A transformer vibration, temperature and moisture in oil synchronous monitoring optical fiber sensing system

Abstract

This invention discloses a fiber optic sensing system for simultaneous monitoring of transformer vibration, temperature, and oil moisture. It includes a vibration and temperature monitoring unit and an oil moisture monitoring unit. The vibration and temperature monitoring unit consists of a Fabry-Perot cavity formed by two identical fiber Bragg gratings (FBGs) deployed on the same optical fiber. The moisture monitoring unit is a phase-shifting grating coated with a multilayer polyimide film. The sensor is packaged as a 2mm thin structure, which can be deployed at the adjacent gaps of transformer windings. The monitoring method uses this fiber optic sensor to simultaneously acquire multiple physical field parameters of transformer winding vibration, temperature, and oil moisture. The temperature and vibration signals are calculated using the Bragg wavelength shift characteristic. The oil moisture content is correlated through the peak spacing of the dual gratings, and the results are calibrated with those from thermocouples, accelerometers, and Karl Fischer titrators. This invention achieves multi-parameter near-winding dynamic monitoring and has advantages such as resistance to electromagnetic interference, small size, and simplified wiring.

Description

Technical Field

[0001] This invention relates to the field of power equipment condition monitoring technology, specifically a fiber optic sensing system for synchronous monitoring of transformer vibration, temperature, and moisture content in oil. Background Technology

[0002] As the core hub for energy transmission and voltage transformation in the power grid, the operational reliability of power transformers directly determines the safety and stability of the power system. Statistics show that approximately 70% of transformer failures are related to insulation system deterioration. Winding vibration, abnormal temperature, and excessive moisture in the oil are the three key contributing factors to insulation aging, mechanical deformation, and even breakdown failures. Winding vibration causes insulation paper wear and winding displacement; excessively high temperatures accelerate insulation thermal decomposition; and moisture significantly reduces the breakdown strength and dielectric loss performance of oil-paper insulation. Therefore, real-time and accurate monitoring of these three parameters is a core requirement for ensuring the long-term safe operation of transformers.

[0003] However, existing transformer parameter monitoring technologies still have many technical bottlenecks, making it difficult to meet the needs of refined monitoring under complex operating conditions. These bottlenecks are manifested in the following aspects:

[0004] First, single-parameter monitoring has blind spots in fault coverage, making it difficult to achieve full-dimensional state perception. Traditional monitoring solutions mostly rely on a single type of sensor, which cannot comprehensively capture the multi-physical field changes of the insulation system. For temperature monitoring, thermocouples or platinum resistance sensors are commonly used, but these sensors need to be deployed on the top or side wall of the oil tank, far from winding hotspots (the temperature of winding hotspots is usually 10-15℃ higher than the top layer oil temperature), resulting in delayed measurements that deviate from actual operating conditions and failing to provide timely warnings of local overheating. For vibration monitoring, piezoelectric accelerometers are mainly used, but they are severely affected by the strong electromagnetic environment inside the transformer (leakage flux, transient overvoltage, etc.), resulting in a low signal-to-noise ratio. Furthermore, they require metal cables for wiring, easily introducing additional electromagnetic noise. Additionally, the accelerometers are large and cannot be deployed in critical areas such as winding gaps. For oil moisture monitoring, traditional methods mainly use Karl Fischer titrators, requiring offline oil sample collection. This cannot reflect real-time dynamic changes in moisture (such as moisture generated by insulation aging or moisture seeping in due to seal failure), resulting in monitoring lag. Moreover, the sampling process is prone to introducing external contamination, affecting measurement accuracy.

[0005] Secondly, electrical sensors have weak anti-interference capabilities and insufficient performance in adapting to the strong electromagnetic environment of transformers. During transformer operation, the windings and core generate strong electromagnetic radiation (power frequency magnetic field strength can reach 10-50 mT, and transient overvoltages can reach several times the rated voltage). The signal transmission and measurement principles of traditional electrical sensors (such as piezoelectric accelerometers and capacitive moisture sensors) are easily affected by electromagnetic interference. On the one hand, electromagnetic radiation induces interference current in the sensor cables, leading to distorted measurement signals; on the other hand, electromagnetic noise can mask weak fault characteristic signals (such as small amplitude changes in early winding vibrations and capacitance shifts caused by trace amounts of moisture), causing a significant decrease in the measurement accuracy of such sensors in the near-winding region, and even resulting in false alarms or missed alarms.

[0006] Third, multi-parameter monitoring systems suffer from complex wiring and a lack of data synchronization mechanisms. Some advanced solutions attempt to integrate multiple types of sensors (such as temperature, vibration, and moisture), but this requires separate cabling (power and signal cables) for each type of sensor. However, the internal space of a transformer tank is limited and requires strict sealing. Dense wiring not only increases the number of openings in the tank (weakening structural strength and increasing the risk of leakage) but also leads to cross-interference between cables. More importantly, the sampling frequencies and time bases of different sensors are inconsistent, making it impossible to achieve spatiotemporal synchronous correlation of vibration, temperature, and moisture parameters. This makes it difficult to analyze the coupling effects between the three types of parameters (such as increased vibration leading to insulation wear, which in turn causes moisture infiltration and temperature rise), and makes it impossible to dynamically reverse the degradation process of the insulation system.

[0007] Fourth, the lack of near-winding monitoring capability leads to significant deviations between measured data and actual operating conditions. Existing sensors are mostly deployed on the outer wall or top of the oil tank, creating a spatial distance from key monitoring objects such as the windings and insulating oil. This results in signal attenuation or distortion during propagation. For example, when winding vibrations travel through the insulating oil to the tank wall, the amplitude decreases by 30%-50%, and the tank's own vibration noise is also incorporated. Furthermore, moisture in the oil exhibits a concentration gradient within the tank, with the moisture content of oil samples at the top differing from that near the windings by as much as 20%-30%. This indirect monitoring method prevents the measured data from accurately reflecting the actual condition around the windings, hindering the precise diagnosis of early faults.

[0008] In summary, existing transformer parameter monitoring technologies face four major problems: incomplete coverage, poor anti-interference capability, difficulty in synchronization, and insufficient proximity. There is an urgent need for a monitoring solution that can achieve near-winding deployment, simultaneous acquisition of multiple parameters, resistance to electromagnetic interference, and simplified wiring, in order to meet the needs of full life-cycle health management of transformer insulation systems. Summary of the Invention

[0009] The purpose of this invention is to provide a fiber optic sensing system for synchronous monitoring of transformer vibration, temperature, and moisture content in oil, including a vibration and temperature monitoring unit and an oil moisture monitoring unit.

[0010] The vibration temperature monitoring unit and the oil moisture monitoring unit are integrated into the same optical fiber carrier;

[0011] The vibration temperature monitoring unit includes two identical fiber Bragg gratings, which together form a Fabry-Perot cavity.

[0012] The oil moisture monitoring unit is a phase-shifting grating coated with a multilayer polyimide film; the polyimide film is hygroscopic.

[0013] Furthermore, the spacing between the two fiber Bragg gratings is set to 5 mm - 10 mm;

[0014] When the center wavelength laser beam is incident, the beams reflected by the two fiber Bragg gratings produce a π phase difference, forming interference fringes.

[0015] Furthermore, the fiber optic sensing system is embedded in the gaps between adjacent transformer windings to monitor transformer vibration, temperature, and moisture in the oil.

[0016] Furthermore, the thickness of the polyimide coating of the oil moisture monitoring unit is adjusted according to the moisture sensitivity requirements.

[0017] A method for multi-parameter monitoring of a transformer based on the aforementioned fiber optic sensor includes the following steps:

[0018] Step 1) Deploy the fiber optic sensor at the adjacent gap of the transformer winding so that the fiber optic sensor is in contact with the winding surface;

[0019] Thermocouples are deployed inside the transformer tank, and the temperature data monitored by the thermocouples is used as the temperature calibration reference.

[0020] Accelerometers are installed on the top and side walls of the transformer tank, and the data monitored by the accelerometers is used as the vibration calibration benchmark.

[0021] Oil samples are periodically collected through the sampling valve in the oil tank, and the water content in the oil samples is measured using a Karl Fischer titrator as a water calibration benchmark.

[0022] Step 2) Use a distributed feedback scanner to send a center wavelength laser beam to the Fabry-Perot cavity of the fiber optic sensor;

[0023] Step 3) Use a distributed feedback scanner to collect the reflection interference signal from the vibration temperature monitoring unit and the phase-shifting grating reflection spectrum from the oil moisture monitoring unit;

[0024] Step 4) Based on the temperature calibration reference, vibration calibration reference, and moisture calibration reference, the reflection interference signal and the phase shift grating reflection spectrum are calculated to obtain the winding temperature, winding vibration, and moisture content in the oil.

[0025] Step 5) When the winding temperature, winding vibration and / or moisture in the oil exceed the corresponding threshold, an early warning is triggered.

[0026] Furthermore, the distance between the thermocouple and the fiber optic sensor is ≤5mm;

[0027] The accelerometer is positioned at the center of the top of the fuel tank and at the midpoint of the side wall.

[0028] Furthermore, the formula for calculating the winding temperature is as follows:

[0029]

[0030]

[0031] Where, α s ζ is the thermal expansion coefficient of optical fiber. s Thermo-optic coefficient; The wavelength is the center wavelength of the Bragg grating; This represents the change in Bragg wavelength caused by temperature changes; This represents the change in winding temperature. This serves as the temperature calibration reference; T represents the winding temperature.

[0032] Furthermore, the winding vibration amplitude is shown below:

[0033]

[0034] Where, ΔI m ΔI represents the measured change in light intensity. v K represents the effective vibrational light intensity change after temperature compensation. T is the temperature-light intensity compensation coefficient. k is the calibration coefficient determined from the experimental data; A is the winding vibration amplitude.

[0035] Furthermore, the water content in the oil is shown below:

[0036]

[0037] In the formula, a, b, and c are the calibration coefficients determined in the experiment; H represents the peak spacing between the phase-shifting grating and the reference FBG. T This represents the corrected water content in the oil at the current oil temperature T.

[0038] The corrected water content H in the oil at the current oil temperature T T As shown below:

[0039]

[0040]

[0041] Among them, Tref For calibration temperature; p o (T) is the water vapor pressure calculated by Buck's empirical formula at temperature T.

[0042] Furthermore, after calculating the winding temperature, winding vibration, and water content in the oil, calibration is performed, and the calibrated winding temperature, winding vibration amplitude, and water content in the oil are output.

[0043] The calibration process is as follows:

[0044] S1) Determine whether parameter correction is required. If yes, correct the parameters and proceed to step S2; otherwise, end the calibration.

[0045] If the relative error between the temperature value calculated by the fiber optic sensor and the measurement value of the nearby thermocouple is greater than or equal to 10%, then the thermo-optic coefficient ζ is corrected. s ;

[0046] The relative errors between the temperature values ​​calculated by the fiber optic sensor and the measurements from nearby thermocouples are shown below:

[0047]

[0048] If the amplitude deviation between the vibration amplitude calculated by the fiber optic sensor and the amplitude measured by the accelerometer is greater than or equal to 5%, then the calibration coefficient k should be adjusted.

[0049] If the correlation coefficient R² of the oil moisture content equation does not meet the requirement of ≥0.99, then the calibration reference correction coefficients a, b, and c are used.

[0050] S2) Recalculate the current sampled data based on the corrected parameters, use the recalculated result as the output of this monitoring, and save the corrected ζ. s k, a, b, and c are used for continuous calibration and iterative updates in subsequent sampling periods.

[0051] The technical effects of this invention are undeniable, and its beneficial effects are:

[0052] The fiber optic sensor and method for simultaneous monitoring of transformer vibration, temperature, and moisture in oil provided by this invention, through systematic and innovative design, brings significant technological progress and engineering application value. Its beneficial effects are mainly reflected in the following aspects:

[0053] 1. Strong anti-electromagnetic interference capability and adaptable to the strong electromagnetic environment of transformers: This invention adopts the fiber optic sensing principle. The fiber itself does not conduct electromagnetic signals, which can effectively resist strong electromagnetic interference such as leakage flux and transient overvoltage inside the transformer. Compared with traditional piezoelectric accelerometers and capacitive moisture sensors, the measurement signal-to-noise ratio is improved by 30%-50%, and the measurement accuracy in the near winding area is significantly improved.

[0054] 2. Enables simultaneous monitoring of multiple parameters near the winding, eliminating monitoring blind spots: The sensor is small in size (30mm×14mm×2mm) and can be directly embedded in the winding gap to directly acquire the temperature, vibration and moisture signals in the surrounding oil of the winding surface, avoiding signal attenuation and distortion of traditional indirect monitoring; at the same time, it integrates three types of parameter monitoring functions, eliminating the need to deploy multiple types of sensors, simplifying the number of openings in the oil tank (only 1 sealing flange is needed), and reducing the risk of leakage.

[0055] 3. Excellent real-time and dynamic response performance: It adopts a DFB scanner and a high-speed data acquisition card, with a temperature response time of ≤1s, a vibration response frequency covering 2Hz-2kHz (capturing dynamic vibration changes of the winding), and a moisture monitoring resolution of 1ppm. Compared with Karl Fischer offline titration (response lag ≥24h), it can reflect the dynamic deterioration process of the insulation system in real time.

[0056] 4. High data accuracy and a sound calibration mechanism: A multi-dimensional calibration system is established through thermocouples, accelerometers, and Karl Fischer titrators. The relative error of temperature measurement is controlled within 0.4%-9.36%, the correlation coefficient R² of moisture measurement with the standard method is ≥0.99, the vibration amplitude deviation is ≤3%, and the data reliability meets the requirements of engineering applications.

[0057] 5. Simplified cabling and convenient operation and maintenance: The single fiber integrates multiple parameter monitoring functions, requiring only one armored fiber to be led out. Compared with the traditional multi-sensor multi-cable cabling, the cabling complexity is reduced by 80%; the sensor's oil-resistant encapsulation design (epoxy resin + stainless steel armor) has a service life of ≥5 years, reducing the frequency of operation and maintenance replacement.

[0058] 6. Support early fault diagnosis and improve transformer reliability: Through multi-parameter synchronous monitoring, the coupled deterioration process of increased vibration-moisture infiltration-temperature rise can be captured, enabling early warning of early faults such as insulation aging and winding loosening. Compared with traditional methods, latent faults can be detected 1-3 months earlier, reducing the probability of unplanned transformer outages.

[0059] In summary, the fiber optic sensor and method for synchronous monitoring of transformer vibration, temperature, and moisture in oil, through the organic combination of multi-parameter collaborative acquisition, dynamic weight adaptation, and intelligent diagnostic decision-making, achieves a comprehensive improvement in the accuracy, real-time performance, reliability, and safety of transformer internal fault diagnosis, and has significant engineering application value and market promotion prospects. Attached Figure Description

[0060] Figure 1 For monitoring flowchart;

[0061] Figure 2This is a schematic diagram of the fiber optic sensor structure of the present invention, including a fiber optic carrier, a Fabry-Perot cavity composed of two identical fiber Bragg gratings, and a polyimide-coated phase-shifting grating for moisture monitoring.

[0062] Figure 3 A schematic diagram of the experimental setup for this invention is provided. Detailed Implementation

[0063] The present invention will be further described below with reference to embodiments, but it should not be construed that the scope of the present invention is limited to the following embodiments. Various substitutions and modifications made based on ordinary technical knowledge and common practices in the art without departing from the above-described technical concept of the present invention should be included within the scope of protection of the present invention.

[0064] Example 1:

[0065] A fiber optic sensing system for synchronous monitoring of transformer vibration, temperature and moisture content in oil includes a vibration and temperature monitoring unit and an oil moisture monitoring unit.

[0066] The vibration temperature monitoring unit and the oil moisture monitoring unit are integrated into the same optical fiber carrier;

[0067] The vibration temperature monitoring unit includes two identical fiber Bragg gratings, which together form a Fabry-Perot cavity.

[0068] The oil moisture monitoring unit is a phase-shifting grating coated with a multilayer polyimide film; the polyimide film is hygroscopic.

[0069] Example 2:

[0070] A fiber optic sensing system for synchronous monitoring of transformer vibration, temperature and moisture in oil, with the same technical content as in Embodiment 1, except that the spacing between the two fiber Bragg gratings is set to 5 mm-10 mm.

[0071] When the center wavelength laser beam is incident, the beams reflected by the two fiber Bragg gratings produce a π phase difference, forming interference fringes.

[0072] Example 3:

[0073] A fiber optic sensing system for synchronous monitoring of transformer vibration, temperature and moisture in oil is provided. The technical content is the same as any one of embodiments 1-2. Furthermore, the fiber optic sensing system is embedded in the adjacent gaps of the transformer windings to monitor transformer vibration, temperature and moisture in oil.

[0074] Example 4:

[0075] A fiber optic sensing system for synchronous monitoring of transformer vibration, temperature and moisture in oil is provided, with the same technical content as any one of embodiments 1-3. Furthermore, the thickness of the polyimide coating of the oil moisture monitoring unit is adjusted according to the moisture sensitivity requirements.

[0076] Example 5:

[0077] A method for multi-parameter monitoring of a transformer based on the aforementioned fiber optic sensor includes the following steps:

[0078] Step 1) Deploy the fiber optic sensor at the adjacent gap of the transformer winding so that the fiber optic sensor is in contact with the winding surface;

[0079] Thermocouples are deployed inside the transformer tank, and the temperature data monitored by the thermocouples is used as the temperature calibration reference.

[0080] Accelerometers are installed on the top and side walls of the transformer tank, and the data monitored by the accelerometers is used as the vibration calibration benchmark.

[0081] Oil samples are periodically collected through the sampling valve in the oil tank, and the water content in the oil samples is measured using a Karl Fischer titrator as a water calibration benchmark.

[0082] Step 2) Use a distributed feedback scanner to send a center wavelength laser beam to the Fabry-Perot cavity of the fiber optic sensor;

[0083] Step 3) Use a distributed feedback scanner to collect the reflection interference signal from the vibration temperature monitoring unit and the phase-shifting grating reflection spectrum from the oil moisture monitoring unit;

[0084] Step 4) Based on the temperature calibration reference, vibration calibration reference, and moisture calibration reference, the reflection interference signal and the phase shift grating reflection spectrum are calculated to obtain the winding temperature, winding vibration, and moisture content in the oil.

[0085] Step 5) When the winding temperature, winding vibration and / or moisture in the oil exceed the corresponding threshold, an early warning is triggered.

[0086] Example 6:

[0087] A method for monitoring multiple parameters of a transformer based on the aforementioned fiber optic sensor, with the same technical content as in Embodiment 5, further wherein the distance between the thermocouple and the fiber optic sensor is ≤5mm;

[0088] The accelerometer is positioned at the center of the top of the fuel tank and at the midpoint of the side wall.

[0089] Example 7:

[0090] A method for multi-parameter monitoring of a transformer based on the aforementioned fiber optic sensor, with the same technical content as any one of embodiments 5-6, further wherein the formula for calculating the winding temperature is as follows:

[0091]

[0092]

[0093] Where, α s ζ is the thermal expansion coefficient of optical fiber. s Thermo-optic coefficient; The wavelength is the center wavelength of the Bragg grating; This represents the change in Bragg wavelength caused by temperature changes; This represents the change in winding temperature. This serves as the temperature calibration reference; T represents the winding temperature.

[0094] Example 8:

[0095] A method for multi-parameter monitoring of a transformer based on the aforementioned fiber optic sensor, with the same technical content as any one of embodiments 5-7, further wherein the winding vibration amplitude is as follows:

[0096]

[0097] Where, ΔI m ΔI represents the measured change in light intensity. v K represents the effective vibrational light intensity change after temperature compensation. T is the temperature-light intensity compensation coefficient. k is the calibration coefficient determined from the experimental data; A is the winding vibration amplitude.

[0098] Specifically, the accelerometer calibration data is mainly used to determine and correct the calibration coefficient k in the vibration calculation, ensuring that the deviation between the vibration amplitude calculated by the fiber optic sensor and the reference vibration amplitude measured by the accelerometer meets preset requirements. Further, the accelerometer data is also used for:

[0099] (1) Verify the optical fiber vibration signal using an external benchmark;

[0100] (2) Compare and verify the vibration frequency response range and amplitude variation trend;

[0101] (3) Update the calibration coefficient k during the periodic calibration process to improve the stability and accuracy of subsequent monitoring results.

[0102] In the vibration calculation process, to eliminate the cross-effect of temperature on the Fabry-Perot cavity operating point drift, the temperature change is first calculated based on the FBG center wavelength drift:

[0103]

[0104] Let the temperature-induced shift in the equivalent light intensity at the operating point be... K TLet be the temperature-light intensity compensation coefficient, then the change in vibrational light intensity after compensation is:

[0105]

[0106] Therefore, the winding vibration amplitude is:

[0107]

[0108] Where, ΔI m ΔI represents the measured change in light intensity. v K represents the effective vibrational light intensity change after temperature compensation. T This is the temperature-light intensity compensation coefficient.

[0109] Example 9:

[0110] A method for multi-parameter monitoring of transformers based on the aforementioned fiber optic sensor, with technical content identical to any one of embodiments 5-8, further, the moisture content in the oil is as follows:

[0111]

[0112] In the formula, a, b, and c are calibration coefficients determined by the experiment; H is the water content in the oil. The peak spacing between the phase-shifting grating and the reference FBG is given.

[0113] Since the solubility of water in oil varies with temperature, a temperature correction term needs to be introduced to ensure the accuracy of water decomposition calculations at different oil temperatures. The above model is then modified as follows:

[0114]

[0115] Among them, H T This represents the corrected water content in the oil at the current oil temperature T. Considering the change in water solubility in the oil with temperature, the temperature-corrected water content model is as follows:

[0116]

[0117] Among them, T ref For calibration temperature; p o (T) is the water vapor pressure calculated by Buck's empirical formula at temperature T:

[0118]

[0119] Example 10:

[0120] A method for multi-parameter monitoring of transformers based on the aforementioned fiber optic sensor, with technical content identical to any one of embodiments 5-9, further comprising: after calculating the winding temperature, winding vibration, and oil moisture content, a calibration is performed, and the calibrated winding temperature, winding vibration amplitude, and oil moisture content are output; specifically, the current parameter ζ is first used... s The parameters ζ are initially calculated using k, a, b, and c. When the error with the corresponding calibration benchmark exceeds a preset threshold, the corresponding parameters are corrected, and the current sampled data is recalculated based on the corrected parameters. The recalculated result is then used as the output of this monitoring. The corrected ζ... s k, a, b, and c are saved simultaneously for continuous calibration and iterative updates in subsequent sampling cycles.

[0121] The calibration process is as follows:

[0122] S1) Determine whether parameter correction is required. If yes, correct the parameters and proceed to step S2; otherwise, end the calibration.

[0123] If the relative error between the temperature value calculated by the fiber optic sensor and the measurement value of the nearby thermocouple is greater than or equal to 10%, then the thermo-optic coefficient ζ is corrected. s ;

[0124] The relative errors between the temperature values ​​calculated by the fiber optic sensor and the measurements from nearby thermocouples are shown below:

[0125]

[0126] If the amplitude deviation between the vibration amplitude calculated by the fiber optic sensor and the amplitude measured by the accelerometer is greater than or equal to 5%, then the calibration coefficient k should be adjusted.

[0127] If the correlation coefficient R² of the oil moisture content equation does not meet the requirement of ≥0.99, then the calibration reference correction coefficients a, b, and c are used.

[0128] S2) Recalculate the current sampled data based on the corrected parameters, use the recalculated result as the output of this monitoring, and save the corrected ζ. s k, a, b, and c are used for continuous calibration and iterative updates in subsequent sampling periods.

[0129] Example 11:

[0130] A fiber optic sensing system for synchronous monitoring of transformer vibration, temperature and moisture in oil includes a vibration and temperature monitoring unit and an oil moisture monitoring unit. The vibration and temperature monitoring unit and the oil moisture monitoring unit are integrated into the same fiber optic carrier and are packaged as a thin structure with a thickness of no more than 2 mm.

[0131] The vibration temperature monitoring unit consists of a Fabry-Perot cavity composed of two identical fiber Bragg gratings (FBGs), with the spacing between the two FBGs set to a preset value to form interference fringes.

[0132] The oil moisture monitoring unit is a phase-shifting grating coated with a multilayer polyimide film. The polyimide film is cured at 300°C using an automated fiber optic coating device and has hygroscopic properties.

[0133] The sensor's operating temperature range covers the typical operating temperature range of transformers, and the vibration monitoring frequency band is 2Hz-2kHz.

[0134] The two FBGs of the vibration temperature monitoring unit are fabricated on the core layer of a single-mode photosensitive fiber using deep ultraviolet laser and phase mask technology. The grating period is optimized by interference fringes. The packaged sensor has a size of 30mm×14mm×2mm and can be embedded in the adjacent gap of the transformer winding.

[0135] The thickness of the polyimide coating of the oil moisture monitoring unit is controlled by the moisture sensitivity requirements. The coating thickness and the fiber radius together affect the moisture response characteristics. The spacing between the phase-shifting grating and the FBG of the vibration temperature monitoring unit is set to a preset value to ensure that the signals of the two are not interfered with.

[0136] Example 12:

[0137] A method for multi-parameter monitoring of a transformer based on the aforementioned fiber optic sensor includes the following steps:

[0138] Step 1, Sensor Deployment: Deploy fiber optic sensors at the adjacent gaps of the transformer windings, so that the sensors are in close contact with the winding surface; deploy thermocouples near the sensor positions inside the transformer tank as temperature calibration references, and deploy accelerometers on the top and side walls of the tank as vibration calibration references.

[0139] Step 2, Multi-parameter acquisition: A center wavelength laser beam is emitted from a distributed feedback (DFB) scanner and incident on the Fabry-Perot cavity of the sensor to receive the reflected interference signal; the phase-shift grating reflection spectrum of the oil moisture monitoring unit is acquired simultaneously.

[0140] Step 3, Signal Decomposition:

[0141] Temperature calculation: Based on the thermo-optic coefficient and thermal expansion coefficient of FBG, the winding temperature is calculated using the Bragg wavelength offset. The formula is as follows:

[0142]

[0143] Where, α s ζ is the thermal expansion coefficient of optical fiber. s Thermo-optic coefficient;

[0144] Vibration calculation: The vibration of the winding is inverted by the change in light intensity caused by the movement of interference fringes. The vibration amplitude is calibrated by the load current sensitivity characteristics. The vibration amplitude is proportional to the square of the load current.

[0145] Water decomposition calculation: Based on the hygroscopic expansion effect of polyimide, the water content in the oil is calculated by the peak spacing between the phase-shifting grating and the reference FBG. The peak spacing is negatively correlated with the water ppm value.

[0146] Step 4, Data Calibration: Compare and calibrate the temperature calculation results with the thermocouple measurements, verify the vibration signal with the accelerometer data, and establish a correlation model between the moisture content and the Karl Fischer titrator measurements;

[0147] Step 5, Status Output: Real-time output of winding temperature, vibration amplitude, and oil moisture content data to generate a transformer insulation system status monitoring report.

[0148] In step three, the vibration calculation sets the operation point at the midpoint of the rising edge of the interference fringes through programming. The left and right movement of the fringes caused by vibration is converted into linear light intensity changes, and temperature cross-interference is eliminated by a self-calibration algorithm.

[0149] In step three, the calculation of water decomposition in oil takes into account the effect of temperature on water solubility, and a three-dimensional calibration model of peak spacing-water content-temperature is established based on the measurement data of Karl Fischer titrator.

[0150] Before deploying the sensors in step one, the transformer needs to be vacuum dried at 105°C for 48 hours and then injected with mineral oil or synthetic ester insulating oil that has been vacuum degassed for 48 hours.

[0151] Example 13:

[0152] A fiber optic sensing system for simultaneous monitoring of transformer vibration, temperature, and moisture content in oil is described below:

[0153] This fiber optic sensor adopts a single-fiber integrated multi-unit architecture, integrating the vibration temperature monitoring unit and the oil moisture monitoring unit into the same single-mode photosensitive fiber. The entire structure is packaged into a thin structure that can be directly embedded into the transformer winding gap. The specific structure is as follows:

[0154] Vibration Temperature Monitoring Unit: This unit uses a Fabry-Perot interferometer cavity as its core and consists of two identical fiber Bragg gratings (FBGs). Furthermore, the two FBGs are etched onto the core layer of a single-mode photosensitive fiber using deep ultraviolet laser (wavelength 248nm) and phase masking technology. During the etching process, the consistency of the grating period Λ is strictly controlled (deviation ≤0.1%) to ensure the center wavelength λ of the two FBGs. BThe initial deviation does not exceed 0.5nm; the spacing between the two FBGs is set to 5-10mm to form a Fabry-Perot interferometer cavity. When the center wavelength laser beam is incident, the beams reflected by the two FBGs will generate a π phase difference, forming dense interference fringes.

[0155] The FBG of this unit is protected by an acrylic coating. The coating process uses a Furukawa Electric S541A fiber optic coating machine, and the coating thickness is controlled at 25-30μm to improve mechanical strength and oil resistance. The unit's operating temperature range covers -20℃ to 150℃ (covering the temperature range of normal transformer operation and short-term overload), and the vibration monitoring frequency band is 2Hz-2kHz (matching the typical frequency range of transformer winding vibration), meeting the monitoring needs under different operating conditions.

[0156] Oil Moisture Monitoring Unit: This unit uses a phase-shift grating as a substrate and achieves moisture sensitivity through multilayer polyimide film coating. The phase-shift grating and the FBG of the vibration temperature monitoring unit are etched onto the same optical fiber in the same batch, with a spacing of 20-30mm between them to avoid signal interference. The polyimide film is coated using a Vytran PRL201 automated optical fiber coating machine, with 3-5 coating layers and a total thickness controlled at 50-80μm. After coating, it is cured at 300℃ for 4 hours to ensure the bonding strength between the film and the grating surface and the oil resistance.

[0157] The hygroscopic properties of polyimide film are the core of moisture monitoring: due to the strong water absorption of polyimide, when the moisture content in the oil changes, the film absorbs moisture and expands in volume, which in turn squeezes the fiber grating, causing the Bragg wavelength of the phase-shifting grating to shift. By adjusting the coating thickness and fiber radius (using 125μm clad fiber), the moisture sensitivity can be optimized, so that the unit's response range to moisture in the oil covers 0-500ppm (meeting the industry standard for transformer oil moisture monitoring).

[0158] Sensor overall encapsulation: To ensure the stability and insulation performance of the sensor in transformer oil, oil-resistant epoxy resin material is used for overall encapsulation.

[0159] Before encapsulation, the optical fiber and grating unit are cleaned (wiped with anhydrous ethanol to remove surface oil). The encapsulation process uses mold forming, and the final encapsulation size is 30mm×14mm×2mm with a thickness of no more than 2mm. It can be directly embedded into the adjacent gap of the transformer winding (typical gap width 3-5mm) to achieve direct monitoring near the winding. The encapsulation surface is smoothed to avoid scratching the winding insulation paper.

[0160] The fiber optic leads at both ends of the sensor are protected by stainless steel armor (armor layer diameter 1.5mm), with a lead length of 1-2m, which facilitates connection to external data acquisition equipment. At the same time, the armor layer has sealing properties to prevent insulating oil from leaking from the lead ends.

[0161] Example 14:

[0162] A method for multi-parameter monitoring of transformers based on the aforementioned fiber optic sensor is proposed. This method achieves simultaneous monitoring and accurate analysis of multiple parameters through five steps: deployment, acquisition, calculation, calibration, and output. The specific steps are as follows:

[0163] Step 1, Sensor Deployment and Transformer Preprocessing:

[0164] Transformer pretreatment: Before sensor deployment, the laboratory or field transformer needs to be vacuum dried and treated with insulating oil—the transformer is placed in a 105℃ vacuum oven for 48 hours to dry, with the vacuum level maintained below -0.095MPa, to remove residual moisture from the core, windings and insulating paper; after drying, insulating oil (mineral oil or synthetic ester) is injected. Before injection, the insulating oil needs to be vacuum degassed for 48 hours. After degassed, the moisture content in the oil is ≤10ppm to avoid initial moisture interfering with the monitoring results.

[0165] Sensor deployment: Open the lifting hole on the top of the transformer tank and fix the encapsulated fiber optic sensor to the adjacent gap of the winding using a special clamp, ensuring that the sensor is in close contact with the winding surface (spacing ≤ 1mm) to directly acquire the vibration and temperature signals of the winding; the armored fiber optic cable led out by the sensor is led out through the sealing flange on the side wall of the tank. The flange is sealed with an O-ring rubber seal to prevent leakage of insulating oil.

[0166] Calibration equipment deployment: A TJ 36-CASS-116G-12-CC type high-temperature thermocouple (temperature measurement range -50℃-400℃, accuracy ±0.5℃) is deployed near the sensor (distance ≤5mm) as a temperature calibration benchmark; piezoelectric accelerometers (sensitivity 100mV / g, frequency band 1Hz-5kHz) are deployed at the center of the top of the oil tank and the midpoint of the side wall as vibration calibration benchmarks; oil samples are collected periodically through the oil tank sampling valve, and the water content in the oil is measured using a Karl Fischer titrator (accuracy ±1ppm) as a water content calibration benchmark.

[0167] Step 2, Simultaneous Acquisition of Multiple Parameters:

[0168] Data Acquisition System Setup: The optical fiber leading from the sensor is connected to a distributed feedback (DFB) laser scanner (output power 10mW, center wavelength matched with the FBG center wavelength, wavelength stability ±0.1nm / ℃). The interference signal and reflection spectrum signal output by the scanner are transmitted to a photoelectric converter (conversion gain 1000V / W, bandwidth 10MHz) to convert the optical signal into an electrical signal. The electrical signal is acquired by an NI DAQ-9174 data acquisition card (sampling rate 100kS / s, resolution 16-bit) and finally stored in a computer.

[0169] Synchronous acquisition and control: Using the power frequency voltage signal (50Hz) as the time reference, the scanner, photoelectric converter, and data acquisition card are triggered to start synchronously. When the amplitude of the ultra-high frequency signal or vibration signal exceeds the preset threshold (0.1g for vibration, ±1℃ for temperature), multi-channel data recording is automatically initiated. All monitored parameters (interference signal, reflection spectrum, thermocouple temperature, accelerometer vibration) are stored with a unified timestamp (1ms accuracy) to ensure the spatiotemporal consistency of multiple parameter data at the same time. Furthermore, the data acquisition frequency is adjusted according to the operating conditions: the acquisition frequency is 1Hz during normal operation, and automatically increases to 10Hz when the load current changes by more than 10% or the temperature changes by more than 2℃, achieving dynamic monitoring.

[0170] Step 3, Multi-parameter signal solution:

[0171] Based on the acquired optical signals, the winding temperature, vibration amplitude, and water content in the oil are calculated respectively. The specific calculation method is as follows:

[0172] Temperature calculation:

[0173] Based on the thermo-optical effect and thermal expansion effect of FBG, the temperature change ΔT will affect the grating period Λ and the effective refractive index n of FBG. eff The change, in turn, causes a shift in the center wavelength Δλ. B The solution formula is:

[0174]

[0175] Where, α s ζ is the thermal expansion coefficient of optical fiber. s The thermo-optic coefficient of FBG; measured by Δλ B (Obtained by analyzing the wavelength shift of the interference fringes using a DFB scanner), ΔT can be calculated by substituting it into the formula. Combined with the initial temperature T0 (the initial value measured by the thermocouple), the real-time winding temperature T = T0 + ΔT can be obtained.

[0176] Vibration solution:

[0177] Vibration causes a change in the cavity length of the Fabry-Perot cavity, resulting in the left and right shift of the interference fringes. By programming and setting the operating point at the midpoint of the rising edge of the interference fringes (the region of optimal linearity), the fringe shift is converted into a linear change in light intensity ΔI.

[0178] The vibration amplitude is linearly related to the change in light intensity ΔI. Based on the calibration data of the accelerometer, the following formula is established: A=k×ΔI (k is the calibration coefficient, which is experimentally measured and has a typical value of 0.01g / mV). At the same time, a self-calibration algorithm is used to eliminate the influence of temperature on the cavity length—by real-time compensation for the temperature offset of the FBG center wavelength, it is ensured that the vibration calculation is not affected by temperature cross-interference.

[0179] Furthermore, in the vibration calculation process, to eliminate the cross-effect of temperature on the drift of the Fabry-Perot cavity operating point, the temperature change is first calculated based on the drift of the FBG center wavelength:

[0180]

[0181] Let the temperature-induced shift in the equivalent light intensity at the operating point be... K T Let be the temperature-light intensity compensation coefficient, then the change in vibrational light intensity after compensation is:

[0182]

[0183] Therefore, the winding vibration amplitude is:

[0184]

[0185] Where, ΔI m ΔI represents the measured change in light intensity. v K represents the effective vibrational light intensity change after temperature compensation. T This is the temperature-light intensity compensation coefficient.

[0186] Furthermore, the relationship between winding vibration and load current is that the vibration amplitude A is proportional to the square of the load current I (A∝I²), and the rationality of the vibration calculation results can be verified by load current data.

[0187] Hydrolysis calculation:

[0188] Changes in the water content of the oil cause the polyimide film to expand, which in turn causes a shift in the center wavelength Δλ of the phase-shifting grating. shift The peak spacing D between the phase-shifting grating and the reference FBG in the vibration temperature unit is calculated by comparing their center wavelengths.

[0189] Based on measured data from a Karl Fischer titrator, a correlation model was established between the peak spacing D and the water content H (ppm) in the oil:

[0190]

[0191] Where a, b, and c are calibration coefficients (obtained from experimental fitting; typical values ​​for mineral oil: a = -2 × 10⁻⁵ nm / (ppm²), b = 0.0002 nm / ppm, c = 0.2545 nm); since the solubility of water in oil varies with temperature, a temperature correction term needs to be introduced to ensure the accuracy of water decomposition calculations at different oil temperatures. The above model is then modified as follows:

[0192]

[0193] Among them, H T This represents the corrected water content in the oil at the current oil temperature T. Considering the change in water solubility in the oil with temperature, the temperature-corrected water content model is as follows:

[0194]

[0195] Among them, T ref For calibration temperature; p o (T) is the water vapor pressure calculated by Buck's empirical formula at temperature T:

[0196]

[0197] Step 4, Data Calibration and Verification:

[0198] Temperature calibration: The temperature value calculated by the fiber optic sensor is compared with the measurement value of a nearby thermocouple in real time, and the relative error is calculated.

[0199]

[0200] When δ T When it exceeds 10%, the thermo-optic coefficient ζ is automatically corrected. s until δ T It should be controlled between 0.4% and 9.36%.

[0201] Vibration calibration: Compare the vibration amplitude calculated by the fiber optic sensor with the measured value of the accelerometer, calculate the amplitude deviation, and when the deviation exceeds 5%, adjust the calibration coefficient k to ensure that the frequency response and amplitude accuracy of the vibration signal meet the requirements (deviation ≤3% within the 2Hz-2kHz frequency band).

[0202] Moisture calibration: Oil samples were collected every 24 hours, and the moisture content (H) was measured using a Karl Fischer titrator. KFT Substitute the values ​​into the water decomposition calculation model, correct the coefficients a, b, and c, and ensure that the correlation coefficient R² of the model is ≥0.99, thereby improving the accuracy of long-term monitoring.

[0203] Step 5: Monitoring Results Output and Status Assessment

[0204] The calculated winding temperature, vibration amplitude (including spectral distribution), and oil moisture content data are output to the monitoring system in real time to form a multi-parameter trend curve. At the same time, thresholds are set according to industry standards (such as temperature threshold of 120℃, vibration amplitude threshold of 0.5g, and moisture threshold of 30ppm). When any parameter exceeds the threshold, an early warning is triggered.

Claims

1. A fiber optic sensing system for simultaneous monitoring of transformer vibration, temperature, and moisture content in oil, characterized in that: Includes a vibration temperature monitoring unit and an oil moisture monitoring unit; The vibration temperature monitoring unit and the oil moisture monitoring unit are integrated into the same optical fiber carrier; The vibration temperature monitoring unit includes two identical fiber Bragg gratings, which together form a Fabry-Perot cavity. The oil moisture monitoring unit is a phase-shifting grating coated with multiple layers of polyimide film; the polyimide film is hygroscopic.

2. The fiber optic sensing system for synchronous monitoring of transformer vibration, temperature, and moisture content in oil according to claim 1, characterized in that: The spacing between the two fiber Bragg gratings is set to 5 mm - 10 mm; When the center wavelength laser beam is incident, the beams reflected by the two fiber Bragg gratings produce a π phase difference, forming interference fringes.

3. The fiber optic sensing system for synchronous monitoring of transformer vibration, temperature, and moisture content in oil according to claim 1, characterized in that: Fiber optic sensing systems are embedded in the gaps between adjacent transformer windings to monitor transformer vibration, temperature, and moisture in the oil.

4. The fiber optic sensing system for synchronous monitoring of transformer vibration, temperature, and moisture content in oil according to claim 1, characterized in that: The thickness of the polyimide coating in the oil moisture monitoring unit is adjusted according to the moisture sensitivity requirements.

5. A method for multi-parameter monitoring of a transformer based on the fiber optic sensor according to any one of claims 1-4, characterized in that: Includes the following steps: Step 1) Deploy the fiber optic sensor at the adjacent gap of the transformer winding so that the fiber optic sensor is in contact with the winding surface; Thermocouples are deployed inside the transformer tank, and the temperature data monitored by the thermocouples is used as the temperature calibration reference. Accelerometers are installed on the top and side walls of the transformer tank, and the data monitored by the accelerometers is used as the vibration calibration benchmark. Oil samples are periodically collected through the sampling valve in the oil tank, and the water content in the oil samples is measured using a Karl Fischer titrator as a water calibration benchmark. Step 2) Use a distributed feedback scanner to send a center wavelength laser beam to the Fabry-Perot cavity of the fiber optic sensor; Step 3) Use a distributed feedback scanner to collect the reflection interference signal from the vibration temperature monitoring unit and the phase-shifting grating reflection spectrum from the oil moisture monitoring unit; Step 4) Based on the temperature calibration reference, vibration calibration reference, and moisture calibration reference, the reflection interference signal and the phase shift grating reflection spectrum are calculated to obtain the winding temperature, winding vibration, and moisture content in the oil. Step 5) When the winding temperature, winding vibration and / or moisture in the oil exceed the corresponding threshold, an early warning is triggered.

6. The monitoring method according to claim 5, characterized in that: The distance between the thermocouple and the fiber optic sensor is ≤5mm; The accelerometer is positioned at the center of the top of the fuel tank and at the midpoint of the side wall.

7. The monitoring method according to claim 5, characterized in that: The formula for calculating the winding temperature is as follows: Where, α s ζ is the thermal expansion coefficient of optical fiber. s Thermo-optic coefficient; The wavelength is the center wavelength of the Bragg grating; This represents the change in Bragg wavelength caused by temperature changes; This represents the change in winding temperature. This serves as the temperature calibration reference; T represents the winding temperature.

8. The monitoring method according to claim 5, characterized in that, The winding vibration amplitude is shown below: Where, ΔI m ΔI represents the measured change in light intensity. v K represents the effective vibrational light intensity change after temperature compensation. T is the temperature-light intensity compensation coefficient; k is the calibration coefficient determined by experimental data; A is the winding vibration amplitude.

9. The monitoring method according to claim 5, characterized in that, The water content in the oil is shown below: In the formula, a, b, and c are the calibration coefficients determined in the experiment; H represents the peak spacing between the phase-shifting grating and the reference FBG. T This represents the corrected water content in the oil at the current oil temperature T. The corrected water content H in the oil at the current oil temperature T T As shown below: Among them, T ref For calibration temperature; p o (T) is the water vapor pressure calculated by Buck's empirical formula at temperature T.

10. The monitoring method according to claim 5, characterized in that, After calculating the winding temperature, winding vibration, and water content in the oil, calibration is performed, and the calibrated winding temperature, winding vibration amplitude, and water content in the oil are output. The calibration process is as follows: S1) Determine whether parameter correction is required. If yes, correct the parameters and proceed to step S2; otherwise, end the calibration. If the relative error between the temperature value calculated by the fiber optic sensor and the measurement value of the nearby thermocouple is greater than or equal to 10%, then the thermo-optic coefficient ζ is corrected. s ; The relative errors between the temperature values ​​calculated by the fiber optic sensor and the measurements from nearby thermocouples are shown below: If the amplitude deviation between the vibration amplitude calculated by the fiber optic sensor and the amplitude measured by the accelerometer is greater than or equal to 5%, then the calibration coefficient k should be adjusted. If the correlation coefficient R² of the oil moisture content equation does not meet the requirement of ≥0.99, then the calibration reference correction coefficients a, b, and c are used. S2) Recalculate the current sampled data based on the corrected parameters, use the recalculated result as the output of this monitoring, and save the corrected ζ. s k, a, b, and c are used for continuous calibration and iterative updates in subsequent sampling periods.

Images