A method for early warning of stator insulation degradation fault of large generator

Abstract

A method for early warning of insulation degradation failure of large-scale generator stator belongs to the field of fault diagnosis of power generation equipment. The present invention obtains original spectrum data based on the online monitoring of partial discharge phase distribution spectrum of large-scale generators, establishes a spectrum feature value extraction algorithm, and establishes early warning information of insulation failure according to the magnitude and trend changes, thereby effectively early warning before insulation degradation failure occurs, avoiding further degradation and serious losses of insulation problems. The present invention is based on mining insulation failure data, obtained through theoretical analysis and actual data law summary, and has the advantages of accuracy and effectiveness, providing an effective method for judging insulation degradation failure of large-scale generator stator.

Description

Technical Field

[0001] The invention relates to the field of large generator fault diagnosis, and in particular to an early warning method for large generator stator insulation degradation fault. Background Art

[0002] Stator insulation failure of large generators is a common problem that is easy to occur and seriously affects the safe operation of the unit. When the insulation failure is serious, it will cause breakdown failure accidents in the stator windings and leads, causing unit damage and shutdown and other adverse conditions. Therefore, it is very important to identify the stator insulation failure of the generator.

[0003] In the monitoring of generator stator insulation, online partial discharge monitoring is one of the most important means. Through online partial discharge monitoring, a partial discharge phase distribution (PRPD) map can be generated, which contains key information about insulation faults. However, it is very difficult to read and identify this information, and it requires professionals. How to establish the relationship between the characteristic values ​​in the PRPD map and the insulation degradation fault, and effectively warn of insulation fault problems, is a technical problem that needs to be solved urgently. Summary of the invention

[0004] The purpose of the present invention is to provide an early warning method for stator insulation degradation failure of a large generator. The technical solution of the present invention comprises the following steps:

[0005] S1. Define Q0 + is the positive discharge matrix at time t0, Q0 – is the negative discharge matrix at time t0, and Q0 is extracted based on the original data of the generator partial discharge phase distribution (PRPD) spectrum at time t0. + and Q0 – , determine Q0 + or Q0 – The number of discharge amplitudes corresponding to each row is n, and the number of phase angles corresponding to each column is m;

[0006] S2. Define k0 + Q0 is at time t0 + The number of non-zero elements in k0 – Q0 is at time t0 – The number of non-zero elements in α0 + is the duty cycle of the positive discharge matrix at time t0, α 0– is the duty cycle of the negative discharge matrix at time t0, and the relationship is established as follows:

[0007] α0 + =k0 + / (m×n)

[0008] α0 – =k0 – / (m×n)

[0009] S3. Definition A m is the duty cycle threshold, and the judgment logic P1 and judgment logic P2 are constructed as follows:

[0010] P1:α0 + ≥A m

[0011] P2:α0 – ≥A m

[0012] S4. Definition of t s is the specified time, t1=t0+t s 、Q1 + is the positive discharge matrix at time t1, Q1 – is the negative discharge matrix at time t1, k1 + Q1 is at time t1 + The number of non-zero elements in k1 – Q1 is at time t1 – The number of non-zero elements in k1 + Q1 is at time t1 + The number of non-zero elements in k1 – Q1 is at time t1 – The number of non-zero elements in , α1 + is the duty cycle of the positive discharge matrix at time t1, α1 – is the duty cycle of the negative discharge matrix at time t1, and the relationship is established as follows:

[0013] α1 + =k1 + / (m×n)

[0014] α1 – =k1 – / (m×n)

[0015] S5. Define G as the duty cycle slope threshold, and construct judgment logic P3 and judgment logic P4 as follows:

[0016] P3:(α1 + –α0 + ) / t s ≥G

[0017] P4:(α1 – –α0 – ) / t s ≥G

[0018] S6. The early warning judgment logic relationship of the stator insulation degradation fault of a large generator is constructed as follows:

[0019] (P1AND P2)OR(P3AND P4)

[0020] When this logic is true, the insulation degradation fault alarm is activated.

[0021] The beneficial effects of the present invention are:

[0022] The present invention proposes an early warning method for stator insulation degradation fault of a large generator. Through this innovative method, the following technical effects can be achieved:

[0023] 1. Deeply explore the potential information of partial discharge maps. By extracting the characteristic values ​​of partial discharge maps, key fault information can be accurately obtained to achieve quantitative evaluation of stator insulation degradation faults of large generators.

[0024] 2. Realize early warning of the development trend of insulation degradation. By analyzing the time trend changes of partial discharge characteristic data, the threshold alarm is activated when the slope of change is large, and the trend of stator insulation status is effectively predicted. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Figure 1 The figure is a flow chart of the early warning method for stator insulation degradation failure of large generators. DETAILED DESCRIPTION

[0026] The present application is further described below in conjunction with the accompanying drawings. The following embodiments are only used to more clearly illustrate the technical solution of the present invention, and cannot be used to limit the protection scope of the present application.

[0027] A method for early warning of stator insulation degradation failure of a large generator, a specific implementation method 1 comprises the following steps:

[0028] S1, such as Figure 1 As shown, define Q0 + is the positive discharge matrix at time t0, Q0 – is the negative discharge matrix at time t0, and Q0 is extracted based on the original data of the generator partial discharge phase distribution (PRPD) spectrum at time t0. + and Q0 – , determine Q0 + or Q0 – The number of discharge amplitudes corresponding to each row is n, and the number of phase angles corresponding to each column is m;

[0029] S2, such as Figure 1 As shown, define k0 + Q0 is at time t0 + The number of non-zero elements in k0 – Q0 is at time t0 – The number of non-zero elements in α0 + is the duty cycle of the positive discharge matrix at time t0, α0 –is the duty cycle of the negative discharge matrix at time t0, and the relationship is established as follows:

[0030] α0 + =k0 + / (m×n)

[0031] α0 – =k0 – / (m×n)

[0032] S3, such as Figure 1 As shown, define A m is the duty cycle threshold, and the judgment logic P1 and judgment logic P2 are constructed as follows:

[0033] P1:α0 + ≥A m

[0034] P2:α0 – ≥A m

[0035] S4, such as Figure 1 As shown, define t s is the specified time, t1=t0+t s 、Q1 + is the positive discharge matrix at time t1, Q1 – is the negative discharge matrix at time t1, k1 + Q1 is at time t1 + The number of non-zero elements in k1 – Q1 is at time t1 – The number of non-zero elements in k1 + Q1 is at time t1 + The number of non-zero elements in k1 – Q1 is at time t1 – The number of non-zero elements in , α1 + is the duty cycle of the positive discharge matrix at time t1, α 1– is the duty cycle of the negative discharge matrix at time t1, and the relationship is established as follows:

[0036] α1 + =k1 + / (m×n)

[0037] α1 – =k1 – / (m×n)

[0038] S5, such as Figure 1 As shown, G is defined as the duty cycle slope threshold, and the judgment logic P3 and judgment logic P4 are constructed as follows:

[0039] P3:(α1 + –α0 + ) / ts ≥G

[0040] P4:(α1 – –α0 – ) / t s ≥G

[0041] S6, such as Figure 1 As shown in the figure, the early warning judgment logic relationship of the stator insulation degradation fault of a large generator is constructed as follows:

[0042] (P1AND P2)OR(P3AND P4)

[0043] When this logic is true, the insulation degradation fault alarm is activated.

[0044] In this implementation, the early warning method for stator insulation degradation fault is universal, and the input conditions can be changed by adjusting parameters such as duty cycle threshold, specified time, duty cycle slope threshold, etc., to ensure the flexibility of fault early warning logic. At the same time, this implementation has strong scalability, and can be combined with other conventional criteria as one of the criteria for insulation degradation trend to further determine the cause of the fault and improve the accuracy of fault diagnosis.

[0045] The present invention is only an exemplary description of the present invention and does not limit its protection scope. Those skilled in the art may also make partial changes thereto, which are within the protection scope of the present invention as long as they do not exceed the spirit of the present invention.

Claims

1. A method for early warning of stator insulation degradation failure of a large generator, the characteristics of which include Follow these steps: S1. Define Q0 + is the positive discharge matrix at time t0, Q0 – is the negative discharge matrix at time t0, and Q0 is extracted based on the original data of the PRPD spectrum of the generator partial discharge phase distribution at time t0. + and Q0 – , determine Q0 + or Q0 – The number of discharge amplitudes corresponding to each row is n, and the number of phase angles corresponding to each column is m; S2. Define k0 + Q0 is at time t0 + The number of non-zero elements in k0 – Q0 is at time t0 – The number of non-zero elements in α0 + is the duty cycle of the positive discharge matrix at time t0, α 0– is the duty cycle of the negative discharge matrix at time t0, and the relationship is established as follows: a0 + =k0 + / (m×n) a0 – =k0 – / (m×n) S3. Definition A m is the duty cycle threshold, and the judgment logic P1 and judgment logic P2 are constructed as follows: P1:a0 + ≥A m P2:a0 – ≥A m S4. Definition of t s is the specified time, t1=t0+t s 、Q1 + is the positive discharge matrix at time t1, Q1 – is the negative discharge matrix at time t1, k1 + Q1 is at time t1 + The number of non-zero elements in k1 – Q1 is at time t1 – The number of non-zero elements in k1 + Q1 is at time t1 + The number of non-zero elements in k1 – Q1 is at time t1 – The number of non-zero elements in α1 + is the duty cycle of the positive discharge matrix at time t1, α1 – is the duty cycle of the negative discharge matrix at time t1, and the relationship is established as follows: a1 + =k1 + / (m×n) a1 – =k1 – / (m×n) S5. Define G as the duty cycle slope threshold, and construct judgment logic P3 and judgment logic P4 as follows: P3:(a1 + –a0 + ) / t s ≥G P4:(a1 – –a0 – ) / t s ≥G S6. The early warning judgment logic relationship of the stator insulation degradation fault of a large generator is constructed as follows: (P1 AND P2) OR (P3 AND P4) When this logic is true, the insulation degradation fault alarm is activated.

Images