Transformer maintenance-free respirator control method integrating temperature, humidity and pressure monitoring and electric heating regeneration and storage medium

Abstract

The invention relates to the field of intelligent operation and maintenance of power equipment, in particular to a transformer maintenance-free respirator control method integrating temperature, humidity and pressure monitoring and electric heating regeneration and a storage medium. The method comprises the following steps: firstly, deploying a sensor in the respirator, collecting humidity and internal temperature and pressure data of a drying agent, and obtaining accurate data after moving average filtering and temperature compensation; and the control module is combined with the data, a humidity threshold value dynamically adjusted along with the load rate of the transformer and a regeneration period to trigger the regeneration system to start. During regeneration, the electric heating piece heats the drying agent, the fan adjusts the rotating speed according to the real-time temperature and humidity to discharge steam, meanwhile, the moisture absorption capacity recovery rate is corrected by combining aging of the drying agent, and starting and stopping or temperature adjusting delay are conducted according to the standard condition. The process data is uploaded to the operation and maintenance cloud platform through the LoRa module encrypted by AES-128, and the platform analyzes the trend through weighted linear regression and gives an alarm; and operation and maintenance personnel optimize parameters according to the parameters and regurgitate a judgment link. Related programs are stored in the storage medium and can be executed by the processor to achieve the method, and support is provided for maintenance-free of the respirator.

Description

A control method and storage medium for a transformer maintenance-free breather integrating temperature, humidity, and pressure monitoring with electric heating regeneration. Technical Field

[0001] This invention relates to the field of intelligent operation and maintenance of power equipment, and in particular to a predictive maintenance method for transformer breathers combined with edge computing and a storage medium. Background Technology

[0002] As the core hub of power transmission and voltage transformation in the power system, the transformer's operational stability directly affects the reliability and quality of power supply. The breather, a key auxiliary device of the transformer, plays a crucial role in balancing the air pressure inside and outside the oil tank and adsorbing moisture and impurities from the air. It effectively prevents contaminants from entering the oil tank, leading to deterioration of the insulating oil and a decrease in insulation performance, thus ensuring the long-term safe operation of the transformer. Traditional transformer breathers mostly adopt a passive structure of "oil cup + silica gel / activated alumina desiccant," relying on manual periodic inspections to determine the desiccant's moisture absorption status and replace or regenerate it. This model has gradually revealed many drawbacks in practical applications, making it difficult to adapt to the intelligent and efficient operation and maintenance requirements of modern power grids. From an operation and maintenance perspective, the maintenance cycle of traditional breathers is fixed and cannot be dynamically adjusted according to differences in ambient temperature and humidity or transformer load fluctuations, easily leading to problems of "over-maintenance" or "under-maintenance." Furthermore, the judgment of the desiccant's color change during manual inspections is easily affected by light and angle, posing a risk of misjudgment. Additionally, the "vacuum period" during replacement, when the breather is disconnected from the oil tank, allows external air to directly intrude, increasing the burden on the insulation system.

[0003] In terms of condition monitoring capabilities, traditional respirators lack sensing and data recording functions, making it impossible to quantitatively monitor key parameters such as desiccant humidity and internal temperature and pressure. They can only judge the condition by the appearance color, making it difficult to accurately grasp the actual moisture absorption rate of the desiccant. When faults such as seal failure or desiccant clumping block the airflow channel occur, they cannot provide timely warnings. Often, the root cause of the fault can only be traced after the transformer insulation becomes abnormal, missing the opportunity for early troubleshooting. Furthermore, they cannot make adaptive adjustments based on differences in altitude, ambient temperature and humidity. In high-altitude areas with low air pressure and reduced moisture absorption efficiency, they are prone to insufficient moisture absorption capacity. From the perspective of operation and maintenance efficiency and cost, with the expansion of the power grid and the surge in the number of transformers, especially in distributed power stations and substations in remote areas, the traditional manual inspection mode requires a large investment of manpower and transportation costs. Breather maintenance accounts for 15%-20% of the operation and maintenance workload, resulting in a heavy cost burden in the long run. Moreover, the operation and maintenance method that relies on paper records cannot form historical data trend analysis, making it difficult to optimize maintenance strategies. It can only maintain a "one-size-fits-all" fixed cycle management, which is seriously out of step with the current trend of power system transformation towards intelligence and digitalization. Therefore, developing a transformer maintenance-free breather control method that integrates real-time monitoring, automatic regeneration, and intelligent control functions has become an inevitable requirement to solve the pain points of traditional breathers and improve the reliability and operation and maintenance efficiency of transformers. Summary of the Invention

[0004] A control method for a transformer maintenance-free breather integrating temperature, humidity, and pressure monitoring with electric heating regeneration includes the following steps:

[0005] S1: Deploy humidity, temperature, and pressure sensors inside the respirator to collect real-time data on desiccant humidity and internal temperature and pressure. The data is then transmitted to the control module, processed by moving average filtering, and outputs accurate monitoring data.

[0006] S2: The control module combines the precise monitoring data, preset humidity threshold and regeneration cycle to make a judgment. If any condition is met, it sends a start command to the regeneration system; otherwise, it continues to receive data and repeats the judgment.

[0007] S3: The control module starts the electric heating element of the regeneration system to heat the desiccant according to the start command, and links the fan and adjusts the speed according to the real-time temperature of S1 to discharge the evaporated water vapor;

[0008] S4: During regeneration, the humidity sensor feeds back the humidity data of the desiccant. The control module determines whether the moisture absorption capacity recovery rate meets the standard and then performs the corresponding operation. If it meets the standard, a stop command is issued; if it does not meet the standard, the temperature is adjusted or the time is delayed.

[0009] S5: The control module uploads the data from S1 to S4 to the power operation and maintenance cloud platform via the LoRa module. The platform stores the data, analyzes trends, and alerts on anomalies.

[0010] S6: Maintenance personnel judge the adaptability of the working conditions based on platform reports, alarms and inspections. If there are problems, upgrade and optimize; if normal, adjust the default parameters of S2 and feed back into the judgment process of S2.

[0011] Preferably, in step S1, the moving average filtering process is combined with ambient temperature compensation, and the algorithm formula is as follows:

[0012] ;

[0013] in, This is the accurate monitoring data after the nth filtering. This is the length of the sliding window, ranging from 5 to 15. For the i-th original data collection, Temperature compensation coefficient, humidity monitoring Pressure monitoring ; The temperature is the external temperature of the respirator, and the data acquisition interval should not exceed 30 seconds.

[0014] Preferably, the preset humidity threshold in step S2 is dynamically adjusted according to the transformer load rate:

[0015] Level 1 threshold: ;

[0016] Secondary threshold: ;

[0017] in , As the baseline threshold, This is the load impact factor. This represents the real-time load power of the transformer. This refers to the transformer's rated power; when the desiccant humidity is... Regeneration should be started immediately when... And the time since the last regeneration exceeds the preset cycle. Regeneration is initiated at the specified time. .

[0018] Preferably, in step S4, the moisture absorption capacity recovery rate incorporates desiccant aging correction, and the calculation method is as follows:

[0019] ;

[0020] in, Moisture absorption capacity recovery rate (standard threshold) ), Humidity before regeneration Real-time humidity during regeneration Standard dry humidity, , is the aging coefficient. This refers to the cumulative usage time of the desiccant.

[0021] Preferably, in step S4, the control module determines whether the moisture absorption capacity recovery rate meets the standard, and then performs the corresponding operation. The specific logic is as follows: If heating temperature rise value Regeneration delay ;like First calculate the pressure deviation. ,like Then according to Calibrate the pressure sensor and then restart the regeneration process. This is the calibration coefficient.

[0022] Preferably, in step S5, the data uploaded by the LoRa module is encrypted using AES-128, and the data frame includes a timestamp, device number, sensor type, monitoring data, checksum, and signal strength value. Upload cycle ,when At this time, data compression and uploading are triggered to ensure data transmission stability in weak signal environments.

[0023] Preferably, in step S3, the speed of the linked fan is adjusted according to two parameters: real-time temperature and humidity. The logic is as follows: set the real-time temperature... Real-time humidity Preset heating temperature range Humidity correction factor ;when At that time, rotational speed ;when hour, ;when hour, ;in , , Units are .

[0024] Preferably, in step S5, the trend analysis uses a weighted linear regression algorithm to predict the time when humidity reaches the threshold.

[0025] ;

[0026] in, The current humidity. Let be the rate of increase in humidity on day i. As time weight, when Hours and At that time, an emergency warning was triggered.

[0027] Preferably, in step S6, parameter optimization takes into account the influence of ambient humidity. If the regeneration frequency increases by more than 15% for every 10% increase in load rate over the past 6 months, then... Lower value , shortening value ;in The average monthly ambient humidity in the maintenance area, when Time to take , .

[0028] A computer read-only storage medium stores one or more programs, which are executed by one or more processors to implement the above-described integrated temperature, humidity, and pressure monitoring and electric heating regeneration transformer maintenance-free respirator control method.

[0029] The beneficial effects of this invention are as follows:

[0030] 1. This invention collects real-time data from the respirator using three types of sensors: humidity, temperature, and pressure. The data is then filtered out by moving average filtering and combined with temperature compensation to correct for environmental influences. The final output is accurate data with small errors, which can accurately reflect the moisture absorption state of the desiccant and the internal temperature and pressure environment of the respirator, avoiding misjudgments of the operating status due to a single parameter or data deviation.

[0031] 2. The humidity threshold of this invention is dynamically adjusted according to the real-time load rate of the transformer, avoiding the problems of delayed regeneration under high load and frequent regeneration under low load under a fixed threshold; the fan speed is adjusted according to the real-time temperature and humidity dual parameters, which reduces ineffective energy consumption and can efficiently discharge evaporated water vapor; during the regeneration process, the aging degree is judged by the cumulative usage time of the desiccant, thereby assessing the moisture absorption capacity recovery rate. If the recovery does not meet the standard, the heating temperature is increased or the regeneration time is extended in time. At the same time, the pressure sensor is calibrated to ensure accurate monitoring and effectively extend the service life of the desiccant.

[0032] 3. The data from the monitoring and regeneration process of this invention is uploaded to the power operation and maintenance cloud platform via LoRa module. The platform analyzes the humidity change trend and predicts the time to reach the threshold through algorithms, and issues timely alarms when abnormalities occur. The operation and maintenance personnel optimize the preset parameters based on the platform report and feed them back into the regeneration judgment process, reducing the frequency and blindness of manual on-site inspections. At the same time, the parameter optimization will be combined with the monthly average environmental humidity of the operation and maintenance area to adapt to the operating needs under different environmental conditions.

[0033] 4. The LoRa module of this invention uses AES-128 encryption to protect uploaded data, and is equipped with a data frame checksum to prevent data theft or tampering, ensuring the security of operation and maintenance data. When a weak signal is detected, data compression and uploading will be automatically triggered to ensure stable data transmission in weak signal environments without losing critical information, and to avoid data loss due to signal problems.

[0034] 5. This invention achieves automated intelligent control from real-time monitoring and intelligent regeneration to cloud analysis and parameter optimization, eliminating the need for frequent manual intervention throughout the entire process and significantly reducing manual operation costs. At the same time, precise control of each link reduces the probability of respirator failure, extends the overall service life of the equipment, and indirectly improves the stability and reliability of transformer operation, truly achieving the goal of maintenance-free design. Attached Figure Description

[0035] Figure 1 is a flowchart of the steps of the transformer maintenance-free breather control method integrating temperature, humidity and pressure monitoring and electric heating regeneration according to the present invention. Detailed Implementation

[0036] A control method for a transformer maintenance-free breather integrating temperature, humidity, and pressure monitoring with electric heating regeneration includes the following steps:

[0037] S1: Deploy humidity, temperature, and pressure sensors inside the respirator to collect real-time data on desiccant humidity and internal temperature and pressure. The data is then transmitted to the control module, processed by moving average filtering, and outputs accurate monitoring data.

[0038] S2: The control module combines accurate monitoring data, preset humidity threshold and regeneration cycle to make judgments. If any condition is met, it sends a start command to the regeneration system; otherwise, it continues to receive data and repeat the judgment.

[0039] S3: The control module starts the electric heating element of the regeneration system to heat the desiccant according to the start command, and links the fan and adjusts the speed according to the real-time temperature of S1 to discharge the evaporated water vapor;

[0040] S4: During regeneration, the humidity sensor feeds back the humidity data of the desiccant. The control module determines whether the moisture absorption capacity recovery rate meets the standard and then performs the corresponding operation. If it meets the standard, a stop command is issued; if it does not meet the standard, the temperature is adjusted or the time is delayed.

[0041] S5: The control module uploads the data from S1 to S4 to the power operation and maintenance cloud platform via the LoRa module. The platform stores the data, analyzes trends, and alerts on anomalies.

[0042] S6: Maintenance personnel judge the adaptability of the working conditions based on platform reports, alarms and inspections. If there are problems, upgrade and optimize; if normal, adjust the default parameters of S2 and feed back into the judgment process of S2.

[0043] In step S1, the moving average filtering process is combined with ambient temperature compensation. The algorithm formula is as follows:

[0044] ;

[0045] in, This is the accurate monitoring data after the nth filtering. This is the length of the sliding window, ranging from 5 to 15. For the i-th original data collection, Temperature compensation coefficient, humidity monitoring Pressure monitoring ; The temperature is the external temperature of the respirator, and the data acquisition interval should not exceed 30 seconds.

[0046] In step S2, the preset humidity threshold is dynamically adjusted according to the transformer load rate:

[0047] Level 1 threshold: ;

[0048] Secondary threshold: ;

[0049] in , As the baseline threshold, This is the load impact factor. This represents the real-time load power of the transformer. This refers to the transformer's rated power; when the desiccant humidity is... Regeneration should be started immediately when... And the time since the last regeneration exceeds the preset cycle. Regeneration is initiated at the specified time. .

[0050] In step S4, the moisture absorption capacity recovery rate is adjusted for desiccant aging, and the calculation method is as follows:

[0051] ;

[0052] in, Moisture absorption capacity recovery rate (standard threshold) ), Humidity before regeneration Real-time humidity during regeneration Standard dry humidity, , is the aging coefficient. This refers to the cumulative usage time of the desiccant.

[0053] In step S4, the control module determines whether the moisture absorption capacity recovery rate meets the standard, and then performs the corresponding operation. The specific logic is as follows: If... heating temperature rise value Regeneration delay ;like First calculate the pressure deviation. ,like Then according to Calibrate the pressure sensor and then restart the regeneration process. This is the calibration coefficient.

[0054] In step S5, the LoRa module uploads data using AES-128 encryption, and the data frame includes a timestamp, device number, sensor type, monitoring data, checksum, and signal strength value. Upload cycle ,when At this time, data compression and uploading are triggered to ensure data transmission stability in weak signal environments.

[0055] In step S3, the speed of the linked fan is adjusted according to the real-time temperature and humidity dual parameters. The logic is as follows: set the real-time temperature... Real-time humidity Preset heating temperature range Humidity correction factor ;when At that time, rotational speed ;when hour, ;when hour, ;in , , Units are .

[0056] In step S5, the trend analysis uses a weighted linear regression algorithm to predict the time when humidity reaches the threshold.

[0057] ;

[0058] in, The current humidity. Let be the rate of increase in humidity on day i. As time weight, when Hours and At that time, an emergency warning was triggered.

[0059] In step S6, parameter optimization takes into account the influence of ambient humidity. If the regeneration frequency increases by more than 15% for every 10% increase in load rate over the past 6 months, then... Lower value , shortening value ; in The average monthly ambient humidity in the maintenance area, when Time to take , .

[0060] A computer read-only storage medium stores one or more programs, which are executed by one or more processors to implement the above-described integrated temperature, humidity, and pressure monitoring and electric heating regeneration transformer maintenance-free respirator control method.

[0061] The embodiments of the present invention are as follows:

[0062] In an application scenario involving a maintenance-free breather for a 110kV outdoor box-type transformer, the following basic parameters are set: the transformer's rated power Prated is 50MVA, the real-time load power Pcurr is 30MVA, and the load rate is 60%; the average monthly ambient humidity Henv in the maintenance area is 65%RH; the external ambient temperature Tenv of the breather is 32℃; the cumulative usage time of the desiccant tuse is 1200h, and the aging coefficient δ is 0.0001 / h; the sliding window length k is 10; the humidity monitoring temperature compensation coefficient α is 0.005 / ℃; the preset heating temperature range is 60℃ to 120℃; the minimum fan speed Vmin is 1200r / min, the maximum speed Vmax is 3000r / min, and the humidity correction coefficient γ is 0.008.

[0063] In step S1, the sensor collects data every 20 seconds, an interval not exceeding 30 seconds. Ten sets of raw humidity data are continuously collected: 38%, 39%, 40%, 37%, 39%, 41%, 38%, 39%, 40%, and 38%RH. Simultaneously, ten sets of raw pressure data are collected: 98.2 kPa, 98.1 kPa, 98.3 kPa, 98.0 kPa, 98.2 kPa, 98.4 kPa, 98.1 kPa, 98.2 kPa, 98.3 kPa, and 98.1 kPa. First, the raw data is processed using the temperature compensation formula Xi×(1+α×(Tenv-25)), where the humidity compensation coefficient is calculated to be 1+0.005×(32-25)=1.035. Then, using the moving average filtering formula, the final output accurate humidity monitoring data is 40.2%RH, and the accurate pressure monitoring data is 98.2 kPa.

[0064] From step S2, it is calculated according to the dynamic threshold formula. The calculation formula for the primary threshold H1 is H1 = H10×(1 + β×(Pcurr / Prated)). Substituting H10 = 30%RH, β = 0.2, and a load rate of 60%, we get H1 = 30×(1 + 0.2×0.6) = 33.6%RH. The calculation formula for the secondary threshold H2 is H2 = H20×(1 + β×(Pcurr / Prated)). Substituting H20 = 50%RH, we get H2 = 50×(1 + 0.2×0.6) = 56%RH. At the same time, it is calculated according to the regeneration cycle formula T = 48×(1 - 0.1×(Pcurr / Prated)), and we get T = 48×(1 - 0.1×0.6) = 45.12h, approximately 45h. The current humidity of 40.2%RH satisfies the condition of H1≤H<H2, and it has been 46h since the last regeneration, exceeding the calculated cycle. Therefore, the control module sends a start instruction to the regeneration system.

[0065] In step S3, the control module starts the electric heating sheet, and the initial heating temperature target is set to 80°C, which is within the preset heating temperature range of 60°C to 120°C. At the same time, the control module联动风机 (the meaning of this part is not clear in Chinese, it may be something like "links with the fan"), and according to the real-time temperature Treal = 80°C and the real-time humidity Hreal = 40.2%RH, calculates the fan speed according to the speed formula. First, calculate the basic speed part:

[0066] 1200 + [(3000 - 1200) / (120 - 60)]×(80 - 60) = 1200 + 600 = 1800r / min. Then, perform humidity correction and calculate:

[0067] 1800×(1 + 0.008×40.2) = 1800×1.3216≈2379r / min, and the value is rounded here. The fan runs at a speed of 2379r / min to discharge the evaporated water vapor.

[0068] Enter step S4. Given that the humidity before regeneration H0 = 40.2%RH and the real-time humidity during regeneration Hreg = 8%RH, this humidity meets the standard drying humidity range of Hstd = 5% - 10%RH. Calculate according to the moisture absorption capacity recovery rate formula. Substitute Hstd = 8%RH into the data and get R = [(40.2 - 8) / ((40.2 - 8)×(1 - 0.0001×1200))]×100% = [32.2 / (32.2×0.88)]×100%≈113.6%. Since R≥85% of the passing threshold, the control module sends a stop instruction. <00,00274>​In the subsequent S5 step, the control module uploads the relevant data from S1 to S4 to the power operation and maintenance cloud platform via the LoRa module. The LoRa module uses AES-128 encryption when uploading data, and the data frame includes a timestamp, device number, sensor type, monitoring data, checksum, and RSSI signal strength value. After receiving the data, the platform stores it and uses a weighted linear regression algorithm to analyze trends. This algorithm calculates the predicted time when humidity will reach a threshold; if abnormal conditions are met, an alarm is triggered.

[0070] Finally, in step S6, maintenance personnel determine the respirator's operational suitability based on platform-generated reports, alarm information, and on-site inspection findings. Upon inspection, the respirator was found to be operating normally. The maintenance personnel then adjusted the preset parameters in step S2 and fed the adjusted parameters back into the S2 judgment process to optimize subsequent regeneration judgment logic.

Claims

1. A control method for a transformer maintenance-free breather integrating temperature, humidity, and pressure monitoring with electric heating regeneration, characterized in that, Includes the following steps: S1: Humidity, temperature, and pressure sensors are deployed inside the respirator to collect real-time data on desiccant humidity and internal temperature and pressure. This data is transmitted to the control module, processed by a moving average filter, and outputs accurate monitoring data. S2: The control module combines the accurate monitoring data with a preset humidity threshold and regeneration cycle. If any condition is met, a start command is sent to the regeneration system; otherwise, data is continuously received and the judgment is repeated. S3: The control module activates the electric heating element of the regeneration system according to the start command to heat the desiccant, and activates the fan, adjusting its speed according to the real-time temperature in S1 to expel evaporated water vapor. S4: During regeneration, the humidity sensor feeds back the humidity data of the desiccant. The control module determines whether the moisture absorption capacity recovery rate meets the standard and then performs the corresponding operation. If it meets the standard, a stop command is issued; otherwise, the temperature is adjusted or the delay is initiated. S5: The control module uploads the data from S1 to S4 to the power operation and maintenance cloud platform via the LoRa module. The platform stores the data, analyzes trends, and alarms for anomalies. S6: Maintenance personnel judge the adaptability of the working conditions based on platform reports, alarms and inspections. If there are problems, upgrade and optimize; if normal, adjust the default parameters of S2 and feed back into the judgment process of S2.

2. The method for controlling a transformer maintenance-free breather integrating integrated temperature, humidity, and pressure monitoring with electric heating regeneration according to claim 1, characterized in that, The moving average filtering process described in step S1, combined with ambient temperature compensation, has the following algorithm formula: ;in, This is the accurate monitoring data after the nth filtering. This is the length of the sliding window, ranging from 5 to 15. This is the original data collected in the i-th time. Temperature compensation coefficient, humidity monitoring Pressure monitoring ； The temperature is the external temperature of the respirator, and the data acquisition interval should not exceed 30 seconds.

3. The method for controlling a transformer maintenance-free breather integrating integrated temperature, humidity, and pressure monitoring with electric heating regeneration according to claim 1, characterized in that, The preset humidity threshold mentioned in step S2 is dynamically adjusted according to the transformer load rate: Level 1 threshold: Secondary threshold: ;in 、 As the baseline threshold, This is the load impact factor. This represents the real-time load power of the transformer. This refers to the transformer's rated power; when the desiccant humidity is... Regeneration should be started immediately when... And the time since the last regeneration exceeds the preset cycle. Regeneration is initiated at the specified time. 。 4. The method for controlling a transformer maintenance-free breather integrating integrated temperature, humidity, and pressure monitoring with electric heating regeneration according to claim 1, characterized in that, In step S3, the linked fan adjusts its speed according to two parameters: real-time temperature and humidity. The logic is as follows: Let the real-time temperature... Real-time humidity Preset heating temperature range Humidity correction factor ;when At that time, rotational speed ;when hour, ； when hour, ;in 、 ， Units are 。 5. The method for controlling a transformer maintenance-free breather integrating integrated temperature, humidity, and pressure monitoring with electric heating regeneration according to claim 1, characterized in that, In step S4, the moisture absorption capacity recovery rate incorporates desiccant aging correction, and the calculation method is as follows: ;in, The moisture absorption capacity recovery rate is the threshold for achieving the target. ， Humidity before regeneration Real-time humidity during regeneration Standard dry humidity, , is the aging coefficient. This refers to the cumulative usage time of the desiccant.

6. The transformer maintenance-free breather control method integrating temperature, humidity, and pressure monitoring with electric heating regeneration according to claim 4, characterized in that, In step S4, the control module determines whether the moisture absorption capacity recovery rate meets the standard, and then performs the corresponding operation. The specific logic is as follows: If heating temperature rise value Regeneration delay ； like First calculate the pressure deviation. ,like Then according to Calibrate the pressure sensor and then restart the regeneration process. This is the calibration coefficient.

7. The control method for a transformer maintenance-free breather integrating integrated temperature, humidity, and pressure monitoring with electric heating regeneration according to claim 1, characterized in that, In step S5, the data uploaded by the LoRa module is encrypted using AES-128, and the data frame includes a timestamp, device number, sensor type, monitoring data, checksum, and signal strength value. Upload cycle ,when At this time, data compression and uploading are triggered to ensure data transmission stability in weak signal environments.

8. The method for controlling a transformer maintenance-free breather integrating temperature, humidity, and pressure monitoring with electric heating regeneration according to claim 1, characterized in that, In step S5, the trend analysis uses a weighted linear regression algorithm to predict the time when humidity reaches the threshold. ;in, The current humidity. Let be the rate of increase in humidity on day i. As time weight, when Hours and At that time, an emergency warning was triggered.

9. The method for controlling a transformer maintenance-free breather integrating integrated temperature, humidity, and pressure monitoring with electric heating regeneration according to claim 1, characterized in that, In step S6, parameter optimization takes into account the influence of ambient humidity. If the regeneration frequency increases by more than 15% for every 10% increase in load rate over the past 6 months, then... Lower value ， shortening value ;in The average monthly ambient humidity in the maintenance area, when Time to take 、 。 10. A computer read storage medium, characterized in that, The computer read storage medium stores one or more programs, which are executed by one or more processors to implement the transformer maintenance-free respirator control method integrating temperature, humidity and pressure monitoring and electric heating regeneration as described in any one of claims 1 to 9.

Images