A transformer winding short circuit state detection method based on pulse frequency response curve characteristics

Abstract

The application discloses a transformer winding short-circuit state detection method based on pulse frequency response curve characteristics, first extracts the anti-resonance point of the amplitude-frequency response curve and the L-C zero-crossing point of the phase-frequency response curve to carry out function fitting, calculates amplitude-phase coexistence fitness TS, judges whether the winding to be detected is faulty or not; then carries out gray scale transformation on the pulse frequency response curve image, calculates image metric relationship TC, distinguishes winding inter-turn short circuit or winding capacitor fault; after judging the winding capacitor fault, calculates the gray scale-gradient coexistence rate TM of the pulse frequency response curve image, and distinguishes winding inter-turn short circuit or pie inter-turn short circuit.

Description

Technical Field

[0001] This application relates to the field of power equipment fault diagnosis technology, specifically to a method for detecting short-circuit conditions of transformer windings based on pulse frequency response curve characteristics. Background Technology

[0002] Transformers, as expensive and crucial equipment in power systems, directly impact power supply quality through their operational safety and reliability. However, due to the limited short-circuit withstand capability and the aging of winding insulation, power transformers are prone to insulation degradation, leading to inter-turn and inter-winding short circuits. This jeopardizes the safe operation of the transformer and, in severe cases, can cause transformer oil temperature rise and fire. Surveys have found that approximately 55% of power transformer failures are caused by winding short circuits. Therefore, timely monitoring of transformer winding conditions is essential for preventing sudden transformer failures.

[0003] Transformer windings can be viewed as passive two-port networks, characterized by distributed parameters such as inductance, resistance, capacitance, and conductance. When the transformer's condition changes, the equivalent inductance, resistance, and capacitance of the windings gradually change, and this change in winding condition can be inferred from the changes in the winding's frequency response. However, current standards related to the pulse frequency response method cannot determine the type of fault and short-circuit condition of the windings. To address this issue, this invention patent introduces a transformer winding short-circuit condition detection method based on pulse frequency response curve characteristics. This method offers advantages such as accurate detection of winding short-circuit conditions and a clear process, which is of great significance for ensuring the operational reliability of power systems. Summary of the Invention

[0004] This application provides a method for detecting the short-circuit state of transformer windings based on the characteristics of pulse frequency response curves. The proposed characteristic parameters can accurately and effectively determine the short-circuit state of transformer windings.

[0005] This application provides a method for detecting the short-circuit state of transformer windings based on pulse frequency response curve characteristics, the method comprising:

[0006] (1) Using a pulse frequency response testing system, obtain the amplitude frequency response curve and phase frequency response curve at 0-1MHz, and select the anti-resonance point of the amplitude frequency response curve. and the phase frequency response curve LC zero crossing point N represents the total number of sampling points for the pulse frequency response curve;

[0007] (2) Using the anti-resonance point X of the amplitude-frequency curve and the zero-crossing point Y of the phase-frequency curve LC as two-dimensional variables, a function fitting is performed to obtain the amplitude-phase co-occurrence function matrix. Calculate the amplitude-phase co-occurrence fitness TS of the winding pulse frequency response:

[0008] (1)

[0009] (2)

[0010] in, , These are the maximum value at the anti-resonance point of the winding pulse amplitude-frequency response curve and the minimum value at the LC zero-crossing point of the winding phase-frequency response curve, respectively. , These represent the total number of anti-resonance points in the winding pulse amplitude-frequency response curve and the total number of LC zero-crossing points in the winding pulse phase-frequency response curve, respectively; when When it is determined that no fault has occurred in the transformer winding; when Timely determination of transformer winding faults; It is a constant related to the transformer's model, size, and capacity;

[0011] (3) After determining that a fault has occurred in the transformer winding, perform grayscale transformation on the pulse amplitude-frequency response curve image and calculate the metric relationship TC of the winding pulse frequency response image:

[0012] (3)

[0013] Where p and q are the gradient number of the pulse amplitude-frequency response curve and the normalized gradient number of the pulse phase-frequency response curve, respectively; when When determining the fault in the winding capacitor of the transformer under test; when Timely detection of inter-turn short circuit faults in transformer windings; It is a constant related to the transformer's model, size, and capacity;

[0014] (4) After determining that a capacitor fault has occurred in the transformer winding, calculate the characteristic descriptor of the amplitude-phase co-occurrence function matrix of the pulse frequency response curve image:

[0015] (4)

[0016] (5)

[0017] Calculate the longitudinal correlation (TR) and lateral correlation (TD) between the amplitude-phase co-occurrence function matrix characteristic descriptors T1 and T2 and the pulse frequency response curve image:

[0018] (6)

[0019] (7)

[0020] (5) Calculate the gray-level-gradient co-occurrence rate TM of the pulse frequency response curve image:

[0021] (8)

[0022] Where p1 and q1 are the first grayscale value and the first gradient value in the mid-frequency band, respectively; p2 and q2 are the first grayscale value and the first gradient value in the high-frequency band, respectively; when When a short circuit is detected between the winding discs of the transformer under test; Timely determination of inter-turn short circuit in transformer windings; It is a constant related to the transformer model, size, and capacity.

[0023] The beneficial effects of this invention are as follows: after obtaining the pulse amplitude-frequency response data of the transformer winding through the pulse frequency response testing system, function fitting is performed to calculate the amplitude-phase co-occurrence fitness TS of the winding amplitude-frequency curve to determine whether a winding fault has occurred; the metric relationship TC is calculated to distinguish between inter-turn short-circuit faults or winding coupling capacitor faults; after determining the winding coupling capacitor fault, the gray-gradient co-occurrence rate TM is calculated to distinguish between inter-pane short circuits or inter-turn short circuits; this provides a reference for transformer winding evaluation and improves the detection accuracy of winding short-circuit states. Attached Figure Description

[0024] Figure 1 This is a flowchart of the method of the present invention;

[0025] Figure 2 This is the pulse frequency response data testing system of the present invention. Detailed Implementation

[0026] The present invention will now be described in further detail with reference to the accompanying drawings:

[0027] like Figure 1 As shown, a method for detecting short-circuit conditions of transformer windings based on pulse frequency response curve characteristics is described, the method comprising:

[0028] (1) Building such Figure 2 The pulse frequency response curve test system shown is mainly composed of an input bushing (1), an output bushing (2), a winding input terminal (3), a winding output terminal (4), a transformer housing (5), and a computer (6). When the system is working, the pulse signal generating circuit generates an excitation signal, which is injected into the winding input terminal (3) through the input bushing (1). The high-pass filter collects the response signal from the winding output terminal (4) through the output bushing (2), filters and denoises the data, and then transmits it to the signal acquisition device for processing and analysis by the computer (6). Through the pulse frequency response test system, the amplitude frequency response curve and phase frequency response curve at 0-1MHz are obtained, and the anti-resonance point of the amplitude frequency response curve is selected. and the phase frequency response curve LC zero crossing point N represents the total number of sampling points for the pulse frequency response curve;

[0029] (2) Using the anti-resonance point X of the amplitude-frequency curve and the zero-crossing point Y of the phase-frequency curve LC as two-dimensional variables, a function fitting is performed to obtain the amplitude-phase co-occurrence function matrix. Calculate the amplitude-phase co-occurrence fitness TS of the winding pulse frequency response:

[0030] (1)

[0031] (2)

[0032] in, , These are the maximum value at the anti-resonance point of the winding pulse amplitude-frequency response curve and the minimum value at the LC zero-crossing point of the winding phase-frequency response curve, respectively. , These represent the total number of anti-resonance points in the winding pulse amplitude-frequency response curve and the total number of LC zero-crossing points in the winding pulse phase-frequency response curve, respectively; when When it is determined that no fault has occurred in the transformer winding; when Timely determination of transformer winding faults; It is a constant related to the transformer's model, size, and capacity;

[0033] (3) After determining that a fault has occurred in the transformer winding, perform grayscale transformation on the pulse amplitude-frequency response curve image and calculate the metric relationship TC of the winding pulse frequency response image:

[0034] (3)

[0035] Where p and q are the gradient number of the pulse amplitude-frequency response curve and the normalized gradient number of the pulse phase-frequency response curve, respectively; when When determining the fault in the winding capacitor of the transformer under test; when Timely detection of inter-turn short circuit faults in transformer windings; It is a constant related to the transformer's model, size, and capacity;

[0036] (4) After determining that a capacitor fault has occurred in the transformer winding, calculate the characteristic descriptor of the amplitude-phase co-occurrence function matrix of the pulse frequency response curve image:

[0037] (4)

[0038] (5)

[0039] Calculate the longitudinal correlation (TR) and lateral correlation (TD) between the amplitude-phase co-occurrence function matrix characteristic descriptors T1 and T2 and the pulse frequency response curve image:

[0040] (6)

[0041] (7)

[0042] (5) Calculate the gray-level-gradient co-occurrence rate TM of the pulse frequency response curve image:

[0043] (8)

[0044] Where p1 and q1 are the first grayscale value and the first gradient value in the mid-frequency band, respectively; p2 and q2 are the first grayscale value and the first gradient value in the high-frequency band, respectively; when When a short circuit is detected between the winding discs of the transformer under test; Timely determination of inter-turn short circuit in transformer windings; It is a constant related to the transformer model, size, and capacity.

Claims

1. A method for detecting the short-circuit state of transformer windings based on pulse frequency response curve characteristics, characterized in that: (1) Using a pulse frequency response testing system, obtain the amplitude frequency response curve and phase frequency response curve at 0-1MHz, and select the anti-resonance point of the amplitude frequency response curve. and the LC zero-crossing point of the phase frequency response curve N represents the total number of sampling points for the pulse frequency response curve; (2) Using the anti-resonance point X of the amplitude-frequency curve and the zero-crossing point Y of the phase-frequency curve LC as two-dimensional variables, a function fitting is performed to obtain the amplitude-phase co-occurrence function matrix. Calculate the amplitude-phase co-occurrence fitness TS of the winding pulse frequency response: （1） （2） in, , These are the maximum value at the anti-resonance point of the winding pulse amplitude-frequency response curve and the minimum value at the LC zero-crossing point of the winding phase-frequency response curve, respectively. , These represent the total number of anti-resonance points in the winding pulse amplitude-frequency response curve and the total number of LC zero-crossing points in the winding pulse phase-frequency response curve, respectively; when When it is determined that no fault has occurred in the transformer winding; when Timely determination of transformer winding faults; It is a constant related to the transformer's model, size, and capacity; (3) After determining that a fault has occurred in the transformer winding, perform grayscale transformation on the pulse amplitude-frequency response curve image and calculate the metric relationship TC of the winding pulse frequency response image: （3） Where p and q are the gradient number of the pulse amplitude-frequency response curve and the normalized gradient number of the pulse phase-frequency response curve, respectively; when When determining the fault in the winding capacitor of the transformer under test; when Timely detection of inter-turn short circuit faults in transformer windings; It is a constant related to the transformer's model, size, and capacity; (4) After determining that a capacitor fault has occurred in the transformer winding, calculate the characteristic descriptor of the amplitude-phase co-occurrence function matrix of the pulse frequency response curve image: （4） （5） Calculate the longitudinal correlation (TR) and lateral correlation (TD) between the amplitude-phase co-occurrence function matrix characteristic descriptors T1 and T2 and the pulse frequency response curve image: （6） （7） (5) Calculate the gray-level-gradient co-occurrence rate TM of the pulse frequency response curve image: （8） Where p1 and q1 are the first grayscale value and the first gradient value in the mid-frequency band, respectively; p2 and q2 are the first grayscale value and the first gradient value in the high-frequency band, respectively; when When a short circuit is detected between the winding discs of the transformer under test; Immediately determine if there is a short circuit between turns in the transformer winding; It is a constant related to the transformer model, size, and capacity.

Images