An ultra-high voltage power transformer winding fault simulation device and diagnosis method based on oscillating wave detection

Abstract

The application discloses an oscillation wave detection-based extra-high voltage power transformer winding fault simulation device and diagnosis method. Firstly, the test transformer winding is tested, the transformer winding connection point is connected with a high-frequency high-voltage switch and a high-frequency high-voltage direct current power supply, the high-frequency high-voltage switch is periodically actuated, thereby exciting the extra-high voltage transformer body structure vibration; according to the test data, an ITD transformation is adopted to establish a signal model X t ; then, an empirical mode decomposition (EMD) method is adopted to decompose the signal into a sum of a plurality of mutually orthogonal intrinsic mode function (IMF) components, a Hilbert transformation is performed on each IMF component to obtain an instantaneous frequency and an instantaneous amplitude, thereby obtaining a Hilbert spectrum of the signal to perform decomposition and reconstruction, improve the signal-to-noise ratio, calculate I(X,Y) to judge and remove most of the redundant noise and interference signals; further reconstruct the required signal, after the EMD decomposition of the non-steady-state signal, the mutual information is utilized to screen the obtained IMF components of each order, and a new x(t) is obtained. Finally, the winding vibration signal after the simulation deformation fault is solved through a signal sequence sample entropy calculation formula, and the winding deformation fault severity is judged, and the transformer fault degree is identified.

Description

Technical Field

[0001] This invention focuses on the application of oscillating wave detection technology in the field of ultra-high voltage power transformers, specifically involving a simulation device and diagnostic method for axial loosening faults in ultra-high voltage power transformer windings based on oscillating wave detection. Background Technology

[0002] In ultra-high voltage (UHV) power systems, UHV transformers bear the core mission of enabling high-voltage power transmission, and their reliability and safety play a decisive role in the stable operation of the entire power system. However, in actual operation, UHV transformers often encounter frequent impacts from instantaneous loads and strong impacts from occasional short-circuit currents. These impacts can easily cause the winding clamps to come loose, leading to winding loosening. Once the windings become loose, their mechanical bearing capacity will decrease significantly, which will not only interfere with the normal operation of the UHV transformer but also pose a serious threat to the stable operation of the UHV power transmission system.

[0003] To effectively prevent potential catastrophic consequences, routine inspections of transformer windings are of irreplaceable importance. Therefore, it is crucial and necessary to utilize an ultra-high voltage (UHV) power transformer winding fault simulation device based on oscillation wave detection to accurately simulate axial loosening faults in the windings and to promptly diagnose the condition of the transformer windings. This device and related diagnostic methods can provide strong technical support and assurance for the stable operation of UHV transformers.

[0004] The ultra-high voltage transformer winding fault simulation device uses a rotating screw to control the rotation of the transformer winding, allowing the high-voltage winding to translate and rotate arbitrarily along its axis to adjust to a suitable position and simulate the fault. Then, the left and right arms of the fault simulation mechanism cause axial loosening of the winding at a designated location. The vibration signal on the transformer tank surface acquired by the high-speed camera exhibits a non-constant curve characteristic. This vibration signal has a close intrinsic relationship with the winding state; therefore, the winding condition can be diagnosed based on the changes in the output vibration signal.

[0005] However, given the numerous unresolved issues in existing technologies, this invention innovatively proposes a simulation device and diagnostic method for UHV power transformer winding faults based on oscillating wave detection. A specially designed mechanical device simulates the loosening and falling of winding clamps at specific locations, and works in conjunction with a high-speed camera testing system. Specifically, this device and system can calculate the vibration signals at various measuring points on the surface of the UHV transformer tank, thereby deriving the entropy expression of vibration signal samples in different regions under different winding conditions. Based on this, the actual state of the winding can be further accurately determined. In summary, this invention, with its unique design and operating mechanism, can stably, efficiently, and reliably achieve accurate diagnosis of axial loosening faults in UHV transformer windings, providing solid technical support and guarantee for the safe and stable operation of UHV power transformers. Summary of the Invention

[0006] A simulation device and diagnostic method for UHV power transformer winding faults based on oscillating wave detection is proposed. This method uses a mechanical device to simulate axial loosening of the UHV transformer winding and combines it with a high-speed camera to achieve non-contact acquisition of transformer vibration signals for winding diagnosis. The device includes: a high-voltage winding of the power transformer, a power transformer winding support platform, a transmission gear for rotating the transformer winding fixed on the power transformer winding support platform, a connecting frame fixed on the power transformer winding support platform, a transformer rotating motor fixed on the power transformer winding connecting frame, and a transmission gear for fixing the transformer winding. Fixed transmission knob, axial slider seat, gear drive motor, transformer clamping drive gear, power transformer winding support frame, robotic arm control motor, power transformer winding support base, gear drive motor, rotating device support frame, power transformer winding rotating screw, power transformer winding rotating screw control frame, base plate fixed on the power transformer winding support platform, front baffle of the support platform, power transformer winding rotation sensor, rear baffle of the support platform, power transformer winding clamping sensor, transformer rotation driven gear, fault simulation mechanical right arm, rotating device support platform, fault simulation mechanical left arm Support platform front baffle, power transformer winding rotating gear support platform, power transformer winding rotating screw connecting shaft, power transformer winding rotating internal gear, power transformer winding rotating planetary gear, transformer winding clamping fixing shim, power transformer winding rotating internal gear support shaft, power transformer winding rotating fixing plate, power transformer winding rotating internal gear rotating column, fault simulation robot arm base, fault simulation robot arm upper arm, fault simulation robot arm clamping device, fault simulation robot arm clamping device motor, fault simulation robot arm lower arm, fault simulation robot arm waist, fault simulation robot arm gripper, fault... The system comprises a fault simulation robotic arm gripper connecting shaft, a fault simulation robotic arm gripper support platform, a fault simulation robotic arm hand transmission device, an upper sleeve for the first high-voltage winding, an upper sleeve for the second low-voltage winding, an upper sleeve for the third high-voltage winding, an upper sleeve for the fourth low-voltage winding, an upper sleeve for the fifth high-voltage winding, an upper sleeve for the sixth low-voltage winding, a high-frequency high-voltage DC power supply controlled by a high-frequency high-voltage switch, a high-speed camera, and a signal acquisition device connected to the high-speed camera. The specific steps include:

[0007] Step 1: Simulate axial loosening faults at different locations in the windings of an ultra-high voltage transformer;

[0008] Step 2: Conduct machine vision vibration testing and fault diagnosis of UHV transformer windings.

[0009] Further, step one includes:

[0010] 1) The high-voltage winding of the power transformer is fixed at the center of the rotating screw connecting shaft of the power transformer winding by placing the rotating internal gear of the power transformer winding on the supporting shaft and the rotating internal gear of the power transformer winding.

[0011] 2) Rotate the clamping transformer winding fixing transmission knob to drive the clamping transformer winding fixing transmission gear and gear drive motor to start working, so that the clamping transformer winding fixing shims on the clamping transformer winding rotation transmission gear expand outward, thereby changing the spatial position of the high voltage winding of the power transformer.

[0012] 3) Rotate the lead screw of the power transformer winding, which drives the connecting shaft of the lead screw of the power transformer winding to rotate. This causes the internal gear and planetary gear of the power transformer winding to rotate axially, allowing the high-voltage winding of the power transformer to translate and rotate arbitrarily in the axial direction to adjust it to a suitable position.

[0013] 4) The fault simulation robot left arm moves along the connecting frame fixed on the power transformer winding support platform via the fault simulation robot arm base to the position where the winding clamping part is loose. The fault simulation robot arm upper arm, fault simulation robot arm lower arm, and fault simulation robot arm clamping device are controlled to move the fault simulation robot arm gripper to the front of the high voltage winding of the power transformer. The fault simulation robot arm gripper connecting shaft pushes out the wedge block and inserts it into the middle of the winding coil at both ends of the high voltage winding of the power transformer, causing loosening. The robot arm retracts, and the fault simulation robot arm upper arm resets along the connecting frame fixed on the power transformer winding support platform via the fault simulation robot arm base.

[0014] 5) The fault simulation robot right arm moves along the connecting frame fixed on the power transformer winding support platform via the fault simulation robot arm base to the position where the winding clamping part is loose. The fault simulation robot arm's upper arm, lower arm, and clamping device are controlled to move the fault simulation robot arm's gripper to the front of the power transformer's high-voltage winding. The fault simulation robot arm's lower arm is driven to feed and clamp the power transformer's high-voltage winding. The robot arm movement completes the extraction of the power transformer's high-voltage winding. The fault simulation robot arm's upper arm resets along the connecting frame fixed on the power transformer winding support platform via the robot arm moving base.

[0015] 6) The fault simulation robot left arm moves along the connecting frame fixed on the power transformer winding support platform via the fault simulation robot arm base to the position where the winding clamping part is loose. The fault simulation robot arm upper arm, fault simulation robot arm lower arm, and fault simulation robot arm clamping device are controlled to move the fault simulation robot arm gripper to the front of the high voltage winding of the power transformer. The fault simulation robot arm gripper retracts the wedge block, the robot arm retracts, and the fault simulation robot arm upper arm resets along the connecting frame fixed on the power transformer winding support platform via the fault simulation robot arm base.

[0016] 7) Rotate the clamping transformer winding fixing transmission knob to start the operation of the clamping transformer winding fixing transmission gear and gear drive motor, so that the clamping transformer winding fixing shims on the clamping transformer winding rotation transmission gear retract inward to control the high voltage winding fixing space position of the power transformer to remain stationary.

[0017] 8) Repeat steps 2) to 7) to simulate axial loosening faults in the windings at different locations.

[0018] Step two includes:

[0019] 1) Define the connection point A for the upper bushings connecting the first high-voltage winding, the third high-voltage winding, and the fifth high-voltage winding; define the connection point B for the upper bushings connecting the second low-voltage winding, the fourth low-voltage winding, and the sixth low-voltage winding.

[0020] 2) Connect transformer winding connection points A and B to a high-frequency high-voltage switch and a high-frequency high-voltage DC power supply. The high-frequency high-voltage switch will perform periodic operations, thereby stimulating the vibration of the UHV transformer body structure.

[0021] When using traditional frequency analysis methods to analyze the components of vibration test signals from multiple vibration sources such as UHV transformer windings, cores, and cooling systems, spurious signals and frequencies may occur. Therefore, it is necessary to use instantaneous frequency to represent the local characteristics of the signal. Thus, the ITD transform is used to establish the signal model.

[0022]

[0023] Where L is the baseline extraction operator, Xt is the input signal, Lt = LXt is the baseline signal, Ht = (1-L)Xt is the intrinsic rotation component, and τ k (k = 1, 2, 3, ...) represents the corresponding time, HLkXt is the inherent rotation component of the (k+1)th layer, and LpXt is the monotonic trend component of the original signal.

[0024] 3) The next step is to use the Empirical Mode Decomposition (EMD) method to decompose the signal into the sum of several mutually orthogonal Intrinsic Mode Function (IMF) components. Then, a Hilbert transform is performed on each IMF component to obtain the instantaneous frequency and instantaneous amplitude, thereby obtaining the Hilbert spectrum of the signal for decomposition and reconstruction, thus improving the signal-to-noise ratio.

[0025]

[0026] Where X and Y are different signal representations, H(X,Y) is the joint entropy, I(X,Y) is the definition of mutual information of the signals, p(x(i)) and p(y(j)) are the probability distributions of X and Y, and p(x(i),y(j)) is the joint probability. By calculating I(X,Y), most of the redundant noise and interference signals can be judged and removed.

[0027] 4) Further reconstruct the required signal and reduce on-site noise interference: Since the IMF can characterize the dependency between two data, mutual information can be used to filter the IMF components of each order after the EMD decomposes the non-steady-state signal.

[0028]

[0029] Where u1(t) and u2(t) are the upper and lower envelopes connecting the local maxima and local minima, respectively, and m(t) is the average of the upper and lower envelopes; h(t) = x(t) - m(t). If h(t) does not satisfy the IMF condition, it is considered as a new x(t). Repeating this process k times yields h1k(t); SD is the standard deviation, used to determine whether the screening process terminates. Finally, we obtain C1 = h1k(t) and r1(t) = x(t) - C1.

[0030] 5) Collect the processed data and set it as the original data X. i ={x1,x2,…,x N}, of length N, with a pre-defined embedding dimension m and similarity capacity r, consider an m-dimensional vector x(i) = [x i ,x i+1 ,…x i+m-1 (i = 1, 2, ..., Nm), and define the distance d[x(i), x(j)] between x(i) and x(j) as the maximum value of the difference between their corresponding elements, i.e.

[0031] 6) Calculate the distance d[x(i),x(j)] between x(i) and the remaining vectors x(j) (j=1,2,...,Nm,j≠i) using the sample entropy method. Count the number of vectors with distances less than r and the ratio of this number to the total number of distances Nm-1, denoted as r. Right now:

[0032]

[0033] 7) Next, the calculations are as follows: Average value: Then, for dimension m+1, i.e., for vector m+1, repeat steps 7) and 8) to obtain... Then, its average value is calculated to obtain B. m+1 (r).

[0034] 8) Finally, the sample entropy calculation formula of this sequence is used to solve the winding vibration signal after simulating deformation fault, and the severity of the winding deformation fault is determined accordingly, as shown below:

[0035]

[0036] When N is a finite number, the expression can be represented as:

[0037]

[0038] 9) The magnitude of the sample entropy is related to the values ​​of m and r. We generally take m = 2 and r = 0.2SD, where SD is the standard deviation of the original data.

[0039] If 0.21≤SamEn(m,r,N)<0.26, the winding is considered to be in a normal state.

[0040] If 0.26≤SamEn(m,r,N)<0.31, it is determined that the winding has undergone slight deformation;

[0041] If 0.31≤SamEn(m,r,N)<0.36, the winding is judged to have undergone severe deformation.

[0042] This invention focuses on a simulation device and diagnostic method for winding faults in ultra-high voltage (UHV) power transformers based on oscillating wave detection. Through a mechanical device, it accurately simulates the occurrence of axial loosening faults in UHV transformer windings. Based on this, it innovatively combines a high-speed camera with a unique detection system built using oscillating wave detection technology to achieve non-contact acquisition of transformer vibration signals, thereby enabling accurate diagnosis of winding conditions and significantly improving the safety and stability of UHV power system operation. Attached Figure Description

[0043] Figure 1This is a schematic diagram of the overall structure of an ultra-high voltage transformer winding fault simulation device according to the present invention;

[0044] Figure 2 This is a perspective view of an ultra-high voltage transformer winding fault simulation device according to the present invention;

[0045] Figure 3 This is a top view of an ultra-high voltage transformer winding fault simulation device according to the present invention;

[0046] Figure 4 This is a magnified view of the high-voltage winding connection rotation of an ultra-high voltage transformer winding fault simulation device according to the present invention.

[0047] Figure 5 This is a schematic diagram of the structure of a fault diagnosis robotic arm for an ultra-high voltage transformer winding fault simulation device according to the present invention;

[0048] Figure 6 This is a schematic diagram of the structure of the fault diagnosis robotic arm gripper of an ultra-high voltage transformer winding fault simulation device according to the present invention.

[0049] Figure 7 This is a schematic diagram of the wiring of an ultra-high voltage transformer winding;

[0050] Figure 8 This is a flowchart of a diagnostic method for an ultra-high voltage transformer winding fault simulation device according to the present invention. Detailed Implementation

[0051] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments:

[0052] like Figure 1As shown in Figures 2, 3, 4, 5, 6, and 7, the UHV power transformer winding fault simulation device based on oscillating wave detection controls the transformer winding to translate and rotate in the horizontal direction. During the translation and rotation process, the selected turns of the winding are driven to undergo compression deformation to achieve winding diagnosis. The device includes: a high-voltage winding of the power transformer (1), a power transformer winding support platform (2), a transmission gear (4) for clamping the transformer winding rotation fixed on the power transformer winding support platform (3), a connecting frame (5) fixed on the power transformer winding support platform, a transformer rotating motor (6) fixed on the power transformer winding connecting frame (5), a transmission gear (7) for clamping the transformer winding fixed, a transmission knob (8) for clamping the transformer winding fixed, and a device located at... Axial slider seat (9), gear drive motor (10), transformer clamping drive gear (11), power transformer winding support frame (12), robotic arm control motor (13), power transformer winding support base (14), gear drive motor (15), rotating device support frame (16), power transformer winding rotating screw (17), power transformer winding rotating screw control frame (18), base plate seat fixed on the power transformer winding support platform (19), front baffle of support platform (20), power transformer winding rotation sensor (21), rear baffle of support platform (22), power transformer winding clamping sensor (23), transformer rotation driven gear (24), fault simulation robotic right arm (25), rotating device Support platform (26), fault simulation mechanical left arm (27), support platform front baffle (28), power transformer winding rotating gear support platform (29), power transformer winding rotating screw connecting shaft (30), power transformer winding rotating internal gear (31), power transformer winding rotating planetary gear (32), clamping transformer winding fixing shim (33), power transformer winding rotating internal gear support shaft (34), power transformer winding rotating fixing plate (35), power transformer winding rotating internal gear rotating column (36), fault simulation mechanical arm base (37), fault simulation mechanical arm upper arm (38), fault simulation mechanical arm clamping device (39), fault simulation mechanical arm clamping device motor (40), fault model The simulated robotic arm includes a forearm (41), a waist section (42), a gripper (43), a connecting shaft (44), a support platform (45), a transmission device (46), an upper sleeve (47) for the first high-voltage winding, an upper sleeve (48) for the second low-voltage winding, an upper sleeve (49) for the third high-voltage winding, an upper sleeve (50) for the fourth low-voltage winding, an upper sleeve (51) for the fifth high-voltage winding, an upper sleeve (52) for the sixth low-voltage winding, a high-frequency high-voltage DC power supply (54) controlled by a high-frequency high-voltage switch (53), a high-speed camera (55), and a signal acquisition device (56) connected to the high-speed camera (55).First high-voltage winding (57), second low-voltage winding (58), third high-voltage winding (59), fourth low-voltage winding (60), fifth high-voltage winding (61), and sixth low-voltage winding (62).

[0053] Figure 8 This is a fault diagnosis method for ultra-high voltage power transformer windings based on oscillating wave detection. Its key feature is the combination of ITD (Inertial Mode Decomposition) to construct a signal model, decomposition and reconstruction using empirical modes, and finally, the sample entropy method to solve for the winding vibration signal after simulating deformation faults, thereby determining the severity of the winding deformation fault. The method specifically includes the following steps:

[0054] 1) Define the connection point of the upper bushing connecting the first high voltage winding (57), the third high voltage winding (59), and the fifth high voltage winding (61) as A; define the connection point of the upper bushing connecting the second low voltage winding (58), the fourth low voltage winding (60), and the sixth low voltage winding (62) as B;

[0055] 2) Connect transformer winding connection points A and B to high-frequency high-voltage switch (53) and high-frequency high-voltage DC power supply (54). The high-frequency high-voltage switch (53) performs periodic operation, thereby stimulating the vibration of the UHV transformer body structure.

[0056] 3) When using traditional frequency analysis methods to analyze the vibration test signal components of non-stationary signals from multiple vibration sources such as UHV transformer windings, cores, and cooling systems, spurious signals and frequencies may occur. Therefore, instantaneous frequency is needed to represent the local characteristics of the signal. Thus, the ITD transform is used to establish the signal model.

[0057]

[0058] Where L is the baseline extraction operator, Xt is the input signal, Lt = LXt is the baseline signal, Ht = (1-L)Xt is the intrinsic rotation component, and τ k (k = 1, 2, 3, ...) represents the corresponding time, HLkXt is the inherent rotation component of the (k+1)th layer, and LpXt is the monotonic trend component of the original signal.

[0059] 4) The next step is to use the Empirical Mode Decomposition (EMD) method to decompose the signal into the sum of several mutually orthogonal Intrinsic Mode Function (IMF) components. Then, a Hilbert transform is performed on each IMF component to obtain the instantaneous frequency and instantaneous amplitude, thereby obtaining the Hilbert spectrum of the signal for decomposition and reconstruction, thus improving the signal-to-noise ratio.

[0060]

[0061] Here, X and Y are different representations of the signal, H(X,Y) is the joint entropy, I(X,Y) is the definition of mutual information of the signal, p(x(i)) and p(y(j)) are the probability distributions of X and Y, and p(x(i),y(j)) is the joint probability. By calculating I(X,Y), most of the redundant noise and interference signals can be judged and removed.

[0062] 5) Further reconstruct the required signal and reduce on-site noise interference: Since the IMF can characterize the dependency between two data, mutual information can be used to filter the IMF components of each order after the EMD decomposes the non-steady-state signal.

[0063]

[0064] Where u1(t) and u2(t) are the upper and lower envelopes connecting the local maxima and local minima, respectively, and m(t) is the average of the upper and lower envelopes; h(t) = x(t) - m(t). If h(t) does not satisfy the IMF condition, it is considered as a new x(t). Repeating this process k times yields h1k(t); SD is the standard deviation, used to determine whether the screening process terminates. Finally, we obtain C1 = h1k(t) and r1(t) = x(t) - C1.

[0065] 6) Collect the processed data and set it as the original data X. i ={x1,x2,…,x N}, of length N, with a pre-defined embedding dimension m and similarity capacity r, consider an m-dimensional vector x(i) = [x i ,x i+1 ,…x i+m-1 (i = 1, 2, ..., Nm), and define the distance d[x(i), x(j)] between x(i) and x(j) as the maximum value of the difference between their corresponding elements, that is:

[0066]

[0067] 7) Calculate the distance d[x(i),x(j)] between x(i) and the other vectors x(j) (j=1,2,...,Nm,j≠i) using the sample entropy method.

[0068] Count the number of values ​​of d[x(i),x(j)] less than r and the ratio of this number to the total distance Nm-1, denoted as r. Right now:

[0069]

[0070] 8) Then calculate to obtain Average value: Then, for dimension m+1, i.e., for vector m+1, repeat steps 7) and 8) to obtain... Then, its average value is calculated to obtain B. m+1 (r).

[0071] 9) Finally, the sample entropy calculation formula of this sequence is used to solve the winding vibration signal after simulating deformation fault, and the severity of the winding deformation fault is determined accordingly, as shown below:

[0072]

[0073] When N is a finite number, the expression can be represented as:

[0074]

[0075] 10) The magnitude of the sample entropy is related to the values ​​of m and r. We generally take m = 2 and r = 0.2SD, where SD is the standard deviation of the original data.

[0076] If 0.21≤SamEn(m,r,N)<0.26, the winding is considered to be in a normal state.

[0077] If 0.26≤SamEn(m,r,N)<0.31, it is determined that the winding has undergone slight deformation;

[0078] If 0.31≤SamEn(m,r,N)<0.36, the winding is judged to have undergone severe deformation.

Claims

1. A fault diagnosis method for an ultra-high voltage power transformer winding fault simulation device based on oscillating wave detection, comprising: The components of a power transformer include: high-voltage winding, power transformer winding support platform, fixed on the power transformer winding support platform, rotating transmission gear for clamping the transformer winding, connecting frame fixed on the power transformer winding support platform, transformer rotating motor fixed on the power transformer winding connecting frame, fixed transmission gear for clamping the transformer winding, fixed transmission knob for clamping the transformer winding, axial slider seat, gear drive motor, transformer clamping drive gear, power transformer winding support frame, robotic arm control motor, power transformer winding support base, gear drive motor, rotating device support frame, power transformer winding rotating screw, power transformer winding rotating screw control frame, base plate fixed on the power transformer winding support platform, front baffle of the support platform, power transformer winding rotation sensor, rear baffle of the support platform, power transformer winding clamping sensor, transformer rotating driven gear, right arm of the fault simulation machine, rotating device support platform, left arm of the fault simulation machine, front baffle of the support platform, power transformer winding rotating gear support platform, power transformer winding rotating screw connecting shaft, and power... Transformer winding rotating internal gear, power transformer winding rotating planetary gear, transformer winding clamping fixing shim, power transformer winding rotating internal gear support shaft, power transformer winding rotating fixing plate, power transformer winding rotating internal gear rotating column, fault simulation robot arm base, fault simulation robot arm upper arm, fault simulation robot arm clamping device, fault simulation robot arm clamping device motor, fault simulation robot arm lower arm, fault simulation robot arm waist, fault simulation robot arm gripper, fault simulation robot arm gripper connecting shaft, fault simulation robot arm gripper The system comprises a support platform, a fault simulation robotic arm transmission device, upper bushings for the first high-voltage winding, the second low-voltage winding, the third high-voltage winding, the fourth low-voltage winding, the fifth high-voltage winding, and the sixth low-voltage winding; a high-frequency high-voltage DC power supply controlled by a high-frequency high-voltage switch; a high-speed camera; a signal acquisition device connected to the high-speed camera; and the first, second, third, fourth, fifth, and sixth low-voltage windings. Specifically, it includes the following steps: Step 1: Simulate axial loosening faults at different locations in the windings of an ultra-high voltage transformer; Step 2: Conduct machine vision vibration testing and fault diagnosis of UHV transformer windings; Further, step one includes: 1) The high-voltage winding of the power transformer is fixed at the center of the rotating screw connecting shaft of the power transformer winding by placing the rotating internal gear of the power transformer winding on the supporting shaft and the rotating internal gear of the power transformer winding. 2) Rotate the clamping transformer winding fixing transmission knob to drive the clamping transformer winding fixing transmission gear and gear drive motor to start working, so that the clamping transformer winding fixing shims on the clamping transformer winding rotation transmission gear expand outward, thereby changing the spatial position of the high voltage winding of the power transformer. 3) Rotate the lead screw of the power transformer winding, which drives the connecting shaft of the lead screw of the power transformer winding to rotate. This causes the internal gear of the power transformer winding and the planetary gear of the power transformer winding to drive the support shaft of the internal gear of the power transformer winding to rotate axially, so that the high voltage winding of the power transformer can be translated and rotated arbitrarily in the axial direction to adjust to a suitable position. 4) The fault simulation robot left arm moves along the connecting frame fixed on the power transformer winding support platform via the fault simulation robot arm base to the position where the winding clamping part is loose. The fault simulation robot arm upper arm, fault simulation robot arm lower arm, and fault simulation robot arm clamping device are controlled to move the fault simulation robot arm gripper to the front of the high voltage winding of the power transformer. The fault simulation robot arm gripper connecting shaft pushes out the wedge block and inserts it into the middle of the winding coil at both ends of the high voltage winding of the power transformer, causing loosening. The robot arm retracts, and the fault simulation robot arm upper arm resets along the connecting frame fixed on the power transformer winding support platform via the fault simulation robot arm base. 5) The fault simulation robot right arm moves along the connecting frame fixed on the power transformer winding support platform via the fault simulation robot arm base to the position where the winding clamping part is loose. The fault simulation robot arm's upper arm, lower arm, and clamping device are controlled to move the fault simulation robot arm's gripper to the front of the power transformer's high-voltage winding. The fault simulation robot arm's lower arm is driven to feed and clamp the power transformer's high-voltage winding. The robot arm movement completes the extraction of the power transformer's high-voltage winding. The fault simulation robot arm's upper arm resets along the connecting frame fixed on the power transformer winding support platform via the robot arm moving base. 6) The fault simulation robot left arm moves along the connecting frame fixed on the power transformer winding support platform via the fault simulation robot arm base to the position where the winding clamping part is loose. The fault simulation robot arm upper arm, fault simulation robot arm lower arm, and fault simulation robot arm clamping device are controlled to move the fault simulation robot arm gripper to the front of the high voltage winding of the power transformer. The fault simulation robot arm gripper retracts the wedge block, the robot arm retracts, and the fault simulation robot arm upper arm resets along the connecting frame fixed on the power transformer winding support platform via the fault simulation robot arm base. 7) Rotate the clamping transformer winding fixing transmission knob to drive the clamping transformer winding fixing transmission gear and gear drive motor to start working, so that the clamping transformer winding fixing shims on the clamping transformer winding rotation transmission gear retract inward to control the high voltage winding fixing space position of the power transformer to remain stationary. 8) Repeat steps 2) to 7) to simulate axial loosening faults in the windings at different locations; Step two includes: 1) Define the connection point A connecting the upper bushings of the first, third, and fifth high-voltage windings as A; define the connection point B connecting the upper bushings of the second, fourth, and sixth low-voltage windings. 2) Connect transformer winding connection points A and B to a high-frequency high-voltage switch and a high-frequency high-voltage DC power supply. The high-frequency high-voltage switch will perform periodic operations, thereby stimulating the vibration of the UHV transformer body structure. When using traditional frequency analysis methods to analyze the components of vibration test signals from multiple vibration sources such as UHV transformer windings, cores, and cooling systems, spurious signals and frequencies may occur. Therefore, it is necessary to use instantaneous frequency to represent the local characteristics of the signal. Thus, the ITD transform is used to establish the signal model. ； Where L is the baseline extraction operator, X t For the input signal, L t =LX t H is the baseline signal. t =(1-L)X t For inherent rotational components, For the corresponding time; 3) The next step is to use Empirical Mode Decomposition (EMD) to decompose the signal into the sum of several mutually orthogonal Intrinsic Mode Function (IMF) components. Then, Hilbert transform is applied to each IMF component to obtain the instantaneous frequency and instantaneous amplitude, thereby obtaining the Hilbert spectrum of the signal for decomposition and reconstruction, thus improving the signal-to-noise ratio. ； ； Where X and Y are different signal representations, H(X,Y) is the joint entropy, I(X,Y) is the definition of mutual information of the signals, p(x(i)) and p(y(j)) are the probability distributions of X and Y, and p(x(i),y(j)) is the joint probability. 4) Further reconstruct the required signal and reduce on-site noise interference: Since the IMF can characterize the dependency between two data, mutual information can be used to filter the obtained IMF components of each order after the EMD decomposes the non-steady-state signal. ； ； Where u1(t) and u2(t) are the upper and lower envelopes connecting the local maxima and local minima, respectively; m(t) is the average of the upper and lower envelopes; h(t) = x(t) - m(t). If h(t) does not satisfy the IMF condition, it is considered as a new x(t), and this is repeated k times to obtain h1k(t); SD is the standard deviation, used to determine whether the screening process terminates. Finally, C1 = h1k(t) and r1(t) = x(t) - C1 are obtained. 5) Collect the processed data and designate it as the original data X. i ={x1,x2,…,x N Given an m-dimensional vector of length N, with a pre-defined embedding dimension m and similarity capacity r, consider the following: Let's define the distance d[x(i),x(j)] between x(i) and x(j) as the maximum difference between their corresponding elements, i.e. ; 6) Calculate the distance d[x(i), x(j)] between x(i) and the remaining vectors x(j) (j = 1, 2, ..., N - m, j ≠ i) using the sample entropy method; count the number of vectors with d[x(i), x(j)] less than r and the ratio of this number to the total number of distances Nm-1, denoted as r. ,Right now: ； 7) Next, the calculations are as follows: Average value: Then, for dimension m+1, i.e., for vector m+1, repeat steps 7) and 8) to obtain... Then, its average value was calculated to obtain ; 8) Finally, the sample entropy calculation formula shown below is used to solve the winding vibration signal after simulating deformation fault, and this is used to determine the severity of the winding deformation fault, as shown below: When N is a finite number, the expression can be represented as: ； 9) The magnitude of the sample entropy is related to the values ​​of m and r. We generally take m=2 and r=0.2SD, where SD is the standard deviation of the original data. like This indicates that the winding is in a normal state. like This indicates that the winding has undergone slight deformation. like This indicates that the winding has undergone severe deformation.

Images