A method for evaluating the uneven aging distribution of oil-immersed transformers

Abstract

The present invention discloses a method for evaluating the uneven aging state distribution of an oil-immersed transformer, belonging to the technical field of power transformer operation and maintenance. The method establishes an axial segmentation evaluation model, sets a dynamic degradation interval, and compares and traverses a measured curve with a calculated data set based on a genetic algorithm to search for the best. The method iterates cyclically within the set degradation interval to determine the actual axial segmentation ratio. The method achieves the positioning of axial aging and the quantitative evaluation of the uneven aging state distribution of the transformer. The method increases the possibility of data combination, performs multi-parameter optimization on laboratory data based on measured data, and inversely solves the insulation state of each insulation layer, thereby eliminating errors in laboratory data that may be caused by temperature, environment, and human factors.

Description

Technical Field

[0001] The present invention belongs to the technical field of power transformer operation and maintenance, and in particular relates to a method for evaluating the uneven aging state distribution of an oil-immersed transformer. Background Art

[0002] Oil-immersed transformers play many important roles in the actual operation of power systems, including voltage conversion, isolation protection, harmonic suppression, and long-distance and efficient power transmission. Therefore, they are indispensable key equipment in modern power grids and are crucial to achieving efficient, safe, and reliable power transportation.

[0003] During the operation of an oil-immersed transformer, the flow pattern of the transformer oil can cause uneven aging of the internal main insulation. This uneven flow path can lead to areas where the oil flow is slow or nearly stagnant, causing the temperature in these areas to be higher than in other areas, forming hot spots. Under rated operating conditions, hot spot temperatures can reach as high as 156°C. High temperatures accelerate the aging of the oil-paper insulation structure, degrading its insulation performance. Hot spots are the weakest points in the oil-immersed transformer's insulation, and their insulation condition can represent the overall insulation condition of the oil-immersed transformer. Therefore, assessing the insulation condition of oil-immersed transformers in light of uneven aging is crucial.

[0004] Traditional insulation condition assessment methods all evaluate the transformer as a whole, without considering the uneven aging of the transformer's oil-paper insulation structure, or failing to detect the uneven distribution of the transformer's aging condition, making the measured data unrepresentative of the true insulation condition of the transformer.

[0005] Therefore, a new technical solution is urgently needed in the existing technology to solve this problem. Summary of the Invention

[0006] In order to overcome the deficiencies in the prior art, the present invention provides a method for evaluating the uneven aging state distribution of an oil-immersed transformer, which is used to solve the problem of accurately evaluating the uneven aging state distribution of the transformer.

[0007] To achieve the above object, the present invention adopts a technical solution: a method for evaluating the uneven aging state distribution of an oil-immersed transformer, comprising the following steps:

[0008] S1. Perform frequency domain dielectric response tests on insulating cardboard samples with different polymerization degrees at different temperatures; perform frequency domain dielectric response tests on insulating oil samples with different aging days at different temperatures;

[0009] S2. Constructing a database of complex relative dielectric constants of insulating paperboards with different degrees of polymerization; constructing a database of complex relative dielectric constants of insulating oils with different aging days;

[0010] S3. Based on the main insulation structure of the transformer, the capacitance expression of the composite multi-oil channel column main insulation structure is obtained by using the relationship between capacitance and complex relative dielectric constant;

[0011] S4. Substitute the actual transformer's structural dimensions and number of radial insulation layers into the database of S2 and use the permutations function to generate all possible combinations, each of which corresponds to the complex relative permittivity of one type of oil and the complex relative permittivity of each insulation layer; substitute all the above combinations into the capacitance expression of the composite multi-oil channel columnar main insulation structure of S3 to obtain a basic database of the equivalent capacitance of the entire transformer; each value in the basic database corresponds to a group of capacitances; the group of capacitances corresponds to the complex relative permittivity of one type of oil and the complex relative permittivity of each insulation layer;

[0012] S5. measuring the actual transformer loss factor at different frequencies to obtain a database of complex capacitance test values ​​of the actual transformer;

[0013] S6. Set the height of the starting point below the actual transformer axial degradation interval as a percentage of the overall height to T, and the percentage of the degradation interval to the overall height to S. Based on the structural dimensions of the transformer and the number of radial insulation layers, use a genetic algorithm to perform a traversal optimization on the complex capacitance test value of the actual transformer in S5 and the basic database of the transformer equivalent capacitance in S4. Perform inverse iteration to obtain the average complex relative dielectric constant and average degree of polymerization of each insulation layer, and determine the values ​​of T and S based on the gradient change of the degree of polymerization.

[0014] S7. Based on the data of S2 and the values ​​of T and S determined in S6, a database of complex relative dielectric constants considering axial non-uniform aging is established, thereby establishing a capacitance expression considering axial non-uniform aging;

[0015] S8. Input the specific geometric parameters and the number of insulation layers of the actual transformer, substitute the database of complex relative dielectric constants in S7 into the basic database of overall transformer equivalent capacitance in S4, and construct an extended database of overall transformer equivalent capacitance that takes into account axial non-uniform aging;

[0016] S9. Based on the complex capacitance test value of the actual transformer of S5, the extended database of the overall transformer equivalent capacitance of S8 is traversed and optimized, and the complex relative dielectric constant distribution matrix at different frequency points is obtained by iterative inversion calculation using the genetic algorithm. The matrix is ​​substituted into the database of S2 to obtain the degree of aggregation distribution matrix at different frequency points; the arithmetic mean of the elements at the same position of the degree of aggregation distribution matrix at different frequency points is calculated to form a degree of aggregation distribution matrix.

[0017] The preferred databases of the complex relative dielectric constant of S2 are all databases that have undergone frequency-temperature shift processing.

[0018] Each capacitance in the basic database of the transformer equivalent capacitance of the preferred S4 is normalized.

[0019] Preferably, the insulating layer described in S4 includes one or more partition-oil gap-strut structural units.

[0020] The specific steps for determining the values ​​of T and S in the preferred S6 are as follows:

[0021] S601. Set a threshold interval for T. When the value of T is at the lower limit of the threshold interval, iterate between samples in the degradation interval S according to the first iteration step to obtain a data set for each sample. Each data set of the sample is compared with the complex relative dielectric constant of each insulating layer obtained by inversion, and the optimal set is selected to obtain the set with the smallest value function value to obtain the average degree of polymerization of the corresponding axial segment of each sample.

[0022] S602: Calculate the relative error between the degree of aggregation of each sample and the degree of aggregation of the previous sample, ensuring that it does not exceed 6% ± 0.8%. Continue iterative calculations until the maximum threshold of the degradation interval is reached. If the relative error exceeds 6% ± 0.8%, output the degradation interval value corresponding to the previous sample to determine the actual axial segmentation ratio at that T value.

[0023] S603, T is iterated according to the second iteration step within the threshold range set in S601, and an iterative cycle from S601 to S602 is performed for each T value; the value function value corresponding to each T value is compared, and the group with the smallest value function value is taken to finally determine the values ​​of T and S of the measured transformer.

[0024] Through the above design scheme, the present invention can bring the following beneficial effects:

[0025] 1. The inversion method based on genetic algorithm increases the possibility of data combination, and optimizes the multi-parameters of laboratory data based on measured data to inversely solve the insulation state of each insulation layer, eliminating the errors that may be caused by temperature, environment and human factors in laboratory data;

[0026] 2. Establish an axial segmentation assessment model, set a dynamic degradation interval, and use a genetic algorithm to compare the measured curve with the calculated data set to find the best solution. Iterate within the set degradation interval to determine the actual axial segmentation ratio. This achieves the location of axial aging and the quantitative assessment of the uneven aging distribution of the transformer.

[0027] 3. Normalize the data in big data to maximize data expansion and save storage space;

[0028] 4. Construct an equivalent capacitance parallel model of the composite multi-oil channel column main insulation structure, which is different from the traditional XY model evaluation model and is more in line with the actual transformer insulation structure. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] Figure 1 The present invention is a flow chart of a method for evaluating the uneven aging state distribution of an oil-immersed transformer.

[0030] Figure 2 This is a schematic diagram of the axial segmented model structure in an embodiment of a method for evaluating the uneven aging state distribution of an oil-immersed transformer according to the present invention.

[0031] Figure 3 This is a comparison diagram of the measured curve and calculated curve of the loss factor of a 220 kV oil-immersed transformer in an embodiment of a method for evaluating the uneven aging state distribution of an oil-immersed transformer of the present invention. DETAILED DESCRIPTION

[0032] The following is a detailed description of the specific embodiments of the present invention with reference to the accompanying drawings.

[0033] To more clearly illustrate the present invention, the present invention is further described below with reference to preferred embodiments. Those skilled in the art will appreciate that the detailed description below is illustrative and non-restrictive, and that the user may make various changes to the following parameters without departing from the mechanism and scope of the invention as set forth in the claims. To avoid obscuring the essence of the present invention, well-known methods and processes are not described in detail.

[0034] By the attached Figures 1 to 3 As shown: A method for evaluating the uneven aging state distribution of an oil-immersed transformer includes the following steps:

[0035] S1. Perform frequency domain dielectric response tests on insulating cardboard samples with different polymerization degrees at different temperatures; perform frequency domain dielectric response tests on insulating oil samples with different aging days at different temperatures;

[0036] S2. Constructing a database of complex relative dielectric constants of insulating paperboards with different degrees of polymerization; constructing a database of complex relative dielectric constants of insulating oils with different aging days;

[0037] S3. Based on the main insulation structure of the transformer, the capacitance expression of the composite multi-oil channel column main insulation structure is obtained by using the relationship between capacitance and complex relative dielectric constant. The specific derivation is as follows:

[0038] The capacitance expressions and required parameters of the partitions, struts, and oil gaps in the composite multi-oil channel column main insulation structure are as follows:

[0039]

[0040] Where r is the inner radius of the high-voltage winding, unit: meter, r0 is the outer radius of the low-voltage winding, unit: meter, and M is the ratio of the difference between the inner radius r of the high-voltage winding and the outer radius r0 of the low-voltage winding to the outer radius r0 of the low-voltage winding.

[0041]

[0042] Where C paper (i) represents the capacitance of the i-th layer of separator, n is the number of layers of the transformer main insulation structure, ε0 is the vacuum dielectric constant, ε0 = 8.854 × 10 -12 Farad / meter, ε b is the relative dielectric constant of the oil-impregnated paperboard, unit: Farad / m, h is the height of the transformer main insulation structure, unit: meter, X is the ratio of the thickness of the brace to the thickness of the partition, M is the ratio of the difference between the inner radius r of the high-voltage winding and the outer radius r0 of the low-voltage winding to the outer radius r0 of the low-voltage winding,

[0043]

[0044] Where W is the length of the brace, unit: meter, r0 is the outer radius of the low-voltage winding, unit: meter, k is the ratio of the brace length W to the outer radius of the low-voltage winding r0,

[0045]

[0046] Where k is the ratio of the length of the stay W to the outer radius r0 of the low-voltage winding, M is the ratio of the difference between the inner radius r of the high-voltage winding and the outer radius r0 of the low-voltage winding to the outer radius r0 of the low-voltage winding, n is the number of layers of the transformer main insulation structure, Y(i) is the ratio of the total length of all stays in the i-th layer to the perimeter of the i-1-th layer partition,

[0047]

[0048] Where: C ba (i) C oil (i) represents the capacitance of the strut and oil gap of the i-th layer, respectively; k is the ratio of the strut length W to the outer radius r0 of the low-voltage winding; Y(i) is the ratio of the total length of all struts in the i-th layer to the perimeter of the i-th layer of the partition; n is the number of layers of the transformer main insulation structure; ε0 is the vacuum dielectric constant, ε0 = 8.854 × 10 -12 Farad / meter, ε b is the relative dielectric constant of oil-impregnated paperboard, unit: Farad / meter, ε oil is the relative dielectric constant of the oil, unit: Farad / m, h is the height of the transformer main insulation structure, unit: meter, X is the ratio of the thickness of the brace to the thickness of the partition, M is the ratio of the difference between the inner radius r of the high-voltage winding and the outer radius r0 of the low-voltage winding to the outer radius r0 of the low-voltage winding,

[0049] The equivalent parallel model is constructed by connecting the capacitances of each layer of partitions and struts, partitions and oil gaps in series, and then connecting the series capacitances in parallel to obtain the parallel equivalent capacitance of each layer. The expression of the equivalent capacitance of each layer is:

[0050]

[0051] Where C paper (i) C ba (i) C oil (i) represents the capacitance of the i-th layer of partition, stay and oil gap, ε b (i) is the relative dielectric constant of the i-th layer of oil-impregnated paperboard, unit: Farad / m, ε oil (i) is the relative dielectric constant of the i-th layer of oil, unit: Farad / m, h is the height of the transformer main insulation structure, unit: meter, Y(i) is the ratio of the total length of all the struts in the i-th layer to the perimeter of the i-1-th layer of partition, and the expressions of the parameters b(i) and d(i) in the i-th layer are:

[0052]

[0053] Where X is the ratio of the thickness of the brace to the thickness of the partition, M is the ratio of the difference between the inner radius r of the high-voltage winding and the outer radius r0 of the low-voltage winding to the outer radius r0 of the low-voltage winding, Y(i) is the ratio of the total length of all braces in the i-th layer to the perimeter of the i-1-th layer of partition, n is the number of layers of the transformer main insulation structure, ε b (i) is the relative dielectric constant of the i-th layer of oil-impregnated paperboard, unit: Farad / m, ε oil (i) is the relative dielectric constant of the i-th layer of oil, unit: Farad / m,

[0054] The equivalent parallel capacitance of each layer is connected in parallel to obtain the equivalent parallel capacitance C of the entire oil-immersed transformer. all :

[0055]

[0056] Where X is the ratio of the thickness of the strut to the thickness of the partition, Y(i) is the ratio of the total length of all struts in the i-th layer to the perimeter of the i-1-th layer of partition, n is the number of layers of the transformer main insulation structure, ε b (i) is the relative dielectric constant of the i-th layer of oil-impregnated paperboard, unit: Farad / m, ε oil (i) is the relative dielectric constant of the i-th layer of oil, unit: Farad / m, h is the height of the transformer main insulation structure, unit: meter, b(i) and d(i) are two parameters of the i-th layer,

[0057] S4. Substitute the actual transformer's structural dimensions and number of radial insulation layers into the database of S2 and use the permutations function to generate all possible combinations, each of which corresponds to the complex relative permittivity of one type of oil and the complex relative permittivity of each insulation layer; substitute all the above combinations into the capacitance expression of the composite multi-oil channel columnar main insulation structure of S3 to obtain a basic database of the equivalent capacitance of the entire transformer; each value in the basic database corresponds to a group of capacitances; the group of capacitances corresponds to the complex relative permittivity of one type of oil and the complex relative permittivity of each insulation layer;

[0058] S5. measuring the actual transformer loss factor at different frequencies to obtain a database of complex capacitance test values ​​of the actual transformer;

[0059] S6. Set the height of the starting point below the actual transformer axial degradation interval as a percentage of the overall height to T, and the percentage of the degradation interval to the overall height to S. Based on the structural dimensions of the transformer and the number of radial insulation layers, use a genetic algorithm to perform a traversal optimization on the complex capacitance test value of the actual transformer in S5 and the basic database of the transformer equivalent capacitance in S4. Perform inverse iteration to obtain the average complex relative dielectric constant and average degree of polymerization of each insulation layer, and determine the values ​​of T and S based on the gradient change of the degree of polymerization.

[0060] S7. Based on the data of S2 and the values ​​of T and S determined in S6, a database of complex relative dielectric constants considering axial non-uniform aging is established, thereby establishing a capacitance expression considering axial non-uniform aging;

[0061] S8. Input the specific geometric parameters and the number of insulation layers of the actual transformer, substitute the database of complex relative dielectric constants in S7 into the basic database of overall transformer equivalent capacitance in S4, and construct an extended database of overall transformer equivalent capacitance that takes into account axial non-uniform aging;

[0062] S9. Based on the complex capacitance test value of the actual transformer of S5, the extended database of the overall transformer equivalent capacitance of S8 is traversed and optimized, and the complex relative dielectric constant distribution matrix at different frequency points is obtained by iterative inversion calculation using the genetic algorithm. The matrix is ​​substituted into the database of S2 to obtain the degree of aggregation distribution matrix at different frequency points; the arithmetic mean of the elements at the same position of the degree of aggregation distribution matrix at different frequency points is calculated to form a degree of aggregation distribution matrix.

[0063] Furthermore, the databases of complex relative dielectric constants of S2 are all databases that have undergone frequency-temperature shift processing.

[0064] In order to further expand the data to the greatest extent and save storage space, each capacitor in the basic database of S4's transformer equivalent capacitance is normalized. The specific formula is as follows:

[0065]

[0066] Where C all is the equivalent parallel capacitance of the oil-immersed transformer under vacuum conditions, in Farads; C0 is the equivalent parallel capacitance of the oil-immersed transformer under vacuum conditions, in Farads; ε0 is the vacuum dielectric constant, ε0 = 8.854 × 10 -12 Farad / meter, h is the height of the transformer main insulation structure, unit: meter,

[0067] Further, the insulating layer described in S4 includes one or more partition-oil gap-support bar structural units.

[0068] The specific steps for determining the values ​​of T and S in S6 are as follows:

[0069] S601. Set a threshold interval for T. When the value of T is at the lower limit of the threshold interval, iterate between samples in the degradation interval S according to the first iteration step to obtain a data set for each sample. Each data set of the sample is compared with the complex relative dielectric constant of each insulating layer obtained by inversion, and the optimal set is selected to obtain the set with the smallest value function value to obtain the average degree of polymerization of the corresponding axial segment of each sample.

[0070] S602: Calculate the relative error between the degree of aggregation of each sample and the degree of aggregation of the previous sample, ensuring that it does not exceed 6% ± 0.8%. Continue iterative calculations until the maximum threshold of the degradation interval is reached. If the relative error exceeds 6% ± 0.8%, output the degradation interval value corresponding to the previous sample to determine the actual axial segmentation ratio at that T value.

[0071] S603, T is iterated according to the second iteration step within the threshold range set in S601, and an iterative cycle from S601 to S602 is performed for each T value; the value function value corresponding to each T value is compared, and the group with the smallest value function value is taken to finally determine the values ​​of T and S of the measured transformer.

[0072] In the specific implementation, multiple oil-paper insulation samples and oil samples were thermally aged under laboratory conditions to obtain samples with varying degrees of aging. The test temperatures selected were -30°C, -10°C, 10°C, 30°C, 60°C, and 90°C. Oil samples with varying degrees of aging, ranging from 1002 to 342 degrees of polymerization, were selected with intervals of 5, and the same number of days of aging were used. The transformer tested was a 220kV oil-immersed transformer with nine main insulation units consisting of a baffle-oil gap-strut structure. Transformer geometric parameters included a main insulation height of 400mm, an outer radius of the main insulation low-voltage winding of 350mm, an outer radius of the main insulation high-voltage winding of 422mm, and a strut thickness of 8mm. To simplify the calculation process and increase computational speed, this example assumes that every three baffle-oil gap-strut units constitute a layer, with a total of three radial layers. This results in a 3x3 distribution matrix for evaluation.

[0073] The dielectric loss factor curve of a 220 kV oil-immersed transformer was measured. The actual temperature was 26 degrees Celsius. The measured curve, the curve calculated by this patent and the curve calculated by the traditional evaluation method are as follows: Figure 3 The frequency range of the curve is: 1000, 470, 220, 110, 70, 40, 20, 10, 4.6, 2.2, 1, 0.46, 0.22, 0.1, 0.046, 0.022, 0.01, 0.0046, 0.0022, 0.001.

[0074] Input the number of insulation layers and geometric parameters of the transformer. Substitute the oil-paper insulation samples and oil samples of different aging degrees into the permutations function to generate 20 lines of 17689 30 Column combinations are introduced into the S3 composite multi-oil channel column main insulation structure equivalent capacitance parallel model to obtain a data set of equivalent capacitance of the entire transformer. Each capacitance in the data set is normalized to construct a large database of the transformer's overall capacitance. The measured transformer curve is read in, and all capacitance values ​​of the curve are compared with the database. The global optimal solution is continuously selected through the genetic algorithm to invert a set of numbers with the minimum value function value. The value function expression is:

[0075] F=min{(y 实际 -y 计算 )}

[0076] Where F is the cost function, which represents the minimum error between the calculated curve and the measured curve after N traversals. The smaller the cost function value, the closer it is to the measured curve. The closer the calculated degree of polymerization and moisture content are to the measured ones, the better the data match. 实际 Indicates the measured value at a certain frequency point. 计算 Indicates the calculated value at a certain frequency point.

[0077] Set the percentage of the height of the starting point below the actual transformer axial degradation interval to T, the threshold range of T to 80% to 90%, the second iteration step to 1%, the degradation interval S corresponding to each T ranges from 1% to 4%, the first iteration step to 0.5%, and iterate seven times; bring the data of S2 into the axial segmented model expression, and repeat the above iterations; the data set obtained from each sample is compared with the complex dielectric constant of each layer obtained by inversion, and the optimal set with the smallest value function is selected. The relative error of the degree of aggregation calculated for all samples does not exceed 6% ± 0.8%, and the dynamic degradation percentage of the transformer is determined to be T = 88%, and the actual degradation range is S = 4%;

[0078] The complex relative permittivity expression of the axial segmented model is:

[0079] ε b =8%ε r (a)+4ε r (b) +88%ε r (c)

[0080] Substitute the data from S2 into the above expression, and then into the capacitance expression of the composite multi-oil channel column main insulation structure in S3. The obtained capacitance is normalized to construct a 3*3 capacitance database. The measured curve is then compared with this 3*3 capacitance database. The global optimal solution is continuously selected through the genetic algorithm to invert a set of numbers with the minimum value function value, which are the complex relative dielectric constant and polymerization degree values ​​of each of the three layers and three segments at each frequency point.

[0081] To simplify the calculation, we take 5 frequency points as an example:

[0082] The calculation results of the distribution matrix of the complex relative dielectric constant at different frequency points are as follows:

[0083] 220Hz:

[0084]

[0085] 110Hz:

[0086]

[0087] 40Hz:

[0088]

[0089] 0.22Hz:

[0090]

[0091] 0.01Hz:

[0092]

[0093] The aggregation degree distribution matrix of different frequency points obtained based on the S2 data is as follows:

[0094] 220Hz:

[0095]

[0096] 110Hz:

[0097]

[0098] 40Hz:

[0099]

[0100] 0.22Hz:

[0101]

[0102] 0.01Hz:

[0103]

[0104] The arithmetic mean of the elements at the same position of the aggregation distribution matrix at different frequency points is calculated to form an aggregation distribution matrix as follows:

[0105]

[0106] The minimum degree of aggregation screened out is 763, which is the degree of aggregation value of the most seriously aged part of the transformer. It can be used to evaluate the actual aging insulation state of the transformer and the quantitative evaluation of the uneven aging state distribution of the transformer.

[0107] The traditional evaluation method that does not consider the specific inversion process of the axial segmentation evaluation model is to grid the transformer segmentation. The evaluation process is based on the iterative calculation of the XY model, and does not consider the impact of the axial segmentation ratio on uneven aging. It cannot locate the actual location where aging is most serious. The minimum aggregation degree obtained by the traditional evaluation method is 831. The comparison between the calculated curve and the measured curve is shown in the figure below. Figure 3 In comparison, the inversion positioning segmentation method proposed in the present invention is more accurate in evaluation.

[0108] Obviously, the embodiments described above are only some of the embodiments of the present invention, rather than all of them. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0109] The above mainly introduces the solution of the embodiment of the present application from the perspective of the execution process of the method side. It is understandable that in order to implement the above functions, the network device includes hardware structures and / or software modules corresponding to the execution of each function. Those skilled in the art should easily realize that, in combination with the units and algorithm steps of each example described in the embodiment provided herein,

[0110] This application can be implemented in hardware or a combination of hardware and computer software. Whether a function is implemented in hardware or by computer software driving hardware depends on the specific application and design constraints of the technical solution. Professionals and technicians may use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0111] If the above-mentioned integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable memory. Based on this understanding, the technical solution of the present application is essentially or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a memory and includes several instructions for enabling a computer device (which can be a personal computer, server or TRP, etc.) to execute all or part of the steps of the various embodiments of the present application. The aforementioned memory includes: U disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), mobile hard disk, magnetic disk or optical disk and other media that can store program code.

Claims

1. A method for evaluating the uneven aging distribution of an oil-immersed transformer, characterized in that: The steps include: S1. Perform frequency domain dielectric response tests on insulating cardboard samples with different polymerization degrees at different temperatures; perform frequency domain dielectric response tests on insulating oil samples with different aging days at different temperatures; S2. Constructing a database of complex relative dielectric constants of insulating paperboards with different degrees of polymerization; constructing a database of complex relative dielectric constants of insulating oils with different aging days; S3. Based on the main insulation structure of the transformer, the capacitance expression of the composite multi-oil channel column main insulation structure is obtained by using the relationship between capacitance and complex relative dielectric constant; S4. Substitute the actual transformer's structural dimensions and number of radial insulation layers into the database of S2 and use the permutations function to generate all possible combinations, each of which corresponds to the complex relative permittivity of one type of oil and the complex relative permittivity of each insulation layer; substitute all the above combinations into the capacitance expression of the composite multi-oil channel columnar main insulation structure of S3 to obtain a basic database of the equivalent capacitance of the entire transformer; each value in the basic database corresponds to a group of capacitances; the group of capacitances corresponds to the complex relative permittivity of one type of oil and the complex relative permittivity of each insulation layer; S5. measuring the actual transformer loss factor at different frequencies to obtain a database of complex capacitance test values ​​of the actual transformer; S6. Set the height of the starting point below the actual transformer axial degradation interval as a percentage of the overall height to T, and the percentage of the degradation interval to the overall height to S. Based on the structural dimensions of the transformer and the number of radial insulation layers, use a genetic algorithm to perform a traversal optimization on the complex capacitance test value of the actual transformer in S5 and the basic database of the transformer equivalent capacitance in S4. Perform inverse iteration to obtain the average complex relative dielectric constant and average degree of polymerization of each insulation layer, and determine the values ​​of T and S based on the gradient change of the degree of polymerization. S7. Based on the data of S2 and the values ​​of T and S determined in S6, a database of complex relative dielectric constants considering axial non-uniform aging is established, thereby establishing a capacitance expression considering axial non-uniform aging; S8. Input the specific geometric parameters and the number of insulation layers of the actual transformer, substitute the database of complex relative dielectric constants in S7 into the basic database of overall transformer equivalent capacitance in S4, and construct an extended database of overall transformer equivalent capacitance that takes into account axial non-uniform aging; S9. Based on the complex capacitance test value of the actual transformer of S5, the extended database of the overall transformer equivalent capacitance of S8 is traversed and optimized, and the complex relative dielectric constant distribution matrix at different frequency points is obtained by iterative inversion calculation using the genetic algorithm. The matrix is ​​substituted into the database of S2 to obtain the degree of aggregation distribution matrix at different frequency points; the arithmetic mean of the elements at the same position of the degree of aggregation distribution matrix at different frequency points is calculated to form a degree of aggregation distribution matrix.

2. The method for evaluating the uneven aging state distribution of an oil-immersed transformer according to claim 1, characterized in that: The complex relative dielectric constant databases of S2 are all databases that have undergone frequency-temperature shift processing.

3. The method for evaluating the uneven aging state distribution of an oil-immersed transformer according to claim 1, characterized in that: Each capacitor in the basic database of S4's transformer equivalent capacitance is normalized.

4. The method for evaluating the uneven aging state distribution of an oil-immersed transformer according to claim 1, characterized in that: The insulating layer described in S4 includes one or more partition-oil gap-support bar structural units.

5. The method for evaluating the uneven aging state distribution of an oil-immersed transformer according to claim 1, characterized in that: The specific steps for determining the values ​​of T and S in S6 are as follows: S601. Set a threshold interval for T. When the value of T is at the lower limit of the threshold interval, iterate between samples in the degradation interval S according to the first iteration step to obtain a data set for each sample. Each data set of the sample is compared with the complex relative dielectric constant of each insulating layer obtained by inversion, and the optimal set is selected to obtain the set with the smallest value function value to obtain the average degree of polymerization of the corresponding axial segment of each sample. S602: Calculate the relative error between the degree of aggregation of each sample and the degree of aggregation of the previous sample, ensuring that it does not exceed 6% ± 0.8%. Continue iterative calculations until the maximum threshold of the degradation interval is reached. If the relative error exceeds 6% ± 0.8%, output the degradation interval value corresponding to the previous sample to determine the actual axial segmentation ratio at that T value. In S603, T is iterated according to the second iteration step within the threshold range set in S601, and an iterative cycle from S601 to S602 is performed for each T value; the value function value corresponding to each T value is compared, and the group with the smallest value function value is taken to finally determine the values ​​of T and S of the actual transformer.

Images