A method and system for online monitoring of distribution transformer windings under short-circuit impact

Abstract

The present invention provides a method and system for online monitoring of distribution transformer windings under short-circuit impact, belonging to the technical field of distribution transformer winding detection. The online monitoring system for distribution transformer windings under short-circuit impact is used to perform online monitoring of the winding deformation state of the distribution transformer during the short-circuit impact process; the online monitoring system for windings is composed of a winding state detection device, an intelligent electric meter, and a data center; the winding state detection device includes a device body, and a vibration acceleration sensor, a high-frequency current sensor, and a temperature sensor that are communicatively connected to the device body; the high-frequency current sensor is connected in series to the low-voltage end, and the vibration acceleration sensor and the temperature sensor are attached to the outer shell of the distribution transformer; the winding state detection device is coupled and powered by the high-frequency current sensor. During operation, the winding state detection device always collects current signals. When the winding state detection device determines that the distribution transformer is in a transient impact state, it starts collecting vibration sensor signals and temperature sensor signals.

Description

Technical Field

[0001] The present invention relates to the technical field of distribution transformer winding detection, and in particular to a method and system for online monitoring of distribution transformer windings under short-circuit impact. Background Art

[0002] After multiple short-circuit shocks, distribution transformers are prone to winding deformation and inter-turn insulation breakdown. In severe cases, the distribution transformer may even be subject to insulation breakdown to the ground. Currently, there are three main methods for testing and analysis, namely, winding deformation testing in a power-off state, infrared thermal imaging detection in a live state, and acoustic imaging detection: The winding deformation test technology is to evaluate the winding deformation state by frequency sweeping when the distribution transformer is in a power-off state. The disadvantages of this method include affecting the power supply reliability of the distribution network, and there is a risk of electric shock for personnel during high-altitude operations. In addition, the defect judgment of this detection technology requires comparison and analysis with the winding deformation test data of the distribution transformer at the factory to give a severity level of the defect. However, the number of distribution transformers is huge, and winding deformation tests are rarely carried out during on-site handover. In addition, the operation and maintenance units have poor data management and do not save or rarely save the corresponding test data, resulting in the winding deformation test method being mainly used in voltages of 110kV and above. Grade transformers; infrared thermal imaging technology, currently mainly based on DL / T664 for distribution transformer detection, this detection method has a good detection effect on external wiring defects of distribution transformers and critical internal defects of distribution transformers, and the defective heating targeted by infrared thermal imaging detection is a steady-state heating, it is difficult to detect heating defects within the impact time of the distribution transformer circuit breaker, and the detection effect on heating defects in local areas inside the distribution transformer after a short-circuit impact is poor; acoustic imaging technology / voiceprint imaging technology has a poor detection effect on transformers with obvious deformation of the windings after a short-circuit impact, and the ultrasonic signals during the circuit breaker impact process or the stable operation process after the impact are not obvious, and the detection effect on ultrasonic signals generated by the lower temperature heating inside the distribution transformer is not obvious. Summary of the Invention

[0003] In view of this, the present invention provides a method and system for online monitoring of distribution transformer windings under short-circuit impact, which can realize online monitoring of winding status, provide an efficient and highly accurate detection solution, and improve the detection and evaluation efficiency of distribution transformer winding status.

[0004] The technical solution adopted by the embodiment of the present invention to solve the technical problem is:

[0005] A method for online monitoring of distribution transformer windings under short-circuit impacts utilizes an online monitoring system for distribution transformer windings under short-circuit impacts to monitor online the deformation state of the windings during the distribution transformer short-circuit impact process. The online monitoring system comprises a winding state detection device, a smart meter, and a data center. The winding state detection device includes a device body, and a vibration acceleration sensor, a high-frequency current sensor, and a temperature sensor that are communicatively connected to the device body.

[0006] The high-frequency current sensor is connected in series to the low-voltage end of the distribution transformer, and the vibration acceleration sensor and the temperature sensor are attached to the outer shell of the distribution transformer;

[0007] The winding state detection device is coupled and powered by a high-frequency current sensor. During operation, the winding state detection device always collects current signals. When the winding state detection device determines that the distribution transformer is in a transient impact state based on the current signal, it starts collecting vibration sensor signals and temperature sensor signals.

[0008] The steps of the online monitoring method include:

[0009] Step S1, installing the vibration acceleration sensor and the temperature sensor at each measuring point of the distribution transformer, and installing the acquisition end of the high-frequency current sensor at the low-voltage side winding position;

[0010] Step S2, the winding state detection device continuously collects the low-voltage side winding current signal and identifies the transient impact current in real time;

[0011] Step S3: When the winding state detection device identifies that the high-frequency current sensor has collected a transient impulse current, the winding state detection device collects vibration sensor signals and temperature sensor signals from each measuring point at a preset collection frequency, and continuously collects the low-voltage side winding current signal;

[0012] Step S4, the winding state detection device plots the waveforms of each group of signals, extracts waveform features, and further determines the defect levels of the three types of signals;

[0013] In step S5, the winding state detection device further calculates the winding state value of the distribution transformer based on the defect levels of the three types of signals. The expression of the winding state value X is:

[0014] X=α1X1+α2X2+α3X3

[0015] Where X1 is the defect level of the current signal, X2 is the defect level of the vibration sensor signal, and X3 is the defect level of the temperature sensor signal; α1 is the current signal weight, α2 is the vibration sensor signal weight, and α3 is the temperature sensor signal weight; α1+α2+α3=1;

[0016] Step S6: The winding state detection device determines the defect degree of the distribution transformer according to the winding state value of the distribution transformer, and determines whether the distribution transformer can continue to operate; wherein X∈[0,0.6) indicates that the distribution transformer can continue to operate, and X∈[0.6,1] indicates that the distribution transformer cannot continue to operate. Specifically, X∈[0.6,0.8) indicates a general defect, X∈[0.8,0.9) indicates a serious defect, and X∈[0.9,1] indicates a critical defect;

[0017] Step S7: when X≥0.6, the winding state detection device sends a judgment result to the high-frequency current sensor;

[0018] Step S8: The high-frequency current sensor generates a 3x frequency signal based on the judgment result by using frequency conversion technology and couples the signal to the low-voltage side line of the distribution transformer. The 3x frequency signal follows the line and is output to the smart meter corresponding to the line. The amplitude of the 3x frequency signal does not exceed 5V.

[0019] Step S9, the smart meter identifies the triple frequency signal, encodes the signal based on the triple frequency signal, and uploads the generated encoded signal to the data center to issue an abnormality alarm upward;

[0020] In step S10, after receiving the coded signal, the data center immediately performs a linear regression analysis on the secondary voltage of the distribution transformer corresponding to the coded signal, and processes the distribution transformer according to the defect level corresponding to the coded signal.

[0021] Preferably, three collection points are set on the outer surface of the side shell of the distribution transformer, which are evenly distributed in the circumferential direction of the winding and located at the middle of the area where the winding is located; at each collection point, one vibration acceleration sensor, one temperature sensor, one detection signal sending device and one protective cover are fixedly installed on the outer surface of the shell, and the vibration acceleration sensor, the detection signal sending device and the temperature sensor are located inside the protective cover.

[0022] Preferably, the step S2 includes:

[0023] Step S21, using the data window to select continuous sampling points, calculate the amplitude of the AC component of the short-circuit impulse current I ac :

[0024]

[0025] Where: I ac is the amplitude of the AC component of the short-circuit impulse current collected; i n+1 ,i n ,i n-1 are three consecutive current sampling points; Ts is the sampling period; ω is the system angular frequency;

[0026] Step S22, I ac When the rate of change increases by 20%, the winding state detection device determines that the distribution transformer enters a transient impact state.

[0027] Preferably, the step S4 includes:

[0028] Step S41: Analyze the current signal to obtain the current signal waveform and waveform amplitude. max It represents the maximum value of the three-phase waveform amplitude, that is, the maximum short-circuit current, and I represents I max The defect level X1 of the current signal corresponding to the rated short-circuit current value of the phase is:

[0029]

[0030] Among them, X1 value of 0.5 indicates a general defect, 0.8 indicates a serious defect, and 1 indicates a critical defect;

[0031] Step S42: vibration sensor signal analysis. Based on the three vibration sensor signal waveforms, the vibration entropy increment and low-frequency power increment are calculated respectively. Let △1 represent the maximum value of the three vibration entropy increments, and △2 represent the maximum value of the three low-frequency power increments. The defect level X2 of the vibration sensor signal is:

[0032]

[0033] Among them, X2 value of 0 indicates no serious defect, 0.7 indicates a serious defect, and 1 indicates a critical defect;

[0034] Step S43: Analyze the temperature sensor signal to obtain the waveforms and amplitudes of the three temperature sensor signals. max Indicates the maximum value among the three waveform amplitudes, t min Represents the minimum value among the three waveform amplitudes. The defect level X3 of the temperature sensor signal is:

[0035]

[0036] Among them, an X3 value of 0.5 indicates a general defect, a value of 0.8 indicates a serious defect, and a value of 1 indicates a critical defect.

[0037] Preferably, step S8 includes: after the winding state detection device sends the judgment result to the high-frequency current sensor, the high-frequency current sensor generates the 3-fold frequency signal through frequency conversion technology, and then couples the 3-fold frequency signal to the line.

[0038] Preferably, step S9 includes: the smart meter converts the 3-fold frequency signal coupled from the line into a binary control signal, and encodes the binary control signal to generate the coded signal; and uploads the coded signal to the data middle station through the limited bandwidth in the smart meter.

[0039] Preferably, in the expression of the winding state value X, α1 = 55%, α2 = 25%, and α3 = 20%.

[0040] Preferably, the specific implementation of the linear regression analysis of the secondary voltage of the distribution transformer corresponding to the coded signal in step S10 includes:

[0041] According to the linear regression analysis expression:

[0042]

[0043] Where I1 and I2 are the currents flowing through the primary and secondary windings of the transformer, N1 and N2 are the turns of the primary and secondary windings of the distribution transformer, and V1 sec 、V2 sec is the voltage value of the primary winding and the secondary winding on the secondary side, I 12 is the neutral point current of the secondary winding, Z1 is the primary winding impedance, Z2 is the secondary winding impedance, Z 12 is the neutral point impedance of the secondary winding;

[0044] During the secondary voltage linear regression analysis of the distribution transformer with abnormal alarm performed by the data center, the first voltage fluctuation amplitude is set to represent the voltage difference between the phase voltages, and the second voltage fluctuation amplitude is set to represent the voltage difference between the single-phase voltage and the voltage during the same load period. If either the first voltage fluctuation amplitude or the second voltage fluctuation amplitude of the distribution transformer exceeds 10%, it is determined that the distribution transformer has a winding breakdown, and the distribution transformer abnormality is registered to arrange a power outage for the distribution transformer; if the first voltage fluctuation amplitude or the second voltage fluctuation amplitude does not exceed 10% in both cases, it is determined that the distribution transformer needs to be tracked and re-measured.

[0045] Preferably, the step S10 of processing the distribution transformer according to the defect level corresponding to the coded signal includes:

[0046] For abnormal alarms of general defects of X∈[0.6,0.8), the tracking and retesting frequency of the distribution transformer is set to one month;

[0047] For abnormal alarms of serious defects X∈[0.8,0.9), set the tracking and retesting frequency of the distribution transformer to one week, generate a maintenance reminder message, and arrange a maintenance within the preset time period;

[0048] For abnormal alarms of critical defects X∈[0.9,1], an immediate maintenance prompt message is generated, and the distribution transformer is arranged to be shut down for maintenance immediately.

[0049] The present invention provides an online monitoring system for distribution transformer windings under short-circuit impact, which is used to implement the above-mentioned method. The system consists of a winding status detection device, a smart meter, and a data center. The winding status detection device includes a device body, and a vibration acceleration sensor, a high-frequency current sensor, and a temperature sensor that are communicatively connected to the device body.

[0050] The high-frequency current sensor is connected in series to the low-voltage end of the distribution transformer, and the vibration acceleration sensor and the temperature sensor are attached to the outer shell of the distribution transformer;

[0051] The winding state detection device is coupled and powered by a high-frequency current sensor. During operation, the winding state detection device always collects current signals. When the winding state detection device determines that the distribution transformer is in a transient impact state based on the current signal, it starts collecting vibration sensor signals and temperature sensor signals.

[0052] It can be seen from the above technical solution that the embodiment of the present invention provides an online monitoring method and system for the windings of a distribution transformer under short-circuit impact. The system consists of a winding state detection device, a smart meter, and a data middle platform; the winding state detection device includes a device body, and a vibration acceleration sensor, a high-frequency current sensor, and a temperature sensor that are communicatively connected to the device body; the high-frequency current sensor is connected in series to the low-voltage end of the distribution transformer, and the vibration acceleration sensor and the temperature sensor are attached to the outer casing of the distribution transformer; the winding state detection device is coupled and powered by the high-frequency current sensor. During operation, the winding state detection device always collects current signals. When the winding state detection device determines that the distribution transformer is in a transient impact state based on the current signal, it starts collecting vibration sensor signals and temperature sensor signals. By implementing this solution, online monitoring of the winding state can be achieved, early warning signals can be sent in a timely manner, and efficient and highly accurate detection can be provided, so that the distribution transformer can be maintained in a timely manner. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] Figure 1 This is the assembly diagram of the online monitoring system for distribution transformer windings under short-circuit impact.

[0054] Figure 2 Schematic diagram of the online monitoring system for distribution transformer windings under short-circuit impact.

[0055] Figure 3 Schematic diagram of the collection points of the online monitoring system for distribution transformer windings under short-circuit impact.

[0056] Figure 4This is a partial schematic diagram of the collection points of the online monitoring system for distribution transformer windings under short-circuit impact.

[0057] In the figure: distribution transformer 1, smart meter 2, signal receiving device 21 of smart meter, high-frequency current sensor 3, vibration acceleration sensor 4, temperature sensor 5, detection signal sending device 6, protective cover 7, collection point 81, collection point 82, collection point 83, low-voltage side 91 of distribution transformer, high-voltage side 92 of distribution transformer. DETAILED DESCRIPTION

[0058] The technical solutions and technical effects of the present invention are further described in detail below with reference to the accompanying drawings of the present invention.

[0059] The present invention provides a method for online monitoring of distribution transformer windings during short-circuit shocks. The method utilizes an online monitoring system for distribution transformer windings during short-circuit shocks to monitor the winding deformation state of the distribution transformer during short-circuit shocks. Signals indicative of winding deformation during a short-circuit shock are primarily current signals, vibration signals, and secondary voltage signals. The primary signals indicative of winding deformation after a short-circuit shock are thermoacoustic signals (i.e., transformer vibration caused by internal heating of the transformer) and abnormal secondary voltage signals.

[0060] refer to Figure 1-4 As shown, the winding online monitoring system of the present invention is composed of a winding state detection device, a smart meter, and a data center; the winding state detection device includes a device body, and a vibration acceleration sensor, a high-frequency current sensor, and a temperature sensor that are communicatively connected to the device body;

[0061] The high-frequency current sensor is connected in series to the low-voltage end of the distribution transformer, and the vibration acceleration sensor and temperature sensor are attached to the outer casing of the distribution transformer;

[0062] The winding state detection device is coupled with a high-frequency current sensor for power supply. During operation, the winding state detection device always collects current signals. When the winding state detection device determines that the distribution transformer is in a transient impact state based on the current signal, it starts collecting vibration sensor signals and temperature sensor signals.

[0063] refer to Figure 3 As shown, three collection points are set on the outer surface of the side shell of the distribution transformer, which are evenly distributed in the circumferential direction of the winding and located in the middle of the winding area; at each collection point, a vibration acceleration sensor, a temperature sensor, a detection signal sending device and a protective cover are fixedly installed on the outer surface of the shell, and the vibration acceleration sensor, the detection signal sending device and the temperature sensor are located inside the protective cover.

[0064] The temperature sensor, vibration sensor, and winding status detection device can be connected wirelessly or wired. When using a wired connection, holes are opened on the surface of the protective cover for wiring. A buffer such as a gasket or sealing ring can be installed between the protective cover and the transformer housing to prevent interference noise.

[0065] The steps of the method include:

[0066] Step S1: Install vibration acceleration sensors and temperature sensors at each measuring point of the distribution transformer, and install the acquisition end of the high-frequency current sensor at the low-voltage side winding position (refer to Figure 1 As shown, specifically, the current clamp is placed on the three phases ABC and the neutral point);

[0067] Step S2: The winding state detection device continuously collects the low-voltage side winding current signal and identifies the transient impact current in real time;

[0068] Step S3: When the winding state detection device identifies that the high-frequency current sensor has collected a transient impulse current, the winding state detection device collects vibration sensor signals and temperature sensor signals at each measuring point according to a preset collection frequency, and continuously collects the low-voltage side winding current signal;

[0069] Step S4: The winding state detection device plots the waveforms of each group of signals, extracts waveform features, and further determines the defect levels of the three types of signals;

[0070] In step S5, the winding state detection device further calculates the winding state value of the distribution transformer based on the defect levels of the three types of signals. The expression of the winding state value X is:

[0071] X=α1X1+α2X2+α3X3 (1)

[0072] Where X1 is the defect level of the current signal, X2 is the defect level of the vibration sensor signal, and X3 is the defect level of the temperature sensor signal; α1 is the current signal weight, α2 is the vibration sensor signal weight, and α3 is the temperature sensor signal weight; α1+α2+α3=1; α1=55%, α2=25%, α3=20%;

[0073] Step S6: The winding state detection device determines the defect degree of the distribution transformer according to the winding state value of the distribution transformer, and determines whether the distribution transformer can continue to operate; wherein X∈[0,0.6) indicates that the distribution transformer can continue to operate, and X∈[0.6,1] indicates that the distribution transformer cannot continue to operate. Specifically, X∈[0.6,0.8) indicates a general defect, X∈[0.8,0.9) indicates a serious defect, and X∈[0.9,1] indicates a critical defect;

[0074] Step S7: When X≥0.6, the winding state detection device sends the judgment result to the high-frequency current sensor;

[0075] Step S8: The high-frequency current sensor generates a 3x frequency signal based on the judgment result through frequency conversion technology and couples it to the low-voltage side line of the distribution transformer. The 3x frequency signal follows the line and is output to the smart meter corresponding to the line. The high-frequency current sensor generates the 3x frequency signal through frequency conversion technology and couples the 3x frequency signal to the line. The amplitude of the 3x frequency signal does not exceed 5V.

[0076] In step S9, the smart meter identifies the 3x frequency signal, encodes the signal based on the 3x frequency signal, and uploads the generated coded signal to the data center to issue an abnormal alarm. The smart meter converts the 3x frequency signal coupled from the line into a binary control signal, and encodes the binary control signal to generate a coded signal. The coded signal is uploaded to the data center as an abnormal alarm through the limited bandwidth in the smart meter.

[0077] In step S10, after receiving the coded signal, the data center immediately performs a linear regression analysis on the secondary voltage of the distribution transformer corresponding to the coded signal, and processes the distribution transformer according to the defect level corresponding to the coded signal.

[0078] The specific implementation of step S2 of identifying the transient impulse current in real time includes:

[0079] Step S21, using the data window to select continuous sampling points, calculate the amplitude of the AC component of the short-circuit impulse current I ac :

[0080]

[0081] Where: I ac is the amplitude of the AC component of the short-circuit impulse current collected; i n+1 ,i n ,i n-1 are three consecutive current sampling points; T s is the sampling period; ω is the system angular frequency;

[0082] Step S22, I ac When the rate of change suddenly increases by 20%, the winding state detection device determines that the distribution transformer enters a transient impact state.

[0083] The specific implementation of step S4 of determining the defect levels of the three types of signals includes:

[0084] Step S41: Analyze the current signal to obtain the current signal waveform and waveform amplitude. max It represents the maximum value of the three-phase waveform amplitude, that is, the maximum short-circuit current, and I represents I maxThe defect level X1 of the current signal corresponding to the rated short-circuit current value of the phase is:

[0085]

[0086] Among them, X1 value of 0.5 indicates a general defect, 0.8 indicates a serious defect, and 1 indicates a critical defect;

[0087] Step S42: vibration sensor signal analysis. Based on the three vibration sensor signal waveforms, the vibration entropy increment and low-frequency power increment are calculated respectively. Let △1 represent the maximum value of the three vibration entropy increments, and △2 represent the maximum value of the three low-frequency power increments. The defect level X2 of the vibration sensor signal is:

[0088]

[0089] Among them, X2 value of 0 indicates no serious defect, 0.7 indicates a serious defect, and 1 indicates a critical defect;

[0090] The calculation of vibration entropy and low-frequency power can refer to the method specified in the "Technical Specifications for Vibration Testing of Power Transformers by Acceleration Method". The vibration entropy H is calculated according to formula (5):

[0091]

[0092] Where B 100i is the frequency proportion corresponding to the i-th harmonic vibration component of 100 Hz, A 100i is the amplitude of the i-th multiple frequency signal of 100 Hz in the vibration spectrum;

[0093] Low frequency power P 100.200 Calculate according to formula (7):

[0094]

[0095] Where A 100 and A 200 100Hz amplitude, 200H Z Amplitude. If the vibration entropy increases by more than 10% compared to historical values ​​(or for transformers of the same model and operating conditions), it can be determined that the structure has winding deformation. If the low-frequency energy increases by more than 20% compared to historical values ​​(or for transformers of the same model and operating conditions), it indicates that the transformer may have a loose fault.

[0096] Step S43: Analyze the temperature sensor signal to obtain the waveforms and amplitudes of the three temperature sensor signals. max Indicates the maximum value among the three waveform amplitudes, t min Represents the minimum value among the three waveform amplitudes. The defect level X3 of the temperature sensor signal is:

[0097]

[0098] Among them, an X3 value of 0.5 indicates a general defect, a value of 0.8 indicates a serious defect, and a value of 1 indicates a critical defect.

[0099] The specific implementation of step S10 of performing linear regression analysis on the secondary voltage of the distribution transformer corresponding to the coded signal includes:

[0100] According to the linear regression analysis expression:

[0101]

[0102] Where I1 and I2 are the currents flowing through the primary and secondary windings of the transformer, N1 and N2 are the turns of the primary and secondary windings of the distribution transformer, and V1 sec 、V2 sec is the voltage value of the primary winding and the secondary winding on the secondary side, I 12 is the neutral point current of the secondary winding, Z1 is the primary winding impedance, Z2 is the secondary winding impedance, Z 12 is the neutral point impedance of the secondary winding;

[0103] During the secondary voltage linear regression analysis of distribution transformers with abnormal alarms, the data center performs a first voltage fluctuation amplitude to represent the phase-to-phase voltage difference, and a second voltage fluctuation amplitude to represent the voltage difference between the single-phase voltage and the voltage during the same load period. If either the first or second voltage fluctuation amplitude of the distribution transformer exceeds 10%, the distribution transformer is determined to have winding breakdown, and the abnormality is registered to arrange a power outage for the distribution transformer. If neither the first or second voltage fluctuation amplitude exceeds 10%, the distribution transformer is determined to require follow-up retesting. Follow-up retesting here refers to on-site status diagnosis and maintenance.

[0104] The specific implementation of step S10 of processing the distribution transformer according to the defect level corresponding to the coded signal includes:

[0105] For abnormal alarms of general defects X∈[0.6,0.8), the middle station registers the corresponding distribution transformer and sets the tracking and retesting frequency of the distribution transformer to one month;

[0106] For abnormal alarms with serious defects X∈[0.8,0.9), the middle station registers the corresponding distribution transformer, sets the tracking and retesting frequency of the distribution transformer to one week, generates a maintenance reminder message, and arranges a maintenance within the preset time period. Here, the preset time period can be one month, indicating that the maintenance time can be flexibly arranged according to actual conditions.

[0107] For abnormal alarms of critical defects X∈[0.9,1], the middle station registers the corresponding distribution transformer, generates an immediate maintenance prompt message, and arranges for the distribution transformer to be shut down for maintenance immediately; immediate maintenance means arranging it within a very short time, such as within 2 days, indicating a very urgent maintenance arrangement.

[0108] The embodiments of the present invention provide a method and system for online monitoring of distribution transformer windings under short-circuit shocks. By implementing this solution, online monitoring of winding status can be achieved, providing efficient and highly accurate detection, improving the efficiency of detection and assessment of distribution transformer winding status, and enabling timely maintenance of distribution transformers.

[0109] The above disclosure is only a preferred embodiment of the present invention, and it is certainly not intended to limit the scope of the present invention. A person skilled in the art can understand that all or part of the processes of the above embodiment and equivalent changes made in accordance with the claims of the present invention are still within the scope of the invention.

Claims

1. A method for online monitoring of distribution transformer windings under short-circuit impact, characterized in that: The online monitoring system for the windings of a distribution transformer under short-circuit shock is used to monitor the deformation state of the windings of the distribution transformer during short-circuit shock. The online monitoring system for the windings consists of a winding state detection device, a smart meter, and a data center. The winding state detection device includes a device body, and a vibration acceleration sensor, a high-frequency current sensor, and a temperature sensor that are communicatively connected to the device body. The high-frequency current sensor is connected in series to the low-voltage end of the distribution transformer, and the vibration acceleration sensor and the temperature sensor are attached to the outer shell of the distribution transformer; The winding state detection device is coupled and powered by a high-frequency current sensor. During operation, the winding state detection device always collects current signals. When the winding state detection device determines that the distribution transformer is in a transient impact state based on the current signal, it starts collecting vibration sensor signals and temperature sensor signals. The steps of the online monitoring method include: Step S1, installing the vibration acceleration sensor and the temperature sensor at each measuring point of the distribution transformer, and installing the acquisition end of the high-frequency current sensor at the low-voltage side winding position; Step S2, the winding state detection device continuously collects the low-voltage side winding current signal and identifies the transient impact current in real time; Step S3: When the winding state detection device identifies that the high-frequency current sensor has collected a transient impulse current, the winding state detection device collects vibration sensor signals and temperature sensor signals from each measuring point at a preset collection frequency, and continuously collects the low-voltage side winding current signal; Step S4, the winding state detection device plots the waveforms of each group of signals, extracts waveform features, and determines the defect levels of the three types of signals; In step S5, the winding state detection device calculates the winding state value of the distribution transformer based on the defect levels of the three types of signals. The expression of the winding state value X is: X=α1X1+α2X2+α3X3 Where X1 is the defect level of the current signal, X2 is the defect level of the vibration sensor signal, and X3 is the defect level of the temperature sensor signal; α1 is the current signal weight, α2 is the vibration sensor signal weight, and α3 is the temperature sensor signal weight; α1+α2+α3=1; Step S6: The winding state detection device determines the defect degree of the distribution transformer according to the winding state value X of the distribution transformer, and determines whether the distribution transformer can continue to operate; wherein X∈[0,0.6) indicates that the distribution transformer can continue to operate, X∈[0.6,1] indicates that the distribution transformer cannot continue to operate, X∈[0.6,0.8) indicates a general defect, X∈[0.8,0.9) indicates a serious defect, and X∈[0.9,1] indicates a critical defect; Step S7: when X≥0.6, the winding state detection device sends a judgment result to the high-frequency current sensor; Step S8, the high-frequency current sensor generates a 3x frequency signal based on the judgment result through frequency conversion technology and couples it to the low-voltage side line of the distribution transformer. The 3x frequency signal follows the line and is output to the smart meter corresponding to the line; Step S9, the smart meter identifies the triple frequency signal, converts the triple frequency signal into a binary control signal, encodes the binary control signal to generate a coded signal, and uploads the generated coded signal to the data center to issue an abnormality alarm upward; In step S10, after receiving the coded signal, the data center immediately performs a linear regression analysis on the secondary voltage of the distribution transformer corresponding to the coded signal, and processes the distribution transformer according to the defect level corresponding to the coded signal.

2. The method for online monitoring of distribution transformer windings under short-circuit impact according to claim 1, characterized in that: Three collection points are set on the outer surface of the side shell of the distribution transformer, which are evenly distributed in the circumferential direction of the winding and located at the middle of the winding area; at each collection point, one vibration acceleration sensor, one temperature sensor, one detection signal sending device and one protective cover are fixedly installed on the outer surface of the shell, and the vibration acceleration sensor, the detection signal sending device and the temperature sensor are located inside the protective cover.

3. The method for online monitoring of distribution transformer windings under short-circuit impact according to claim 2, characterized in that: The step S2 comprises: Step S21, using the data window to select continuous sampling points, calculate the amplitude of the AC component of the short-circuit impulse current I ac : ; Where: i n+1 ,i n ,i n-1 are three consecutive current sampling points; T s is the sampling period; ω is the system angular frequency; Step S22, I ac When the rate of change increases by 20%, the winding state detection device determines that the distribution transformer enters a transient impact state.

4. The method for online monitoring of distribution transformer windings under short-circuit impact according to claim 3, characterized in that: The step S4 comprises: Step S41: Analyze the current signal to obtain the current signal waveform and waveform amplitude. max It represents the maximum value of the three-phase waveform amplitude, that is, the maximum short-circuit current, and I represents I max The defect level X1 of the current signal corresponding to the rated short-circuit current value of the phase is: ； Among them, X1 value of 0.5 indicates a general defect, 0.8 indicates a serious defect, and 1 indicates a critical defect; Step S42: vibration sensor signal analysis. Based on the three vibration sensor signal waveforms, the vibration entropy increment and low-frequency power increment are calculated respectively. Let △1 represent the maximum value of the three vibration entropy increments, and △2 represent the maximum value of the three low-frequency power increments. The defect level X2 of the vibration sensor signal is: ； Among them, X2 value of 0 indicates no serious defect, 0.7 indicates a serious defect, and 1 indicates a critical defect; Step S43: Analyze the temperature sensor signal to obtain the waveforms and amplitudes of the three temperature sensor signals. max Indicates the maximum value among the three waveform amplitudes, t min Represents the minimum value among the three waveform amplitudes. The defect level X3 of the temperature sensor signal is: ； Among them, an X3 value of 0.5 indicates a general defect, a value of 0.8 indicates a serious defect, and a value of 1 indicates a critical defect.

5. The method according to claim 4, wherein Step S8 includes: after the winding state detection device sends the judgment result to the high-frequency current sensor, the high-frequency current sensor generates the 3-fold frequency signal through frequency conversion technology, and then couples the 3-fold frequency signal to the line; the signal amplitude of the 3-fold frequency signal does not exceed 5V.

6. The method according to claim 5, wherein Step S9 includes: uploading the encoded signal to the data middle station through the limited bandwidth in the smart meter.

7. The method according to claim 6, wherein In the expression of the winding state value X, α1 = 55%, α2 = 25%, and α3 = 20%.

8. The method for online monitoring of distribution transformer windings under short-circuit impact according to claim 7, characterized in that: The specific implementation of the linear regression analysis of the secondary voltage of the distribution transformer corresponding to the coded signal in step S10 includes: According to the linear regression analysis expression: ； Where I1 and I2 are the currents flowing through the primary and secondary windings of the transformer, N1 and N2 are the turns of the primary and secondary windings of the distribution transformer, and V1 sec 、V2 sec is the voltage value of the primary winding and the secondary winding on the secondary side, I 12 is the neutral point current of the secondary winding, Z1 is the primary winding impedance, Z2 is the secondary winding impedance, Z 12 is the neutral point impedance of the secondary winding; During the secondary voltage linear regression analysis of the distribution transformer with abnormal alarm performed by the data center, the first voltage fluctuation amplitude is set to represent the phase voltage difference, and the second voltage fluctuation amplitude is set to represent the voltage difference of the single-phase voltage during the same load period. If either the first voltage fluctuation amplitude or the second voltage fluctuation amplitude of the distribution transformer exceeds 10%, it is determined that the distribution transformer has a winding breakdown, and the distribution transformer abnormality is registered to arrange a power outage for the distribution transformer; if the first voltage fluctuation amplitude or the second voltage fluctuation amplitude does not exceed 10% in both cases, it is determined that the distribution transformer needs to be tracked and re-measured.

9. The method for online monitoring of distribution transformer windings under short-circuit impact according to claim 8, characterized in that: The step S10 of processing the distribution transformer according to the defect level corresponding to the coded signal includes: For abnormal alarms of general defects of X∈[0.6,0.8), the tracking and retesting frequency of the distribution transformer is set to one month; For abnormal alarms of serious defects X∈[0.8,0.9), set the tracking and retesting frequency of the distribution transformer to one week, generate a maintenance reminder message, and arrange a maintenance within the preset time period; For abnormal alarms of critical defects X∈[0.9,1], an immediate maintenance prompt message is generated, and the distribution transformer is arranged to be shut down for maintenance immediately.

10. An online monitoring system for distribution transformer windings under short-circuit impact, characterized in that: For implementing the method according to any one of claims 1 to 9, the system comprises a winding state detection device, a smart meter, and a data center; the winding state detection device comprises a device body, and a vibration acceleration sensor, a high-frequency current sensor, and a temperature sensor that are communicatively connected to the device body; The high-frequency current sensor is connected in series to the low-voltage end of the distribution transformer, and the vibration acceleration sensor and the temperature sensor are attached to the outer shell of the distribution transformer; The winding state detection device is coupled and powered by a high-frequency current sensor. During operation, the winding state detection device always collects current signals. When the winding state detection device determines that the distribution transformer is in a transient impact state based on the current signal, it starts collecting vibration sensor signals and temperature sensor signals.

Images