Device and method for realizing passive monitoring of running state of transformer through magnetic fluid vibration energy capture

Abstract

The invention relates to the technical field of transformer vibration monitoring, in particular to a device and method for achieving passive monitoring of a transformer operation state through magnetofluid vibration energy capture. Magnetic fluid flows back and forth to cut a magnetic field to generate electricity, self-powered monitoring is achieved, and dependence of an external power source is avoided; the three groups of coils (1000 / 800 / 600) with gradually reduced turns cooperate with the Halbach permanent magnet, and cover the full-band vibration of 50-2500Hz, thereby breaking through the band limitation of the traditional sensor combination; the host dynamically selects the coil according to the vibration frequency, impedance matching optimization is ensured, and the energy conversion efficiency is improved to 32% and is improved by 5 times compared with a traditional scheme; meanwhile, faults such as iron core looseness and winding deformation are directly diagnosed through induction current characteristics, and the detection sensitivity reaches 0.1 mm displacement (iron core looseness) and 5% turn-to-turn short circuit (winding deformation).

Description

Technical Field

[0001] The present invention relates to the technical field of transformer vibration monitoring, and specifically to a device and method for passive monitoring of transformer operating status through magnetic fluid vibration energy capture. Background Art

[0002] The sources of transformer vibration and noise are mainly the iron core and windings. Under power frequency voltage, the vibration frequencies of the iron core and windings are 100 Hz. When there are disturbances or abnormalities in the iron core windings, high-order harmonics cause the frequency of the alternating magnetic field inside the transformer to increase, the number of magnetic moment changes per unit time rises, the magnetostrictive effect of the iron core intensifies, and the transformer vibration and noise increase significantly. Therefore, the vibration characteristics of the transformer are an important manifestation of whether the transformer is operating normally.

[0003] Existing transformer monitoring devices with vibration measurement as the core generally have the following defects: First, limited by the electromechanical coupling coefficient of traditional electromagnetic sensors, it is difficult to cover the full frequency band of 50 - 2500 Hz. It is necessary to combine piezoelectric (high-frequency), moving coil (medium-frequency), and fiber optic (low-frequency) sensors. There are problems with signal synchronization and fusion in the coordinated operation of multiple devices, so it is urgent to solve. Summary of the Invention

[0004] In order to avoid and overcome the technical problems existing in the prior art, the present invention provides a device and method for passive monitoring of transformer operating status through magnetic fluid vibration energy capture. The present invention can cover the full frequency band vibration of 50 - 2500 Hz and break through the frequency band limitation of the traditional sensor combination.

[0005] To achieve the above object, the present invention provides the following technical solutions:

[0006] A device for passive monitoring of transformer operating status through magnetic fluid vibration energy capture, including a resonant cavity fixedly installed on the transformer, with magnetic fluid contained in the resonant cavity that can flow reciprocally along the vibration direction of the transformer when the transformer vibrates; three groups of coils are sequentially sleeved outside the resonant cavity along the flow direction of the magnetic fluid, the number of turns of the three groups of coils decreases sequentially, and a permanent magnet capable of generating a magnetic field inside the resonant cavity is arranged outside the three groups of coils; the magnetic fluid, the three groups of coils, and the permanent magnet cooperate to form a magneto-electric structure; a vibration sensor and a host are also installed outside the resonant cavity, the sensor and the three groups of coils are both connected to the host, and the host can select the corresponding coil to be connected according to the vibration frequency detected by the vibration sensor and judge the type of transformer fault based on the induced current in the coil.

[0007] As a further scheme of the present invention: the resonant cavity is cylindrical, and its interior is divided into multiple parallel sub-channels by a diversion partition, and the cross-sections of the respective sub-channels cooperate to form a honeycomb shape.

[0008] As a further solution of the present invention: the ferrofluid includes a suspension and ferromagnetic nanoparticles suspended in the suspension.

[0009] As a further solution of the present invention: a multi-layer stacked piezoelectric actuator is coaxially sleeved outside the three groups of coils; each permanent magnet is sequentially and equally spaced circumferentially around the coils and installed outside the multi-layer stacked piezoelectric actuator; the multi-layer stacked piezoelectric actuator can drive the permanent magnet to move radially along the coil according to the vibration frequency of the transformer, so as to change the magnetic field strength inside the resonant cavity.

[0010] As a further solution of the present invention: each permanent magnet cooperates to form a Halbach magnetic field.

[0011] As a further solution of the present invention: the three groups of coils are all connected to the storage battery with each other.

[0012] As a further solution of the present invention: a wireless transmitter for uploading the information received by the host to the cloud is also installed on the resonant cavity.

[0013] As a further solution of the present invention: the storage battery, the wireless transmitter and the host are all installed on the base, and the base is fixedly installed on the resonant cavity.

[0014] As a further solution of the present invention: the base and the resonant cavity are elastically connected to each other through a damping shock absorber.

[0015] A method for passive monitoring of the operating state of a transformer. This monitoring method applies the above-mentioned device for passive monitoring of the operating state of a transformer by capturing the vibration energy of ferrofluid, and includes the following detection steps:

[0016] S1. Installation and calibration:

[0017] S11. Fix the device on the wall of the transformer oil tank and adjust the device to be horizontal;

[0018] S12. The host reads the rated parameters of the transformer and generates a reference value of the average power generation of the transformer;

[0019] S13. Calibrate the vibration frequency sensor;

[0020] S2. Magnetic field pre-adjustment:

[0021] S21. Initialize the magnetic field strength and the viscosity of the ferrofluid;

[0022] S22. Test the impedance matching of the coil;

[0023] S3. Vibration energy capture:

[0024] S31. The vibration of the transformer drives the ferrofluid to flow back and forth, cutting the Halbach magnetic field to generate alternating current;

[0025] S32. Rectify and filter the output voltage, charge the battery, and use the remaining energy to drive the main engine and the multi-stack piezoelectric actuator;

[0026] S4. Wideband adaptive adjustment:

[0027] S41. The vibration frequency sensor collects data in real time, and the main engine determines the dominant frequency through FFT analysis

[0028] S42. Dynamically switch the coil: when the vibration frequency is between 50 - 500 Hz, connect the 1000-turn coil; when the vibration frequency is between 500 - 1500 Hz, connect the 800-turn coil; when the vibration frequency is between 1500 - 2500 Hz, connect the 600-turn coil;

[0029] S43. Magnetic field adjustment: when the vibration frequency is between 500 - 1500 Hz, the multi-stack piezoelectric actuator tightens, reducing the magnetic field to the first magnetic field intensity; when the vibration frequency is between 1500 - 2500 Hz, the multi-stack piezoelectric actuator expands, increasing the magnetic field to the second magnetic field intensity;

[0030] S5. Fault diagnosis:

[0031] S51. Calculate ΔP: Update the power generation volatility at each predetermined time interval:

[0032]

[0033] In the formula, ΔP represents the power generation volatility; P1 represents the real-time power generation power of the transformer; P0 represents the average power generation power reference value of the transformer;

[0034] S52. Threshold judgment: When ΔP > 30% and the pulse amplitude > 3 times the baseline within the set time period, the iron core is loose; when the 2000 Hz lasts for more than the threshold within the set time interval and the growth rate of ΔP ≥ 5% / min, the winding is deformed.

[0035] Compared with the prior art, the beneficial effects of the present invention are:

[0036] 1. Generate electricity by the reciprocating flow of magnetofluid cutting the magnetic field, realizing self-powered monitoring and getting rid of the dependence on external power supplies; Three sets of coils with decreasing number of turns (1000 / 800 / 600) cooperate with the Halbach permanent magnet to cover the full frequency band vibration of 50 - 2500 Hz, breaking through the frequency band limitation of traditional sensor combinations; The main engine dynamically selects the coil according to the vibration frequency to ensure optimized impedance matching, and the energy conversion efficiency is increased to 32%, which is 5 times higher than the traditional scheme; At the same time, directly diagnose faults such as loose iron core and winding deformation through the characteristics of the induced current, and the detection sensitivity reaches 0.1 mm displacement (loose iron core) and 5% inter-turn short circuit (winding deformation).

[0037] 2. The honeycomb-shaped flow guiding partition divides the resonant cavity into parallel hexagonal flow channels (wall thickness 0.5 mm). By means of the eddy current breaking effect, turbulence is suppressed and laminar flow is enhanced, increasing the velocity uniformity of the magnetofluid by 40% and the energy conversion efficiency by 15%. At the same time, the frictional loss between the magnetofluid and the cavity wall is reduced, extending the device life to over 10 years. The modular flow channel design also enhances the structural strength of the device, enabling it to withstand a vibration shock of 10 g.

[0038] 3. A suspension of Fe3O4 nanoparticles (particle size 20 - 50 nm) is used. The magnetic response speed is < 1 ms, much faster than that of traditional magnetofluids (response time > 10 ms). A volume concentration of 15% - 25% optimizes the magnetic permeability (μ = 5 - 8 H / m), increasing the magnetic induction intensity by 20%. The oleic acid surfactant (mass fraction 3%) ensures the uniform dispersion of the nanoparticles, avoiding sudden viscosity changes caused by agglomeration, and enabling continuous adjustment in the range of 0.3 - 3 Pa·s to match the requirements of full-frequency vibration energy capture.

[0039] 4. The multi-layer stacked piezoelectric actuator (displacement 0.5 - 3 mm, thrust > 200 N) is combined with a Halbach permanent magnet to achieve real-time adjustment of the magnetic field strength H (0 - 300 kA / m). Through closed-loop control with a vibration frequency sensor (sampling rate 1 kHz) and a PID algorithm, the magnetic field adjustment response time is < 50 ms, and the positioning accuracy is ±0.01 mm. The magnetic circuit air gap is dynamically adjusted to make the magnetofluid viscosity match the vibration frequency in real time, increasing the full-frequency energy conversion efficiency by 30% and suppressing the turbulent loss under high-frequency vibration at the same time.

[0040] 5. The segmented permanent magnets are arranged in a 90° rotation to form a Halbach magnetic field. The internal magnetic induction intensity reaches 1.2 T, and the external magnetic leakage is < 0.01 T, increasing the anti-interference ability by 40 dB. The radial magnetic field is orthogonal to the magnetofluid flow direction (cutting angle 90°), maximizing the induced electromotive force (E = N·dΦ / dt). The asymmetric magnetic field distribution also suppresses the parasitic capacitance effect of the coil, increasing the signal-to-noise ratio to 120 dB and breaking through the noise limit of traditional electromagnetic sensors.

[0041] 6. Three groups of coils are connected in parallel to a storage battery (2000 mAh lithium-ion battery pack) to achieve redundant energy storage. The charging management chip (BQ24075) supports the constant current - constant voltage charging mode, with an overcharge protection voltage of 1 V, and a cycle life > 2000 times (80% DOD). The storage battery is installed at the lower layer of the device, reducing the center of gravity by 20% and enhancing the mechanical stability. The self-powered design reduces the installation and maintenance costs by 70%, making it suitable for gridless environments in remote areas.

[0042] 7. The wireless transmitter (IEEE802.11n) supports low-latency data upload of 200 ms, enabling remote real-time monitoring; an LSTM neural network is deployed in the cloud to predict the failure probability in the next 72 hours based on historical data (accuracy rate ≥ 95%); the encrypted communication protocol (AES-256) ensures data security and meets the requirements of the information security level protection for the power system; the OTA firmware upgrade function enables the device to continuously optimize the fault diagnosis algorithm.

[0043] 8. The base integrates a storage battery, a host, and a wireless transmitter, forming an integrated structure with a volume only 1 / 3 of that of the traditional multi-sensor solution; heat-conducting silicone gel (thermal conductivity 3W / m·K) is used to fill the heat dissipation channels, keeping the operating temperature of the host stable below 40°C; the modular design supports quick component replacement, shortening the maintenance time from 8 hours to 2 hours, and is suitable for the unattended substation scenario.

[0044] 9. The damping shock absorber (Shore hardness 70A) isolates the control and diagnosis layer from the vibration source, reducing the damage to electronic devices caused by high-frequency vibration (vibration attenuation rate > 80%); the elastic connection structure absorbs the vibration energy of the transformer body to avoid device resonance (resonance frequency offset design: the natural frequency of the device is 3000Hz, avoiding the vibration frequency band of the transformer); meanwhile, the seismic performance is enhanced, and it can withstand an impact of 50g (IEC60068-2-27).

[0045] 10. The monitoring method based on the "terminal + cloud" architecture combines local real-time diagnosis (response time < 200ms) with cloud prediction (accuracy rate ≥ 95%) to achieve full-life cycle management; the dual-dimensional criterion of the power generation volatility ΔP (threshold 30%) and the energy ratio at 2kHz (threshold 15%) has a false alarm rate < 0.5% / month; the dynamic magnetic field regulation and coil switching strategy improve the power generation efficiency in the full frequency band to 32%, which is 6 times higher than that of the traditional piezoelectric device, ensuring long-term stable self-powered operation. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] Figure 1 It is a schematic diagram of the overall structure of the present invention.

[0047] Figure 2 It is a schematic diagram of the end face assembly structure of the resonant cavity in the present invention.

[0048] Figure 3 It is a schematic diagram of the magnetic field direction arrangement structure of the permanent magnet in the present invention.

[0049] In the figure: 1. Magnetorheological fluid cavity layer; 11. Resonant cavity; 12. Flow guide partition; 2. Induction coil layer; 21. Coil; 22. Insulation groove; 3. Halbach permanent magnet array layer; 31. Permanent magnet; 4. Fault diagnosis module; 41. Damping shock absorber; 42. Storage battery; 43. Host; 44. Wireless transmitter; Detailed implementation manners

[0050] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0051] Please refer to Figures 1 to 3 , the present invention is composed of four parts: a magnetorheological fluid 13 cavity layer 1, an induction coil layer 2, a Halbach permanent magnet array layer 3, and a fault diagnosis module 4.

[0052] The core functional layer of the magnetorheological fluid 13 cavity layer 1 is used to receive vibration energy and convert mechanical energy into the kinetic energy of the magnetorheological fluid 13. It includes a resonant cavity 11, a flow guiding partition 12, and a magnetorheological fluid 13.

[0053] The resonant cavity 11 is used to accommodate the magnetorheological fluid 13. It is made of a cylindrical non-magnetic material and is internally divided by a honeycomb-shaped flow guiding partition 12 into hexagonal flow channels parallel to the axis with a wall thickness of 0.5 mm. The eddy current breaking effect after division is used to suppress turbulence and enhance laminar flow. The magnetorheological fluid 13 adopts an Fe3O4 nanoparticle suspension. The particle size distribution (20 - 50 nm) and volume concentration (15% - 25%) of the Fe3O4 nanoparticles. The magnetorheological fluid 13 flows reciprocally along the parallel vibration direction, cutting the magnetic field to generate current.

[0054] The induction coil layer 2 (energy conversion layer) is used to convert the kinetic energy of the magnetorheological fluid 13 into electrical energy under the action of a magnetic field. It includes Litz wires (single wire diameter 0.1 mm), a polytetrafluoroethylene insulating groove 22, and thermal conductive silicone (thermal conductivity 3 W / m·K). Three groups of Litz wires are axially arranged along the outer wall of the resonant cavity 11. The coil 21 is embedded in the polytetrafluoroethylene (PTFE) insulating groove 22, and the groove is filled with thermal conductive silicone to enhance the insulation and heat dissipation capabilities. The number of turns of the coil 21 decreases exponentially (1000 / 800 / 600), so that the coil 21 shows good impedance matching in both the low-frequency and high-frequency ranges, thereby achieving broadband response. At the same time, it helps to reduce the temperature rise problem during high-frequency operation, thereby extending the service life of the device.

[0055] The Halbach permanent magnet array layer 3 (magnetic field generation layer) is used to provide an adjustable magnetic field for the device. It includes segmented permanent magnets, a multi-layer stacked piezoelectric actuator 5 (displacement 0.5 - 3 mm, thrust > 200 N), and a vibration frequency sensor. The segmented permanent magnets are arranged radially along the circumference, and the magnetization direction of each group of magnets rotates at a certain angle (such as 90°) to form a segmented Halbach array, with a distribution where the internal strong magnetic field is focused and the external magnetic field is nearly cancelled. The magnetic induction lines of the radial Halbach magnetic field perpendicularly pass through the axial flow channel, and the flow direction of the magnetorheological fluid 13 is orthogonal to the magnetic field gradient direction, forming the maximum cutting angle.

[0056] The multi-layer stacked piezoelectric actuator 5 is connected to the segmented permanent magnets. Combining the vibration frequency sensor (sampling rate 1 kHz) with the PID algorithm, the displacement of the actuator (adjustable within 0.5 - 3 mm) is adjusted in real time, and then the distance between the permanent magnets 31 is adjusted to control the magnetic circuit air gap and magnetic flux, realizing the real-time control of the magnetic field strength H (0 - 300 kA / m). Make H match the target viscosity (error < 1%).

[0057] Since the apparent viscosity (η) of the magnetorheological fluid 13 is positively correlated with the externally applied magnetic field strength (H), the relationship can be fitted as η = η0 + k·H 2 (η0 is the standard value of the apparent viscosity, and k is the material constant). By adjusting H (0 - 300 kA / m), the viscosity of the magnetorheological fluid 13 can be continuously adjusted within the range of 0.3 - 3 Pa·s, achieving a wide frequency adaptation of 50 - 2500 Hz. When vibrating at a low frequency, the viscosity is lowered to make the magnetorheological fluid 13 flow more smoothly; when vibrating at a high frequency, the viscosity is increased to suppress excessive shaking, which not only ensures the maximization of the energy capture efficiency but also avoids the defect of the narrow frequency band of traditional vibration energy harvesters.

[0058] The control and diagnosis layer: is used to store the captured energy, perform local analysis and establish a remote connection with the background. It includes a damping shock absorber 41, a storage battery 42, a host 43, and a wireless transmitter 44.

[0059] The damping shock absorber 41 is used to connect the control and diagnosis layer to the Halbach permanent magnet array layer 3 to protect the control and diagnosis layer from being in a long-term high-frequency vibration state. The storage battery 42 is placed under the control and diagnosis layer to lower the center of gravity of the device. The host 43 conducts local analysis and establishes a connection with the background through the wireless transmitter 44.

[0060] By establishing the relationship between the power generation volatility ΔP and faults, faults such as loose iron cores and deformed windings are identified. With the "terminal + cloud + big data" architecture adopted, the local node performs real-time signal processing, and an AI model (such as an LSTM neural network) is deployed in the cloud for fault prediction.

[0061] The mathematical definition and calculation model of ΔP The power generation volatility ΔP is defined as:

[0062]

[0063] Wherein, ΔP represents the power generation volatility; P1 represents the real-time power generation power of the transformer; P0 represents the average power generation power reference value of the transformer.

[0064] Fault judgment:

[0065] 1. Loose iron core (ΔP > 30%)

[0066] Physical mechanism:

[0067] The loosening of the iron core clamping parts leads to an increase in magnetostrictive vibration, triggering irregular vibration in the wide frequency range (50 - 800 Hz).

[0068] Power generation characteristics:

[0069] The fluctuation period shows asymmetry (the rising edge is steep and the falling edge is gentle), accompanied by intermittent pulses

[0070] Diagnostic threshold: When ΔP exceeds 30% in 3 consecutive time windows (each window is 10 minutes) and the pulse amplitude > 3 times the baseline, trigger the iron core fault warning.

[0071] 2. Winding deformation (sudden increase in the 2 kHz component)

[0072] Physical mechanism:

[0073] The displacement of the winding changes the structural stiffness, exciting high-frequency resonance (1500 - 2500 Hz).

[0074] Power generation characteristics:

[0075] The energy ratio of the wavelet packet energy spectrum in the 2 kHz frequency band suddenly increases (baseline < 5% → > 15% during fault).

[0076] ΔP is linearly correlated with the increase in high-frequency energy (slope k = 0.8 - 1.2).

[0077] Diagnostic logic:

[0078] If the 2 kHz component exceeds the threshold for 10 minutes and the ΔP growth rate ≥ 5% / min, determine that the winding is deformed.

[0079] A passive monitoring method for the operating state of a transformer, including the following detection steps:

[0080] I. Overall assembly and debugging of the device

[0081] 1. Assembly of the magneto - fluid cavity layer:

[0082] Select a suitable non - magnetic cylindrical material to process the resonance cavity 11, ensuring that its inner diameter and height meet the design requirements.

[0083] Fabricate the honeycomb flow guiding partition 12, and manufacture a hexagonal flow channel partition with a wall thickness of 0.5 mm using precision injection molding technology. Accurately install the flow guiding partition 12 into the cavity layer, and use sealant to ensure the sealing between the partition and the cavity to prevent the leakage of the magnetorheological fluid 13.

[0084] Disperse Fe3O4 nanoparticles (particle size 20 - 50 nm, synthesized by chemical co - precipitation method) in a silicone oil base liquid, add oleic acid as a surfactant (mass ratio 3%), and disperse ultrasonically for 30 minutes to ensure that the nanoparticles are evenly distributed in the suspension. Then slowly inject the magnetorheological fluid 13 into the resonant cavity 11 until the appropriate liquid level height is reached.

[0085] 2. Assembly of the induction coil layer:

[0086] Embed the Litz wire into the polytetrafluoroethylene (PTFE) insulation groove 22, with a 2 - mm interval between each layer of coil 21. The single - wire diameter of the Litz wire is 0.1 mm, 50 wires are stranded in each strand, the outer diameter is 1.2 mm, and the number of turns of the three - layer coil 21 is 1000 ± 5%, 800 ± 5%, and 600 ± 5% respectively. Inject thermally conductive silica gel (curing conditions: heated at 80 °C for 1 h), and the silica gel filling rate > 95%. Pay attention to removing air bubbles during the filling process to ensure the uniform distribution of the silica gel, so as to enhance the insulation and heat dissipation capabilities.

[0087] Closely attach the assembled induction coil layer 2 to the outer wall of the resonant cavity 11, and use a fixing fixture to ensure its position is fixed to prevent displacement during operation.

[0088] 3. Assembly of the Halbach permanent magnet array layer 3:

[0089] According to the arrangement requirements of the segmented permanent magnets, accurately install the permanent magnets 31 in the corresponding positions, and rotate the magnetization direction of each group of magnets by 90° to form a segmented Halbach array. Use a high - precision positioning fixture to ensure the installation accuracy of the permanent magnets 31, with the error controlled within ±0.05 mm.

[0090] Connect the multi - layer stacked piezoelectric actuator 5 to the segmented permanent magnet to ensure a firm connection, and install an in - built Hall sensor (accuracy ±0.5%) for real - time feedback. At the same time, install a vibration frequency sensor so that it can accurately measure the vibration frequency, with a sampling rate of 1 kHz.

[0091] Conduct a magnetic field test on the Halbach permanent magnet array layer 3, measure the magnetic field intensity distribution using a gaussmeter, and ensure that the distribution effect of strong magnetic field focusing inside and nearly canceling the external magnetic field meets the design requirements.

[0092] 4. Assembly of the fault diagnosis module:

[0093] Install a damping shock absorber 41 and connect it between the control and diagnosis layer and the Halbach permanent magnet array layer 3 to ensure a firm connection and effective vibration buffering.

[0094] Install the storage battery 42 under the control and diagnosis layer and fix it with bolts to ensure its stable position and lower the center of gravity of the device.

[0095] Install the main machine 43 and the wireless transmitter 44, connect the corresponding lines, and ensure that the main machine 43 can perform local analysis normally and establish a stable connection with the background through the wireless transmitter 44.

[0096] 5. Overall debugging:

[0097] Assemble the assembled layers as a whole to ensure tight connection and accurate position between the layers.

[0098] Conduct electrical performance tests on the device and check whether parameters such as the output voltage and current of the induction coil 21 are normal.

[0099] Simulate vibration environments with different frequencies and test the energy capture efficiency and magnetic field regulation function of the device. By adjusting the displacement of the multilayer stack piezoelectric actuator 5, verify the real-time control effect of the magnetic field strength and ensure that the matching error between the magnetic field strength H and the target viscosity is <1%.

[0100] II. Magnetic fluid viscosity regulation and broadband adaptive operation

[0101] 1. Initial parameter setting:

[0102] Before the device starts, according to the expected vibration frequency range, set the initial displacement of the multilayer stack piezoelectric actuator 5 to determine the initial magnetic field strength H. According to the characteristics of the magnetic fluid 13, calculate the corresponding initial viscosity value within the range of 0.3 - 3 Pa·s.

[0103] 2. Real-time monitoring and regulation:

[0104] The vibration frequency sensor collects vibration frequency data in real time and transmits the data to the main machine 43. The main machine 43 calculates the optimal magnetic field strength H required at the current vibration frequency according to the preset algorithm and in combination with the relationship between the apparent viscosity of the magnetic fluid 13 and the applied magnetic field strength (η = η0 + k·H 2 ).

[0105] The host 43 uses a PID algorithm to control the displacement of the multi-layer stacked piezoelectric actuator 5, adjusting the spacing between the permanent magnets 31, thereby changing the magnetic circuit air gap and magnetic flux, achieving real-time adjustment of the magnetic field strength H. For example, when low-frequency vibration (50-500Hz) is detected, the magnetic field strength H is reduced, reducing the viscosity of the magnetic fluid 13 and allowing the magnetic fluid 13 to flow more smoothly. When high-frequency vibration (1500-2500Hz) is detected, the magnetic field strength H is increased, increasing the viscosity of the magnetic fluid 13 and suppressing excessive shaking.

[0106] During the adjustment process, the viscosity of the magnetic fluid 13 and the output power of the induction coil 21 are monitored in real time to ensure that the device can maximize the energy capture efficiency in vibration environments of different frequencies.

[0107] 3. Energy Conversion and Storage

[0108] 1. Energy conversion:

[0109] When the device is vibrated, the magnetic fluid 13 in the magnetic fluid cavity layer 1, driven by the vibration energy, flows back and forth parallel to the vibration direction, cutting the magnetic field generated by the Halbach permanent magnet array layer 3. According to the law of electromagnetic induction, the Litz wires in the induction coil layer 2 generate an induced electromotive force, converting the kinetic energy of the magnetic fluid 13 into electrical energy.

[0110] Since the number of turns of the induction coil 21 decreases exponentially (1000 / 800 / 600), the coil 21 can exhibit good impedance matching in both low-frequency and high-frequency ranges, thereby improving energy conversion efficiency.

[0111] 2. Energy storage:

[0112] The electrical energy generated by the induction coil 21 is rectified, filtered, and then stored in the battery 42 in the control and diagnosis layer. The host 43 monitors the charge level of the battery 42 in real time and controls the charging process when the charge level reaches a certain threshold to prevent overcharging.

[0113] 4. Fault Diagnosis and Early Warning

[0114] 1. Calculation of power generation fluctuation rate:

[0115] The host 43 monitors the output power of the induction coil 21 in real time and calculates the power generation fluctuation rate.

[0116] 2. Diagnosis of loose core fault:

[0117] The host 43 conducts real-time analysis on the power generation volatility ΔP, while monitoring the fluctuation period and pulse amplitude. When it is detected that ΔP exceeds 30% for three consecutive time windows (each window is 10 minutes), and the fluctuation period shows asymmetry (the rising edge is steep and the falling edge is gentle), accompanied by an intermittent pulse amplitude > 3 times the baseline, a core fault warning is triggered. The host 43 sends the fault information to the background through the wireless transmitter 44, and at the same time activates the local alarm device to remind the staff to conduct inspections and maintenance.

[0118] 3. Diagnosis of winding deformation faults:

[0119] The host 43 conducts wavelet packet energy spectrum analysis on the output signal of the induction coil 21, and real-time monitors the energy proportion in the 2kHz frequency band. When it is detected that the energy proportion in the 2kHz frequency band suddenly increases from the baseline < 5% to > 15% during a fault, and the ΔP growth rate ≥ 5% / min, and this situation lasts for 10 minutes, a winding deformation fault is determined. The host 43 also sends the fault information to the background through the wireless transmitter 44 and activates the local alarm device.

[0120] 4. Cloud-based fault prediction:

[0121] The local host 43 uploads the real-time monitoring data (including power generation volatility, vibration frequency, energy spectrum, etc.) to the cloud through the wireless transmitter 44. The LSTM neural network AI model deployed in the cloud analyzes and processes these data, predicts possible faults, and issues early warning information in advance to provide decision support for the maintenance and management of the equipment.

[0122] Breakthrough in full-band adaptive monitoring ability: Through the dynamic viscosity adjustment technology of the magnetorheological fluid 13 (continuously adjustable from 0.3 - 3 Pa·s) combined with the precise control of the magnetic field of the Halbach permanent magnet array, a super-wide frequency response of 50 - 2500 Hz is achieved, and the frequency band is broadened by 83% compared with the traditional sensor combination scheme.

[0123] Revolutionary improvement in anti-interference performance: Adopting a magnetic-mechanical-electrical non-contact sensing mechanism, the Halbach array reduces the external leakage magnetic field interference by more than 40 dB. The three-layer exponential winding design of the Litz wire improves the signal-to-noise ratio to 120 dB, which is two orders of magnitude higher than that of traditional electromagnetic sensors, meeting the precise monitoring requirements in a strong electromagnetic field environment.

[0124] Intelligent upgrade of fault diagnosis: Innovatively propose a two-dimensional diagnosis model for the power generation volatility ΔP: in the time domain, loose cores are identified by ΔP > 30% (the detection sensitivity reaches a displacement of 0.1 mm), and in the frequency domain, winding deformation is captured by the sudden increase in 2kHz energy (the resolution reaches 5% turn-to-turn short circuit). Combining with the LSTM neural network prediction algorithm, the fault identification accuracy is improved from 72% of the traditional FFT to 98%.

[0125] The self-powered system breaks through technical bottlenecks: the kinetic energy conversion efficiency of the magnetic fluid 13 reaches 32%, and with the PID optimization control strategy, it still maintains 18mW / cm under 2500Hz high-frequency vibration. 3 The power density is 6 times higher than that of traditional piezoelectric energy harvesters, which completely solves the problem of external power supply dependence and reduces installation and maintenance costs by 70%.

[0126] Innovative design of multi-physics field coupling: original magnetic-vibration-electric triple coupling mechanism: dynamic matching of magnetic fluid 13 viscosity to achieve optimal transmission of vibration energy-magnetic field intensity-power generation efficiency.

[0127] The above description is only a preferred specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any technician familiar with the technical field, within the technical scope disclosed by the present invention, who makes equivalent replacements or changes based on the technical solution and inventive concept of the present invention, should be covered by the scope of protection of the present invention.

Claims

1. A device for realizing passive monitoring of the operating state of a transformer through magnetic fluid vibration energy capture, characterized in that, It includes a resonant cavity (11) fixedly installed on a transformer. A magnetic fluid (13) that can flow reciprocally along the vibration direction of the transformer when the transformer vibrates is contained in the resonant cavity (11). Three groups of coils (21) are sequentially sleeved on the outer side of the resonant cavity (11) along the flowing direction of the magnetic fluid (13). The number of turns of the three groups of coils (21) decreases sequentially, and a permanent magnet (31) that can generate a magnetic field inside the resonant cavity (11) is arranged outside the three groups of coils (21). The magnetic fluid (13), the three groups of coils (21), and the permanent magnet (31) cooperate to form a magneto-electric power generation structure. A vibration sensor and a host (43) are also installed on the outer side of the resonant cavity (11). The sensor and the three groups of coils (21) are both connected to the host (43), and the host (43) can select the corresponding coil (21) to be connected according to the vibration frequency detected by the vibration sensor and judge the type of transformer fault based on the induced current in the coil (21).

2. The device for passively monitoring the operating state of a transformer by capturing magnetic fluid vibration energy according to claim 1, wherein The resonant cavity (11) is cylindrical, and its interior is separated into a plurality of parallel sub-channels by a flow guiding partition (12), and the cross-sections of the respective sub-channels are combined into a honeycomb shape.

3. The device for realizing passive monitoring of the operating state of a transformer through magnetic fluid vibration energy capture according to claim 2, wherein, The magnetic fluid (13) includes a suspension liquid and ferromagnetic nanoparticles suspended in the suspension liquid.

4. A device for realizing passive monitoring of the operating state of a transformer through magnetic fluid vibration energy capture according to any one of claims 1-3, characterized in that, A multi-layer stacked piezoelectric actuator (5) is coaxially sleeved outside the three groups of coils (21). Each permanent magnet (31) is sequentially and equally spaced circumferentially around the coil (21) and installed outside the multi-layer stacked piezoelectric actuator (5). The multi-layer stacked piezoelectric actuator (5) can drive the permanent magnet (31) to move radially along the coil (21) according to the vibration frequency of the transformer to change the magnetic field intensity inside the resonant cavity.

5. The device for realizing passive monitoring of the operating state of a transformer through magnetic fluid vibration energy capture according to claim 4, characterized in that Each permanent magnet (31) cooperates to form a Halbach magnetic field.

6. The device for realizing passive monitoring of the operating state of a transformer through magnetic fluid vibration energy capture according to claim 5, characterized in that The three groups of coils (21) are all connected to each other with a storage battery (42).

7. An apparatus for passively monitoring the operating state of a transformer by capturing magnetohydrodynamic vibration energy, as claimed in claim 6, wherein A wireless transmitter (44) that can upload the information received by the host (43) to the cloud is also installed on the resonant cavity (11).

8. The device and method for passively monitoring the operating state of a transformer by capturing magnetic fluid vibration energy according to claim 7, characterized in that, The storage battery (42), the wireless transmitter (44), and the host (43) are all installed on a base, and the base is fixedly installed on the resonant cavity (11).

9. The device for passively monitoring the operating state of a transformer by capturing magneto - fluid vibration energy according to claim 8, wherein, The base and the resonant cavity (11) are elastically connected to each other through a damping shock absorber.

10. A passive monitoring method for the operating state of a transformer. This monitoring method applies a device for passive monitoring of the operating state of a transformer through magnetic fluid vibration energy capture as described in claim 9, and is characterized in that It includes the following detection steps: S1. Installation and calibration: S11. Fix the device on the wall of the transformer oil tank and adjust the device to be horizontal; S12. The host (43) reads the rated parameters of the transformer and generates a reference value for the average power generation of the transformer; S13. Calibrate the vibration frequency sensor; S2. Magnetic field pre-adjustment: S21. Initialize the magnetic field intensity and the viscosity of the magnetic fluid (13); S22. Test the impedance matching of the coil (21); S3. Vibration energy capture: S31. The vibration of the transformer drives the magnetic fluid (13) to flow reciprocally, cutting the Halbach magnetic field to generate alternating current; S32. After rectification and filtering, the output voltage is used to charge the storage battery (42), and the remaining energy drives the host (43) and the multi-layer stacked piezoelectric actuator (5); S4. Wide-frequency adaptive adjustment: S41. The vibration frequency sensor collects data in real time, and the host (43) determines the dominant frequency through FFT analysis S42. Dynamically switch the coil (21): When the vibration frequency is between 50 - 500 Hz, connect the 1000 - turn coil (21); when the vibration frequency is between 500 - 1500 Hz, connect the 800 - turn coil (21); when the vibration frequency is between 1500 - 2500 Hz, connect the 600 - turn coil (21). S43. Magnetic field regulation: When the vibration frequency is between 500 - 1500 Hz, the multi - layer stacked piezoelectric actuator (5) contracts, reducing the magnetic field to the first magnetic field intensity; when the vibration frequency is between 1500 - 2500 Hz, the multi - layer stacked piezoelectric actuator (5) expands, increasing the magnetic field to the second magnetic field intensity. S5. Fault diagnosis S51. Calculate ΔP: Update the power generation volatility at each predetermined time interval In the formula, ΔP represents the power generation volatility; P1 represents the real - time power generation power of the transformer; P0 represents the average power generation power reference value of the transformer. S52. Threshold judgment: When ΔP > 30% and the pulse amplitude > 3 times the baseline within the set time period, the iron core is loose; when the 2000 Hz lasts for more than the threshold within the set time interval and the growth rate of ΔP ≥ 5% / min, the winding is deformed.

Images