A transformer fault status assessment method, system, medium and device

Abstract

The present invention provides a transformer fault state assessment method. First, based on the correlation characteristics of transformer faults, a transformer fault evolution network diagram is established. Secondly, considering the insufficient monitoring data caused by the complexity of the fault mechanism, simulation software is used to perform coupled simulation analysis on multivariate physical fields; at the same time, combined with fuzzy number theory, a transformer fault evolution probability calculation model is established. Finally, risk entropy is introduced to characterize the uncertainty of fault evolution, and the maximum probability expression of the fault evolution path is given and converted into a linear programming problem, so as to obtain the maximum probability evolution path of the transformer fault under different initial risk factors. This design fully considers the inherent connection of transformer faults and can evaluate the transformer fault state as a whole.

Description

Technical Field

[0001] The present invention relates to the technical field of fault state assessment, and in particular to a transformer fault state assessment method, system, medium and equipment. Background Art

[0002] Transformers are an important part of the power system. With the expansion of the scale and increase of complexity of my country's power system, the safety and stability of system operation are facing challenges. Therefore, it is of great significance to select a reasonable transformer fault model for fault state analysis to carry out fault prevention and life management of transformers.

[0003] In existing research, various fault states of transformers are usually studied as independent evaluation objects, without considering the evolution and connection between the various fault states of the transformer. Often, major faults evolve from minor faults, and the fault evolution risk is not considered in existing research. Summary of the invention

[0004] In order to make the transformer status assessment more in line with the actual situation of power grid operation, it is necessary to further improve the problems mentioned in the background technology. The purpose of the present invention is to provide a transformer fault status assessment method based on multi-physical field coupling that is more in line with the operation mechanism of the power system and has practical significance in view of the above problems existing in the existing transformer status assessment.

[0005] In order to achieve the above object, in a first aspect, the present invention provides a transformer fault state assessment method, comprising the following steps:

[0006] S1. Based on the correlation characteristics of transformer faults and the complex system theory, a transformer fault evolution network diagram is established;

[0007] S2. Obtaining the corresponding failure probability by establishing a failure evolution probability model and / or by using fuzzy number theory;

[0008] S3. By combining the corresponding fault probability with the fault evolution network diagram, the maximum probability fault evolution path of the transformer under different initial risk factors is obtained, and the fault state of the transformer is evaluated.

[0009] According to the transformer fault state assessment method provided by the present invention, through the complex system theory, the parameters characterizing the correlation characteristics of the transformer fault include one or more of the characteristic path length and the clustering coefficient;

[0010] The formula for the characteristic path length is:

[0011] Where n is the total number of network nodes, d ijis the shortest distance between two nodes i and j in the network, and L is the characteristic path length;

[0012] The formula for the clustering coefficient is:

[0013] Where n is the total number of network nodes, C i is the local clustering coefficient of node i, k i is the number of nodes adjacent to node i, e i For this k i The actual number of edges between nodes.

[0014] According to the transformer fault state assessment method provided by the present invention, the multivariate physical field under the fault condition is analyzed by multi-physical field simulation software to establish the fault evolution probability model.

[0015] According to the transformer fault state assessment method provided by the present invention, the corresponding fault probability obtained by establishing a fault evolution probability model includes one or more of overheating fault probability, short circuit fault probability and external cause fault probability.

[0016] According to the transformer fault state assessment method provided by the present invention, the formula for the overheating fault probability is:

[0017] Among them, β t is the shape parameter, which can be obtained by data fitting, T t is the expected life time of transformer insulation, T e is the equivalent operating time under the benchmark hot spot temperature, P T0 is the overheating failure probability;

[0018] The formula for the short circuit failure probability is:

[0019] Among them, the current magnitude I is the converted value of the simulation data, f(I) is the intermediate value of the first short-circuit fault probability, F(I) is the intermediate value of the second short-circuit fault probability, P T1 is the short circuit failure probability;

[0020] The formula for the probability of external failure is: P T2 =m1exp(n1Y)+b

[0021] Among them, m1, n1, and b are dimensionless coefficients fitted by historical fault data, and Y is the transformer failure probability based on historical fault data.

[0022] According to the transformer fault state assessment method provided by the present invention, the corresponding fault probability obtained by fuzzy number theory is the other fault probability that cannot be obtained by the fault evolution probability model;

[0023] The formula for the other failure probability is:

[0024] It first calculates the objective fuzzy number and then converts the objective fuzzy number into the quantified other fault probability through defuzzification, μ i (x) is the membership function, x is the output variable, P T3 is the other failure probability.

[0025] According to the transformer fault state assessment method provided by the present invention, the maximum probability fault evolution path is obtained by calculating the minimum risk entropy. The specific calculation formula is:

[0026]

[0027]

[0028] Where V is the node set corresponding to any path in the fault evolution scenario, p i is the failure probability of the transformer under the initial risk factor numbered i; the formula is calculated using the Dijkstra algorithm.

[0029] In order to achieve the above objective, in a second aspect, the present invention provides a transformer fault state assessment system, which can implement the steps of any of the above fault state assessment methods.

[0030] In order to achieve the above-mentioned purpose, in a third aspect, the present invention provides a non-transitory computer-readable storage medium having a computer program stored thereon, and when the computer program is executed by a processor, the steps of any of the above-mentioned fault state assessment methods can be implemented.

[0031] In order to achieve the above-mentioned purpose, in a fourth aspect, the present invention provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor can implement the steps of any of the above-mentioned fault state assessment methods when executing the computer program.

[0032] The present invention provides a transformer fault state assessment method. First, based on the correlation characteristics of transformer faults, a transformer fault evolution network diagram is established to evaluate the transformer state from the fault mechanism, which is more in line with the actual situation. Secondly, considering the insufficient monitoring data caused by the complexity of the fault mechanism, multi-physics field simulation software is used to perform coupled simulation analysis on multivariate physical fields. At the same time, combined with fuzzy number theory, a transformer fault evolution probability calculation model is established, which greatly improves the reliability of the analysis. Finally, risk entropy is introduced to characterize the uncertainty of fault evolution, the maximum probability expression of the fault evolution path is given and converted into a linear programming problem, and the Dijkstra algorithm is used to obtain the shortest path of fault evolution, and the maximum probability evolution path of transformer faults under different initial risk factors is obtained, which realizes the quantitative evaluation of the transformer fault evolution process, and can provide an important reference for the quantitative evaluation of accident risks in modern complex process systems. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] In order to more clearly illustrate the technical solution of the present invention, the following is a brief introduction to the drawings required for use in the description of the embodiments. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0034] Figure 1 is a flow chart of a transformer fault state assessment method in an embodiment of the present invention;

[0035] Figure 2 is a transformer fault evolution network diagram in an embodiment of the present invention;

[0036] Figure 3 is a graph showing the relationship between the risk entropy of the transformer fault evolution path and the fault probability in an embodiment of the present invention. DETAILED DESCRIPTION

[0037] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be clearly and completely described below in conjunction with the accompanying drawings and tables. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0038] like Figure 1 As shown, in one embodiment, a transformer fault state assessment method of the present invention includes the following steps S1-S3:

[0039] S1. Based on the correlation characteristics of transformer faults and the complex system theory, a transformer fault evolution network diagram is established.

[0040] According to the complex system theory, the parameters characterizing the correlation characteristics of the transformer fault include one or more of characteristic path length and clustering coefficient;

[0041] The characteristic path length mainly represents the degree of connection between nodes. The characteristic path length can be defined as:

[0042] Where n is the total number of network nodes, d ij is the shortest distance between two nodes i and j in the network, and L is the characteristic path length;

[0043] The clustering coefficient is an important parameter used to measure the degree of clustering of nodes in a network. The clustering coefficient can be defined as:

[0044] Where n is the total number of network nodes, C i is the local clustering coefficient of node i, k i is the number of nodes adjacent to node i, e i For this k i The actual number of edges between nodes.

[0045] S2. Obtain the corresponding failure probability by establishing a failure evolution probability model and / or by using fuzzy number theory.

[0046] Based on COMSOL multi-physics field simulation software, the multivariate physical field under fault conditions is analyzed and a fault evolution probability model is established. Other fault probabilities that cannot be obtained through the fault evolution probability model can be obtained through fuzzy number theory.

[0047] The corresponding fault probabilities obtained by establishing the fault evolution probability model include one or more of overheating fault probability, short circuit fault probability and external cause fault probability.

[0048] (1) The steps for deducing the overheating failure probability are as follows:

[0049] The expected life of transformer insulation and hot spot temperature conform to the Arrhenius law, and the calculation formula is as follows:

[0050]

[0051] In the formula, C t With B t are all empirical constants, θ H is the hot spot temperature (in °C), which can be obtained through COMSOL finite element simulation.

[0052] According to the operation conditions at different temperatures, the operation time of the transformer at different temperatures can be converted to the equivalent operation time T under the benchmark hot spot temperature operation.e :

[0053]

[0054] Where G is the total number of operating life intervals; θ Hi is the actual operating temperature of the transformer during its service life; t i is the operating time of the transformer at temperature (in h); θ0 is the reference hot spot temperature.

[0055] Based on the Weibull distribution, the insulation failure probability model of the transformer at high temperature is established:

[0056]

[0057] In the formula, β t is the shape parameter, which can be obtained by data fitting, T t is the expected life time of transformer insulation, T e is the equivalent operating time under the benchmark hot spot temperature, P T0 is the overheat failure probability.

[0058] (2) The steps for deducing the short-circuit fault probability are as follows:

[0059] The three-parameter Weibull distribution is used to fit the transformer short-circuit fault probability:

[0060]

[0061]

[0062] Where m, θ, and I0 are all Weibull distribution parameters, which are obtained by data fitting.

[0063] According to the definition of failure probability, the short circuit failure probability P T1 for:

[0064]

[0065] Where, the current magnitude I is the converted value of the COMSOL finite element simulation data, f(I) is the intermediate value of the first short-circuit fault probability, F(I) is the intermediate value of the second short-circuit fault probability, and P T1 is the short circuit failure probability.

[0066] (3) The steps for deducing the probability of external failure are as follows:

[0067] Transformer fault probability modeling based on historical fault information:

[0068]

[0069] Among them, Di is the number of failures that occurred in the i-th time period in the historical failure data; D total The total number of failures that occurred on the same day in history.

[0070] The model of the failure probability caused by external factors of transformer is constructed based on the Poisson regression model:

[0071] P T2 =m1exp(n1Y)+b

[0072] Where m1, n1, and b are dimensionless coefficients fitted by historical fault data, and Y is the transformer failure probability based on historical fault data.

[0073] (4) The deduction steps for other failure probabilities are as follows:

[0074] Assume that the judgment value E of any two experts i and j on the weight of a certain fault evolution path is i and E j The similarity is S ij (E i ,E j ), the corresponding trapezoidal fuzzy numbers are A = (a1, a2, a3, a4), B = (b1, b2, b3, b4), then the similarity between the two is:

[0075]

[0076] in

[0077]

[0078] y B * Calculation and y A * Similar, S ij The larger the value of (A,B), the greater the similarity between the two fuzzy numbers.

[0079] On this basis, the expert opinion E i The average similarity AA(E i ) and relative similarity RA(E i ) are:

[0080]

[0081]

[0082] Expert OpinionE i The consistency coefficient is:

[0083] CC(E i )=βW(E i)+(1-β)RA(E i )

[0084] In the formula, W(E i ) is the expert opinion E i The influence weight of W(E i ), all expert opinions can be expressed as follows after calculation:

[0085] R AG =CC(E1)×E1+CC(E2)×E2+…+CC(E M )×E M

[0086] To convert the expert's subjective evaluation level into the corresponding objective probability level, first obtain the objective fuzzy number FPr:

[0087]

[0088] After obtaining the objective fuzzy number, the objective fuzzy inference value is converted into a quantitative probability value P by defuzzification. T3 :

[0089]

[0090] In the formula, μ i (x) is the membership function, x is the output variable, P T3 is the other failure probability.

[0091] S3. By combining the corresponding fault probability with the fault evolution network diagram, the maximum probability fault evolution path of the transformer under different initial risk factors is obtained, and the fault state of the transformer is evaluated.

[0092] The maximum probability failure evolution path introduces risk entropy quantification, which is characterized as follows:

[0093] I i =-lnp i

[0094] In the formula, p i is the failure probability of the transformer under the initial risk factor numbered i.

[0095] Based on the risk entropy value, assume that the node set V corresponding to any path in the fault evolution scenario is {v1,v2,…,v k}, the risk entropy I(V) of the fault evolution path can be expressed as:

[0096]

[0097] The search for the maximum probability failure evolution path is transformed into a linear programming problem for seeking the minimum risk entropy:

[0098]

[0099]

[0100] The Dijkstra weighted directed graph shortest path algorithm (Dijkstra algorithm) is used for calculation. The algorithm traverses and expands outward from the initial node to the final node, and is suitable for shortest path calculation without negative weight loops.

[0101] In one embodiment, the three steps of the transformer fault state assessment method of the present invention are further described in detail:

[0102] Step S1: According to the "Guidelines for Power Transmission and Transformation Status Maintenance" issued by the State Grid Corporation of China, considering the possible fault forms of transformers in actual operation, oil-immersed transformers are selected for analysis. Based on the correlation characteristics of transformer faults, a transformer fault evolution network diagram is established based on complex network theory, such as Figure 2 shown. Figure 2 The risk factors are shown in Table 1.

[0103] Table 1 Transformer failure risk factors

[0104] Table 1 Risk factors for transformer failure

[0105]

[0106]

[0107] Step S2: Based on the historical transformer failure cases of a provincial company of the State Grid in the past ten years and the most representative data of a certain type of 110KV main transformer, the transformer model is modeled by COMSOL software. The transformer related data are shown in Table 2 and Table 3. After the fault probability evolution model is established, the risk entropy of the transformer fault evolution path is calculated based on the simulation data and fuzzy number theory, as shown in Table 2 and Table 3. Figure 3 shown.

[0108] Table 2 Transformer related structural data

[0109] Table 2 Transformer size data

[0110]

[0111] Table 3 Transformer material data

[0112] Table 3 Transformer material data

[0113]

[0114] Step S3: Dijkstra method is used to solve the shortest fault evolution path characterized by risk entropy, and the shortest fault path evolved by various initial risk factors is shown in Table 4. In actual maintenance, the example model can provide the highest probability fault development direction of the transformer, and through targeted prevention and maintenance of key links of fault evolution, the occurrence of serious transformer faults can be effectively prevented.

[0115] Table 4 Shortest path of failure evolution for each initial risk factor

[0116] Table 4 Shortest path of failure evolution

[0117]

[0118] In a second aspect, an embodiment of the present invention provides a transformer fault state assessment system, which can implement the steps of any of the above-mentioned fault state assessment methods.

[0119] In a third aspect, an embodiment of the present invention provides a non-transitory computer-readable storage medium having a computer program stored thereon, and when the computer program is executed by a processor, the steps of any of the above-mentioned fault state assessment methods can be implemented.

[0120] In a fourth aspect, an embodiment of the present invention provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor can implement any step of the above-mentioned fault state assessment method when executing the computer program.

[0121] It will be easily understood by those skilled in the art that the above are only some preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A transformer fault status assessment method, characterized in that: The steps include: S1. Based on the correlation characteristics of transformer faults and the complex system theory, a transformer fault evolution network diagram is established; S2. Obtaining the corresponding failure probability by establishing a failure evolution probability model and / or by using fuzzy number theory; S3. The maximum probability fault evolution path of the transformer under different initial risk factors is obtained by combining the corresponding fault probability with the fault evolution network diagram, and the fault state of the transformer is evaluated; the fault evolution probability model is established by analyzing the multivariate physical field under the fault condition through multi-physics field simulation software; wherein, the corresponding fault probability obtained by establishing the fault evolution probability model includes one or more of the overheating fault probability, the short circuit fault probability and the external cause fault probability; the formula of the overheating fault probability is: in, β t is the shape parameter, which can be obtained by data fitting. is the expected life time of transformer insulation, is the equivalent operating time under the benchmark hot spot temperature, is the overheating failure probability; The formula for the short circuit failure probability is: Among them, the current size I is the converted value of the simulation data, f ( I ) is the intermediate value of the first short-circuit fault probability, F ( I ) is the second short-circuit fault probability median value, is the short circuit failure probability; The formula for the probability of external failure is: in, m 1. n 1. b are dimensionless coefficients fitted by historical fault data. Y is the transformer failure probability based on historical failure data; The corresponding failure probability obtained through fuzzy number theory is the other failure probability that cannot be obtained through the failure evolution probability model; The formula for the other failure probability is: It first calculates the objective fuzzy number and then converts the objective fuzzy number into the quantified other fault probability through defuzzification. μ i ( x ) is the membership function, x is the output variable, is the other failure probability; The maximum probability fault evolution path is obtained by calculating the minimum risk entropy. The specific calculation formula is: in, V is the node set corresponding to any path in the fault evolution scenario, p i For the number i The failure probability of the transformer under the initial risk factor is calculated by using the Dijkstra algorithm to calculate the minimum risk entropy to obtain the maximum probability fault evolution path.

2. The transformer fault status assessment method according to claim 1, characterized in that: According to the complex system theory, the parameters characterizing the correlation characteristics of the transformer fault include one or more of characteristic path length and clustering coefficient; The formula for the characteristic path length is: in, n is the total number of network nodes, d ij For two nodes in the network i and j The shortest distance between L is the characteristic path length; The formula for the clustering coefficient is: in, n is the total number of network nodes, C i For Node i The local clustering coefficient of k i For the node i The number of adjacent nodes, e i For this k i The actual number of edges between nodes.

3. A transformer fault status assessment system, characterized in that: The method can implement the steps of the fault status assessment method described in any one of claims 1 to 2.

4. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the fault status assessment method according to any one of claims 1 to 2 can be implemented.

5. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the computer program, the steps of the fault status assessment method described in any one of claims 1 to 2 can be implemented.

Images