Discharge monitoring device and method for transformer tap changer oil chamber based on ultraviolet imaging

Abstract

The application belongs to the technical field of electric power engineering, and discloses a transformer on-load tap-changer internal discharge detection system and method, which comprises a monitoring and sensing module, a wireless transmission module, an information analysis module and a correction and early warning module. The monitoring and sensing module collects tap-changer internal discharge and external disturbance data, and is connected with the information analysis module. The information analysis module evaluates the equipment state according to current tap-changer ultraviolet discharge data. The information analysis module is connected with the correction and early warning module, which analyzes and early warns the defects existing in the tap-changer and the ultraviolet light sensor according to the monitoring data of the ultraviolet light sensor. The application realizes the evaluation of the running state of the tap-changer through the transformer on-load tap-changer discharge state detection.

Description

Technical Field

[0001] This invention belongs to the field of transformer condition monitoring technology, and relates to a transformer tap changer oil chamber discharge monitoring device and method based on ultraviolet imaging. Background Technology

[0002] In power systems, tap changers are critical components of power transformers. They are voltage regulating devices that can adjust the tap position of the transformer windings under transformer excitation or load conditions. With increased frequency of tap changer adjustments, the probability of tap changer failure also increases. Internal discharge is a common cause of tap changer failures; however, current monitoring systems are ineffective at detecting tap discharge. Therefore, to improve the efficiency of maintenance personnel, monitoring tap changers and early detection of internal discharge defects is of paramount importance for the safe and stable operation of power equipment. Summary of the Invention

[0003] To detect the discharge status of transformer tap changers during operation, this invention provides a transformer tap changer oil chamber discharge monitoring device and method based on ultraviolet imaging. An ultraviolet light sensor is preset according to the tap changer size to acquire historical data on the number of ultraviolet photons in the tap changer under historical operating conditions. The data is corrected based on the differences in the number of ultraviolet photons received by different ultraviolet light sensors to monitor the current discharge status inside the tap changer.

[0004] The present invention is implemented as follows. A transformer tap changer oil chamber discharge monitoring device based on ultraviolet imaging includes a monitoring and sensing module, a wireless transmission module, an information analysis module, and a correction and early warning module. The monitoring and sensing module collects data on internal discharge and external disturbances of the tap changer. The monitoring and sensing module is connected to the information analysis module. The information analysis module evaluates the device status based on the current ultraviolet discharge data of the tap changer. The information analysis module is connected to the correction and early warning module, which analyzes and issues early warnings about defects in the tap changer and the ultraviolet sensor itself based on the monitoring data from the ultraviolet light sensor.

[0005] In a further preferred embodiment, the monitoring and sensing module includes an optical fiber, an optical signal integrator, and a certain number of ultraviolet light sensors. The ultraviolet light sensors are arranged inside the tap changer oil chamber, pointing radially towards the oil chamber. The optical signal integrator is installed at the bottom of the tap changer oil chamber, and each ultraviolet light sensor is connected to the optical signal integrator via an oil-resistant optical fiber.

[0006] Further preferably, the ultraviolet light sensors of the monitoring and sensing module are evenly arranged according to the size of the tap changer oil chamber, and the number of ultraviolet light sensors is:

[0007]

[0008] In the formula, r is the maximum radial distance from the tap changer oil chamber to the center point, l is the maximum axial length of the tap changer oil chamber, α is the effective measurement angle of the ultraviolet light sensor, and k tp This is the dispersion correction coefficient for the ultraviolet light sensor, and its value depends on the sensitivity of the ultraviolet light sensor.

[0009] Further preferably, the optical signal integrator of the monitoring and sensing module reads the instantaneous ultraviolet discharge photon count collected by the ultraviolet light sensor, integrates and collects signals from each ultraviolet light sensor, and the collection interval is calculated according to the following formula:

[0010]

[0011] In the formula, T UV f is the data collection interval. H k is the highest sampling and transmission frequency. r Let k be the external disturbance coefficient. c The monitoring correction coefficient is affected by the historical ultraviolet photon data of the tap changer being monitored; k s k represents the state perception coefficient of the information analysis module in the sleep state. s =1000, k under monitoring status s =p s / 1000, p s This indicates the number of photons emitted during ultraviolet discharge.

[0012] Further optimization involves the information analysis module analyzing the discharge data of the transformer tap changer, specifically the number of ultraviolet discharge photons p monitored by all ultraviolet sensors over a continuous time period T0. s All are less than the monitoring photon threshold p a When the number of ultraviolet discharge photons p is monitored, the information analysis module enters sleep mode; when the number of ultraviolet discharge photons p is monitored, the information analysis module enters sleep mode. s Exceeding the threshold 10p a At that time, the information analysis module enters the monitoring state.

[0013] A method for monitoring transformer tap changer oil chamber discharge based on ultraviolet imaging includes:

[0014] The number of ultraviolet photons detected by different ultraviolet sensors is normalized and calibrated to a 10m reference distance to form an ultraviolet sensor monitoring dataset G; G = [G1, G2, ..., G k ,…,G n ], where n is the number of ultraviolet light sensors;

[0015]

[0016] In the formula G k Let x be the number of ultraviolet photons after calibration for the k-th ultraviolet sensor. kLet y be the distance from the k-th ultraviolet light sensor to the center of the tap changer oil chamber. k The number of ultraviolet photons detected by the k-th ultraviolet light sensor.

[0017] The ultraviolet light sensor monitoring dataset is dynamically normalized in the time domain to remove outliers. The average value of the ultraviolet light sensor monitoring dataset is then calculated. The standard error S of the ultraviolet light sensor monitoring dataset:

[0018]

[0019] When the ultraviolet light sensor monitors the data set containing At that time, G j For the j-th ultraviolet photon count after calibration, the outlier value G is... j Remove the data and perform data cleaning again.

[0020] Based on the ultraviolet light sensor corresponding to the outlier, the number of times F is cleared after correction of the ultraviolet photon count of the k-th ultraviolet light sensor. k Satisfy F k >10k c F sum When / n, it indicates that the ultraviolet light sensor is malfunctioning, where k c To monitor the correction factor, F sum This represents the total number of data clears. For ultraviolet (UV) sensors exhibiting abnormalities, a warning is issued indicating a defect in the UV sensor; data from that UV sensor will not be collected again until it is replaced.

[0021] The cleaned and corrected ultraviolet photon count data is used to generate a time-scaled discharge information dataset. Based on data clustering principles, the ultraviolet photon count data from different historical operating stages of the tap changer are classified and aggregated.

[0022] A database of different types of tap changers was established to collect historical operating data of the tap changers and their corresponding main transformers. Based on data clustering principles, the historical operating data of the tap changers was classified and aggregated, forming a dataset of on-load tap changer discharge influencing factors, Ultr=[Ultr1,Ultr2,…,Ultr…]. i ,…,Ultr m ]; where Ultr i The i-th element of the dataset of factors affecting on-load tap changer discharge is represented, i = 1 to m, where m is the total number of elements. The dataset of factors affecting on-load tap changer discharge is a subset of time series data including factors such as the number of switch switching, the number of times the polarity changes, the overload time of the main transformer, short circuit impact, and DC bias.

[0023] By mapping the ultraviolet light sensor monitoring dataset with the Ultrar dataset of on-load tap changer discharge influencing factors, a time-series data subset including influencing factors such as undisturbed operation, switch tap adjustment, short-circuit impact, and external fluctuations is formed. The weighted eigenvalue E of different time-series discharge photon data subsets is then calculated. s The characteristics of tap changer discharge photon data include feature parameters such as average value, rate of change, weighted value, abrupt change, and torsion coefficient. Among them, the weighted feature value E of the discharge photon data subset is... s The mapping relationship with the ultraviolet light sensor monitoring dataset is as follows:

[0024]

[0025] In the formula t k denoted as the continuous discharge time monitored by the k-th ultraviolet light sensor, r is the maximum radial distance from the tap changer oil chamber to the center point, and T0 is the monitoring duration.

[0026] To assess the severity of tap changer discharge under different UV sensor monitoring values ​​at the current moment, the XGBoost model parameters are initialized, and a dataset of factors affecting on-load tap changer discharge is generated; Ultr = [Ultr1, Ultr2, ..., Ultr i ,…,Ultr n Import the XGBoost model for training, and then determine the maximum depth, number of iterations, shrinkage step size, non-split score, and complexity penalty coefficient of the optimal decision tree.

[0027] The weighted feature value Es of the evaluated discharge photon data subset is input into the trained XGBoost model, which outputs weighted feature values ​​of different time series data subsets and predicts the tap changer's position in the current discharge photon data subset weighted feature value E. s The following are the conditions and targeted maintenance strategies.

[0028] The beneficial effects of this invention are as follows: This invention acquires historical data on the number of ultraviolet photons in the tap changer of a transformer under historical operating conditions; corrects the data based on the differences in the number of ultraviolet photons received by different ultraviolet sensors, and monitors the internal discharge status of the tap changer at the current stage. It can analyze and provide early warnings of defects in the tap changer and the ultraviolet sensors themselves. Attached Figure Description

[0029] Figure 1 This is a schematic diagram of a transformer tap changer oil chamber discharge monitoring device based on ultraviolet imaging. Detailed Implementation

[0030] The invention will now be explained in further detail with reference to the accompanying drawings.

[0031] like Figure 1As shown, a transformer tap changer oil chamber discharge monitoring device based on ultraviolet imaging includes a monitoring and sensing module, a wireless transmission module, an information analysis module, and a correction and early warning module. The monitoring and sensing module collects data on internal discharge and external disturbances of the tap changer and is connected to the information analysis module. The information analysis module evaluates the device status based on the current ultraviolet discharge data of the tap changer. The information analysis module is connected to the correction and early warning module, which analyzes and warns of defects in the tap changer and the ultraviolet sensor itself based on the monitoring data from the ultraviolet light sensor.

[0032] In a preferred embodiment, the monitoring and sensing module includes an optical fiber, an optical signal integrator, and a number of ultraviolet (UV) sensors. The UV sensors are arranged inside the tap changer oil chamber, pointing radially towards the oil chamber. The optical signal integrator is installed at the bottom of the tap changer oil chamber, and each UV sensor is connected to the optical signal integrator via an oil-resistant optical fiber.

[0033] In a preferred embodiment, the ultraviolet light sensors of the monitoring and sensing module are evenly arranged according to the size of the tap changer oil chamber, and the number of ultraviolet light sensors is:

[0034]

[0035] In the formula, r is the maximum radial distance from the tap changer oil chamber to the center point, l is the maximum axial length of the tap changer oil chamber, α is the effective measurement angle of the ultraviolet light sensor, and k tp This is the dispersion correction coefficient for the ultraviolet light sensor, and its value depends on the sensitivity of the ultraviolet light sensor.

[0036] In a preferred embodiment, the optical signal integrator of the monitoring and sensing module reads the instantaneous ultraviolet discharge photon count collected by the ultraviolet light sensor, integrates and collects signals from each ultraviolet light sensor, and calculates the collection interval using the following formula:

[0037]

[0038] In the formula, T UV f is the data collection interval. H k is the highest sampling and transmission frequency. r Let k be the external disturbance coefficient. c The monitoring correction coefficient is affected by the historical ultraviolet photon data of the tap changer being monitored; k s k represents the state perception coefficient of the information analysis module in the sleep state. s =1000, k under monitoring status s =p s / 1000, p s This indicates the number of photons emitted during ultraviolet discharge.

[0039] In a preferred embodiment, the information analysis module analyzes the discharge data of the transformer tap changer, and calculates the number p of ultraviolet discharge photons monitored by all ultraviolet sensors over a continuous time period T0. s All are less than the monitoring photon threshold p a When the number of ultraviolet discharge photons p is monitored, the information analysis module enters sleep mode; when the number of ultraviolet discharge photons p is monitored, the information analysis module enters sleep mode. s Exceeding the threshold 10p a At that time, the information analysis module enters the monitoring state.

[0040] A method for monitoring transformer tap changer oil chamber discharge based on ultraviolet imaging includes:

[0041] The number of ultraviolet photons detected by different ultraviolet sensors is normalized and calibrated to a 10m reference distance to form an ultraviolet sensor monitoring dataset G; G = [G1, G2, ..., G k ,…,G n ], where n is the number of ultraviolet light sensors;

[0042]

[0043] In the formula G k Let x be the number of ultraviolet photons after calibration for the k-th ultraviolet sensor. k Let y be the distance from the k-th ultraviolet light sensor to the center of the tap changer oil chamber. k The number of ultraviolet photons detected by the k-th ultraviolet light sensor.

[0044] The ultraviolet light sensor monitoring dataset is dynamically normalized in the time domain to remove outliers. The average value of the ultraviolet light sensor monitoring dataset is then calculated. The standard error S of the ultraviolet light sensor monitoring dataset:

[0045]

[0046] When the ultraviolet light sensor monitors the data set containing At that time, G j For the j-th ultraviolet photon count after calibration, the outlier value G is... j Remove the data and perform data cleaning again.

[0047] Based on the ultraviolet light sensor corresponding to the outlier, the number of times F is cleared after correction of the ultraviolet photon count of the k-th ultraviolet light sensor. k Satisfy F k >10k c F sum When / n, it indicates that the ultraviolet light sensor is malfunctioning, where k c To monitor the correction factor, F sumThis represents the total number of data clears. For ultraviolet (UV) sensors exhibiting abnormalities, a warning is issued indicating a defect in the UV sensor; data from that UV sensor will not be collected again until it is replaced.

[0048] The cleaned and corrected ultraviolet photon count data is used to generate a time-scaled discharge information dataset. Based on data clustering principles, the ultraviolet photon count data from different historical operating stages of the tap changer are classified and aggregated.

[0049] A database of different types of tap changers was established to collect historical operating data of the tap changers and their corresponding main transformers. Based on data clustering principles, the historical operating data of the tap changers was classified and aggregated, forming a dataset of on-load tap changer discharge influencing factors, Ultr=[Ultr1,Ultr2,…,Ultr…]. i ,…,Ultr m ]; where Ultr i The i-th element of the dataset of factors affecting on-load tap changer discharge is represented, i = 1 to m, where m is the total number of elements. The dataset of factors affecting on-load tap changer discharge is a subset of time series data including factors such as the number of switch switching, the number of times the polarity changes, the overload time of the main transformer, short circuit impact, and DC bias.

[0050] By mapping the ultraviolet light sensor monitoring dataset with the Ultrar dataset of on-load tap changer discharge influencing factors, a time-series data subset including influencing factors such as undisturbed operation, switch tap adjustment, short-circuit impact, and external fluctuations is formed. The weighted eigenvalue E of different time-series discharge photon data subsets is then calculated. s The characteristics of tap changer discharge photon data include feature parameters such as average value, rate of change, weighted value, abrupt change, and torsion coefficient. Among them, the weighted feature value E of the discharge photon data subset is... s The mapping relationship with the ultraviolet light sensor monitoring dataset is as follows:

[0051]

[0052] In the formula t k denoted as the continuous discharge time monitored by the k-th ultraviolet light sensor, r is the maximum radial distance from the tap changer oil chamber to the center point, and T0 is the monitoring duration.

[0053] To assess the severity of tap changer discharge under different UV sensor monitoring values ​​at the current moment, the XGBoost model parameters are initialized, and a dataset of factors affecting on-load tap changer discharge is generated; Ultr = [Ultr1, Ultr2, ..., Ultr i ,…,Ultr n Import the XGBoost model for training, and then determine the maximum depth, number of iterations, shrinkage step size, non-split score, and complexity penalty coefficient of the optimal decision tree.

[0054] The weighted feature value Es of the evaluated discharge photon data subset is input into the trained XGBoost model, which outputs weighted feature values ​​of different time series data subsets and predicts the tap changer's position in the current discharge photon data subset weighted feature value E. s The following are the conditions and targeted maintenance strategies.

[0055] The above description is merely a preferred embodiment of the present invention and is not intended to limit other embodiments of the invention. Those skilled in the art can modify the above disclosure to create equivalent embodiments. However, any simple modifications, substitutions, and simplifications of the above embodiments without departing from the scope of the present invention should be within the protection scope of the present invention.

Claims

1. A method for monitoring discharge in the oil chamber of a transformer tap changer based on ultraviolet imaging, characterized in that, include: The number of ultraviolet photons detected by different ultraviolet sensors is normalized and corrected to a reference distance to form an ultraviolet sensor monitoring dataset G; G=[G1,G2,…,G k ,…,G n ], where n is the number of ultraviolet light sensors; ； In the formula G k Let x be the number of ultraviolet photons after calibration for the k-th ultraviolet sensor. k Let y be the distance from the k-th ultraviolet light sensor to the center of the tap changer oil chamber. k The number of ultraviolet photons detected by the k-th ultraviolet light sensor; Dynamic time-domain normalization of the ultraviolet light sensor monitoring dataset was performed to remove outlier data. For the ultraviolet light sensor corresponding to the removed outlier data, the number of times F is removed after correction of the ultraviolet photon count of the k-th ultraviolet light sensor. k satisfy When this occurs, it indicates that the ultraviolet light sensor is malfunctioning, where k c To monitor the correction factor, F sum This represents the total number of data clearing attempts. For ultraviolet (UV) sensors exhibiting abnormalities, a warning is issued indicating a defect in the UV sensor, and data from that UV sensor will not be collected again until it is replaced. The cleaned and corrected ultraviolet photon count data is used to generate a discharge information dataset based on time. Ultraviolet photon count data from different historical operating stages of the tap changer are classified and aggregated based on data clustering principles. Establish a database of different types of tap changers and collect historical operating data of tap changers and corresponding main transformers; classify and collect historical operating data of tap changers based on the principle of data clustering, and form a dataset of on-load tap changer discharge influencing factors according to different attributes. By mapping the ultraviolet light sensor monitoring dataset with the on-load tap changer discharge influencing factor dataset, a time series data subset of influencing factors is formed. The weighted eigenvalue E of different time series discharge photon data subsets is then calculated. s The characteristics of tap changer discharge photon data include feature parameters such as average value, rate of change, weighted value, abrupt change, and torsion coefficient. Among them, the weighted feature value E of the discharge photon data subset is... s The mapping relationship with the ultraviolet light sensor monitoring dataset is as follows: ； In the formula The continuous discharge time monitored by the k-th ultraviolet light sensor is denoted as r, which is the maximum radial distance from the center point of the tap changer oil chamber, and T0 is the monitoring duration. To assess the severity of tap changer discharge under different ultraviolet light sensor monitoring values ​​at the current moment, the XGBoost model parameters were initialized. The dataset of factors affecting on-load tap changer discharge was imported into the XGBoost model for training. The maximum depth, number of iterations, shrinkage step size, non-split score, and complexity penalty coefficient of the optimal decision tree were determined in turn. The weighted feature value Es of the evaluated discharge photon data subset is input into the trained XGBoost model, which outputs weighted feature values ​​of different time series data subsets and predicts the tap changer's position in the current discharge photon data subset weighted feature value E. s The following are the conditions and targeted maintenance strategies.

2. The method for monitoring transformer tap changer oil chamber discharge based on ultraviolet imaging according to claim 1, characterized in that, The method for removing outlier data is as follows: Average value of ultraviolet light sensor monitoring dataset The standard error S of the ultraviolet light sensor monitoring dataset: ； When the ultraviolet light sensor monitors the data set containing At that time, G j Let G be the number of ultraviolet photons after calibration for the j-th ultraviolet sensor. j Remove the data and perform data cleaning again.

3. The method for monitoring transformer tap changer oil chamber discharge based on ultraviolet imaging according to claim 1, characterized in that, Data set of factors affecting discharge of on-load tap changers: Ultr = [Ultr1, Ultr2, ..., Ultr] i ,…, Ultr m ]; where Ultr i This represents the i-th element of the dataset of factors affecting on-load tap changer discharge, where i = 1 to m, and m is the total number of elements. The dataset of factors affecting on-load tap changer discharge is a subset of the time series data of the factors affecting the discharge.

4. The method for monitoring transformer tap changer oil chamber discharge based on ultraviolet imaging according to claim 3, characterized in that, Influencing factors include the number of switch switching times, the number of times the polarity changes, the overload time of the main transformer, short-circuit impact, and DC bias.

5. A transformer tap changer oil chamber discharge monitoring device based on ultraviolet imaging, used to implement the transformer tap changer oil chamber discharge monitoring method according to any one of claims 1-4, characterized in that, It includes a monitoring and sensing module, a wireless transmission module, an information analysis module, and a correction and early warning module. The monitoring and sensing module collects data on internal discharge and external disturbances of the tap changer. The monitoring and sensing module is connected to the information analysis module. The information analysis module evaluates the equipment status based on the current ultraviolet discharge data of the tap changer. The information analysis module is connected to the correction and early warning module. The correction and early warning module analyzes and issues early warnings about defects in the tap changer and the ultraviolet sensor itself based on the monitoring data from the ultraviolet light sensor.

6. The transformer tap changer oil chamber discharge monitoring device based on ultraviolet imaging according to claim 5, characterized in that, The monitoring and sensing module includes an optical fiber, an optical signal integrator, and a number of ultraviolet (UV) sensors. The UV sensors are arranged inside the tap changer oil chamber, pointing radially towards the oil chamber. The optical signal integrator is installed at the bottom of the tap changer oil chamber, and each UV sensor is connected to the optical signal integrator via an oil-resistant optical fiber.

7. The transformer tap changer oil chamber discharge monitoring device based on ultraviolet imaging according to claim 5, characterized in that, The ultraviolet light sensors of the monitoring and sensing module are evenly arranged according to the size of the tap changer oil chamber, and the number of ultraviolet light sensors is: ； In the formula, r is the maximum radial distance from the tap changer oil chamber to the center point. The maximum axial length of the tap changer oil chamber. The effective measurement angle of the ultraviolet light sensor, This is the dispersion correction coefficient for the ultraviolet light sensor, and its value depends on the sensitivity of the ultraviolet light sensor.

8. The transformer tap changer oil chamber discharge monitoring device based on ultraviolet imaging according to claim 5, characterized in that, The optical signal integrator of the monitoring and sensing module reads the instantaneous ultraviolet discharge photon count collected by the ultraviolet light sensor, integrates and collects signals from each ultraviolet light sensor, and calculates the collection interval using the following formula: ； In the formula, T UV f is the data collection interval. H k is the highest sampling and transmission frequency. r Let k be the external disturbance coefficient. c The monitoring correction coefficient is affected by the historical ultraviolet photon data of the tap changer being monitored; k s k represents the state perception coefficient of the information analysis module in the sleep state. s =1000, k under monitoring status s =p s / 1000, p s This indicates the number of photons emitted during ultraviolet discharge.

9. The transformer tap changer oil chamber discharge monitoring device based on ultraviolet imaging according to claim 5, characterized in that, The information analysis module analyzes the discharge data of the transformer tap changer, and measures the number of ultraviolet discharge photons p monitored by all ultraviolet sensors over a continuous time T0. s All are less than the monitoring photon threshold p a When the number of ultraviolet discharge photons p is detected, the information analysis module enters sleep mode; when the number of ultraviolet discharge photons detected is p s Exceeding the threshold of 10p a At that time, the information analysis module enters the monitoring state.

Images