A method and system for monitoring the automatic oil extraction process of a transformer

Abstract

The present invention relates to the field of transformer oil extraction, and discloses a transformer automatic oil extraction process monitoring method and system, which detects whether the movement of the push rod is smooth by performing time series analysis on the collected voltage signal data, detects whether there is a time delay in the switching time of the solenoid valve by performing delay analysis on the switching time of the solenoid valve, analyzes the oil sample state by electrical signals and temperature signals, and analyzes the oil extraction amount in combination with the number of triggering times of the stroke switch of the needle oil extraction to detect whether the oil extraction amount reaches a predetermined standard, adjusts the oil extraction amount detected not meeting the predetermined standard based on the detection result, monitors and analyzes the state of the oil extraction pipeline by oil extraction flow and pipeline pressure, detects whether there is an abnormal risk in the oil extraction pipeline, and finally feeds back the oil extraction process analysis and detection flow to the detection personnel terminal in real time, which is conducive to optimizing the oil extraction method, improving the accuracy of the oil extraction amount, and ensuring the safety and stability of the oil extraction process.

Description

Technical Field

[0001] The present invention relates to the technical field of transformer oil extraction, and more particularly to a method and system for monitoring a transformer automatic oil extraction process. Background Art

[0002] The traditional oil extraction method mainly relies on manual operation using a syringe barrel to connect with the sampling valve of the equipment to extract oil. However, this method has some problems, such as the impact of operational errors on analysis results, large workload, waste of oil samples, environmental pollution risks, and problems with sampling accuracy and cleanliness. Therefore, a method for monitoring the automatic oil extraction process of transformers has emerged. The existing automatic oil extraction method of transformers uses a transformer oil sampling robot. Through a complex mechanical structure and control system, it achieves precise automatic oil extraction, which can effectively improve the accuracy of oil sampling and reduce interference from air and impurities.

[0003] As described in the Chinese patent application CN115728096A, a transformer multi-mode layered automatic oil sampling monitoring method is provided. The present invention provides a transformer multi-mode automatic oil sampling monitoring method, comprising: an oil sampling system is installed on the side of the transformer casing, the oil sampling port of the transformer is connected to a pneumatic filling device for cooperating with the oil sampling system to extract oil; a controller has a data transmission function, which is triggered according to the received instructions or the mode set by itself, or the current detection data to control the pneumatic filling device and various components in the oil sampling system to start / stop in a timed manner; the method of the present invention can realize unmanned monitoring, automatically monitor the transformer and regularly take oil samples. After the samples are taken, the user can go to the site to take them away, realizing digital intelligent duty.

[0004] However, the above process still has the following disadvantages:

[0005] First, the current transformer automatic oil extraction process monitoring technology mainly relies on manual inspections, which cannot detect minor faults such as push rod jams and solenoid valve delays in real time;

[0006] Second, the current transformer automatic oil extraction process monitoring technology only monitors and calibrates the oil extraction volume through a single signal from a flow meter, failing to consider the impact of electrical signals, temperature signals, and the number of times the travel switch is triggered on the oil extraction volume, resulting in oil extraction deviation.

[0007] Third, the current transformer automatic oil extraction process monitoring technology mainly relies on manual experience to judge the pipeline status, and is unable to promptly detect whether there are abnormal risks in the oil extraction pipeline, resulting in the inability to ensure the safety and stability of the oil extraction process. Summary of the Invention

[0008] In order to overcome the above-mentioned defects in the prior art, the present invention provides a method and system for monitoring the automatic oil extraction process of a transformer to solve the problems existing in the above-mentioned background technology.

[0009] The present invention provides the following technical solution: a method for monitoring the automatic oil extraction process of a transformer, comprising:

[0010] S1: used to collect and record the voltage signal data corresponding to the push rod feedback resistance signal during oil extraction, the solenoid valve status data and the oil extraction flow rate;

[0011] S2: Based on real-time monitoring of the actuator's movement status, the collected voltage signal data is analyzed in time series to detect whether the actuator's movement is smooth;

[0012] S3: Based on the recorded solenoid valve status data, a delay analysis is performed on the switching time of the solenoid valve to detect whether there is a time delay in the switching time of the solenoid valve;

[0013] S4: used to detect the amount of oil taken. The oil sample status is analyzed through the electrical signal and temperature signal during oil taking to obtain the oil sample status evaluation coefficient. The oil amount is analyzed in combination with the number of times the travel switch of the needle oil taking is triggered to detect whether the oil amount reaches the predetermined standard.

[0014] S5: Based on the oil extraction amount detection and analysis results, a calibration analysis is performed on the oil extraction amount that does not meet the predetermined standard, and an oil amount calibration coefficient is calculated to adjust the oil extraction amount that does not meet the predetermined standard in real time;

[0015] S6: Used to monitor the pipeline status in real time during the oil extraction process. It monitors and analyzes the oil extraction pipeline status through oil extraction flow and pipeline pressure to detect whether there is any abnormal risk in the oil extraction pipeline;

[0016] S7: It is used to provide real-time feedback of the oil extraction process analysis and detection process to the detection personnel terminal, and to display the oil extraction process analysis results in a visual manner.

[0017] Preferably, the specific method of collecting the voltage signal data corresponding to the push rod feedback resistance signal in S1 is: using a high-precision data acquisition card to collect the voltage signal converted from the push rod feedback resistance signal, setting a sampling frequency, and collecting the voltage signal data based on the set sampling frequency;

[0018] The specific method for collecting the solenoid valve status data is as follows: monitoring and recording the on / off status of the solenoid valve through the digital input / output port, and recording the timestamp of the status change;

[0019] The specific method for collecting the oil extraction time is as follows: when the push rod starts to move, the current time is recorded as the oil extraction start time; when the push rod moves to completion or the flow sensor shows the flow rate returns to zero, the current time is recorded as the oil extraction end time;

[0020] The specific method for collecting the oil extraction flow is: by installing a flow sensor on the oil extraction pipeline to monitor the flow of the oil ball valve, when the flow is greater than zero, it is considered that oil extraction has begun, and the oil extraction flow value is recorded.

[0021] Preferably, the S2 performs noise reduction and normalization processing on the collected voltage signal data, uses a time series analysis method to analyze the change state of the voltage signal over time, and draws the analysis results of the voltage signal over time into a time series chart, and calculates the statistical characteristics of the voltage signal, including the mean and standard deviation. Based on the analysis of the statistical characteristics of the voltage signal, a stability index is calculated, and a stability threshold is set. The stability index is compared with the stability threshold to determine whether the movement of the push rod is smooth.

[0022] Preferably, after the analysis result of the push rod action state is based on the S3, the actual time of the solenoid valve switching is delayed analyzed according to the timestamp data of the solenoid valve state change, and the time delay of each solenoid valve is calculated, and the calculated time delay of each solenoid valve is compared with the expected time delay to determine whether there is a time delay in the switching time of each solenoid valve.

[0023] Preferably, the S4 collects electrical signals and temperature signals during the oil extraction process, pre-processes and extracts features of the collected electrical signals and temperature, extracts electrical signal features and temperature signal features reflecting changes in the state of the oil sample, uses statistical methods to analyze the relationship between the electrical signal features and the temperature signal features, identifies the features most relevant to the changes in the state of the oil sample, and calculates an oil sample state evaluation coefficient based on the results of the feature analysis to reflect the quality of the oil, including oil viscosity, oil density, and oil moisture content;

[0024] By recording the number of times the needle moves to trigger the switch each time oil is taken, a machine learning algorithm is used to construct an oil extraction quantity prediction model, the oil sample state evaluation coefficient and the number of times the travel switch is triggered during needle oil extraction form a data set, and the data set is divided into a training set, a validation set, and a test set. The constructed oil extraction quantity prediction model is trained using the training set, and the model parameters are adjusted until a satisfactory prediction effect is achieved. The model performance is evaluated using the test set, and the oil sample state evaluation coefficient and the number of times the travel switch is triggered are input into the trained model to obtain an oil extraction quantity prediction value. A predetermined value is set, and the predicted oil extraction quantity is compared with the predetermined value to determine whether the oil extraction quantity meets the predetermined standard. If the predicted oil extraction quantity reaches or exceeds the predetermined value, it is considered that the oil extraction quantity meets the predetermined standard and there is no need to perform a calibration operation on the oil quantity. If the predicted oil extraction quantity does not reach the predetermined value, it is considered that the oil extraction quantity does not meet the predetermined standard, and the detection result that does not meet the predetermined standard is transmitted to S5 for oil quantity calibration analysis.

[0025] Preferably, when receiving the oil extraction amount detection result that does not meet the predetermined standard, S5 calculates the oil extraction amount calibration coefficient according to the oil extraction amount data that does not meet the predetermined standard. The specific calculation formula is: , where Y represents the original predicted value of oil extraction, represents the actual amount of oil taken at the i-th time, Represents the predicted value of the oil extraction for the i-th time, N represents the number of data points that do not meet the predetermined standards; by setting an error threshold , combined with the oil extraction calibration coefficient, the oil extraction amount that does not meet the predetermined standard is calibrated, and when the error of the oil extraction amount for M consecutive times is less than the error threshold, that is, , stop oil quantity calibration.

[0026] Preferably, the S6 constructs a time series prediction model by using a machine learning algorithm, takes flow and pressure as characteristic parameters of the input model, and then the model outputs the pipeline flow prediction value and pipeline status. When the oil extraction pipeline status output by the model is abnormal, the real-time flow rate at which the abnormality occurs is recorded, and the pipeline flow prediction value is analyzed with the real-time flow rate to detect abnormal conditions of the pipeline status, and the abnormal results are given as early warning feedback to remind the inspection personnel that there is a risk of abnormality in the oil extraction pipeline.

[0027] Preferably, the S7 generates a test report according to the oil extraction process analysis and detection process and sends it to the detection personnel terminal, and displays the oil extraction process analysis results in a visual manner on the TFT liquid crystal screen.

[0028] To achieve the above object, the present invention provides the following technical solution: a transformer automatic oil extraction process monitoring system, which implements the above transformer automatic oil extraction process monitoring method, comprising:

[0029] Data acquisition module: used to collect and record the voltage signal data corresponding to the push rod feedback resistance signal during oil extraction, solenoid valve status data and oil extraction flow rate;

[0030] Actuator status detection module: Based on real-time monitoring of the actuator's movement status, the module detects whether the actuator's movement is smooth by performing time series analysis on the collected voltage signal data.

[0031] Solenoid valve time delay detection module: Based on the recorded solenoid valve status data, it performs delay analysis on the solenoid valve switching time to detect whether there is a time delay in the solenoid valve switching time;

[0032] Oil extraction detection and analysis module: used to detect the oil extraction volume, analyze the oil sample status through the electrical signal and temperature signal during oil extraction, obtain the oil sample status evaluation coefficient, and analyze the oil extraction volume in combination with the number of triggering times of the travel switch of the needle oil extraction to detect whether the oil extraction volume meets the predetermined standard;

[0033] Oil volume calibration module: Based on the oil volume detection and analysis results, the module calibrates and analyzes the oil volume that does not meet the predetermined standards, and calculates the oil volume calibration coefficient, which is used to adjust the oil volume that does not meet the predetermined standards in real time; the oil volume calibration operation is stopped until the oil volume meets the predetermined standards;

[0034] Pipeline status analysis module: used to monitor the pipeline status in real time during the oil extraction process, monitor and analyze the oil extraction pipeline status through oil extraction flow and pipeline pressure, and detect whether the oil extraction pipeline is blocked;

[0035] Oil extraction result feedback module: used to feed back the oil extraction process analysis and detection process to the detection personnel terminal, and display the oil extraction process analysis results in a visual manner.

[0036] Technical effects and advantages of the present invention:

[0037] Based on time series analysis, the push rod action status is monitored in real time and the switching time of the solenoid valve is accurately analyzed. Minor faults such as push rod sticking and solenoid valve delay can be detected in real time, helping to prevent oil extraction failures caused by abnormal push rod action and solenoid valve switching delay.

[0038] The oil sample status is analyzed based on the electrical signal and temperature signal during oil extraction. The oil extraction volume is analyzed in combination with the number of times the travel switch of the needle oil extraction is triggered. It is detected whether the oil extraction volume reaches the predetermined standard. The oil extraction volume is calibrated according to the test results. This can more comprehensively analyze the oil sample status and accurately detect whether the oil extraction volume reaches the predetermined standard. It is conducive to realizing real-time adjustment of the oil extraction volume and improving the accuracy of the oil extraction volume.

[0039] Based on the monitoring of oil extraction flow and pipeline pressure, it is possible to promptly detect whether there are abnormal risks in the oil extraction pipeline, such as blockage, leakage, etc., to ensure the safety and stability of the oil extraction process. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 A diagram showing the steps of the method of the present invention.

[0041] Figure 2 This is a structural block diagram of the device of the present invention.

[0042] Figure 3 Schematic diagram of the system of the present invention.

[0043] Figure 4 This is a flow chart of an automatic transformer oil extraction device according to this embodiment. DETAILED DESCRIPTION

[0044] The technical solutions of the present invention will be clearly and completely described below in conjunction with the accompanying drawings of the present invention. In addition, the forms of the various structures described in the following embodiments are merely examples. The method and system for monitoring the automatic oil extraction process of a transformer involved in the present invention are not limited to the various structures described in the following embodiments. All other embodiments obtained by ordinary technicians in this field without making creative work fall within the scope of protection of the present invention.

[0045] like Figure 1 This embodiment provides a method for monitoring the automatic oil extraction process of a transformer, including:

[0046] S1: used to collect and record the voltage signal data corresponding to the push rod feedback resistance signal during oil extraction, the solenoid valve status data and the oil extraction flow rate.

[0047] In this embodiment, the specific method of collecting the voltage signal data corresponding to the push rod feedback resistance signal by S1 is: using a high-precision data acquisition card to collect the voltage signal converted from the push rod feedback resistance signal, setting a sampling frequency, and collecting the voltage signal data based on the set sampling frequency;

[0048] The specific method for collecting the solenoid valve status data is as follows: monitoring and recording the on / off status of the solenoid valve through the digital input / output port, and recording the timestamp of the status change;

[0049] The specific method for collecting the oil extraction time is as follows: when the push rod starts to move, the current time is recorded as the oil extraction start time; when the push rod moves to completion or the flow sensor shows the flow rate returns to zero, the current time is recorded as the oil extraction end time;

[0050] The specific method for collecting the oil extraction flow is: by installing a flow sensor on the oil extraction pipeline to monitor the flow of the oil ball valve, when the flow is greater than zero, it is considered that oil extraction has begun, and the oil extraction flow value is recorded.

[0051] It should be specifically explained that a resistive displacement sensor is installed on the push rod to convert the displacement of the push rod into a resistance change, and the resistance signal output by the sensor is converted into a voltage signal through a signal conditioning circuit. A data acquisition card is used to connect the output of the signal conditioning circuit, and a suitable sampling rate is set to ensure that the details of the push rod movement can be captured. The DAQ system is started to start collecting voltage signal data and store it in the database; by connecting the on / off state of the solenoid valve to the system through a digital input interface, the closed or open state of the solenoid valve contacts, the number of changes of the solenoid valve contacts and the timestamp of each state change are recorded, and the recorded data are stored in the database; the start time and end time of oil extraction are recorded according to the start and completion time of the push rod movement. At the same time, a flow sensor is installed on the oil extraction pipeline, and the output signal of the flow sensor is connected to the DAQ system. The set sampling rate collects flow data from the start time to the end time of oil extraction, and the start time, end time and flow data of oil extraction are stored together in the database.

[0052] S2: Based on real-time monitoring of the action state of the push rod, time series analysis of the collected voltage signal data is performed to detect whether the action of the push rod is smooth.

[0053] In this embodiment, the S2 performs noise reduction and normalization processing on the collected voltage signal data, uses a time series analysis method to analyze the change state of the voltage signal over time, and plots the analysis results of the voltage signal over time into a time series chart, and calculates the statistical characteristics of the voltage signal, including the mean and standard deviation. Based on the analysis of the statistical characteristics of the voltage signal, a stability index is calculated, and a stability threshold is set. The stability index is compared with the stability threshold to determine whether the movement of the push rod is smooth.

[0054] It should be noted that the voltage signal is analyzed by using the autoregressive model. The specific model calculation formula is: ,in, represents the voltage signal value at time t, c represents the constant term, represents the autoregressive coefficient, represents the error term at time t, Respectively indicate time The voltage signal value is plotted based on the model analysis results using a chart tool to plot the voltage signal over time. The mean and standard deviation of the voltage signal over time are calculated from the time series chart. and mean The stability index is calculated by the ratio of , by analyzing the stability of the push rod under different working conditions through experimental data to set the stability threshold , compare the calculated stability index with the set stability threshold to detect whether the push rod action is smooth when taking oil. , it is considered that the push rod action is in a stable state when taking oil. , it is considered that the push rod action during oil extraction is unstable, and an early warning prompt must be immediately sent to the inspection personnel terminal to prompt the inspection personnel to adjust and maintain the push rod, and the inspection results at this time are displayed on the TFT LCD screen.

[0055] S3: Based on the recorded state data of the solenoid valve, a delay analysis is performed on the switching time of the solenoid valve to detect whether there is a time delay in the switching time of the solenoid valve.

[0056] In this embodiment, after analyzing the push rod action state, S3 performs a delay analysis on the actual switching time of the solenoid valve according to the timestamp data of the solenoid valve state change, calculates the time delay of each solenoid valve, and compares the calculated time delay of each solenoid valve with the expected time delay to determine whether there is a time delay in the switching time of each solenoid valve.

[0057] It should be noted that by collecting the timestamp data of the solenoid valve state changes, including the time when the solenoid valve is opened and closed, the actual switching time delay of each solenoid valve is calculated based on the timestamp data. The specific calculation formula is: ,in, Indicates the timestamp of the solenoid valve opening, The timestamp indicating when the solenoid valve is closed is used to determine the expected switching time delay based on the solenoid valve design parameters and application scenarios. , respectively compare the actual time delay of each solenoid valve with the expected time delay to detect whether there is a time delay in the switching time of each solenoid valve, and adjust the switching time of each solenoid valve. , then there is no time delay in the switching time of the solenoid valve. If , it detects that there is a time delay in the switching time of the solenoid valve and automatically issues a delay warning notification signal.

[0058] S4: Used to detect the amount of oil taken. The oil sample status is analyzed through the electrical signal and temperature signal during oil taking to obtain the oil sample status evaluation coefficient. The oil amount is analyzed in combination with the number of times the travel switch of the needle oil taking is triggered to detect whether the oil amount reaches the predetermined standard.

[0059] In this embodiment, S4 collects electrical signals and temperature signals during the oil extraction process, preprocesses and extracts features from the collected electrical signals and temperature, extracts electrical signal features and temperature signal features that reflect changes in the state of the oil sample, uses statistical methods to analyze the relationship between the electrical signal features and the temperature signal features, identifies the features most relevant to the changes in the state of the oil sample, and calculates an oil sample state assessment coefficient based on the results of the feature analysis to reflect the quality of the oil, including oil viscosity, oil density, and oil moisture content;

[0060] By recording the number of times the needle moves to trigger the switch each time oil is taken, a machine learning algorithm is used to construct an oil extraction quantity prediction model, the oil sample state evaluation coefficient and the number of times the travel switch is triggered during needle oil extraction form a data set, and the data set is divided into a training set, a validation set, and a test set. The constructed oil extraction quantity prediction model is trained using the training set, and the model parameters are adjusted until a satisfactory prediction effect is achieved. The model performance is evaluated using the test set, and the oil sample state evaluation coefficient and the number of times the travel switch is triggered are input into the trained model to obtain an oil extraction quantity prediction value. A predetermined value is set, and the predicted oil extraction quantity is compared with the predetermined value to determine whether the oil extraction quantity meets the predetermined standard. If the predicted oil extraction quantity reaches or exceeds the predetermined value, it is considered that the oil extraction quantity meets the predetermined standard and there is no need to perform a calibration operation on the oil quantity. If the predicted oil extraction quantity does not reach the predetermined value, it is considered that the oil extraction quantity does not meet the predetermined standard, and the detection result that does not meet the predetermined standard is transmitted to S5 for oil quantity calibration analysis.

[0061] It should be specifically noted that the corresponding characteristic parameters are extracted from the electrical signals and temperature signals collected during the oil extraction process. The electrical signal characteristics include the root mean square value, peak form factor, frequency component, etc., and the temperature signal characteristics include the average value, standard deviation, maximum value, minimum value, etc. The Pearson correlation coefficient is then used to analyze the correlation between the characteristics and the oil sample state changes, and the correlation coefficient between each feature and the oil sample state is calculated. The specific calculation formula is: ,in, represents the correlation coefficient between feature j and oil sample state, represents the value of feature j, represents the average value of feature j, Indicates the viscosity value of the oil sample state, Represents the average value of the viscosity of the oil sample state. The weight coefficient of each feature is analyzed and calculated based on the correlation coefficient between each feature and the oil sample state. The specific calculation formula is: ,in, represents the sum of the absolute values ​​of all characteristic correlation coefficients, and the oil sample condition evaluation coefficient is ;

[0062] Record the number of times the travel switch is triggered each time oil is taken, and combine the oil sample state evaluation coefficient and the number of travel switch triggers to form a data set. The data set is divided into a training set, a validation set, and a test set. Select a machine learning algorithm as the constructed oil extraction volume prediction model, such as random forest, gradient boosting machine, etc. Use the training set to train the model, use the validation set to adjust the model parameters, such as decision tree depth, learning rate, etc., and use the test set to evaluate the model performance. The oil sample state evaluation coefficient and the number of travel switch triggers of the current oil extraction are used as the feature parameters of the input model, and the model outputs the oil extraction volume prediction value. Set the predetermined value of the oil extraction volume according to demand, and compare the oil extraction volume prediction value with the predetermined value to determine whether the oil extraction volume meets the predetermined standard and calibration.

[0063] S5: Based on the oil extraction amount detection and analysis results, a calibration analysis is performed on the oil extraction amount that does not meet the predetermined standard, and an oil amount calibration coefficient is calculated to adjust the detected oil extraction amount that does not meet the predetermined standard in real time.

[0064] In this embodiment, when receiving the oil volume detection result that does not meet the predetermined standard, S5 calculates the oil volume calibration coefficient based on the oil volume data that does not meet the predetermined standard. The specific calculation formula is: , where Y represents the original predicted value of oil extraction, represents the actual amount of oil taken at the i-th time, Represents the predicted value of the oil extraction for the i-th time, N represents the number of data points that do not meet the predetermined standards; by setting an error threshold , combined with the oil extraction calibration coefficient, the oil extraction amount that does not meet the predetermined standard is calibrated, and when the error of the oil extraction amount for M consecutive times is less than the error threshold, that is, , stop oil quantity calibration.

[0065] S6: It is used to monitor the pipeline status in real time during the oil extraction process. It monitors and analyzes the oil extraction pipeline status through oil extraction flow and pipeline pressure, and detects whether there is any abnormal risk in the oil extraction pipeline.

[0066] In this embodiment, the S6 constructs a time series prediction model by using a machine learning algorithm, takes flow and pressure as characteristic parameters of the input model, and then outputs the pipeline flow prediction value and pipeline status from the model. When the oil extraction pipeline status output by the model is abnormal, the real-time flow rate at which the abnormality occurs is recorded, and the pipeline flow prediction value is analyzed with the real-time flow rate to detect abnormal conditions in the pipeline status, and the abnormal results are given as early warning feedback to remind the inspection personnel that there is a risk of abnormality in the oil extraction pipeline.

[0067] It should be specifically explained that by extracting characteristic parameters such as mean, standard deviation, peak-to-peak value, instantaneous flow, pressure fluctuation rate, etc. from the flow and pressure data, a long short-term memory network is used to construct a time series prediction model. The instantaneous flow and pressure fluctuation rate characteristics of the oil pipeline are then used as characteristic parameters of the input model, and the model outputs the pipeline flow prediction value and pipeline status label ("normal", "leakage" or "blockage"). When the pipeline status label output by the model shows an abnormality detection, the normalized mean square error between the real-time flow and the predicted value is calculated. If the normalized mean square error between the real-time flow and the predicted value for three consecutive time points exceeds the dynamic threshold based on the distribution of historical data, it is marked as a potential leakage anomaly and the manual review process is triggered.

[0068] S7: Used to provide real-time feedback of the oil extraction process analysis and testing process to the testing personnel terminal, and to display the oil extraction process analysis results in a visual manner

[0069] In this embodiment, the S7 generates a test report according to the oil extraction process analysis and detection process and sends it to the detection personnel terminal, and displays the oil extraction process analysis results in a visual manner on the TFT liquid crystal screen.

[0070] like Figure 2 The embodiment shown provides an implementation system corresponding to the transformer automatic oil extraction process monitoring method, including a data acquisition module, a push rod state detection module, a solenoid valve time delay detection module, an oil extraction quantity detection and analysis module, an oil extraction quantity calibration module, a pipeline state analysis module and an oil extraction result feedback module; the data acquisition module is connected to the push rod state detection module, the push rod state detection module is connected to the solenoid valve time delay detection module, the solenoid valve time delay detection module is connected to the oil extraction quantity detection and analysis module, the oil extraction quantity detection and analysis module is connected to the oil extraction quantity calibration module, the solenoid valve time delay detection module is connected to the pipeline state analysis module, and the oil extraction result feedback module is connected to the push rod state detection module, the solenoid valve time delay detection module, the oil extraction quantity calibration module and the pipeline state analysis module.

[0071] The data acquisition module is used to collect and record the voltage signal data corresponding to the push rod feedback resistance signal during oil extraction, the solenoid valve status data and the oil extraction flow rate;

[0072] The push rod state detection module monitors the action state of the push rod in real time and detects whether the action of the push rod is smooth by performing time series analysis on the collected voltage signal data;

[0073] The solenoid valve time delay detection module performs a delay analysis on the switching time of the solenoid valve based on the recorded solenoid valve status data to detect whether there is a time delay in the switching time of the solenoid valve;

[0074] The oil extraction detection and analysis module is used to detect the oil extraction amount, analyze the oil sample state through the electrical signal and temperature signal during oil extraction, obtain the oil sample state evaluation coefficient, and analyze the oil extraction amount in combination with the number of triggering times of the travel switch of the needle oil extraction to detect whether the oil extraction amount meets the predetermined standard;

[0075] The oil quantity calibration module performs calibration analysis on the oil quantity that does not meet the predetermined standard based on the oil quantity detection and analysis results, and calculates the oil quantity calibration coefficient, which is used to adjust the oil quantity that does not meet the predetermined standard in real time; the oil quantity calibration operation is stopped until the oil quantity meets the predetermined standard;

[0076] The pipeline status analysis module is used to monitor the pipeline status in real time during the oil extraction process, monitor and analyze the oil extraction pipeline status through the oil extraction flow rate and pipeline pressure, and detect whether the oil extraction pipeline is blocked;

[0077] The oil extraction result feedback module is used to feed back the oil extraction process analysis and detection process to the detection personnel terminal, and to display the oil extraction process analysis results in a visual manner.

[0078] like Figure 3 The embodiment shown provides an automatic oil extraction system for a transformer, including the above-mentioned terminal device and network device.

[0079] like Figure 4 The embodiment shown provides a process of an automatic oil extraction device for a transformer, including:

[0080] The three-way valve is used to move the oil from the converter to the waste oil barrel, and the oil-clearing ball valve is opened to discharge the dead oil in the oil extraction pipe. The three-way valve is used to move the oil from the converter to the needle tube, and the oil-clearing ball valve is opened to clean the needle tube. It is judged whether the travel switch is triggered. If the travel switch is triggered, the three-way valve is used to move the oil from the needle tube to the waste oil barrel. If the travel switch is not triggered, the oil-clearing ball valve is reopened to clean the needle tube. After the three-way valve is used to move the oil from the needle tube to the waste oil barrel, the push rod empties the waste oil in the needle tube, and the three-way valve is used to move the oil from the converter to the needle tube. The oil-clearing ball valve is opened to remove oil from the needle tube. It is judged whether the travel switch is triggered. If the travel switch is triggered, the process ends. If the travel switch is not triggered, the oil-clearing ball valve is opened again to remove oil from the needle tube.

[0081] Finally: The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

[0082] The above description is merely a specific embodiment of the present application, but the scope of protection of the present application is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in this application should be included in the scope of protection of this application. Therefore, the scope of protection of this application should be based on the scope of protection of the claims.

Claims

1. A method for monitoring the automatic oil extraction process of a transformer, characterized in that: include: S1: Collect and record the voltage signal data corresponding to the push rod feedback resistance signal during oil extraction, the solenoid valve status data and the oil extraction flow rate; S2: Based on real-time monitoring of the actuator's movement status, the collected voltage signal data is analyzed in time series to detect whether the actuator's movement is smooth; S3: Based on the recorded solenoid valve status data, a delay analysis is performed on the switching time of the solenoid valve to detect whether there is a time delay in the switching time of the solenoid valve; S4: Detecting the amount of oil taken out. The oil sample status is analyzed through the electrical signal and temperature signal during oil taking out. The oil sample status evaluation coefficient is obtained. The oil sample status evaluation coefficient is combined with the number of times the travel switch of the needle oil taking out is triggered to analyze the amount of oil taken out and detect whether the amount of oil taken out reaches the predetermined standard. S5: Based on the oil extraction amount detection and analysis results, the oil extraction amount that does not meet the predetermined standards is calibrated and analyzed, and an oil amount calibration coefficient is calculated. The oil extraction amount that does not meet the predetermined standards is adjusted in real time according to the oil amount calibration coefficient; S6: Monitor and analyze the oil extraction pipeline status through oil extraction flow and pipeline pressure to detect whether there is any abnormal risk in the oil extraction pipeline, so as to monitor the pipeline status in real time during the oil extraction process; S7: The push rod action status detection results, the solenoid valve switch time delay detection results, the oil extraction amount calibration adjustment results and the pipeline status monitoring results are fed back to the inspection personnel terminal in real time, and the oil extraction process analysis results are displayed in a visual manner.

2. A transformer automatic oil extraction process monitoring method according to claim 1, characterized in that: The specific method of collecting the voltage signal data corresponding to the push rod feedback resistance signal by S1 is: using a high-precision data acquisition card to collect the voltage signal converted from the push rod feedback resistance signal, setting a sampling frequency, and collecting the voltage signal data based on the set sampling frequency; The specific method for collecting the solenoid valve status data is as follows: monitoring and recording the on / off status of the solenoid valve through the digital input / output port, and recording the timestamp of the status change; The specific method for collecting the oil extraction time is as follows: when the push rod starts to move, the current time is recorded as the oil extraction start time; when the push rod moves to completion or the flow sensor shows the flow rate returns to zero, the current time is recorded as the oil extraction end time; The specific method for collecting the oil extraction flow is: by installing a flow sensor on the oil extraction pipeline to monitor the flow of the oil ball valve, when the flow is greater than zero, it is considered that oil extraction has begun, and the oil extraction flow value is recorded.

3. A transformer automatic oil extraction process monitoring method according to claim 2, characterized in that: The S2 performs noise reduction and normalization processing on the collected voltage signal data, uses a time series analysis method to analyze the change state of the voltage signal over time, plots the analysis results of the voltage signal over time into a time series chart, and calculates the statistical characteristics of the voltage signal, including the mean and standard deviation. Based on the analysis of the statistical characteristics of the voltage signal, a stability index is calculated, and a stability threshold is set. The stability index is compared with the stability threshold to determine whether the movement of the push rod is smooth.

4. A transformer automatic oil extraction process monitoring method according to claim 3, characterized in that: Based on the analysis results of the push rod action state, S3 performs delay analysis on the actual switching time of the solenoid valve according to the timestamp data of the solenoid valve state change, calculates the time delay of each solenoid valve, compares the calculated time delay of each solenoid valve with the expected time delay, and determines whether there is a time delay in the switching time of each solenoid valve.

5. A transformer automatic oil extraction process monitoring method according to claim 4, characterized in that: The S4 collects electrical signals and temperature signals during the oil extraction process, pre-processes and extracts features of the collected electrical signals and temperature, extracts electrical signal features and temperature signal features reflecting changes in the state of the oil sample, uses statistical methods to analyze the relationship between the electrical signal features and the temperature signal features, identifies the features most relevant to the changes in the state of the oil sample, and calculates an oil sample state assessment coefficient based on the results of the feature analysis to reflect the quality of the oil, including oil viscosity, oil density, and oil moisture content; By recording the number of times the needle moves to trigger the switch each time oil is taken, a machine learning algorithm is used to construct an oil extraction quantity prediction model, the oil sample state evaluation coefficient and the number of times the travel switch is triggered during needle oil extraction form a data set, and the data set is divided into a training set, a validation set, and a test set. The constructed oil extraction quantity prediction model is trained using the training set, and the model parameters are adjusted until a satisfactory prediction effect is achieved. The model performance is evaluated using the test set, and the oil sample state evaluation coefficient and the number of times the travel switch is triggered are input into the trained model to obtain an oil extraction quantity prediction value. A predetermined value is set, and the predicted oil extraction quantity is compared with the predetermined value to determine whether the oil extraction quantity meets the predetermined standard. If the predicted oil extraction quantity reaches or exceeds the predetermined value, it is considered that the oil extraction quantity meets the predetermined standard and there is no need to perform a calibration operation on the oil quantity. If the predicted oil extraction quantity does not reach the predetermined value, it is considered that the oil extraction quantity does not meet the predetermined standard, and the detection result that does not meet the predetermined standard is transmitted to S5 for oil quantity calibration analysis.

6. A transformer automatic oil extraction process monitoring method according to claim 5, characterized in that: When receiving the oil volume detection result that does not meet the predetermined standard, S5 calculates the oil volume calibration coefficient based on the oil volume data that does not meet the predetermined standard. The specific calculation formula is: , where Y represents the original predicted value of oil extraction, represents the actual amount of oil taken at the i-th time, Represents the predicted value of the oil extraction for the i-th time, N represents the number of data points that do not meet the predetermined standards; by setting an error threshold , combined with the oil extraction calibration coefficient, the oil extraction amount that does not meet the predetermined standard is calibrated, and when the error of the oil extraction amount for M consecutive times is less than the error threshold, that is, , stop oil quantity calibration.

7. A transformer automatic oil extraction process monitoring method according to claim 4, characterized in that: The S6 uses a machine learning algorithm to build a time series prediction model, takes flow and pressure as characteristic parameters of the input model, and then the model outputs the pipeline flow prediction value and pipeline status. When the oil pipeline status output by the model is abnormal, the real-time flow rate at the time of the abnormality is recorded, and the pipeline flow prediction value is analyzed with the real-time flow rate to detect abnormal conditions of the pipeline status, and the abnormal results are given as early warning feedback to remind the inspection personnel of the risk of abnormality in the oil pipeline.

8. The method for monitoring the automatic oil extraction process of a transformer according to claim 1, characterized in that: The S7 generates a test report according to the oil extraction process analysis and detection process and sends it to the detection personnel terminal, and displays the oil extraction process analysis results in a visual manner on the TFT liquid crystal screen.

9. A transformer automatic oil extraction process monitoring system, implementing a transformer automatic oil extraction process monitoring method according to any one of claims 1 to 8, characterized in that: include: Data acquisition module: used to collect and record the voltage signal data corresponding to the push rod feedback resistance signal during oil extraction, solenoid valve status data and oil extraction flow rate; Actuator status detection module: Based on real-time monitoring of the actuator's movement status, the module detects whether the actuator's movement is smooth by performing time series analysis on the collected voltage signal data. Solenoid valve time delay detection module: Based on the recorded solenoid valve status data, it performs delay analysis on the solenoid valve switching time to detect whether there is a time delay in the solenoid valve switching time; Oil extraction detection and analysis module: used to detect the oil extraction volume, analyze the oil sample status through the electrical signal and temperature signal during oil extraction, obtain the oil sample status evaluation coefficient, and analyze the oil extraction volume in combination with the number of triggering times of the travel switch of the needle oil extraction to detect whether the oil extraction volume meets the predetermined standard; Oil volume calibration module: Based on the oil volume detection and analysis results, the module calibrates and analyzes the oil volume that does not meet the predetermined standards, and calculates the oil volume calibration coefficient, which is used to adjust the oil volume that does not meet the predetermined standards in real time; the oil volume calibration operation is stopped until the oil volume meets the predetermined standards; Pipeline status analysis module: used to monitor the pipeline status in real time during the oil extraction process, monitor and analyze the oil extraction pipeline status through oil extraction flow and pipeline pressure, and detect whether the oil extraction pipeline is blocked; Oil extraction result feedback module: used to feed back the oil extraction process analysis and detection process to the detection personnel terminal, and display the oil extraction process analysis results in a visual manner.

Images