A transformer oil chromatographic multi-channel detection system

Abstract

The application discloses a transformer oil chromatographic multi-channel detection system, relates to the technical field of chromatographic detection, and is used for solving the problem of reduced detection precision of transformer oil dissolved gas and monitoring the pressure in the transformer oil dissolved gas cavity, analyzing the altitude characteristics, and deciding whether to start the altitude compensation mode according to the analysis result, performing multi-channel parallel detection on the oil gas, collecting the main detection channel and the reference compensation channel signals, extracting the peak half-height width information of the chromatographic peaks, evaluating the detection state, judging whether to collect the inlet and outlet pressures of each chromatographic column, calculating the flow trend and linear velocity of the carrier gas in each chromatographic column based on the pressure data, correcting the linear velocity by using the flow trend, calling the gas retention time of the theoretical database for stretching compensation, and performing secondary sampling on the oil gas based on the compensation parameters, so that the influence of the air pressure change is automatically eliminated, and the detection precision under different environments is ensured.

Description

Technical Field

[0001] This invention relates to the field of chromatographic detection technology, and more specifically, to a multi-channel chromatographic detection system for transformer oil. Background Technology

[0002] In existing multi-channel chromatographic detection technology for dissolved gases in transformer oil, the detection results are easily affected by changes in ambient air pressure and altitude. Especially in high-altitude or low-pressure environments, the peak shape parameters of dissolved gases will drift, leading to a decrease in analytical accuracy. To address this issue, some systems perform local compensation by adding a reference channel or adjusting the column flow rate, but these methods still cannot achieve comprehensive dynamic correction of multi-channel signals and carrier gas flow rates.

[0003] The existing technology has the following shortcomings: Currently, existing technologies typically employ fixed chromatographic conditions for detection, lacking an automatic compensation mechanism for different altitudes and air pressures. This makes it difficult to accurately correct the peak shape and carrier gas flow rate of multi-channel chromatographic signals, resulting in reduced detection accuracy of dissolved gases in transformer oil and an inability to operate stably in complex environments. Therefore, a multi-channel chromatographic detection system for transformer oil is proposed.

[0004] The information disclosed in the background section is only intended to enhance the understanding of the background of this disclosure, and therefore may include information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0005] To overcome the aforementioned deficiencies of the prior art, embodiments of the present invention provide a transformer oil chromatography multi-channel detection system, which solves the problems mentioned in the background art by employing environmental analysis based on altitude characteristics, automatic compensation mode, multi-channel peak half-width extraction, column pressure detection, and carrier gas flow rate correction and theoretical retention time stretching compensation methods.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a transformer oil chromatography multi-channel detection system, comprising an altitude identification module, a channel acquisition module, a flow rate detection module, and a compensation and correction module, the functions of each module being as follows: Before detecting the dissolved gas in the transformer oil, the altitude recognition module sets an environmental analysis cycle. During the environmental analysis cycle, it monitors the intracavitary pressure of the dissolved gas in the transformer oil and analyzes the altitude characteristics. Based on the altitude characteristics, it identifies whether to activate the altitude compensation mode. The channel acquisition module performs multi-channel detection and processing of the dissolved gas in the current transformer oil in the altitude compensation mode and acquires multi-channel output data. Based on the multi-channel output data, it extracts the peak half-width information of the dissolved gas in the current transformer oil. The flow rate detection module assesses the current multi-channel detection status of dissolved gases in transformer oil based on peak half-width information and determines whether to collect pressure data from different chromatographic columns. Based on the pressure data, it calculates the carrier gas flow trend and detects the carrier gas linear velocity of different chromatographic columns. The compensation and correction module corrects the linear velocity of the carrier gas by utilizing the flow trend of the carrier gas, retrieves the theoretical retention time of the gas from the theoretical database, performs stretching compensation on the theoretical retention time of the gas based on the corrected linear velocity of the carrier gas, and resamples the dissolved gas in the current transformer oil.

[0007] In a preferred embodiment, the altitude identification module is configured with an environmental analysis cycle, and the cavity pressure of the current transformer oil dissolved gas is monitored by a high-precision pressure sensor installed at the gas inlet of the cavity. The pressure of the dissolved gas in the transformer oil cavity is compared with the standard atmospheric pressure to determine the following: If the internal pressure of the dissolved gas in the transformer oil is greater than or equal to 1.1 times the standard atmospheric pressure, then the altitude characteristic is determined to be a high-pressure characteristic. If the pressure of dissolved gas in the transformer oil cavity is greater than or equal to 0.9 times the standard atmospheric pressure and less than 1.1 times the standard atmospheric pressure, then the altitude characteristic is determined to be a normal pressure characteristic. If the internal pressure of the dissolved gas in the transformer oil is less than 0.9 times the standard atmospheric pressure, then the altitude characteristic is determined to be a low-pressure characteristic. If the altitude characteristics are low-pressure characteristics, then the altitude compensation mode is activated; Conversely, the altitude compensation mode will not be activated.

[0008] In a preferred embodiment, when the altitude compensation mode is activated in the channel acquisition module, multi-channel detection and processing are performed on the current transformer oil dissolved gas. The dissolved gas in the current transformer oil is introduced into the distributor, and the dissolved gas in the current transformer oil is distributed to the main detection channel and the reference compensation channel according to the preset distribution ratio. The main detection channel is equipped with a non-polar or weakly polar chromatographic column, and the reference compensation channel is equipped with a medium polar chromatographic column. The columns of the two channels are placed in the same constant temperature oven to maintain the same temperature. Multi-channel output data refers to the set of raw chromatographic signals that are synchronously acquired and output by the channel acquisition module through multiple parallel detection channels, including the main detection channel signal sequence and the reference compensation channel signal sequence.

[0009] In a preferred embodiment, in the channel acquisition module, the acquisition time corresponding to the right half-width point and the acquisition time corresponding to the left half-width point are obtained by traversing the main detection channel signal sequence. Subtract the acquisition time corresponding to the right half-width point from the acquisition time corresponding to the left half-width point to obtain the peak half-width information of the dissolved gas in the transformer oil. If the reference compensation channel signal corresponding to the acquisition time of the right half-width point is less than the preset reference channel abnormal fluctuation threshold, and the reference compensation channel signal corresponding to the acquisition time of the left half-width point is less than the preset reference channel abnormal fluctuation threshold, then it is determined that the peak half-width information of the current transformer oil dissolved gas is not interfered data. Conversely, the peak half-width and height information of the dissolved gas in the transformer oil is determined to be interfered data.

[0010] In a preferred embodiment, in the flow rate detection module, if the peak half-width information of the current transformer oil dissolved gas is less than the preset broadening threshold, and the peak half-width information of the current transformer oil dissolved gas is not interfered with, then the multi-channel detection state of the current transformer oil dissolved gas is determined to be normal, and pressure data of different chromatographic columns are not collected. Conversely, if the current multi-channel detection status of dissolved gases in transformer oil is determined to be abnormal, pressure data from different chromatographic columns will be collected. The pressure data for different chromatographic columns refers to the pressure values ​​at the inlet and outlet of each chromatographic column installed in the main detection channel and the reference compensation channel during the carrier gas flow process, including the inlet and outlet pressures of the main channel and the inlet and outlet pressures of the reference channel.

[0011] In a preferred embodiment, in the flow velocity detection module, the pressure difference between the inlet pressure and the outlet pressure of the main channel is obtained by subtracting the absolute value. Subtract the inlet pressure and outlet pressure of the reference channel from each other and take the absolute value to obtain the pressure difference of the reference channel. Divide the pressure difference of the main channel and the pressure difference of the reference channel by the preset column length to obtain the carrier gas flow trend of the main channel and the carrier gas flow trend of the reference channel, respectively. The carrier gas linear velocities of the main channel and the reference channel are obtained by using thermal mass flow sensors located at the outlet of the main detection channel column and the outlet of the reference compensation channel column.

[0012] In a preferred embodiment, in the compensation and correction module, the main channel carrier gas flow trend, the reference channel carrier gas flow trend, the main channel carrier gas linear velocity, and the reference channel carrier gas linear velocity are used to form a carrier gas linear velocity feature set, which is then used to form an input feature matrix. The input feature matrix is ​​input into the pre-trained carrier gas linear velocity compensation model to obtain the corrected main channel carrier gas linear velocity and the reference channel carrier gas linear velocity.

[0013] In a preferred embodiment, the theoretical retention time of the gas is retrieved from the theoretical database in the compensation and correction module. Add the corrected main channel carrier gas linear velocity and the reference channel carrier gas linear velocity together, divide by 2, and use the ratio of the result to the preset standard carrier gas linear velocity as the stretching compensation coefficient. Multiply the theoretical gas retention time by the tensile compensation coefficient to obtain the theoretical gas retention time after tensile compensation. The dissolved gas in the transformer oil is sampled a second time to obtain the main detection channel signal sequence and the reference compensation channel signal sequence.

[0014] In a preferred embodiment, in the compensation and correction module, the peak shape correlation coefficient of the main and auxiliary channels is calculated by combining the main detection channel signal sequence of the secondary sampling and the reference compensation channel signal sequence. If the correlation coefficient of the main and auxiliary channels is greater than or equal to the preset correlation coefficient threshold, it is determined that the peak shape of the main and auxiliary channels is consistent, and the detection result after secondary sampling is directly output. If the correlation coefficient of the peak shape of the main and auxiliary channels is less than the preset correlation coefficient threshold, it is determined that the peak shape consistency of the main and auxiliary channels has decreased, and the peak shape reconstruction algorithm is triggered. Peak shape reconstruction algorithm refers to a data processing method that uses interference information provided by the reference compensation channel to perform baseline correction, noise filtering, and peak shape fitting on contaminated chromatographic peaks in the main detection channel, and finally reconstructs a chromatographic peak shape that is close to the real situation.

[0015] The technical effects and advantages of this invention are as follows: This invention establishes an environmental analysis cycle before detection, during which the pressure of dissolved gas in transformer oil within the detection chamber is monitored, and altitude characteristics are analyzed. Based on these altitude characteristics, an altitude compensation mode is activated. When activated, the dissolved gas is detected in parallel via multiple channels, and the output signals of the main detection channel and the reference compensation channel are collected. The peak half-width at half-maximum (HWHM) information of the chromatographic peaks is extracted from the output signals. The current detection status is assessed based on the HWHM information, determining whether to collect the inlet and outlet pressures of different chromatographic columns. The flow trend of the carrier gas within each column is calculated based on the pressure data, thus obtaining the carrier gas linear velocity for each column. The carrier gas linear velocity is corrected using the flow trend. The theoretical retention time of the gas is retrieved from a theoretical database, and the theoretical retention time is stretched and compensated based on the corrected carrier gas linear velocity. The dissolved gas in the transformer oil is then sampled a second time based on the compensated parameters. By analyzing altitude characteristics before detection and automatically activating the compensation mode, the influence of gas pressure changes on the detection results is eliminated, ensuring detection accuracy under different environments. Attached Figure Description

[0016] Figure 1 This is a flowchart illustrating the implementation of a multi-channel chromatographic detection system for transformer oil according to the present invention.

[0017] Figure 2 This is a schematic diagram of a multi-channel detection system for transformer oil chromatography according to the present invention. Detailed Implementation

[0018] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0019] This invention establishes an environmental analysis cycle before detection, during which the pressure of dissolved gas in transformer oil within the detection chamber is monitored, and altitude characteristics are analyzed. Based on these altitude characteristics, an altitude compensation mode is activated. When activated, the dissolved gas is detected in parallel via multiple channels, and the output signals of the main detection channel and the reference compensation channel are collected. The peak half-width at half-maximum (HWHM) information of the chromatographic peaks is extracted from the output signals. The current detection status is assessed based on the HWHM information, determining whether to collect the inlet and outlet pressures of different chromatographic columns. The flow trend of the carrier gas within each column is calculated based on the pressure data, thus obtaining the carrier gas linear velocity for each column. The carrier gas linear velocity is corrected using the flow trend. The theoretical retention time of the gas is retrieved from a theoretical database, and the theoretical retention time is stretched and compensated based on the corrected carrier gas linear velocity. The dissolved gas in the transformer oil is then sampled a second time based on the compensated parameters. By analyzing altitude characteristics before detection and automatically activating the compensation mode, the influence of pressure changes on the detection results is eliminated.

[0020] Example 1, as Figures 1 to 2 As shown, a transformer oil chromatography multi-channel detection system includes an altitude identification module, a channel acquisition module, a flow rate detection module, and a compensation and correction module. The functions of each module are as follows: Before detecting the dissolved gas in the transformer oil, the altitude recognition module sets an environmental analysis cycle. During the environmental analysis cycle, it monitors the intracavitary pressure of the dissolved gas in the transformer oil and analyzes the altitude characteristics. Based on the altitude characteristics, it identifies whether to activate the altitude compensation mode. The channel acquisition module performs multi-channel detection and processing of the dissolved gas in the current transformer oil in the altitude compensation mode and acquires multi-channel output data. Based on the multi-channel output data, it extracts the peak half-width information of the dissolved gas in the current transformer oil. The flow rate detection module assesses the current multi-channel detection status of dissolved gases in transformer oil based on peak half-width information and determines whether to collect pressure data from different chromatographic columns. Based on the pressure data, it calculates the carrier gas flow trend and detects the carrier gas linear velocity of different chromatographic columns. The compensation and correction module corrects the linear velocity of the carrier gas by utilizing the flow trend of the carrier gas, retrieves the theoretical retention time of the gas from the theoretical database, performs stretching compensation on the theoretical retention time of the gas based on the corrected linear velocity of the carrier gas, and resamples the dissolved gas in the current transformer oil.

[0021] The specific implementation is as follows: In the altitude identification module, transformer oil dissolved gas detection is an important means of power equipment fault diagnosis. By detecting the composition and concentration of dissolved gases in the oil, it is possible to determine whether there are latent faults such as overheating or discharge inside the transformer. The difference in environmental pressure in different altitude areas will affect the accuracy of gas detection. Before detection, the environmental characteristics should be identified to determine whether compensation for altitude is required to ensure the reliability of the detection data. Set an environmental analysis cycle and monitor the internal pressure of the dissolved gas in the transformer oil by using a high-precision pressure sensor installed at the gas inlet of the cavity; It should be noted that the environmental analysis cycle can be set according to the pressure sensor response time, gas sample stabilization time, and detection efficiency requirements; a high-precision pressure sensor refers to a sensing device that monitors and outputs gas pressure values ​​in real time with high sensitivity and low error, used to monitor the intracavitary pressure of dissolved gas in transformer oil; standard atmospheric pressure refers to the atmospheric pressure value at sea level, latitude 45°, and temperature 0°C, and is a benchmark reference pressure used for scientific research and engineering calculations.

[0022] The pressure of the dissolved gas in the transformer oil cavity is compared with the standard atmospheric pressure to determine the following: If the internal pressure of the dissolved gas in the transformer oil is greater than or equal to 1.1 times the standard atmospheric pressure, then the altitude characteristic is determined to be a high-pressure characteristic. If the pressure of dissolved gas in the transformer oil cavity is greater than or equal to 0.9 times the standard atmospheric pressure and less than 1.1 times the standard atmospheric pressure, then the altitude characteristic is determined to be a normal pressure characteristic. If the internal pressure of the dissolved gas in the transformer oil is less than 0.9 times the standard atmospheric pressure, then the altitude characteristic is determined to be a low-pressure characteristic. If the altitude characteristics are low-pressure characteristics, then the altitude compensation mode is activated; Conversely, the altitude compensation mode will not be activated.

[0023] It should be explained that high pressure characteristic refers to an environmental state where the internal pressure of the dissolved gas in the transformer oil is significantly higher than the standard atmospheric pressure, usually corresponding to low-altitude areas or closed pressurized environments; normal pressure characteristic refers to an environmental state where the internal pressure of the dissolved gas in the transformer oil is close to the standard atmospheric pressure, corresponding to near sea level or conventional laboratory environments; low pressure characteristic refers to an environmental state where the internal pressure of the dissolved gas in the transformer oil is significantly lower than the standard atmospheric pressure, usually corresponding to high-altitude areas; altitude compensation mode refers to a parameter correction mechanism that is automatically activated when the detection environment is determined to be low pressure characteristic, used to eliminate changes in carrier gas linear velocity and retention time offset caused by the decrease in gas pressure, ensuring the accuracy of the detection results.

[0024] By setting an environmental analysis cycle, the system monitors the pressure inside the chamber in real time and compares it with the standard atmospheric pressure. It automatically identifies high pressure, normal pressure, and low pressure characteristics, and only activates the altitude compensation mode when the pressure is low. This enables the monitoring system to have intelligent environmental perception and adaptive adjustment capabilities without the need for manual intervention.

[0025] In the channel acquisition module, in the altitude compensation mode, the dissolved gases in transformer oil are detected. Accurate separation and identification of the fault characteristic gases dissolved in the oil are the key to judging the latent faults inside the transformer. In the actual detection process, the separation effect of gas samples in the chromatographic column is easily affected by a variety of factors, such as column aging, carrier gas flow rate fluctuation, ambient temperature change and detector noise. These factors will cause the chromatographic peak shape to broaden and the retention time to shift, thus affecting the accuracy of qualitative and quantitative analysis of gas components. When the altitude compensation mode is activated, multi-channel detection and processing are performed on the dissolved gas in the current transformer oil. The specific processing procedure is as follows: The dissolved gas in the current transformer oil is introduced into the distributor, and the dissolved gas in the current transformer oil is distributed to the main detection channel and the reference compensation channel according to the preset distribution ratio. The main detection channel is equipped with a non-polar or weakly polar chromatographic column for separating light hydrocarbon gas components such as hydrogen, methane, and ethane. A medium-polarity column is configured in the reference compensation channel to simultaneously monitor interfering components such as moisture and oxygen, as well as baseline drift caused by column bleed. The two chromatographic columns are placed in the same constant temperature oven to maintain a consistent temperature; High-purity nitrogen or helium is used as the carrier gas in the chromatographic column. It passes through the chromatographic column at a constant flow rate to propel the gas sample to be distributed and separated within the column. It should be noted that a splitter is a microfluidic gas distribution device installed between the gas outlet of the detection chamber and the inlet of the chromatographic column, used to precisely distribute the current transformer oil dissolved gas sample to the main detection channel and the reference compensation channel according to a preset ratio. The preset split ratio can be set according to the gas sample concentration, detector sensitivity, column capacity, and interference monitoring requirements. The main detection channel refers to the chromatographic analysis pathway in a multi-channel parallel detection structure, specifically used for separating and detecting target gas components, and is the core channel for obtaining qualitative and quantitative results of transformer oil dissolved gas. The reference compensation channel refers to the chromatographic analysis pathway in a multi-channel parallel detection structure specifically used for separating and detecting target gas components. An auxiliary analytical pathway is used to simultaneously monitor environmental interference, baseline drift, and common-mode noise, providing a reference benchmark for calibrating the main detection channel data. Non-polar or weakly polar chromatographic columns refer to gas chromatography separation columns with low stationary phase polarity. Their stationary phase is mainly composed of dimethyl polysiloxane, which has good separation ability for non-polar compounds. Medium polar chromatographic columns refer to gas chromatography separation columns with a certain polar functional group in the stationary phase, and the polarity is between that of non-polar and strongly polar columns. Their stationary phase is usually copolymerized from dimethyl polysiloxane and a certain proportion of polar functional groups, which can simultaneously achieve compound separation based on boiling point differences and polar interactions.

[0026] Multi-channel output data refers to the set of raw chromatographic signals that the channel acquisition module synchronously acquires and outputs through multiple parallel detection channels, including the main detection channel signal sequence and the reference compensation channel signal sequence; The sampling period is preset and multiple sampling times are divided. The gas component concentrations of the main detection channel and the reference compensation channel are obtained by using flame ionization detectors placed at the exits of the main detection channel and the reference compensation channel. The gas component concentrations of the main detection channel and the reference compensation channel are converted into electrical signals, which are used as the main detection channel signal sequence and the reference compensation channel signal sequence. Traverse the main detection channel signal sequence to obtain the maximum signal value and its corresponding acquisition time; Divide the maximum value of the main detection channel signal sequence by 2 to obtain the half-peak height. Traverse the main detection channel signal sequence, search for the first main detection channel signal value that is equal to the half peak height, record the acquisition time corresponding to the signal value, and use it as the acquisition time corresponding to the left half width point; Traverse the main detection channel signal sequence, search for the signal value of the second main detection channel signal that is equal to the half peak height, record the acquisition time corresponding to the signal value, and use it as the acquisition time corresponding to the right half width point; Subtract the acquisition time corresponding to the right half-width point from the acquisition time corresponding to the left half-width point to obtain the peak half-width information of the dissolved gas in the transformer oil. Peak half-height width information refers to the set of data extracted from the chromatographic peaks in the main detection channel to characterize the width features of the chromatographic peaks at the half-height position. It is a core indicator for judging the chromatographic separation effect and detection status. The reference compensation channel signal corresponding to the acquisition time of the right half-width point and the acquisition time of the left half-width point is compared with the preset abnormal fluctuation threshold of the reference channel for judgment. If the reference compensation channel signal corresponding to the acquisition time of the right half-width point is less than the preset reference channel abnormal fluctuation threshold, and the reference compensation channel signal corresponding to the acquisition time of the left half-width point is less than the preset reference channel abnormal fluctuation threshold, then it is determined that the peak half-width information of the current transformer oil dissolved gas is not interfered data. Conversely, the peak half-width and height information of the dissolved gas in the transformer oil is determined to be interfered data.

[0027] It should be explained that the preset acquisition cycle can be set according to the target gas retention time, chromatographic peak width, sampling frequency requirements, and detection efficiency requirements; the flame ionization detector refers to a detection device that uses the high-temperature energy generated by the combustion of a hydrogen flame to ionize organic compounds into ions, and generates an electrical signal by collecting the ion flow, which is used to obtain the gas component concentrations of the main detection channel and the reference compensation channel; the preset abnormal fluctuation threshold of the reference channel can be set according to the baseline noise level, detector sensitivity, and interference identification sensitivity requirements; the interfered data refers to the data in the peak half-width information extracted from the main detection channel whose reliability is reduced due to environmental interference factors, and which needs to be marked or further processed.

[0028] The gas sample is distributed to the main detection channel and the reference compensation channel according to a preset ratio by the splitter. The main channel focuses on the separation of the target gas, while the reference channel monitors environmental interference in real time, providing a benchmark for subsequent signal correction and significantly improving the detection system's ability to identify environmental interference.

[0029] In the flow rate detection module, the broadening of chromatographic peaks is an important indicator for measuring separation efficiency and detection status in the detection of dissolved gases in transformer oil. When chromatographic peaks are too broad, it may be due to various factors such as abnormal carrier gas flow rate, decreased column performance, or environmental interference, which directly affects the separation efficiency and quantitative accuracy of gas components. As a key parameter affecting retention time and peak shape, the accurate acquisition of carrier gas linear velocity is crucial for subsequent compensation and correction. Establishing a detection status evaluation mechanism based on peak half-width information and collecting column pressure data in real time during abnormal situations to calculate carrier gas flow trend and linear velocity has become an important technical means to improve detection reliability. The peak half-width at half-maximum (HWHM) of the dissolved gas in the transformer oil is compared with the preset broadening threshold for determination. If the peak half-width information of the current transformer oil dissolved gas is less than the preset broadening threshold, and the peak half-width information of the current transformer oil dissolved gas is not interfered with, then the current multi-channel detection status of the transformer oil dissolved gas is determined to be normal, and pressure data of different chromatographic columns are not collected. Conversely, if the current multi-channel detection status of dissolved gases in transformer oil is determined to be abnormal, pressure data from different chromatographic columns will be collected. The pressure data for different chromatographic columns refers to the pressure values ​​at the inlet and outlet of each chromatographic column installed in the main detection channel and the reference compensation channel during the carrier gas flow process, including the inlet and outlet pressures of the main channel and the inlet and outlet pressures of the reference channel. The main channel inlet and outlet pressures and the reference channel inlet and outlet pressures are obtained by high-precision pressure sensors arranged at the inlet and outlet ends of the main detection channel chromatographic column and the inlet and outlet ends of the reference compensation channel chromatographic column. Subtract the inlet pressure from the outlet pressure of the main channel and take the absolute value to obtain the pressure difference of the main channel; Subtract the inlet pressure and outlet pressure of the reference channel from each other and take the absolute value to obtain the pressure difference of the reference channel. Divide the pressure difference of the main channel and the pressure difference of the reference channel by the preset column length to obtain the carrier gas flow trend of the main channel and the carrier gas flow trend of the reference channel, respectively. The carrier gas linear velocities of the main channel and the reference channel are obtained by using thermal mass flow sensors located at the outlet of the main detection channel column and the outlet of the reference compensation channel column.

[0030] It should be explained that the preset broadening threshold can be set based on the theoretical plate number of the chromatographic column, the retention time of the target gas, the resolution requirements, and the statistics of historical detection data; the preset column length can be set based on the resolution requirements, analysis time limits, carrier gas pressure conditions, and the properties of the target gas.

[0031] By comparing peak half-width and height information with a preset broadening threshold and combining this with markers of disturbed data, the system intelligently determines whether the current multi-channel detection status is normal. Stress data is collected only under abnormal conditions to avoid unnecessary resource consumption, while ensuring timely capture of critical anomalies.

[0032] In the compensation and correction module, the carrier gas linear velocity is the core parameter affecting the retention time and separation effect of chromatographic peaks in the detection of dissolved gases in transformer oil. Its accuracy directly determines the qualitative and quantitative analysis results of gas components. In actual detection, the carrier gas linear velocity is affected by a variety of factors, including changes in environmental pressure, column aging, carrier gas viscosity fluctuations, and temperature changes, which lead to deviations between the measured linear velocity and the theoretical value, resulting in retention time shifts and peak shape distortion. Establishing a carrier gas linear velocity compensation model based on deep learning to achieve collaborative correction of the linear velocity of the two channels, and quantitatively evaluating the secondary sampling results through peak shape correlation coefficients, has become a key technical means to improve the accuracy and reliability of transformer oil dissolved gas detection. The main channel carrier gas flow trend, the reference channel carrier gas flow trend, the main channel carrier gas linear velocity, and the reference channel carrier gas linear velocity constitute a carrier gas linear velocity feature set, forming an input feature matrix. The input feature matrix is ​​input into the pre-trained carrier gas linear velocity compensation model to obtain the corrected main channel carrier gas linear velocity and the reference channel carrier gas linear velocity. The carrier gas linear velocity compensation model consists of multiple fully connected layers, batch normalization layers, and activation function layers, which can automatically extract the nonlinear mapping relationship between the carrier gas flow trend and the carrier gas linear velocity. The model training process includes the following steps: Data preparation: Collect a large amount of historical detection data on carrier gas flow trend characteristics and corresponding measured carrier gas linear velocity under different column conditions, and match each set of carrier gas flow trend characteristics with the measured carrier gas linear velocity values ​​to form input label pairs; Data preprocessing: The input feature matrix is ​​standardized to normalize the carrier gas flow trend features and carrier gas linear velocity data to the same scale, eliminating the impact of dimensional differences on model training; Training strategy: Supervised learning method is used, with the carrier gas flow trend feature matrix as the model input, the measured carrier gas linear velocity as the label, and the mean squared error loss function used to measure the difference between the model-predicted carrier gas linear velocity and the measured carrier gas linear velocity. Model optimization: The weight parameters of the carrier gas linear velocity compensation neural network are updated through backpropagation algorithm to continuously reduce the loss function value. During training, a learning rate decay strategy is adopted to improve training effect and generalization ability and avoid overfitting. Model validation: Evaluate model performance on an independent validation dataset, compare the corrected carrier gas linear velocity output by the model with the actual collected carrier gas linear velocity, and ensure that the model can accurately predict the carrier gas linear velocity compensation amount under different flow trend conditions. Model update: By continuously adding new historical detection data for retraining or fine-tuning, the model's adaptability and prediction accuracy for gas linear velocity compensation under complex flow conditions are optimized. It should be noted that standardization methods include, but are not limited to, standard linear transformation based on interval scaling, statistical Z-Score standardization, or normalization based on nonlinear mapping functions. The specific methods of standardization will not be elaborated upon here. Supervised learning is a machine learning method used to train a model using existing input-output samples, enabling the model to predict or classify new data. Mean squared error is a commonly used error metric used to assess the difference between predicted and true values. Backpropagation is the most commonly used optimization algorithm for training artificial neural networks, used to calculate the gradient of the loss function relative to the network weights and update the weights to reduce prediction error. Learning rate decay is a commonly used optimization method for training neural networks, used to dynamically adjust the learning rate during training to improve convergence stability and final model accuracy. Access the theoretical database to retrieve the theoretical retention time of the gas; Add the corrected main channel carrier gas linear velocity and the reference channel carrier gas linear velocity together, divide by 2, and use the ratio of the result to the preset standard carrier gas linear velocity as the stretching compensation coefficient. Multiply the theoretical gas retention time by the tensile compensation coefficient to obtain the theoretical gas retention time after tensile compensation. The dissolved gas in the transformer oil is sampled a second time to obtain the main detection channel signal sequence and the reference compensation channel signal sequence of the second sampling. The correlation coefficient of the peak shape of the main and auxiliary channels is calculated by combining the main detection channel signal sequence and the reference compensation channel signal sequence after secondary sampling. The calculation formula is as follows: ,in, This represents the total number of members in the main detection channel signal sequence after secondary sampling. For the first The main detection channel signal sequence members are sampled twice. For the first Each secondary sampled reference compensation channel signal sequence member, The correlation coefficient of peak shape between primary and secondary channels; The correlation coefficient between the main and auxiliary channels reflects the similarity between the chromatographic peak shapes of the main detection channel and the reference compensation channel after secondary sampling. The closer the correlation coefficient between the main and auxiliary channels is to 1, the more consistent the peak shapes of the two channels are, the less the detection results of the main channel are affected by environmental interference, and the higher the data reliability. The closer the correlation coefficient between the main and auxiliary channels is to 0 or less than the preset threshold, the more the consistency of the peak shapes of the two channels decreases, the detection results of the main channel may be affected by environmental interference, and the peak shape reconstruction algorithm needs to be triggered for correction. The correlation coefficients of the primary and secondary channels are compared with a preset correlation coefficient threshold for judgment. If the correlation coefficient of the main and auxiliary channels is greater than or equal to the preset correlation coefficient threshold, it is determined that the peak shape of the main and auxiliary channels is consistent, and the detection result after secondary sampling is directly output. If the correlation coefficient of the peak shape of the main and auxiliary channels is less than the preset correlation coefficient threshold, it is determined that the consistency of the peak shape of the main and auxiliary channels has decreased, and the current detection result is greatly affected by environmental interference, thus triggering the peak shape reconstruction algorithm.

[0033] It should be explained that the theoretical database refers to a dedicated data storage set that stores the theoretical retention times and related physical property parameters of various gas components under standard detection conditions, used to obtain the theoretical retention time of gases; the preset standard carrier gas linear velocity can be set according to the carrier gas flow rate corresponding to the optimal separation efficiency of the chromatographic column, the detection efficiency requirements, and industry testing standards; the preset correlation coefficient threshold can be set according to the intrinsic correlation of the two channel signals, the interference identification sensitivity requirements, and the reliability requirements of the detection results; the peak shape reconstruction algorithm refers to a data processing method that, based on the interference information provided by the reference compensation channel, performs baseline correction, noise filtering, and peak shape fitting on the contaminated chromatographic peaks in the main detection channel, and finally reconstructs a chromatographic peak shape that is close to the real situation.

[0034] By constructing a carrier gas linear velocity feature set and inputting it into a pre-trained deep learning model, the collaborative correction of the carrier gas linear velocity in the main channel and the reference channel is achieved. The model automatically extracts the nonlinear mapping relationship between the carrier gas flow trend and the linear velocity, effectively eliminating the linear velocity deviation caused by factors such as changes in environmental pressure and column aging, and providing accurate linear velocity input for retention time compensation.

[0035] Finally, it should be noted that in this paper, relational terms such as first and second are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations.

[0036] Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0037] In this document, the singular forms “a,” “an,” and “the” may also include the plural forms unless the context clearly indicates otherwise. It should also be understood that terms such as “comprising / including” or “having” specify the presence of the stated features, integrals, steps, operations, components, parts, or combinations thereof, but do not preclude the possibility of the presence or addition of one or more other features, integrals, steps, operations, components, parts, or combinations thereof. Meanwhile, the term “and / or” as used in this specification includes any and all combinations of the associated listed items.

[0038] The various embodiments in this specification are described in a progressive manner. Each embodiment focuses on the differences from other embodiments. The various embodiments can be combined as needed, and the same or similar parts can be referred to each other.

[0039] The above description of the disclosed embodiments will enable those skilled in the art to make or use various modifications to these embodiments. It will be readily apparent to those skilled in the art that the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of this application. Therefore, this application is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A multi-channel chromatographic detection system for transformer oil, characterized in that: It includes an altitude recognition module, a channel acquisition module, a flow velocity detection module, and a compensation and correction module. The functions of each module are as follows: Before detecting the dissolved gas in the transformer oil, the altitude recognition module sets an environmental analysis cycle. During the environmental analysis cycle, it monitors the intracavitary pressure of the dissolved gas in the transformer oil and analyzes the altitude characteristics. Based on the altitude characteristics, it identifies whether to activate the altitude compensation mode. The channel acquisition module performs multi-channel detection and processing of the dissolved gas in the current transformer oil in the altitude compensation mode and acquires multi-channel output data. Based on the multi-channel output data, it extracts the peak half-width information of the dissolved gas in the current transformer oil. The flow rate detection module assesses the current multi-channel detection status of dissolved gases in transformer oil based on peak half-width information and determines whether to collect pressure data from different chromatographic columns. Based on the pressure data, it calculates the carrier gas flow trend and detects the carrier gas linear velocity of different chromatographic columns. The compensation and correction module corrects the linear velocity of the carrier gas by utilizing the flow trend of the carrier gas, retrieves the theoretical retention time of the gas from the theoretical database, performs stretching compensation on the theoretical retention time of the gas based on the corrected linear velocity of the carrier gas, and resamples the dissolved gas in the current transformer oil.

2. The transformer oil chromatography multichannel detection system according to claim 1, characterized in that: In the altitude identification module, an environmental analysis cycle is set, and the pressure inside the cavity of the current transformer oil dissolved gas is monitored by a high-precision pressure sensor installed at the gas inlet of the cavity. The pressure of the dissolved gas in the transformer oil cavity is compared with the standard atmospheric pressure to determine the following: If the internal pressure of the dissolved gas in the transformer oil is greater than or equal to 1.1 times the standard atmospheric pressure, then the altitude characteristic is determined to be a high-pressure characteristic. If the pressure of dissolved gas in the transformer oil cavity is greater than or equal to 0.9 times the standard atmospheric pressure and less than 1.1 times the standard atmospheric pressure, then the altitude characteristic is determined to be a normal pressure characteristic. If the internal pressure of the dissolved gas in the transformer oil is less than 0.9 times the standard atmospheric pressure, then the altitude characteristic is determined to be a low-pressure characteristic. If the altitude characteristics are low-pressure characteristics, then the altitude compensation mode is activated; Conversely, the altitude compensation mode will not be activated.

3. The transformer oil chromatography multichannel detection system according to claim 1, characterized in that: In the channel acquisition module, when the altitude compensation mode is activated, multi-channel detection and processing are performed on the dissolved gas in the current transformer oil. The dissolved gas in the current transformer oil is introduced into the distributor, and the dissolved gas in the current transformer oil is distributed to the main detection channel and the reference compensation channel according to the preset distribution ratio. The main detection channel is equipped with a non-polar or weakly polar chromatographic column, and the reference compensation channel is equipped with a medium polar chromatographic column. The columns of the two channels are placed in the same constant temperature oven to maintain the same temperature. Multi-channel output data refers to the set of raw chromatographic signals that are synchronously acquired and output by the channel acquisition module through multiple parallel detection channels, including the main detection channel signal sequence and the reference compensation channel signal sequence.

4. The transformer oil chromatography multichannel detection system according to claim 3, characterized in that: In the channel acquisition module, the acquisition time corresponding to the right half-width point and the acquisition time corresponding to the left half-width point are obtained by traversing the main detection channel signal sequence; Subtract the acquisition time corresponding to the right half-width point from the acquisition time corresponding to the left half-width point to obtain the peak half-width information of the dissolved gas in the transformer oil. If the reference compensation channel signal corresponding to the acquisition time of the right half-width point is less than the preset reference channel abnormal fluctuation threshold, and the reference compensation channel signal corresponding to the acquisition time of the left half-width point is less than the preset reference channel abnormal fluctuation threshold, then it is determined that the peak half-width information of the current transformer oil dissolved gas is not interfered data. Conversely, the peak half-width and height information of the dissolved gas in the transformer oil is determined to be interfered data.

5. The transformer oil chromatography multichannel detection system according to claim 4, characterized in that: In the flow rate detection module, if the peak half-width information of the current transformer oil dissolved gas is less than the preset broadening threshold, and the peak half-width information of the current transformer oil dissolved gas is not interfered with, then the multi-channel detection status of the current transformer oil dissolved gas is determined to be normal, and pressure data of different chromatographic columns are not collected. Conversely, if the current multi-channel detection status of dissolved gases in transformer oil is determined to be abnormal, pressure data from different chromatographic columns will be collected. The pressure data for different chromatographic columns refers to the pressure values ​​at the inlet and outlet of each chromatographic column installed in the main detection channel and the reference compensation channel during the carrier gas flow process, including the inlet and outlet pressures of the main channel and the inlet and outlet pressures of the reference channel.

6. The transformer oil chromatography multichannel detection system according to claim 5, characterized in that: In the flow velocity detection module, the pressure difference between the inlet pressure and the outlet pressure of the main channel is obtained by subtracting the absolute value. Subtract the inlet pressure and outlet pressure of the reference channel from each other and take the absolute value to obtain the pressure difference of the reference channel. Divide the pressure difference of the main channel and the pressure difference of the reference channel by the preset column length to obtain the carrier gas flow trend of the main channel and the carrier gas flow trend of the reference channel, respectively. The carrier gas linear velocities of the main channel and the reference channel are obtained by using thermal mass flow sensors located at the outlet of the main detection channel column and the outlet of the reference compensation channel column.

7. The transformer oil chromatography multichannel detection system according to claim 6, characterized in that: In the compensation and correction module, the main channel carrier gas flow trend, the reference channel carrier gas flow trend, the main channel carrier gas linear velocity, and the reference channel carrier gas linear velocity are combined to form a carrier gas linear velocity feature set, which forms an input feature matrix. The input feature matrix is ​​input into the pre-trained carrier gas linear velocity compensation model to obtain the corrected main channel carrier gas linear velocity and the reference channel carrier gas linear velocity.

8. The transformer oil chromatography multichannel detection system according to claim 7, characterized in that: In the compensation and correction module, the theoretical retention time of the gas is retrieved from the theoretical database. Add the corrected main channel carrier gas linear velocity and the reference channel carrier gas linear velocity together, divide by 2, and use the ratio of the result to the preset standard carrier gas linear velocity as the stretching compensation coefficient. Multiply the theoretical gas retention time by the tensile compensation coefficient to obtain the theoretical gas retention time after tensile compensation. The dissolved gas in the transformer oil is sampled a second time to obtain the main detection channel signal sequence and the reference compensation channel signal sequence.

9. A transformer oil chromatography multichannel detection system according to claim 8, characterized in that: In the compensation and correction module, the peak shape correlation coefficient of the main and auxiliary channels is calculated by combining the main detection channel signal sequence of the secondary sampling and the reference compensation channel signal sequence; If the correlation coefficient of the main and auxiliary channels is greater than or equal to the preset correlation coefficient threshold, it is determined that the peak shape of the main and auxiliary channels is consistent, and the detection result after secondary sampling is directly output. If the correlation coefficient of the peak shape of the main and auxiliary channels is less than the preset correlation coefficient threshold, it is determined that the peak shape consistency of the main and auxiliary channels has decreased, and the peak shape reconstruction algorithm is triggered. Peak shape reconstruction algorithm refers to a data processing method that uses interference information provided by the reference compensation channel to perform baseline correction, noise filtering, and peak shape fitting on contaminated chromatographic peaks in the main detection channel, and finally reconstructs a chromatographic peak shape that is close to the real situation.

Images