Online monitoring method, device and system for moisture in transformer oil and storage medium

Abstract

The invention relates to an online monitoring method, device and equipment for moisture in transformer oil and a storage medium, and the method comprises the steps: obtaining a capacitance signal and real-time oil temperature data outputted by sensing equipment in the transformer oil, and converting the capacitance signal into a frequency signal; performing moisture calculation on the frequency signal and the real-time oil temperature according to the accumulated running time of the sensing equipment to obtain a moisture content parameter; carrying out period segmentation on the accumulated operation duration to obtain unit periods, and carrying out period change calculation on the moisture content parameter according to the unit periods to obtain a change rate; and performing classification comparison on the moisture content parameter and the change rate based on a preset early warning rule, and outputting an early warning signal. According to the method, quantitative capture of the dynamic change trend of the water in the oil can be realized, so that different risk development modes such as slow permeation and sudden water inflow are distinguished.

Description

Technical Field

[0001] This invention relates to the technical field of transformers, and in particular to an online monitoring method, device, system, and storage medium for moisture in transformer oil. Background Technology

[0002] Transformers are core equipment in power systems, responsible for voltage transformation and power distribution. The quality of their internal insulating oil directly affects their safe, stable operation and service life. Continuous and accurate online monitoring of moisture content in transformer oil is crucial for condition-based maintenance and preventing major accidents. Existing online monitoring methods primarily focus on directly measuring instantaneous moisture content, but often treat the sensors as ideal, neglecting the impact of long-term performance drift on measurement accuracy. Furthermore, most of these methods only set fixed thresholds based on absolute moisture values ​​for alarms, failing to effectively differentiate between slow increases and rapid short-term increases in moisture content. This results in insufficient timeliness of warnings and inadequate risk level differentiation, failing to provide more accurate trend data and tiered guidance for maintenance decisions. Summary of the Invention

[0003] The main objective of this invention is to provide an online monitoring method, device, system, and storage medium for moisture in transformer oil, which can quantitatively capture the dynamic changes in moisture in the oil, thereby distinguishing different risk development modes such as slow seepage and sudden water ingress.

[0004] To achieve the above objectives, the present invention provides an online monitoring method for moisture in transformer oil, comprising: Acquire the capacitance signal and real-time oil temperature data output by the sensing device in the transformer oil, and convert the capacitance signal into a frequency signal; Based on the cumulative operating time of the sensing device, the moisture content parameter is obtained by calculating the moisture content of the frequency signal and the real-time oil temperature. The cumulative runtime is divided into periods to obtain unit periods, and the periodic change of the moisture content parameter is calculated based on the unit periods to obtain the rate of change. The moisture content parameter and the rate of change are compared and classified according to preset early warning rules, and an early warning signal is output.

[0005] Further, the step of acquiring the capacitance signal and real-time oil temperature data output by the sensing device in the transformer oil, and converting the capacitance signal into a frequency signal, includes: The capacitance signal and the real-time oil temperature data are acquired periodically by the sensing device. The capacitor signal is frequency-converted based on a preset capacitor frequency conversion table to obtain an initial frequency pulse signal. The initial frequency pulse signal is bound to the real-time oil temperature data by sampling points according to the preset capacitor frequency conversion table to obtain the frequency signal.

[0006] Further, the step of calculating the moisture content parameter by analyzing the frequency signal and the real-time oil temperature data based on the cumulative operating time of the sensing device includes: The initial humidity parameters are obtained by querying the pre-stored first mapping relationship based on the frequency signal; Based on the real-time oil temperature data, the pre-stored temperature compensation parameter table is queried to obtain the temperature compensation coefficient; The initial humidity parameter is initially compensated based on the temperature compensation coefficient to obtain the temperature-compensated humidity value. Based on the accumulated runtime, the time drift compensation parameter table is queried to obtain the time drift correction coefficient; The temperature-compensated humidity value is corrected for time drift based on the time drift correction coefficient to obtain the corrected humidity value; The moisture content parameter is obtained by querying a pre-stored oil-water balance table based on the corrected humidity value and the real-time oil temperature data.

[0007] Further, the step of querying a pre-stored oil-water balance table based on the corrected humidity value and the real-time oil temperature data to obtain the moisture content parameter includes: The pre-stored oil-water balance relationship table is recorded and matched according to the corrected humidity value and the corrected humidity value; If a balance record is found that matches the corrected humidity value and the real-time oil temperature data, the moisture content parameter in the balance record is extracted. If no balance record is found, the humidity reference value that is closest to the corrected humidity value and the temperature reference value that is closest to the real-time oil temperature data are found in the pre-stored oil-water balance relationship table. Based on the humidity reference value and the temperature reference value, the corresponding moisture content reference value is located in the pre-stored oil-water balance relationship table; The moisture content parameter is obtained by performing bilinear interpolation on the moisture content reference value based on the corrected humidity value and the real-time oil temperature.

[0008] Further, the step of correcting the temperature-compensated humidity value for time drift based on the time drift correction coefficient to obtain the corrected humidity value includes: Multiply the temperature-compensated humidity value by the time-drift correction coefficient to obtain the first intermediate humidity value; The difference between the first intermediate humidity value and the pre-stored baseline humidity value is calculated to obtain the humidity difference. Determine whether the humidity difference exceeds the preset time drift correction range; If the humidity difference does not exceed the preset time drift correction range, then the first intermediate humidity value is used as the corrected humidity value; If the humidity difference exceeds the time drift correction threshold, the first intermediate humidity value is iteratively adjusted according to the time drift correction coefficient and the preset correction step size to obtain the corrected humidity value.

[0009] Further, the step of periodically dividing the cumulative runtime to obtain a unit period, and calculating the periodic change rate of the moisture content parameter based on the unit period, includes: The cumulative runtime is divided into multiple unit periods according to a preset time window length; The cumulative runtime is associated with each of the moisture content parameters, and the corresponding moisture content parameter is obtained as a characterization value based on the unit period. The content difference between two adjacent characterization content values ​​is calculated sequentially, and the content difference is divided by the unit period to obtain the rate of change.

[0010] Furthermore, the step of classifying and comparing the moisture content parameter and the rate of change based on preset early warning rules, and outputting an early warning signal, includes: The moisture content parameter is compared with a first moisture threshold and a second moisture threshold of the preset warning rule, wherein the first moisture threshold is less than the second moisture threshold; If the moisture content parameter is less than the first moisture threshold, the current state is determined to be normal and no alarm is triggered. If the moisture content parameter is not less than the first moisture threshold but less than the second moisture threshold, then it is determined whether the rate of change exceeds the rate threshold of the preset warning rule. When the rate of change does not exceed the rate threshold, a first-level warning signal is triggered; When the rate of change exceeds the rate threshold, a second-level warning signal is triggered; If the moisture content parameter is not less than the second moisture threshold, the third-level alarm signal is directly triggered.

[0011] This invention also provides an online monitoring device for moisture in transformer oil, applied to the online monitoring method for moisture in transformer oil described in any one of the above claims, comprising: The acquisition module is used to acquire the capacitance signal and real-time oil temperature data output by the sensing device in the transformer oil, and convert the capacitance signal into a frequency signal. The analysis module is used to calculate the moisture content parameter by analyzing the frequency signal and the real-time oil temperature based on the cumulative running time of the sensing device. The association module is used to divide the cumulative running time into periods to obtain unit periods, and to calculate the periodic change of the moisture content parameter based on the unit periods to obtain the rate of change. The processing module is used to perform hierarchical comparison of the moisture content parameter and the rate of change based on preset early warning rules, and output early warning signals.

[0012] This invention also provides an online monitoring system for moisture in transformer oil, comprising: Memory, used to store programs; A processor is used to execute the program to implement the various steps of the online monitoring method for moisture in transformer oil as described in any of the above-mentioned methods.

[0013] The present invention also provides a storage medium storing computer instructions for causing a computer to perform any of the methods described above.

[0014] The present invention provides an online monitoring method, device, system, and storage medium for moisture in transformer oil, which has the following beneficial effects: By calculating moisture content based on frequency signals and oil temperature data according to the cumulative operating time of the sensing device, performance drift caused by long-term sensor operation can be dynamically compensated, thereby significantly improving the measurement accuracy and long-term reliability of moisture content parameters throughout the entire equipment lifecycle. By periodically segmenting the cumulative operating time and calculating the rate of change of moisture content parameters, the dynamic trend of moisture changes in the oil can be quantitatively captured, thus distinguishing between different risk development modes such as slow seepage and sudden water ingress. Based on this, by classifying and jointly judging moisture content parameters and their rate of change according to preset early warning rules, more discriminative early warning signals can be output, enabling earlier and more accurate identification and graded early warning of potential fault risks. Attached Figure Description

[0015] Figure 1 A flowchart of an online monitoring method for moisture in transformer oil provided by the present invention; Figure 2 A structural diagram of an online monitoring device for moisture in transformer oil provided by the present invention; Figure 3 This invention provides a structural diagram of an online monitoring system for moisture in transformer oil.

[0016] The objectives, features, and advantages of this invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0017] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0018] The present invention will now be further described in conjunction with the accompanying drawings and specific embodiments.

[0019] Reference Figure 1 As shown, the present invention provides an online monitoring method for moisture in transformer oil, comprising: Step S1: Obtain the capacitance signal and real-time oil temperature data output by the sensing device in the transformer oil, and convert the capacitance signal into a frequency signal; Specifically, sensing devices deployed in the transformer oil output a capacitance signal characterizing the oil's dielectric constant at fixed intervals. This capacitance signal is directly affected by both the water content and temperature of the oil. Periodic sampling and analog-to-digital conversion of the original capacitance signal form a discrete sequence of capacitance sample values. Synchronously, a temperature sensor located at the same measurement point of the sensing devices collects the real-time temperature of the oil, generating a timestamp-aligned oil temperature data sequence. A preset capacitance-to-frequency conversion table is used for conversion, defining the mapping relationship between capacitance values ​​and standard frequency values ​​at a specific temperature reference point. For each capacitance sample value, based on the synchronously collected real-time oil temperature data, the closest temperature reference point is matched in the conversion table, and the corresponding mapping relationship is applied to convert the capacitance value into an initial frequency pulse signal. By binding sampling points, it is ensured that each converted frequency pulse strictly corresponds to a precise real-time oil temperature data point, together forming the frequency signal.

[0020] Step S2: Calculate the moisture content parameter by analyzing the frequency signal and the real-time oil temperature based on the cumulative operating time of the sensing device. Specifically, based on the frequency signal, a pre-stored mapping relationship is queried. This relationship represents the correspondence between frequency and humidity parameters under standard conditions, thus obtaining an initial humidity parameter. A temperature compensation parameter table is then consulted using real-time oil temperature data to obtain a temperature compensation coefficient for the current oil temperature. This temperature compensation coefficient is used to correct the initial humidity parameter, resulting in a temperature-compensated humidity value. The cumulative operating time of the sensor since its commissioning is read. Based on the cumulative operating time, a pre-stored time-drift compensation parameter table is consulted. This table records the drift pattern of the sensor's response characteristics over time due to factors such as long-term oil immersion and material aging, and outputs a time-drift correction coefficient. This time-drift correction coefficient is applied to further correct the temperature-compensated humidity value, resulting in a corrected humidity value. This corrected humidity value and real-time oil temperature data are used as a set of joint conditions to query a pre-stored oil-water balance relationship table. This table, based on the thermodynamic principle of oil-water balance, describes the absolute water content corresponding to the saturation equilibrium of dissolved water in oil under different temperatures and relative humidity conditions. Through this joint query, the corrected humidity value is finally mapped to an absolute water content parameter, i.e., the mass or concentration of water per unit volume of oil.

[0021] Step S3: Divide the cumulative running time into periods to obtain unit periods, and calculate the periodic change of the moisture content parameter based on the unit periods to obtain the rate of change; Specifically, the cumulative runtime serves as a time scale, uniformly divided by a fixed-length time window, with each time window defined as a unit period. The length of the unit period is set according to monitoring needs, such as 24 hours or one week, with the aim of creating a stable time benchmark for trend analysis. Based on the timestamps of the moisture content parameters, they are categorized into the corresponding unit periods. For each unit period, a characteristic content value representing the overall moisture level for that time period is generated by calculating the average of all moisture content parameters within the unit period or selecting a specific value at the end of the period. The content difference between the current characteristic content value and the characteristic content value of the previous unit period is calculated sequentially. Dividing the content difference by the unit period spanned by the two consecutive periods yields a rate of change. A positive rate indicates an upward trend in moisture content, while a negative rate indicates a decrease or stabilization.

[0022] Step S4: Based on preset early warning rules, classify and compare the moisture content parameter and the rate of change, and output an early warning signal.

[0023] Specifically, the moisture content parameter is compared with a first moisture threshold of a preset early warning rule. If the moisture content parameter is below this first moisture threshold, the oil condition is determined to be normal, and no early warning signal is triggered. If the moisture content parameter is equal to or exceeds the first threshold, the tiered judgment process begins. At this point, the judgment further diverges: if the moisture content parameter has reached or exceeded the second moisture threshold of the preset early warning rule, this is considered a severely excessive state, and trend analysis will be bypassed, directly triggering the highest-level third-level alarm signal, indicating the need for immediate intervention. The first moisture threshold is less than the second moisture threshold.

[0024] If the moisture content parameter is between the first and second thresholds, it indicates an abnormal situation but has not yet reached an emergency level. The current rate of change is compared to a preset rate threshold. When the rate of change does not exceed the rate threshold, it indicates that the moisture increase is relatively slow, possibly due to long-term infiltration, triggering a lower-level first-level warning signal. When the rate of change exceeds the rate threshold, it means that the moisture is increasing rapidly, posing a higher risk of sudden influx, triggering a more urgent second-level warning signal, indicating the need for immediate inspection or intervention.

[0025] This invention provides an online monitoring method for moisture in transformer oil. By calculating moisture content based on frequency signals and oil temperature data according to the cumulative operating time of the sensing device, it can dynamically compensate for performance drift caused by long-term sensor operation, thereby significantly improving the measurement accuracy and long-term reliability of moisture content parameters throughout the entire equipment lifecycle. By periodically segmenting the cumulative operating time and calculating the rate of change of moisture content parameters, it achieves quantitative capture of the dynamic trend of moisture changes in the oil, thus distinguishing between different risk development modes such as slow seepage and sudden water ingress. Based on this, by classifying and jointly judging the moisture content parameters and their rate of change according to preset early warning rules, it can output more discriminative early warning signals, thereby achieving earlier and more accurate identification and graded early warning of potential fault risks.

[0026] In some embodiments, acquiring the capacitance signal and real-time oil temperature data output by the sensing device in the transformer oil, and converting the capacitance signal into a frequency signal, includes: The capacitance signal and the real-time oil temperature data are acquired periodically by the sensing device. The capacitor signal is frequency-converted based on a preset capacitor frequency conversion table to obtain an initial frequency pulse signal. The preset capacitance-frequency conversion table defines the correspondence between the capacitance value (or its digital code) output by the capacitance sensing unit and the standard frequency value under one or more specific reference temperature conditions.

[0027] Specifically, each discrete capacitance value in the capacitance sampling value sequence obtained in the previous step is read. For each capacitance value to be converted, the preset capacitance frequency conversion table is accessed. Logically, the conversion table uses the capacitance value as the index key and the standard frequency value as the mapping result. A lookup operation is performed to find the index key in the table that is equal to or closest to the current capacitance value, and the predetermined frequency value associated with that index key is retrieved. If the current capacitance value happens to match a certain index key in the table, the frequency value corresponding to that key is directly output. If the current capacitance value is between two adjacent index keys in the table, a corresponding frequency value is calculated using linear interpolation. By traversing the entire capacitance sampling value sequence and repeating this lookup and calculation process, each capacitance sampling value is converted into a corresponding frequency value one by one. These frequency values ​​are organized into a new digital sequence in which the timing of each data point is completely consistent with the original capacitance sampling sequence. This sequence constitutes the initial frequency pulse signal.

[0028] The initial frequency pulse signal is bound to the real-time oil temperature data by sampling points according to the preset capacitor frequency conversion table to obtain the frequency signal.

[0029] Specifically, the binding operation relies on a shared sampling time reference established during the initial data acquisition phase. For the Nth frequency value in the initial frequency pulse sequence, its source is the Nth capacitance value in the capacitance sampling sequence, and the acquisition time of this capacitance value is exactly the same as the acquisition time of the Nth temperature value in the oil temperature data sequence. Based on this one-to-one timing relationship, the Nth frequency value and the Nth temperature value are extracted from their respective sequences and combined to form a data structure with a common timestamp. Traversing the entire sequence length, the same pairing operation is performed on the frequency and temperature values ​​at all sequential positions, thereby generating a series of frequency-temperature data pairs arranged in chronological order. Encapsulating this series of frequency-temperature data pairs, the output is the frequency signal.

[0030] The method provided in this embodiment calculates moisture content by incorporating the cumulative operating time of the sensing device into the frequency signal and oil temperature data. This dynamically compensates for performance drift caused by long-term sensor operation, thereby significantly improving the measurement accuracy and long-term reliability of the moisture content parameter throughout its entire lifespan. By periodically segmenting the cumulative operating time and calculating the rate of change of the moisture content parameter, the method achieves quantitative capture of the dynamic trend of moisture changes, effectively distinguishing between different risk modes such as slow seepage and sudden water ingress.

[0031] In some embodiments, the step of calculating the moisture content parameter by analyzing the frequency signal and the real-time oil temperature data based on the cumulative operating time of the sensing device includes: The initial humidity parameters are obtained by querying the pre-stored first mapping relationship based on the frequency signal; Based on the real-time oil temperature data, the pre-stored temperature compensation parameter table is queried to obtain the temperature compensation coefficient; The initial humidity parameter is initially compensated based on the temperature compensation coefficient to obtain the temperature-compensated humidity value. Specifically, two intermediate variables corresponding to the current calculation cycle are extracted: the initial humidity parameter and the temperature compensation coefficient. The compensation operation is performed using a multiplicative model, i.e., temperature-compensated humidity value = initial humidity parameter × temperature compensation coefficient. In this model, when the real-time oil temperature equals the reference temperature used when establishing the first mapping relationship, the obtained temperature compensation coefficient should theoretically be 1.0. At this time, the temperature-compensated humidity value equals the initial humidity parameter and no correction is needed. When the real-time oil temperature is higher than the reference temperature, the oil dielectric constant itself changes due to temperature. The initial humidity parameter corresponding to the frequency signal output by the sensor will contain this temperature interference. At this time, the temperature compensation coefficient is usually a coefficient less than 1.0, used to pull the reading back to the reference temperature condition. Conversely, when the oil temperature is lower than the reference temperature, the temperature compensation coefficient is usually greater than 1.0. Through this multiplicative operation, the physical meaning represented by the temperature-compensated humidity value is normalized to the equivalent humidity indication under the reference temperature condition, thereby removing the main influence of instantaneous oil temperature fluctuations on the moisture sensing signal. This operation is performed on each set of initial humidity parameters (temperature compensation coefficients) in the data sequence to generate a new sequence of temperature-compensated humidity values ​​that is in the same order as the original data stream.

[0032] Based on the accumulated runtime, the time drift compensation parameter table is queried to obtain the time drift correction coefficient; Specifically, the latest cumulative runtime value is read from the device's non-volatile storage unit. Using this cumulative runtime as input, the time drift compensation parameter table is accessed. This parameter table logically uses a series of incremental time nodes as index bases, such as 0 hours, 1000 hours, 5000 hours, etc., with each time node index associated with a fixed time drift correction coefficient determined through prior calibration. The processing unit searches the table to locate the specific time interval of the current cumulative runtime, i.e., finding two adjacent time indices that satisfy the condition: previous time node index ≤ current cumulative runtime < next time node index, and simultaneously obtaining the time drift correction coefficients corresponding to these two time indices. Using the coefficient corresponding to the previous time node as the base value, the difference between the coefficient of the next node and the coefficient of the previous node is added, and then multiplied by the product of the difference between the current cumulative runtime and the previous time node and the quotient of the time interval between the two nodes, finally yielding the time drift correction coefficient matching the current precise cumulative runtime. If the cumulative runtime is exactly equal to a certain time node index in the table, the correction coefficient pre-stored for that node is directly used without interpolation.

[0033] The temperature-compensated humidity value is corrected for time drift based on the time drift correction coefficient to obtain the corrected humidity value; The moisture content parameter is obtained by querying a pre-stored oil-water balance table based on the corrected humidity value and the real-time oil temperature data.

[0034] The method provided in this embodiment obtains the initial humidity parameters by querying a pre-stored first mapping relationship based on the frequency signal, and performs temperature compensation based on real-time oil temperature. This effectively eliminates the direct impact of temperature fluctuations on the dielectric measurement results of the capacitive sensor, thereby significantly improving the accuracy of the moisture indication value at a single measurement point. By introducing a cumulative running time query for a time drift compensation parameter table and performing a secondary correction on the temperature-compensated humidity value, the method can dynamically offset the systematic performance drift of the sensor caused by long-term operation, material aging, and oil contamination, thus ensuring the long-term consistency and reliability of the measurement results throughout the entire life cycle of the monitoring system.

[0035] In some embodiments, obtaining the moisture content parameter by querying a pre-stored oil-water balance table based on the corrected humidity value and the real-time oil temperature data includes: The pre-stored oil-water balance relationship table is recorded and matched according to the corrected humidity value and the corrected humidity value; If a balance record is found that matches the corrected humidity value and the real-time oil temperature data, the moisture content parameter in the balance record is extracted. If no balance record is found, the humidity reference value that is closest to the corrected humidity value and the temperature reference value that is closest to the real-time oil temperature data are found in the pre-stored oil-water balance relationship table. Specifically, when an exact match fails, the humidity index sequence in the relation table is traversed and compared for the corrected humidity value. By calculating the absolute difference between the corrected humidity value and each humidity reference value in the sequence, the humidity reference value with the smallest difference that is less than the corrected humidity value is identified and recorded as the first lower humidity reference value; simultaneously, the humidity reference value with the smallest difference that is greater than the corrected humidity value is identified and recorded as the first upper humidity reference value. The same operation is performed on the real-time oil temperature data, identifying the temperature reference value closest to and less than this temperature value in the temperature index sequence as the first lower temperature reference value, and the temperature reference value closest to and greater than this temperature value as the first upper temperature reference value. This search process ensures that the four selected reference values ​​(two humidity values ​​and two temperature values) numerically closely surround the actual measurement point, constructing a basic rectangular interpolation region for the next step of local approximation in the two-dimensional plane.

[0036] Based on the humidity reference value and the temperature reference value, the corresponding moisture content reference value is located in the pre-stored oil-water balance relationship table; Specifically, based on the obtained lower humidity limit reference value, upper humidity limit reference value, lower temperature limit reference value, and upper temperature limit reference value, the oil-water balance relationship table can be logically viewed as a two-dimensional matrix. Its rows are indexed by the temperature reference value sequence, and its columns are indexed by the humidity reference value sequence. The lower temperature limit reference value is used to locate the corresponding matrix row. Then, within that row, the lower and upper humidity limit reference values ​​are used to locate the corresponding two matrix columns, thereby extracting the two moisture content reference values ​​located at the intersection of that row and these two columns. These are denoted as the first benchmark value corresponding to (lower temperature limit, lower humidity limit) and the second benchmark value corresponding to (lower temperature limit, upper humidity limit), respectively. Similarly, the upper temperature limit reference value is used to locate another matrix row, and the moisture content reference values ​​at the intersection of that row and the same two humidity reference values ​​are extracted, denoted as the third benchmark value corresponding to (upper temperature limit, lower humidity limit) and the fourth benchmark value corresponding to (upper temperature limit, upper humidity limit), respectively. Thus, the standard moisture content data at the four vertices of the rectangular area enclosed by the four reference coordinate points are successfully obtained.

[0037] The moisture content parameter is obtained by performing bilinear interpolation on the moisture content reference value based on the corrected humidity value and the real-time oil temperature.

[0038] Specifically, the coordinates of the punctuated points, i.e., the corrected humidity value and the real-time oil temperature value; and the coordinates and data of the four vertices of the rectangular interpolation area determined by the previous steps, including the lower humidity reference value, upper humidity reference value, lower temperature reference value, upper temperature reference value, and the moisture content reference values ​​corresponding to these four vertices, are respectively denoted as the first reference value corresponding to (lower temperature reference value, lower humidity reference value), the second reference value corresponding to (lower temperature reference value, upper humidity reference value), the third reference value corresponding to (upper temperature reference value, lower humidity reference value), and the fourth reference value corresponding to (upper temperature reference value, upper humidity reference value). On the horizontal line determined by the lower temperature reference value, a first linear interpolation is performed for the humidity dimension. Using the first and second reference values, a preliminary estimated moisture content is calculated when the temperature is fixed at the lower temperature reference value and the humidity is the corrected humidity value of the actual value. This value can be called the temporary estimated value on the low-temperature side. Next, on another horizontal line determined by the upper temperature reference value, the above linear interpolation process for the humidity dimension is repeated. Using the third and fourth baseline values, another preliminary estimate of moisture content is calculated when the temperature is fixed at the upper limit reference value and the humidity is the actual value corrected for humidity. This value can be called the temporary estimate for the high-temperature side. Finally, a second, final linear interpolation is performed along the temperature dimension. Using the current real-time oil temperature as the target point, linear interpolation is performed using the temporary estimates for the low-temperature side and the temporary estimates for the high-temperature side obtained in the previous step. This yields a continuous and smooth moisture content parameter corresponding to the target coordinate point (real-time oil temperature value, corrected humidity value).

[0039] The method provided in this embodiment, through precise record matching and querying, can directly extract the corresponding standard moisture content parameters when the corrected humidity value and real-time oil temperature data coincide with the experimental calibration points pre-stored in the oil-water balance relationship table, thereby completely avoiding any secondary errors introduced by calculation. By automatically finding the closest humidity and temperature reference values ​​and locating the corresponding four moisture content reference values ​​when a precise match is not possible, it ensures that the estimation process is always based on the most relevant original experimental data, laying the foundation for accurate calculation.

[0040] In some embodiments, the step of correcting the temperature-compensated humidity value for time drift based on the time drift correction coefficient to obtain the corrected humidity value includes: Multiply the temperature-compensated humidity value by the time-drift correction coefficient to obtain the first intermediate humidity value; The difference between the first intermediate humidity value and the pre-stored baseline humidity value is calculated to obtain the humidity difference. Determine whether the humidity difference exceeds the preset time drift correction range; If the humidity difference does not exceed the preset time drift correction range, then the first intermediate humidity value is used as the corrected humidity value; If the humidity difference exceeds the time drift correction threshold, the first intermediate humidity value is iteratively adjusted according to the time drift correction coefficient and the preset correction step size to obtain the corrected humidity value.

[0041] Specifically, the current first intermediate humidity value and the time-drift correction coefficient are used as the initial state for the iteration. The goal of the adjustment is to ensure that the new difference between the calculated humidity value after the new coefficient correction and the pre-stored baseline humidity value falls within the preset time-drift correction range. The adjustment logic is based on the correction of the time-drift correction coefficient: if the original humidity difference exceeds the positive threshold (i.e., the first intermediate humidity value is much higher than the baseline), it means that the original time-drift correction coefficient may be too large and needs to be appropriately reduced; conversely, if it exceeds the negative threshold, it may need to be increased. Based on the direction of the deviation, the original time-drift correction coefficient is fine-tuned in reverse with a small preset correction step size, generating a new correction coefficient. This new coefficient is then multiplied again with the original temperature-compensated humidity value to obtain a new, adjusted intermediate humidity value. The difference calculation and range judgment process is repeated. If the newly calculated humidity difference still does not fall within a reasonable range, the next fine-tuning is performed based on the new correction coefficient obtained from the previous iteration, either in the same direction or with a more refined strategy, and the result is verified again. This cyclical process continues until the humidity difference calculated after a certain iteration falls within a preset reasonable range, or until the preset maximum number of iterations is reached. When the iteration successfully meets the conditions, the intermediate humidity value calculated in the last iteration is finally determined as the corrected humidity value for that calculation cycle.

[0042] The method provided in this embodiment achieves rapid and effective basic compensation for linear performance drift caused by long-term operation of the sensor by directly multiplying the temperature-compensated humidity value by the time-drift correction coefficient derived from the cumulative operating time. This allows for efficient output of a pre-corrected, reasonable humidity value under most normal aging conditions. By calculating the difference between this pre-corrected result and the pre-stored reference humidity value and determining whether it exceeds a preset safety range, a crucial safety verification mechanism is introduced into the correction process. This effectively identifies abnormal correction results caused by sudden sensor degradation or calibration parameter inaccuracies, preventing severe data distortion in the output.

[0043] In some embodiments, the step of periodically dividing the cumulative runtime to obtain a unit period, and calculating the periodic change rate of the moisture content parameter based on the unit period, includes: The cumulative runtime is divided into multiple unit periods according to a preset time window length; The cumulative runtime is associated with each of the moisture content parameters, and the corresponding moisture content parameter is obtained as a characterization value based on the unit period. The content difference between two adjacent characterization content values ​​is calculated sequentially, and the content difference is divided by the unit period to obtain the rate of change.

[0044] Specifically, the generated ordered sequence of characterization content values ​​is accessed. Starting with the second characterization content value in the sequence, the characterization content value of the current period is subtracted from the characterization content value of the previous period, resulting in a series of sequentially arranged content differences. Each content difference uniquely corresponds to a transition interval of an adjacent period pair. For each calculated content difference, it is divided by the unit period length used as the standard time base (e.g., if the unit period is 24 hours, then divide by 24). The absolute change is converted into the average rate of change per unit time, i.e., the rate of change. For example, if the characterization content value increases from 30 ppm to 33 ppm in two adjacent periods (in days), the content difference is 3 ppm, which, divided by the unit period of 1 day, yields a rate of change of 3 ppm / day. This calculation is repeated for all adjacent period pairs in the sequence, generating a rate of change equal to the number of adjacent period pairs.

[0045] The method provided in this embodiment establishes a time-based framework for dynamic trend analysis by dividing the continuous cumulative runtime into discrete unit periods according to a preset time window. This organizes the disordered continuous monitoring data stream into ordered data units that can be processed in batches periodically. By associating each moisture content parameter with the corresponding unit period based on the timestamp and extracting the characteristic content value within each period, effective dimensionality reduction and feature processing of high-frequency monitoring data is achieved. While preserving the overall moisture level characteristics of each period, random fluctuations and measurement noise within the period are filtered out, highlighting the macroscopic state evolution across time scales. By sequentially calculating the difference in characteristic content values ​​between adjacent periods and dividing it by the unit period length, the average rate of change of moisture content between different analysis time windows can be accurately quantified, i.e., the rate of change. This transforms static moisture level information into a dynamic trend intensity indicator.

[0046] In some embodiments, the step of classifying and comparing the moisture content parameter and the rate of change based on preset early warning rules, and outputting an early warning signal, includes: The moisture content parameter is compared with a first moisture threshold and a second moisture threshold of the preset warning rule, wherein the first moisture threshold is less than the second moisture threshold; If the moisture content parameter is less than the first moisture threshold, the current state is determined to be normal and no alarm is triggered. If the moisture content parameter is not less than the first moisture threshold but less than the second moisture threshold, then it is determined whether the rate of change exceeds the rate threshold of the preset warning rule. When the rate of change does not exceed the rate threshold, a first-level warning signal is triggered; When the rate of change exceeds the rate threshold, a second-level warning signal is triggered; If the moisture content parameter is not less than the second moisture threshold, the third-level alarm signal is directly triggered.

[0047] The first-level warning signal indicates that the moisture content parameter in the oil has exceeded the permissible normal baseline (first moisture threshold), but has not yet reached a high-risk level (second moisture threshold), and the rate of change is gradual, without showing an accelerating upward trend. The monitoring system marks the transformer's status as "attention," automatically increasing the frequency of recording and tracking moisture data for this equipment.

[0048] The Level 2 warning signal indicates that the moisture content parameter is within the same range as above, but the rate of change has exceeded the safety threshold, meaning that the moisture is increasing rapidly and there may be a risk of sudden water ingress or accelerated deterioration of insulation. An immediate maintenance pre-notification should be issued through the monitoring system, initiating preliminary preparations for the emergency response plan.

[0049] The Level 3 alarm signal indicates that the moisture content parameter has exceeded the absolute safety limit (the second moisture threshold, which typically corresponds to the critical point of significant decrease in insulation strength), and an insulation fault may occur at any time. The highest level alarm is immediately triggered, and an emergency alarm message is forcibly sent through the monitoring system and communication network.

[0050] This embodiment compares the moisture content parameter with preset first and second moisture thresholds in a tiered manner, enabling early identification of abnormal conditions when the moisture level just exceeds the normal range but has not yet reached a dangerous level. By further introducing a comparison between the rate of change and a rate threshold within the aforementioned abnormal range, it can effectively distinguish between two fundamentally different risk modes: a slow increase in moisture and a rapid, sharp increase. This overcomes the limitation of a single absolute value threshold alarm, which cannot assess the urgency of the risk. Finally, based on the dual judgment results of the moisture content level and its changing trend, differentiated signals are triggered, ranging from low-level warnings to high-level alarms, significantly improving the accuracy and operability of alarm information.

[0051] Reference Figure 2 As shown, the present invention also provides an online monitoring device for moisture in transformer oil, applied to the online monitoring method for moisture in transformer oil described in any of the above-mentioned claims, comprising: The acquisition module is used to acquire the capacitance signal and real-time oil temperature data output by the sensing device in the transformer oil, and convert the capacitance signal into a frequency signal. The analysis module is used to calculate the moisture content parameter by analyzing the frequency signal and the real-time oil temperature based on the cumulative running time of the sensing device. The association module is used to divide the cumulative running time into periods to obtain unit periods, and to calculate the periodic change of the moisture content parameter based on the unit periods to obtain the rate of change. The processing module is used to perform hierarchical comparison of the moisture content parameter and the rate of change based on preset early warning rules, and output early warning signals.

[0052] This invention provides an online monitoring device for moisture in transformer oil. By calculating moisture content based on frequency signals and oil temperature data according to the cumulative operating time of the sensing device, it can dynamically compensate for performance drift caused by long-term sensor operation, thereby significantly improving the measurement accuracy and long-term reliability of moisture content parameters throughout the entire equipment lifecycle. By periodically segmenting the cumulative operating time and calculating the rate of change of moisture content parameters, it achieves quantitative capture of the dynamic trend of moisture changes in the oil, thus distinguishing different risk development modes such as slow seepage and sudden water ingress. Based on this, by classifying and jointly judging the moisture content parameters and their rate of change according to preset early warning rules, it can output more discriminative early warning signals, thereby achieving earlier and more accurate identification and graded early warning of potential fault risks.

[0053] Reference Figure 3 As shown, the present invention also provides an online monitoring system for moisture in transformer oil, comprising: Memory, used to store programs; A processor is used to execute the program to implement the various steps of the online monitoring method for moisture in transformer oil as described in any of the above-mentioned methods.

[0054] In this embodiment, the processor and memory can be connected via a bus or other means. The memory may include volatile memory, such as random access memory; the memory may also include non-volatile memory, such as read-only memory, flash memory, hard disk, or solid-state drive. The processor may be a general-purpose processor, such as a central processing unit, digital signal processor, application-specific integrated circuit, or one or more integrated circuits configured to implement embodiments of the present invention.

[0055] The present invention also provides a storage medium storing computer instructions for causing a computer to perform any of the methods described above.

[0056] It should be noted that those skilled in the art will understand that, for the sake of convenience and brevity, the specific working processes of the system and each module described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0057] The above description is only a preferred embodiment of the present invention and does not limit the patent scope of the present invention. Any equivalent structural or procedural transformations made based on the content of the present invention specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of the present invention.

Claims

1. A method for online monitoring of moisture in transformer oil, characterized in that, include: Acquire the capacitance signal and real-time oil temperature data output by the sensing device in the transformer oil, and convert the capacitance signal into a frequency signal; Based on the cumulative operating time of the sensing device, the moisture content parameter is obtained by calculating the moisture content of the frequency signal and the real-time oil temperature. The cumulative runtime is divided into periods to obtain unit periods, and the periodic change of the moisture content parameter is calculated based on the unit periods to obtain the rate of change. The moisture content parameter and the rate of change are compared and classified according to preset early warning rules, and an early warning signal is output.

2. The online monitoring method for moisture in transformer oil according to claim 1, characterized in that, The process of acquiring the capacitance signal and real-time oil temperature data output by the sensing device in the transformer oil, and converting the capacitance signal into a frequency signal, includes: The capacitance signal and the real-time oil temperature data are acquired periodically by the sensing device. The capacitor signal is frequency-converted based on a preset capacitor frequency conversion table to obtain an initial frequency pulse signal. The initial frequency pulse signal is bound to the real-time oil temperature data by sampling points according to the preset capacitor frequency conversion table to obtain the frequency signal.

3. The online monitoring method for moisture in transformer oil according to claim 1, characterized in that, The step of calculating moisture content parameters by analyzing the frequency signal and the real-time oil temperature data based on the cumulative operating time of the sensing device includes: The initial humidity parameters are obtained by querying the pre-stored first mapping relationship based on the frequency signal; Based on the real-time oil temperature data, the pre-stored temperature compensation parameter table is queried to obtain the temperature compensation coefficient; The initial humidity parameter is initially compensated based on the temperature compensation coefficient to obtain the temperature-compensated humidity value. Based on the accumulated runtime, the time drift compensation parameter table is queried to obtain the time drift correction coefficient; The temperature-compensated humidity value is corrected for time drift based on the time drift correction coefficient to obtain the corrected humidity value; The moisture content parameter is obtained by querying a pre-stored oil-water balance table based on the corrected humidity value and the real-time oil temperature data.

4. The online monitoring method for moisture in transformer oil according to claim 3, characterized in that, The step of querying a pre-stored oil-water balance table based on the corrected humidity value and the real-time oil temperature data to obtain the moisture content parameter includes: The pre-stored oil-water balance relationship table is recorded and matched according to the corrected humidity value and the corrected humidity value; If a balance record is found that matches the corrected humidity value and the real-time oil temperature data, the moisture content parameter in the balance record is extracted. If no balance record is found, the humidity reference value that is closest to the corrected humidity value and the temperature reference value that is closest to the real-time oil temperature data are found in the pre-stored oil-water balance relationship table. Based on the humidity reference value and the temperature reference value, the corresponding moisture content reference value is located in the pre-stored oil-water balance relationship table; The moisture content parameter is obtained by performing bilinear interpolation on the moisture content reference value based on the corrected humidity value and the real-time oil temperature.

5. The online monitoring method for moisture in transformer oil according to claim 3, characterized in that, The step of correcting the temperature-compensated humidity value for time drift based on the time drift correction coefficient to obtain the corrected humidity value includes: Multiply the temperature-compensated humidity value by the time-drift correction coefficient to obtain the first intermediate humidity value; The difference between the first intermediate humidity value and the pre-stored baseline humidity value is calculated to obtain the humidity difference. Determine whether the humidity difference exceeds the preset time drift correction range; If the humidity difference does not exceed the preset time drift correction range, then the first intermediate humidity value is used as the corrected humidity value; If the humidity difference exceeds the time drift correction threshold, the first intermediate humidity value is iteratively adjusted according to the time drift correction coefficient and the preset correction step size to obtain the corrected humidity value.

6. The online monitoring method for moisture in transformer oil according to claim 1, characterized in that, The step of dividing the cumulative runtime into periodic segments to obtain unit periods, and calculating the periodic change rate of the moisture content parameter based on the unit periods, includes: The cumulative runtime is divided into multiple unit periods according to a preset time window length; The cumulative runtime is associated with each of the moisture content parameters, and the corresponding moisture content parameter is obtained as a characterization value based on the unit period. The content difference between two adjacent characterization content values ​​is calculated sequentially, and the content difference is divided by the unit period to obtain the rate of change.

7. The online monitoring method for moisture in transformer oil according to claim 1, characterized in that, The step of classifying and comparing the moisture content parameter and the rate of change based on preset early warning rules, and outputting an early warning signal, includes: The moisture content parameter is compared with a first moisture threshold and a second moisture threshold of the preset warning rule, wherein the first moisture threshold is less than the second moisture threshold; If the moisture content parameter is less than the first moisture threshold, the current state is determined to be normal and no alarm is triggered. If the moisture content parameter is not less than the first moisture threshold but less than the second moisture threshold, then it is determined whether the rate of change exceeds the rate threshold of the preset warning rule. When the rate of change does not exceed the rate threshold, a first-level warning signal is triggered; When the rate of change exceeds the rate threshold, a second-level warning signal is triggered; If the moisture content parameter is not less than the second moisture threshold, the third-level alarm signal is directly triggered.

8. An online monitoring device for moisture in transformer oil, characterized in that, The online monitoring method for moisture in transformer oil according to any one of claims 1-7 includes: The acquisition module is used to acquire the capacitance signal and real-time oil temperature data output by the sensing device in the transformer oil, and convert the capacitance signal into a frequency signal. The analysis module is used to calculate the moisture content parameter by analyzing the frequency signal and the real-time oil temperature based on the cumulative running time of the sensing device. The association module is used to divide the cumulative running time into periods to obtain unit periods, and to calculate the periodic change of the moisture content parameter based on the unit periods to obtain the rate of change. The processing module is used to perform hierarchical comparison of the moisture content parameter and the rate of change based on preset early warning rules, and output early warning signals.

9. An online monitoring system for moisture in transformer oil, characterized in that, include: Memory, used to store programs; A processor is configured to execute the program to implement the various steps of the online monitoring method for moisture in transformer oil as described in any one of claims 1-7.

10. A storage medium, characterized in that, The computer contains computer instructions for causing the computer to perform the method according to any one of claims 1 to 7.

Images