Transformer winding fault diagnosis method considering time-frequency domain multi-periodic component

Abstract

The invention discloses a transformer winding fault diagnosis method considering time-frequency domain multi-periodic components, which comprises the following steps: firstly, building a transformer frequency response excitation and measurement experiment platform, and obtaining time-domain multi-periodic component response signals of a transformer winding under different frequencies through computer control; calculating time domain multi-response period effective factor characteristics and limit factor characteristics under different frequencies according to the obtained response signals; further calculating the multi-period correlation coefficient and the multi-period complementary angle similarity of the frequency response signal; and finally, calculating a transformer winding state judgment factor and obtaining an evaluation factor gamma to judge whether the transformer winding breaks down or not.

Description

Technical Field

[0001] The present invention relates to the technical field of transformer winding fault diagnosis, and in particular to a transformer winding fault diagnosis method considering multi-periodic components in time and frequency domains. Background Art

[0002] In the power system, the transformer undertakes the core task of power conversion and transmission, and the health status of the transformer winding, as a key component of the transformer, directly affects the safe operation and power supply reliability of the power system. Due to the complex structure and continuous high-load operation, the transformer winding is extremely susceptible to factors such as overload, environmental degradation, and material aging, which gradually lead to a series of potential faults. These faults may cause the transformer to shut down, resulting in power outages, causing serious interference to the power supply of industrial production and important facilities, and even causing economic losses. In addition, winding faults may also cause problems such as internal short circuits or partial discharges, further exacerbating equipment losses, and even causing fires or explosions, posing a safety threat. Therefore, in order to ensure the stable operation of the power system, the status monitoring and fault warning of the transformer windings are crucial.

[0003] As one of the most effective means to detect transformer winding faults, the frequency response analysis method has been widely used in transformer winding fault detection. It can perform fault analysis by measuring the response of the winding under different frequency conditions. The principle of the frequency response method is to apply signals of different frequencies to the winding and record its amplitude and phase response. When the winding state changes (such as insulation damage, short circuit, winding deformation and other faults inside the winding), its corresponding response will change to varying degrees, thereby providing key fault information to discover the potential fault location and nature. However, the analysis and interpretation of the frequency response method depends on rich electrical engineering knowledge, and ordinary people have certain difficulties in understanding and applying it. To solve this problem, the patent of this invention introduces a transformer winding fault diagnosis method that considers multi-periodic components in the time-frequency domain. While improving the detection sensitivity, this method can intuitively determine whether the transformer winding has a fault, thereby further enhancing the evaluation of the transformer winding state and providing reliable technical support for the safety of the power system. Summary of the invention

[0004] The present application provides a transformer winding fault diagnosis method considering multi-periodic components in the time-frequency domain. According to the proposed method, transformer winding faults can be accurately and effectively determined.

[0005] The present application provides a transformer winding fault diagnosis method considering multi-periodic components in the time-frequency domain. The test platform mainly comprises: a transformer high-voltage winding (1), a medium-voltage winding (2), a low-voltage winding (3), a high-voltage bushing (4), a medium-voltage bushing (5), a low-voltage bushing (6), a transformer core (7), transformer oil (8), a transformer box (9), a temperature sensor (10), a humidity sensor (11), a data acquisition system (12), a frequency control system (13), a data analysis system (14), an AC signal injection source (15), a computer a (16), and a computer b (17). The specific test method comprises the following steps:

[0006] Step 1: Obtain the multi-periodic component output response of the transformer winding in the time domain at different frequencies

[0007] The output response of the multi-periodic components of the transformer winding in the time domain at different frequencies is measured. The high-voltage winding (1) is connected to the high-voltage bushing (4), the medium-voltage winding (2) is connected to the medium-voltage bushing (5), and the low-voltage winding (3) is connected to the low-voltage bushing (6). The computer a (16) is connected to the frequency control system (13) for issuing instructions and displaying curves. The computer a (16) is connected to the data acquisition system (12) for monitoring the state of the transformer oil (8) to ensure the accuracy of the test. The computer b (17) is connected to the data analysis system (14) for setting analysis strategies and displaying data. First, The temperature sensor (10) and the humidity sensor (11) collect transformer oil data, and keep the transformer oil temperature and humidity within the allowable range. The computer a (16) sets the frequency control AC signal injection source (15) to inject a high-frequency signal into the high-voltage winding (1). The signal is received at the low-voltage winding (3). The computer b (17) sets the control strategy to obtain the required multi-cycle time domain output signal response, stops the injection and measurement of related equipment, and continues to monitor the transformer oil temperature and humidity. When it is within the allowable range, repeat the above steps, measure the next frequency, and obtain the injection signal U in (f)=[u f1 u f2 …u fk ], the output signal first response cycle Q out (f) = [q f1 q f2 …q fk ], output signal second response cycle E out (f) = [e f1 e f2 …e fk ], output signal third response cycle W out (f) = [w f1 w f2 …w fk ], where f ranges from [1kHz, 1MHz], u fk,q fk 、e fk 、w fk Respectively represent the k-th point amplitude of the time signal at the f-th test frequency of the injection signal, the first response cycle, the second response cycle, and the third response cycle;

[0008] Step 2: Obtaining the periodic frequency response curve

[0009] (1) Calculate the offset correction factor

[0010]

[0011] Among them, C Qf_j , C Ef_j , C Wf_j denote the offset correction factor of the jth point of the time signal at the fth test frequency in the first response period, the second response period, and the third response period, respectively, where j is [1, 2, 3, …, k];

[0012] (2) Calculate the effective factor characteristics of the response period

[0013]

[0014] Among them, R f 、M f , S f Respectively represent the effective factor characteristics of the first response period, the second response period, and the third response period;

[0015] (3) Calculate the response cycle limit factor characteristics

[0016]

[0017]

[0018] Among them, J f , K f , L f Respectively represent the limiting factor characteristics of the first response period, the second response period, and the third response period;

[0019] (4) Calculate the response cycle frequency response

[0020]

[0021] Among them, YI f , ER f ,SA f Respectively represent the frequency responses of the first response period, the second response period, and the third response period under the effective factor characteristics, JA f , JB f , JC fThe frequency responses under the limiting factor characteristics of the first response period, the second response period, and the third response period respectively;

[0022] Step 3: Transformer winding fault determination

[0023] (1) Calculate the multi-period correlation coefficient of the frequency response signal

[0024]

[0025] Among them, YIO f It represents the response amplitude under the effective factor characteristic of the first response cycle frequency at the fth test frequency under normal conditions, ERO f It represents the response amplitude under the effective factor characteristic of the second response cycle at the fth test frequency under normal conditions, SAO f It represents the response amplitude under the effective factor characteristic of the third response period at the fth test frequency under normal conditions, and N represents the number of frequency points in the taken frequency band [1kHz, 1MHz];

[0026] (2) Calculate the multi-cycle complementary angle similarity of the frequency response signal

[0027]

[0028] Among them, JAO f Indicates the response amplitude under the first response cycle frequency limit factor characteristic under normal conditions at the fth test frequency, JBO f Indicates the response amplitude under the second response cycle limit factor characteristic at the fth test frequency under normal conditions, JCO f It represents the response amplitude under the third response cycle limit factor characteristic at the fth test frequency under normal conditions;

[0029] (3) Calculate the transformer winding state evaluation factor γ

[0030]

[0031] If γ≥G, it is judged that the transformer winding has a fault. If γ<G, it is judged that the transformer winding has no fault. G is a constant related to the transformer model, insulating oil state, and winding connection method. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] Figure 1 The test device diagram of the method of the present invention is

[0033] Figure 2 It is a flow chart of the method of the present invention Specific implementation methods

[0035] The present invention is further described in detail below in conjunction with the accompanying drawings:

[0036] like Figure 1 As shown, a time domain multi-periodic component response signal test platform of a transformer at different frequencies mainly includes: a transformer high voltage winding (1), a medium voltage winding (2), a low voltage winding (3), a high voltage bushing (4), a medium voltage bushing (5), a low voltage bushing (6), a transformer core (7), transformer oil (8), a transformer box (9), a temperature sensor (10), a humidity sensor (11), a data acquisition system (12), a frequency control system (13), a data analysis system (14), an AC signal injection source (15), a computer a (16), and a computer b (17);

[0037] Figure 2 The figure shows a flow chart of a transformer winding fault diagnosis method considering multi-periodic components in the time-frequency domain, which is characterized in that the response signals of multi-periodic components in the time domain at different frequencies are combined to comprehensively complete the fault diagnosis of the transformer winding, and specifically includes the following steps:

[0038] Step 1: Obtain the multi-periodic component output response of the transformer winding in the time domain at different frequencies

[0039] The output response of the multi-periodic components of the transformer winding in the time domain at different frequencies is measured. The high-voltage winding (1) is connected to the high-voltage bushing (4), the medium-voltage winding (2) is connected to the medium-voltage bushing (5), and the low-voltage winding (3) is connected to the low-voltage bushing (6). The computer a (16) is connected to the frequency control system (13) for issuing instructions and displaying curves. The computer a (16) is connected to the data acquisition system (12) for monitoring the state of the transformer oil (8) to ensure the accuracy of the test. The computer b (17) is connected to the data analysis system (14) for setting analysis strategies and displaying data. First, The temperature sensor (10) and the humidity sensor (11) collect transformer oil data, and keep the transformer oil temperature and humidity within the allowable range. The computer a (16) sets the frequency control AC signal injection source (15) to inject a high-frequency signal into the high-voltage winding (1). The signal is received at the low-voltage winding (3). The computer b (17) sets the control strategy to obtain the required multi-cycle time domain output signal response, stops the injection and measurement of related equipment, and continues to monitor the transformer oil temperature and humidity. When it is within the allowable range, repeat the above steps, measure the next frequency, and obtain the injection signal U in (f)=[u f1 u f2 …u fk ], the output signal first response cycle Q out (f) = [q f1 q f2 …q fk ], output signal second response cycle E out (f) = [e f1 e f2…e fk ], output signal third response cycle W out (f) = [w f1 w f2 …w fk ], where f ranges from [1kHz, 1MHz], u fk ,q fk 、e fk 、w fk Respectively represent the k-th point amplitude of the time signal at the f-th test frequency of the injection signal, the first response cycle, the second response cycle, and the third response cycle;

[0040] Step 2: Obtaining the periodic frequency response curve

[0041] (1) Calculate the offset correction factor

[0042]

[0043] Among them, C Qf_j , C Ef_j , C Wf_j denote the offset correction factor of the jth point of the time signal at the fth test frequency in the first response period, the second response period, and the third response period, respectively, where j is [1, 2, 3, …, k];

[0044] (2) Calculate the effective factor characteristics of the response period

[0045]

[0046] Among them, R f 、M f , S f Respectively represent the effective factor characteristics of the first response period, the second response period, and the third response period;

[0047] (3) Calculate the response cycle limit factor characteristics

[0048]

[0049] Among them, J f , K f , L f Respectively represent the limiting factor characteristics of the first response period, the second response period, and the third response period;

[0050] (4) Calculate the response cycle frequency response

[0051]

[0052] Among them, YI f , ER f ,SAf Respectively represent the frequency responses of the first response period, the second response period, and the third response period under the effective factor characteristics, JA f , JB f , JC f The frequency responses under the limiting factor characteristics of the first response period, the second response period, and the third response period respectively;

[0053] Step 3: Transformer winding fault determination

[0054] (1) Calculate the multi-period correlation coefficient of the frequency response signal

[0055]

[0056] Among them, YIO f It represents the response amplitude under the effective factor characteristic of the first response cycle frequency at the fth test frequency under normal conditions, ERO f It represents the response amplitude under the effective factor characteristic of the second response cycle at the fth test frequency under normal conditions, SAO f It represents the response amplitude under the effective factor characteristic of the third response period at the fth test frequency under normal conditions, and N represents the number of frequency points in the taken frequency band [1kHz, 1MHz];

[0057] (2) Calculate the multi-cycle complementary angle similarity of the frequency response signal

[0058]

[0059] Among them, JAO f Indicates the response amplitude under the first response cycle frequency limit factor characteristic under normal conditions at the fth test frequency, JBO f Indicates the response amplitude under the second response cycle limit factor characteristic at the fth test frequency under normal conditions, JCO f It represents the response amplitude under the third response cycle limit factor characteristic at the fth test frequency under normal conditions;

[0060] (3) Calculate the transformer winding state evaluation factor γ

[0061]

[0062] If γ≥G, it is judged that the transformer winding has a fault. If γ<G, it is judged that the transformer winding has no fault. G is a constant related to the transformer model, insulating oil state, and winding connection method.

Claims

1. A transformer winding fault diagnosis method considering multi-periodic components in the time and frequency domain, characterized in that: Transformer high voltage winding (1), medium voltage winding (2), low voltage winding (3), high voltage bushing (4), medium voltage bushing (5), low voltage bushing (6), transformer core (7), transformer oil (8), transformer box (9), temperature sensor (10), humidity sensor (11), data acquisition system (12), frequency control system (13), data analysis system (14), AC signal injection source (15), computer a (16), computer b (17), the specific testing method comprises the following steps: Step 1: Obtain the multi-periodic component output response of the transformer winding in the time domain at different frequencies The output response of the multi-periodic components of the transformer winding in the time domain at different frequencies is measured. The high-voltage winding (1) is connected to the high-voltage bushing (4), the medium-voltage winding (2) is connected to the medium-voltage bushing (5), and the low-voltage winding (3) is connected to the low-voltage bushing (6). The computer a (16) is connected to the frequency control system (13) for issuing instructions and displaying curves. The computer a (16) is connected to the data acquisition system (12) for monitoring the state of the transformer oil (8) to ensure the accuracy of the test. The computer b (17) is connected to the data analysis system (14) for setting analysis strategies and displaying data. First, The temperature sensor (10) and the humidity sensor (11) collect transformer oil data, and keep the transformer oil temperature and humidity within the allowable range. The computer a (16) sets the frequency control AC signal injection source (15) to inject a high-frequency signal into the high-voltage winding (1). The signal is received at the low-voltage winding (3). The computer b (17) sets the control strategy to obtain the required multi-cycle time domain output signal response, stops the injection and measurement of related equipment, and continues to monitor the transformer oil temperature and humidity. When it is within the allowable range, repeat the above steps, measure the next frequency, and obtain the injection signal U in (f)=[u f1 u f2 …u fk ], the output signal first response cycle Q out (f) = [q f1 q f2 …q fk ], output signal second response cycle E out (f) = [e f1 e f2 …e fk ], output signal third response cycle W out (f) = [w f1 w f2 …w fk ], where f ranges from [1kHz, 1MHz], u fk ,q fk 、e fk 、w fk Respectively represent the k-th point amplitude of the time signal at the f-th test frequency of the injection signal, the first response cycle, the second response cycle, and the third response cycle; Step 2: Obtaining the periodic frequency response curve (1) Calculate the offset correction factor Among them, C Qf_j , C Ef_j , C Wf_j denote the offset correction factor of the jth point of the time signal at the fth test frequency in the first response period, the second response period, and the third response period, respectively, where j is [1, 2, 3, …, k]; (2) Calculate the effective factor characteristics of the response period Among them, R f 、M f , S f Respectively represent the effective factor characteristics of the first response period, the second response period, and the third response period; (3) Calculate the response cycle limit factor characteristics Among them, J f , K f , L f Respectively represent the limiting factor characteristics of the first response period, the second response period, and the third response period; (4) Calculate the response cycle frequency response Among them, YI f , ER f ,SA f Respectively represent the frequency responses of the first response period, the second response period, and the third response period under the effective factor characteristics, JA f , JB f , JC f The frequency responses under the limiting factor characteristics of the first response period, the second response period, and the third response period respectively; Step 3: Transformer winding fault determination (1) Calculate the multi-period correlation coefficient of the frequency response signal Among them, YIO f It represents the response amplitude under the effective factor characteristic of the first response cycle frequency at the fth test frequency under normal conditions, ERO f It represents the response amplitude under the effective factor characteristic of the second response cycle at the fth test frequency under normal conditions, SAO f It represents the response amplitude under the effective factor characteristic of the third response period at the fth test frequency under normal conditions, and N represents the number of frequency points in the taken frequency band [1kHz, 1MHz]; (2) Calculate the multi-cycle complementary angle similarity of the frequency response signal Among them, JAO f Indicates the response amplitude under the first response cycle frequency limit factor characteristic under normal conditions at the fth test frequency, JBO f Indicates the response amplitude under the second response cycle limit factor characteristic at the fth test frequency under normal conditions, JCO f It represents the response amplitude under the third response cycle limit factor characteristic at the fth test frequency under normal conditions; (3) Calculate the transformer winding state evaluation factor γ If γ≥G, it is judged that the transformer winding has a fault. If γ<G, it is judged that the transformer winding has no fault. G is a constant related to the transformer model, insulating oil state, and winding connection method.

Images