method for extracting the initial phase point of the shaft center orbit of a rotating machine. Background Technique

[0002] The shaft center orbit diagram can describe the planar motion orbit of the rotor center relative to the bearing housing, and can intuitively and clearly reflect the dynamic behavior of the rotor. Its shape characteristics are closely related to the operating state of the unit. Many researchers at home and abroad have applied the shaft center orbit characteristics to the detection and fault diagnosis of the operating state of rotating machinery to improve the accuracy of equipment fault diagnosis. According to the information provided in the literature (Wen Guangrui, Yin Jian'an, Zhang Xining, etc. Identification of the rotating frequency fault before balancing using the initial phase point of the holographic spectrum [J]. Journal of Vibration and Shock, 2008, 27(S): 180-183), the definition of the initial phase point of the shaft center orbit can be obtained. It refers to the position of the rotor center on the shaft center orbit when the key phase sensor is aligned with the keyway (or reflective tape) position. As an important characteristic parameter of the rotor shaft center orbit, the initial phase point is often used for quantitative analysis and evaluation of the repeatability and stability of the orbit, and is an important index for accurately distinguishing equipment faults. For example, oil film oscillation and stator-rotor rubbing are two common factors that induce the rotor sub-synchronous fault of the equipment. Since their characteristic frequencies are very close, it is very difficult to accurately distinguish them only relying on spectral or time-frequency characteristic information. However, their repeatability indexes of the shaft center orbit often have obvious differences. Therefore, the initial phase point characteristic of the shaft center orbit has important value for accurately identifying these two types of faults.

[0003] At present, in the field of equipment monitoring and diagnosis, the conventional method for extracting the initial phase point of the shaft center trajectory using data analysis software is as follows: First, synchronously collect the vibration displacements in the X and Y directions and the coaxial key phase signal at a certain measurement section of the rotor. Second, manually determine the key phase pulse trigger direction and shape according to the selected key phase sensor model, and set a fixed amplitude threshold. Finally, when the amplitude of the key phase signal exceeds the set amplitude threshold, determine this moment as the pulse trigger moment, and the corresponding shaft center trajectory coordinate point is the initial phase point. It can be seen that the acquisition of the initial phase point of the equipment shaft center trajectory is significantly affected by the key phase signal, which makes the conventional initial phase point extraction method have the following disadvantages: First, the pulse directions generated by different types of key phase sensors are different, and the conventional method cannot automatically determine the pulse trigger direction. Second, the pulse amplitudes generated by the same type of key phase sensor also vary. The conventional method uses a fixed threshold as the standard for detecting the pulse trigger position and cannot adaptively adjust the detection threshold. Finally, the speed of the equipment, the width of the keyway or the reflective tape will directly affect the pulse signal width, and the pulse shape of the key phase signal is diverse, which will seriously affect the accuracy of pulse trigger position recognition. Facing such complex and variable on-site test conditions, how to accurately achieve the adaptive extraction of the initial phase point of the equipment shaft center trajectory in the data analysis software for key phase signals with different trigger directions, trigger amplitudes and pulse widths is one of the key problems that need to be solved urgently to improve the accuracy of rotating machinery fault diagnosis. Summary of the Invention

[0004] To overcome the deficiencies of the above-mentioned prior art, the purpose of the present invention is to provide an adaptive extraction method for the initial phase point of the shaft center trajectory of a rotating machine to adapt to the extraction of the initial phase point characteristics of the shaft center trajectory under different pulse trigger directions, pulse amplitudes and pulse widths, so as to provide richer diagnostic information for the accurate diagnosis of rotating equipment faults.

[0005] To achieve the above purpose, the technical solution adopted by the present invention is as follows:

[0006] An adaptive extraction method for the initial phase point of the shaft center trajectory of a rotating machine includes the following steps:

[0007] 1) Select the shaft vibration displacement sensors X and Y installed perpendicular to each other and the coaxial key phase sensor K at the same measurement section of the equipment as the analysis objects, and obtain the vibration sample signals x(i), y(i) and the key phase signal k(i) through signal sampling, where i = 1, 2,..., N, and N represents the sampling length of a single sample;

[0008] 2) Calculate the characteristic parameters of the key phase signal k(i), including the maximum value k max , the minimum value k min , and the mean value

[0009] 3) According to the maximum value k in step 2) max , the minimum value k min and the mean value k mean , automatically determine the trigger direction of the key phase signal, and the judgment rule is as follows: If k max -k mean >k mean -k min , it is determined as a positive pulse signal, and mark s = 1; if k max -k mean <k mean -k min , it is determined as a negative pulse signal, and mark s = -1;

[0010] 4) First, perform a first-order backward difference operation on the key phase signal k(i), and then multiply it by the coefficient s to construct a new sequence k1(i). The operation is as follows: Let k1(1) = 0, k1(i) = s × [k(i) - k(i - 1)], i = 2, 3,..., N; Search for all local maximum points k1(j) of the sequence k1(i), and the array j(l) stores the position indices of the local maxima of k1(i);

[0011] 5) Set the judgment threshold σ0 using the pulse amplitude of the key phase signal k(i), and find the pulse trigger position in k1(j(l)). The judgment method is: If k1(j(l)) > σ0, then judge that this local maximum point is the pulse trigger position, and successively store the index values corresponding to each pulse trigger position into the array p(m), m = 1, 2, 3,..., M, where M represents all the pulse numbers contained in the key phase signal k(i);

[0012] 6) Draw the waveform diagrams of the signals x(i), y(i), and k(i) respectively, and mark the signal points x(p(m)), y(p(m)) corresponding to the pulse trigger moments with special symbols in the vibration waveform diagram to obtain the coordinates (x(p(m)), y(p(m))) of the initial phase points of the shaft center trajectory; Use the vibration signals x(i), y(i) to draw the shaft center trajectory diagram of this measurement section, and finally mark the initial phase points (x(p(m)), y(p(m))) of the shaft center trajectory with special symbols.

[0013] In step 1), the key phase sensor includes but is not limited to eddy current sensors, photoelectric sensors, etc., and there are no special requirements for the pulse direction, pulse width, and amplitude of the key phase signal.

[0014] In step 2), the judgment threshold σ0 is determined by the formula σ0 = ε(k max -k min ), and the value range of ε is 0.2 to 0.8.

[0015] The beneficial effects of the present invention are:

[0016] The present invention automatically determines the pulse trigger direction of the key phase signal, and adaptively determines the pulse detection threshold and trigger position according to the characteristics of the key phase signal itself, which well solves the problem that the traditional method cannot accurately extract the initial phase point of the shaft center trajectory under different pulse trigger directions, pulse amplitudes and pulse widths, realizes the adaptive extraction of the initial phase point characteristics of the shaft center trajectory, provides key information for the quantitative evaluation of the stability and complexity of the shaft center trajectory, and helps to achieve the accurate diagnosis of rotating equipment faults. Description of the Drawings

[0017] Figure 1 is the flowchart of the embodiment of the present invention.

[0018] Figure 2 is the layout diagram of the Bently rotor test bench and sensors in the embodiment.

[0019] Figure 3 is the vibration waveform diagram of the displacement sensors 3X and 3Y in Embodiment 1.

[0020] Figure 4 is the waveform diagram of the key phase sensor K in Embodiment 1.

[0021] Figure 5 is the new sequence k1(i) constructed based on the key phase signal k(i) in Embodiment 1.

[0022] Figure 6 is the waveform diagram of 3X, 3Y and the key phase sensor K after marking the key phase trigger position in Embodiment 1.

[0023] Figure 7 is the shaft center trajectory diagram and its initial phase point of the 3# bearing measurement section in Embodiment 1.

[0024] Figure 8 is the vibration waveform diagram of the displacement sensors 3X and 3Y in Embodiment 2.

[0025] Figure 9 is the waveform diagram of the key phase sensor K in Embodiment 2.

[0026] Figure 10 is the new sequence k1(i) constructed based on the key phase signal k(i) in Embodiment 2.

[0027] Figure 11 is the waveform diagram of 3X, 3Y and the key phase sensor K after marking the key phase trigger position in Embodiment 2.

[0028] Figure 12 is the shaft center trajectory diagram and its initial phase point of the 3# bearing measurement section in Embodiment 2. Detailed Implementation Manner

[0029] The present invention will be further described in detail below with reference to the accompanying drawings and embodiments:

[0030] In Embodiment 1, the vibration displacement signal and the key phase signal are both collected relying on the Bently high-speed rotor test bench, and its structural schematic diagram is as Figure 2 shown: This test bench includes motor A and rotor B. The rotor of motor A is supported by bearings 1# and 2#, and rotor B is supported by bearings 3# and 4#. The two rotors are connected by a flexible coupling C. Along the axial direction symmetrically arranged within the span of rotor B are balance disks 1# and 2#. The shaft diameter is 10 mm, and the span between the front and rear supports is 325 mm. Two mutually perpendicular vibration displacement sensors are installed in the vibration measuring belts of bearings 3# and 4# in the X and Y directions, marked as 3X, 3Y, 4X, and 4Y; a key phase sensor K is installed at bearing 2# of motor A, aligned with the keyway on motor A. The vibration displacement sensors and the key phase sensor K both adopt Bently eddy current sensors with a diameter of 5 mm specification. During the experiment, the working speed of the motor is set to 240 r / min, and a negative pulse key phase signal with a certain pulse width is simulated and generated.

[0031] Referring to Figure 1 , an adaptive extraction method for the initial phase point of the axis orbit of a rotating machine, includes the following steps:

[0032] 1) Select the measurement section of bearing 3# in Figure 2 as the monitoring object, and the corresponding eddy current sensors are 3X and 3Y. Set the sampling frequency to 1024 Hz and the number of sampling points N to 2048. After the rotational speed stabilizes to 240 r / min, use the portable monitoring system to discretely sample and select a sample at a certain moment, and its corresponding vibration displacement signals x(i), y(i) and key phase signal k(i), i = 1, 2,..., 2048, as Figure 3 , Figure 4 shown;

[0033] 2) Calculate the characteristic parameters of the key phase signal k(i), including: the maximum value k max = 101.8, the minimum value k min = -375.2, the average value k mean = 1.2, as Figure 4 shown;

[0034] 3) Calculate from step 2): k max - k mean = 100.6, k mean - k min = 376.4. According to k max - k mean < k mean - k min , determine that the pulse type is a negative pulse, and mark s = -1;

[0035] 4) With s = -1 obtained in step 3), let k1(1) = 0, k1(i) = s × [k(i) - k(i - 1)], i = 2, 3,..., N, and construct a new sequence k1(i), as shown below; Search for all local maximum points k1(j) of the sequence k1(i), store the position indices of the local maxima of k1(i) into the array j(l), and a total of 605 extreme points are searched; Figure 5 As shown; Search for all local maximum points k1(j) of the sequence k1(i), store the position indices of the local maxima of k1(i) into the array j(l), and a total of 605 extreme points are searched;

[0036] 5) Set ε = 0.25, then the threshold σ0 = ε(k max -k min ) = 119.3. According to k1(j(l)) > 119.3, determine that this point is the pulse trigger position, and store the pulse position index into the array p(m). A total of 7 pulse trigger positions are recorded, as shown in the second column of Table 1;

[0037] Table 1

[0038] Serial number p(m) x(p(m)) y(p(m)) 1 252 -32.4 -21.9 2 516 -32.4 -22.3 3 780 -33.8 -24.1 4 1046 -32.1 -23.9 5 1310 -32.0 -21.8 6 1575 -31.4 -21.1 7 1840 -33.8 -23.8

[0039] 6) Plot the waveforms of the signals x(i), y(i), and k(i), and mark the signal points corresponding to the pulse trigger moments with solid circles "●" in the vibration waveform diagram, as shown below. Obtain the coordinates (x(p(m)), y(p(m))) of the initial phase points of the shaft center orbit, as shown in the third and fourth columns of Table 1; Use the vibration signals x(i) and y(i) to plot the shaft center orbit diagram of the measurement section of bearing 3#. Finally, mark the initial phase points (x(p(m)), y(p(m))) of the shaft center orbit of this section with solid circles "●", as shown below. Figure 6 As shown; Obtain the coordinates (x(p(m)), y(p(m))) of the initial phase points of the shaft center orbit, as shown in the third and fourth columns of Table 1; Use the vibration signals x(i) and y(i) to plot the shaft center orbit diagram of the measurement section of bearing 3#. Finally, mark the initial phase points (x(p(m)), y(p(m))) of the shaft center orbit of this section with solid circles "●", as shown below. Figure 7 As shown.

[0040] The test bench and vibration displacement sensors used in Example 2 are exactly the same as those in Example 1. The difference is that in Example 2, a ROS-P type photoelectric sensor is used as the key phase sensor, and a reflective tape is pasted on the surface of rotor B, and the photoelectric sensor is aligned with the reflective tape. During the experiment, the operating speed of the motor is set to 900 r / min to simulate the generation of a positive pulse key phase signal.

[0041] Referring to Figure 1 , an adaptive extraction method for the initial phase points of the shaft center orbit of a rotating machine includes the following steps:

[0042] 1) Select Figure 2Taking the bearing measurement section 3# as the monitoring object, the corresponding eddy current sensors are 3X and 3Y. The sampling frequency is set to 2048 Hz, and the number of sampling points N is 2048. After the rotational speed is stabilized to 900 r / min, a sample at a certain moment is selected after discrete sampling using a portable monitoring system. The corresponding vibration displacement signals x(i), y(i) and key phase signal k(i), where i = 1, 2,..., 2048, are as Figure 8 shown;

[0043] 2) Calculate the characteristic parameters of the key phase signal k(i), including: the maximum value k max =-9.1, the minimum value k min =-2570, the mean value k mean =-2550, as Figure 9 shown;

[0044] 3) Calculate from step 2): k max -k mean =2560.9, k mean -k min =20. According to k max -k mean >k mean -k min , it is determined that the pulse type is a positive pulse, and mark s = 1;

[0045] 4) Given s = 1 obtained from step 3), let k1(1) = 0, k1(i) = s × [k(i)-k(i - 1)], i = 2, 3,..., N, to construct a new sequence k1(i), as Figure 10 shown; Search for all local maximum points k1(j) of the sequence k1(i). The array j(l) stores the position indices of the local maxima of k1(i), and 15 extreme points are searched;

[0046] 5) Set ε = 0.25, then the threshold σ0 = ε(k max -k min ) = 640.2. When k1(j(l))>640.2, it is determined that this point is the pulse trigger position, and the pulse position indices are stored in the array p(m). A total of 15 pulse trigger positions are recorded, and the results are shown in the second column of Table 2;

[0047] Table 2

[0048]

[0049]

[0050] 6) Plot the waveform diagrams of the signals x(i), y(i), k(i), and mark the signal points corresponding to the pulse trigger moments with solid circles "●" in the vibration waveform diagram, as Figure 11As shown, the coordinates (x(p(m)), y(p(m))) of the initial phase point of the shaft center orbit are obtained, as shown in the 3rd and 4th columns of Table 2; the shaft center orbit diagram of the 3# bearing measurement section is drawn using the vibration signals x(i) and y(i), and finally the initial phase point (x(p(m)), y(p(m))) of the shaft center orbit of this section is marked with a solid circle "●", as Figure 12 shown.

Claims

1. An adaptive extraction method for the initial phase point of the shaft center orbit of a rotating machine, characterized in that, It includes the following steps: 1) Select the vibration displacement sensors X and Y of two mutually perpendicular rotating shafts and the coaxial key-phase sensor K at the same measurement section of the device as the analysis objects, and obtain the vibration sample signals x(i), y(i) and the key-phase signal k(i) through signal sampling, where i = 1, 2,..., N, and: N represents the sampling length of a single sample; 2) Calculate the characteristic parameters of the key phase signal k(i), including the maximum value k max , the minimum value k min , the mean value 3) According to the maximum value k in step 2) max , the minimum value k min and the mean value k mean 's mutual relationship, automatically determine the triggering direction of the key phase signal, and the judgment rules are as follows: If k max -k mean >k mean -k min , it is determined as a positive pulse signal, and mark s = 1; If k max -k mean <k mean -k min , it is determined as a negative pulse signal, and mark s = -1; 4) First perform a first-order backward difference operation on the key-phase signal k(i), and then multiply it by the coefficient s to construct a new sequence k1(i). The operation is as follows: Let k1(1) = 0, k1(i) = s × [k(i) - k(i - 1)], i = 2, 3,..., N; Search for all local maximum points k1(j) of the sequence k1(i), and the array j(l) stores the position indices of the local maxima of k1(i); 5) Set the judgment threshold σ0 using the pulse amplitude size of the key-phase signal k(i), and find the pulse trigger position in k1(j(l)). The judgment method is: If k1(j(l)) > σ0, then judge that this local maximum point is the pulse trigger position, and successively store the index values corresponding to each pulse trigger position into the array p(m), where m = 1, 2, 3,..., M, and M represents all the pulses contained in the key-phase signal k(i); 6) Respectively plot the waveform diagrams of the signals x(i), y(i), and k(i), and mark the signal points x(p(m)), y(p(m)) corresponding to the pulse trigger moments with symbols in the vibration waveform diagram to obtain the coordinates (x(p(m)), y(p(m))) of the initial phase points of the shaft center trajectory; Use the vibration signals x(i), y(i) to plot the shaft center trajectory diagram of this measurement section, and finally mark the initial phase points (x(p(m)), y(p(m))) of the shaft center trajectory with symbols.

2. The method according to claim 1, wherein: In step 1) described above, the key-phase sensor includes but is not limited to an eddy current sensor and an optoelectronic sensor, and there are no special requirements for the pulse direction, pulse width, and amplitude of the key-phase signal.

3. The method according to claim 1, wherein: In step 2), the judgment threshold value σ0 is determined by the formula σ0 = ε(k max -k min ), where the value range of ε is 0.2 to 0.8.

Images