Transformer multi-time scale state evaluation method based on improved D-S evidence theory

Abstract

The invention discloses a transformer multi-time-scale state evaluation method based on an improved D-S evidence theory, and the method comprises the steps: carrying out the daily evaluation of a transformer based on the improved DS evidence theory according to a membership vector matrix of real-time monitoring data on that day and a transformer operation environment; transformer monthly evaluation is carried out based on the improved DS evidence theory according to the transformer daily evaluation result and the preventive test result; and performing transformer annual evaluation based on the improved DS evidence theory according to the transformer monthly evaluation result. Therefore, the transformer day-month-year comprehensive evaluation provided by the invention can reflect the comprehensive state grade of the equipment in a short period, a middle period and a long period, timely reflect the health change of the equipment, discover the fault information of the equipment and provide a certain theoretical basis for operation and maintenance personnel to formulate a maintenance strategy, thereby promoting the development from planned maintenance to state maintenance.

Description

Technical Field

[0001] The present invention relates to a transformer multi-time scale state assessment method, device, equipment and medium based on improved DS evidence theory, and relates to the field of electric power technology. Background Art

[0002] The safe and stable operation of power equipment is an important factor in ensuring the stability of the power system. When power equipment fails, it is very easy to cause power outages and other accidents, which will not only cause great economic losses, but also cause safety problems. As one of the most important key equipment in the power system, the operating status of the power transformer is greatly related to the stability of the power system, and its importance is self-evident. However, the power transformer is in load operation for a long time and is in an exposed environment, and it is difficult to avoid failure. However, the failure of the power transformer will not only affect the stability of the power system, but also may cause the collapse of the power system and even cause casualties. Therefore, in order to reduce the various losses caused by power equipment failure, the operating status of the power transformer should be understood in real time, the equipment failure information should be discovered in time, a reasonable maintenance plan should be formulated, the service life of the transformer should be paid attention to, and the maintenance and management during operation should be strengthened to improve the reliability of the equipment operation, which will play an important and positive role in maintaining the stability of the power system. Transformer status assessment can timely understand the equipment operating status based on the transformer operation data, can timely discover equipment failure information, and is the basis for formulating maintenance plans. Therefore, it is very important to explore reasonable, scientific and effective status assessment methods.

[0003] At present, conventional condition assessment methods can only perform short-term assessments, that is, condition assessments are performed based on transformer operating data within a few days. This method can only reflect the recent status of the transformer, but cannot reflect the comprehensive trend of transformer status changes. In addition, the assessment results may be affected by many factors, such as data quality, algorithm selection, etc., and there is a certain degree of uncertainty.

[0004] Therefore, there is an urgent need for an assessment method that can reflect the transformer status from multiple time scales to improve the reliability of transformer assessment. Summary of the invention

[0005] The present invention aims to solve at least one of the technical problems existing in the prior art. To this end, in view of the above problems, the purpose of the present invention is to provide a transformer multi-time scale state assessment method, device, equipment and medium based on improved DS evidence theory, so as to solve the problem that the prior art is insufficient in research on transformer state assessment.

[0006] In order to achieve the above-mentioned invention object, the technical solution adopted by the present invention is:

[0007] In a first aspect, the present invention provides a transformer multi-time scale state assessment method based on an improved DS evidence theory, comprising:

[0008] According to the membership vector matrix of the real-time monitoring data of the day and the transformer operating environment, the transformer daily evaluation is carried out based on the improved DS evidence theory;

[0009] According to the daily transformer evaluation results and preventive test results, monthly transformer evaluation is carried out based on the improved DS evidence theory;

[0010] According to the monthly evaluation results of transformers, annual evaluation of transformers is carried out based on the improved DS evidence theory.

[0011] In one possible implementation, according to the membership vector matrix of the real-time monitoring data of the day and the transformer operating environment, the transformer daily evaluation is performed based on the improved DS evidence theory, including:

[0012] The membership vectors of the monitoring data obtained by the real-time evaluation of the day are sorted to obtain the membership vector matrix B;

[0013] Normalize the membership vector matrix B according to the time sequence of real-time evaluation to obtain vector B0 as the evidence source of evidence theory;

[0014] The improved DS evidence theory is used to fuse the daily real-time evaluation results in chronological order to obtain the transformer daily evaluation probability vector C, where: C1 is the monitoring data state discrimination matrix; a is the expected score matrix of each state, a=[95,75,55,35,15], i=1,2,3,4,5 represents the state is excellent, good, caution, abnormal, severe;

[0015] Based on the transformer daily evaluation probability vector C, the fused daily evaluation score score_day is obtained:

[0016] Assign environmental factor score factor S according to the actual operating environment on that day 11 And the fusion daily evaluation score score_day, get the final daily evaluation score of the transformer: Score_day=score_day×S 11 ;

[0017] The daily comprehensive status assessment grade is obtained based on the final daily assessment score of the transformer and the assessment grade division interval.

[0018] In a possible implementation, the calculation process of the membership vector matrix B is:

[0019] Firstly, quantitative data preprocessing is performed on each monitoring data to serve as cloud model input value;

[0020] Secondly, the transformer health index is calculated according to the equipment operating years and the number of overhauls to correct the digital eigenvalues ​​of the cloud model;

[0021] Then, the five evaluation results are converted into digital features of the cloud model using the health index;

[0022] Finally, the indicator membership discrimination method based on the combination of trapezoidal cloud and normal cloud is adopted to obtain the membership vector matrix B corresponding to the five states of each indicator of the monitoring data. Among them, the membership vector matrix B divides the real-time evaluation indicators into five qualitative concepts: excellent, good, attention, abnormal and serious.

[0023] In a possible implementation, the calculation formula for obtaining the fused daily evaluation score based on the transformer daily evaluation probability vector C is:

[0024]

[0025] Where Scorej is the monitoring data score.

[0026] In a possible implementation, a monthly transformer assessment is performed based on the transformer daily assessment results and preventive test results, including:

[0027] The probability vector C obtained from the daily comprehensive evaluation results within the month is normalized to obtain a probability vector as the basic evidence source D1 of the monthly evaluation model;

[0028] Data from preventive trials with set times were selected as supplementary evidence based on the evaluation period;

[0029] The normalized data of the supplementary evidence is used as the input value of the cloud model membership discrimination algorithm. The membership vector matrix B2 of each indicator is obtained through the normal cloud + trapezoidal cloud model membership discrimination, and the membership vector matrix B2 is weighted and summed in an equal weighted manner to obtain the supplementary evidence source D2;

[0030] The basic evidence source D1 is fused in chronological order based on the improved DS evidence theory fusion rule to obtain the basic support vector D3;

[0031] The basic support vector D3 and the supplementary evidence source D2 are fused based on the improved DS evidence theory to obtain the monthly evaluation probability vector D;

[0032] Based on the monthly evaluation probability vector D, the fused monthly evaluation score score_month is obtained;

[0033] The monthly comprehensive status assessment grade is obtained based on the final score of the transformer and the assessment grade division interval.

[0034] In a possible implementation manner, the preventive test index test values ​​include insulation resistance absorption ratio, polarization coefficient, volume resistivity, winding DC resistance mutual difference and / or winding insulation dielectric loss.

[0035] In a possible implementation, according to the monthly transformer evaluation results, an annual transformer evaluation is performed based on the improved DS evidence theory, including:

[0036] Normalize the monthly evaluation probability vectors D of the 12 months in the year respectively to obtain the basic evidence source vector E1;

[0037] The annual evaluation probability vector E is obtained by fusing in chronological order based on the improved DS evidence theory;

[0038] Based on the annual evaluation probability vector E, the fused annual evaluation score score_year is obtained;

[0039] The comprehensive status assessment grade within the year is obtained based on the transformer annual assessment score and the assessment grade division interval.

[0040] In a second aspect, the present invention further provides a transformer multi-time scale state assessment device based on an improved DS evidence theory, comprising:

[0041] The transformer daily evaluation unit is configured to perform transformer daily evaluation based on the improved DS evidence theory according to the membership vector matrix of the real-time monitoring data of the day and the transformer operating environment;

[0042] The transformer monthly evaluation unit is configured to perform monthly transformer evaluation based on the improved DS evidence theory according to the transformer daily evaluation results and preventive test results;

[0043] The transformer annual evaluation unit is configured to perform transformer annual evaluation based on the improved DS evidence theory according to the transformer monthly evaluation result.

[0044] In a third aspect, the present invention also provides an electronic device, comprising: at least one processor; and a memory communicatively connected to the processor; wherein the memory stores instructions executable by the processor, and the instructions are executed by the processor so that the processor can execute any one of the methods described.

[0045] In a fourth aspect, the present invention further provides a computer-readable storage medium storing one or more programs, characterized in that the one or more programs include computer instructions, and the computer instructions are used to enable a computer to execute any of the methods described.

[0046] The present invention adopts the above technical solution, and has the following characteristics:

[0047] 1. The present invention relies on the membership vector matrix obtained by the transformer assessment cloud model as the fusion evidence source, and uses the improved evidence theory to fuse the membership vector matrix of the results within the day in chronological order to obtain the intra-day comprehensive assessment results; the monthly assessment and annual assessment respectively fuse the daily result probability vector within the month and the monthly assessment result probability vector within the year, and then combine preventive tests, equipment operating environment and other indicators to realize the transformer multi-time scale status assessment.

[0048] 2. The daily, monthly and annual comprehensive evaluation of the transformer proposed in the present invention can reflect the comprehensive status level of the equipment in the short, medium and long term, timely reflect the health changes of the equipment, discover equipment fault information, and provide a certain theoretical basis for operation and maintenance personnel to formulate maintenance strategies, thereby promoting the development of "planned maintenance" to "status" maintenance.

[0049] In summary, the present invention can be widely used in transformer status assessment. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] Various other advantages and benefits will become apparent to those of ordinary skill in the art by reading the detailed description of the preferred embodiments below. The accompanying drawings are only for the purpose of illustrating the preferred embodiments and are not to be considered as limiting the present invention. Throughout the accompanying drawings, the same reference numerals are used to represent the same components. In the accompanying drawings:

[0051] Figure 1 It is the daily, monthly and annual evaluation data source of the embodiment of the present invention;

[0052] Figure 2 A daily assessment flow chart of an embodiment of the present invention;

[0053] Figure 3 This is a monthly evaluation flow chart of an embodiment of the present invention;

[0054] Figure 4 4 is an evaluation flow chart of an embodiment of the present invention. DETAILED DESCRIPTION

[0055] It should be understood that the terms used herein are only for the purpose of describing specific example embodiments and are not intended to be limiting. Unless the context clearly indicates otherwise, the singular forms "one", "an" and "said" as used herein may also be meant to include plural forms. The terms "include", "comprise", "contain", and "have" are inclusive, and therefore specify the existence of stated features, steps, operations, elements and / or parts, but do not exclude the existence or addition of one or more other features, steps, operations, elements, parts, and / or combinations thereof. The method steps, processes, and operations described herein are not interpreted as necessarily requiring them to be performed in the specific order described or illustrated, unless the execution order is clearly indicated. It should also be understood that additional or alternative steps may be used.

[0056] Although the terms first, second, third, etc. can be used in the text to describe multiple elements, components, regions, layers and / or sections, these elements, components, regions, layers and / or sections should not be limited by these terms. These terms can only be used to distinguish an element, component, region, layer or section from another region, layer or section. Unless the context clearly indicates, terms such as "first", "second" and other numerical terms do not imply order or sequence when used in the text. Therefore, the first element, component, region, layer or section discussed below can be referred to as the second element, component, region, layer or section without departing from the teaching of the example embodiments.

[0057] For ease of description, spatially relative terms may be used herein to describe the relationship of one element or feature relative to another element or feature as shown in the figures, such as "inside", "outside", "inner side", "outer side", "below", "above", etc. Such spatially relative terms are intended to include different orientations of the device in use or operation in addition to the orientation depicted in the figures.

[0058] Since the current conventional status assessment method can only perform short-term assessment, that is, the status assessment is performed based on the transformer operation data within a few days, this method can only reflect the recent status of the transformer, and cannot reflect the comprehensive trend of the transformer status change, and the assessment result may be affected by multiple factors. The transformer multi-time scale status assessment method, device, equipment and medium based on the improved DS evidence theory provided by the present invention include: based on the membership vector matrix of the real-time monitoring data of the day and the transformer operating environment, the transformer daily assessment is performed based on the improved DS evidence theory; based on the transformer daily assessment results and preventive test results, the transformer monthly assessment is performed based on the improved DS evidence theory; based on the transformer monthly assessment results, the transformer annual assessment is performed based on the improved DS evidence theory. Therefore, the transformer daily, monthly and annual comprehensive assessment proposed by the present invention can reflect the comprehensive status level of the equipment in the short, medium and long term, timely reflect the health changes of the equipment, discover equipment fault information and provide a certain theoretical basis for the operation and maintenance personnel to formulate maintenance strategies.

[0059] The exemplary embodiments of the present invention will be described in more detail below with reference to the accompanying drawings. Although the exemplary embodiments of the present invention are shown in the accompanying drawings, it should be understood that the present invention can be implemented in various forms and should not be limited by the embodiments described herein. On the contrary, these embodiments are provided in order to enable a more thorough understanding of the present invention and to fully convey the scope of the present invention to those skilled in the art.

[0060] The daily, monthly and annual status assessment of the transformer refers to the comprehensive reflection of the daily, monthly and annual operating status of the transformer. The daily assessment is mainly based on the real-time assessment results within the day and combined with the equipment operating environment on that day; the monthly assessment is mainly combined with the daily preventive tests of the equipment and the daily assessment results of the equipment in the month; the annual assessment results are used to reflect the health of the equipment in the current year of operation. The present invention can timely discover potential problems through multi-time scale transformer status assessment, take preventive maintenance, and improve the reliability and life of the transformer.

[0061] Embodiment 1: The transformer multi-time scale state assessment method based on the improved DS evidence theory provided by the present invention comprises:

[0062] S1. Set a cloud model membership determination method based on health index.

[0063] In this embodiment, the cloud model membership determination method based on the health index includes:

[0064] S11. Calculation of transformer aging health index.

[0065] The health index of a transformer mainly reflects the health of the equipment. When the initial state remains unchanged during commissioning, the health index of the transformer is only related to its service life and design life. However, based on actual operating conditions, the calculation of the health index is too idealistic. In daily use of the transformer, the environment in which the equipment is located, historical maintenance records, etc. will cause changes in the health index, which will affect the actual life of the transformer. Therefore, this embodiment considers the equipment's operating life while also considering the number of equipment overhauls when calculating the transformer health index. The specific calculation formula is as follows:

[0066]

[0067] Where HI is the current transformer health index; ΔT is the transformer operating life; T des S is the design life of the transformer; 11 In the above formula, it is assumed that the initial health index of the transformer is 0.5, and when the design life is reached, the health index of the transformer is 6.5. The environmental factor score factor S 11 The determination of is shown in Table 1 below. For the environmental index quantification, it is impossible to directly obtain its degradation degree through a specific formula. The present invention establishes a qualitative index scoring model and converts the environmental factor scoring factor S 11 and overhaul score S 12 The quantification process is performed according to Tables 1 and 2 below.

[0068] Table 1 Transformer operating environment scores

[0069]

[0070] Table 2 Transformer overhaul frequency score

[0071]

[0072] S12. Cloud model membership determination.

[0073] In this embodiment, in the transformer fuzzy evaluation, fuzzy theory is usually used to determine the index membership function, so as to determine the index membership and realize fuzzy comprehensive evaluation. Common models include triangular membership function, trapezoidal membership function and triangular + trapezoidal membership function. However, in actual use, the boundary is too clear and cannot meet the characteristics of fuzziness and randomness of transformer evaluation indicators. According to the analysis, the cloud model can realize the mutual conversion between qualitative and quantitative, and the spacing handles the characteristics of fuzziness and randomness, so the cloud model (normal cloud + trapezoidal cloud) is used to realize the index membership discrimination. Index membership discrimination refers to the probability degree of a certain qualitative concept of the membership of a certain index value. For example, in the transformer comprehensive evaluation of this embodiment, the qualitative concept refers to the equipment evaluation level (excellent, good, attention, abnormal, serious). In this embodiment, the membership of the index is solved by using an X-condition cloud generator, which is specifically divided into a normal cloud membership algorithm and a trapezoidal cloud membership algorithm.

[0074] Furthermore, the implementation steps of the membership discrimination algorithm of the normal cloud generator are as follows:

[0075] a. Take the digital feature En in the cloud model as the expectation and He as the standard deviation to generate a normally distributed random number EN;

[0076] EN=NORM(En,He 2 );

[0077] b. Calculate the membership degree based on the expected value Ex and the input value x;

[0078]

[0079] c. Repeat steps a and b n times to calculate the arithmetic average and get the final membership degree.

[0080]

[0081] Furthermore, the trapezoidal cloud is disassembled in the middle position and simplified into a left half trapezoidal cloud and a right half trapezoidal cloud. The membership discrimination algorithm of the normal cloud + trapezoidal cloud includes:

[0082] The membership algorithm of the left half trapezoidal cloud model is as follows:

[0083] a. Determine the specific input value x and the expected value E x If x ≥ E x , then the final membership degree f Ai (x)' = 1;

[0084] b. If x < Ex, then according to the membership degree algorithm of the normal cloud model, the final membership degree f Ai (x)' is obtained according to the above formula.

[0085] The membership degree algorithm of the right semi-trapezoidal cloud model is as follows:

[0086] a. Judge the size relationship between the specific input value x and the expected value E x If x ≤ E x then the final membership degree f Ai (x)' = 1;

[0087] b. If x > E x , then according to the membership degree algorithm of the normal cloud model, the final membership degree f Ai (x)' is obtained according to the formula.

[0088] The real-time evaluation indexes of the present invention are divided into five levels of qualitative concepts: excellent, good, attention, abnormal, and serious. Among them, normal is the left semi-trapezoidal cloud, serious is the right semi-trapezoidal cloud, and good, attention, and abnormal are all normal clouds; each state level is a double-constrained space [x min , x max , as shown in Table 3.

[0089] According to the cloud theory and the above grade intervals, the cloud models of different state levels can be converted into digital characteristics (E x , E n , He), and the specific steps are as follows:

[0090] a. The expectation E x is the center of the sample interval and can best reflect the sample mean, so the middle value of each grade interval is used as the expectation;

[0091]

[0092] b. The entropy E n is a measure of the randomness and fuzziness of the sample, indicating the randomness and fuzziness of the sample values. The entropy is determined according to the 50% principle for both fuzziness and randomness:

[0093]

[0094] c. The size of the hyper-entropy He is adjusted according to experience and is a constant, so:

[0095] He = Ci (Ci is a constant).

[0096] Table 3

[0097]

[0098] S2, DS evidence theory.

[0099] In this embodiment, the DS evidence theory was originally proposed by Dempster in the late 1960s, but it was only a preliminary model at first. Later, in 1978, his student Shafer further expanded the theory on the preliminary model, and now it has developed into a widely used data reasoning model. DS evidence theory is an important method for dealing with uncertainty problems. First, it can use uncertain evidence sources for reasoning, thereby improving the accuracy of the results; secondly, the theory can effectively deal with the contradictions and conflicts between evidence sources and can adapt to various different reasoning scenarios. DS evidence theory mainly includes: evidence identification framework, basic probability distribution function and Dempster fusion rule. The core idea of ​​this theory is to synthesize different evidence sources through certain fusion rules to achieve the trust in the hypothesis. The evidence identification framework is essentially a hypothesis of the decision problem, and the basic probability distribution function represents the uncertainty of the hypothesis. The most important is the Dempster fusion rule, which is to obtain comprehensive support by fusing the support of different evidences, thereby obtaining a comprehensive and relatively accurate inference result, so that decision makers can make better decisions. Specifically:

[0100] S21. Basic concepts of DS evidence theory.

[0101] DS evidence theory mainly consists of three parts: evidence identification framework, basic probability distribution function and Dempster fusion rule.

[0102] 1) Evidence theory identification framework Θ:

[0103] The identification framework is a hypothesis about the structure of the problem, which forms a set of all possible outcomes. In this set of assumptions, each assumption is mutually exclusive. The subset consisting of all elements in the identification framework is called a power set, denoted by 2 Θ Therefore, the recognition framework containing m hypothesis elements can be expressed as:

[0104] Θ={h1,h2,…,h m}.

[0105] Then its power set can be expressed as:

[0106]

[0107] 2) Basic Probability Allocation Function (BPA):

[0108] BPA is a mapping function, which can be expressed as: 2Θ→[0,1]. Each subset A in the power set is assigned a quality function m(A) under the distribution function, which indicates the trust level of the evidence source in each subset. Its value represents the basic newcomer distribution value of the subset. At the same time, BPA also needs to meet the following two verification conditions:

[0109]

[0110] In the formula, the trust degree of the empty set in the power set is 0, and the sum of the trust values ​​of all subsets except the empty set is equal to 1. If the trust degree of subset A is greater than 0, then subset A is called the focal element of the evidence theory.

[0111] 3) Dempster fusion rule:

[0112] The Dempster fusion rule is the most important part of evidence theory. It is mainly based on the orthogonal operation of the probability distribution function between each evidence source and then sums them up. The new confidence level for each hypothesis can be obtained by calculating the Dempster fusion rule. The definition and expression are as follows.

[0113] Assume that there are two sets of evidence sources in the same evidence identification framework, with probability distribution functions m1 and m2 respectively, and focal elements A and i and B j , then the Dempster fusion rule can be expressed as:

[0114]

[0115] In the above formula, m(A) is the basic probability distribution function after combination, and k is the conflict coefficient between evidences. The larger its value, the greater the conflict. When the value is 1, the combination rule cannot be calculated and has no meaning.

[0116] S22. DS evidence theory with improved fusion rules.

[0117] The classic Dempster fusion rule has certain defects. When there is a great conflict between the same evidence or different evidences, the fusion result will be contrary to the factual basis, thus the evidence theory will be invalid and the scope of application will be limited. In transformer status assessment, it is often necessary to fuse different sources of evidence. The sources of evidence are independent of each other but may be very different, resulting in greater conflict. Obviously, it is difficult for the traditional Dempster fusion rule to accurately evaluate the results. Therefore, this embodiment adopts a synthesis formula, which can effectively solve the problem of too strong conflict between evidence sources. The specific steps are as follows:

[0118] a. Assume that there are n sources of evidence, and the corresponding basic probability distribution functions are m1, m2, …, m n ;

[0119]

[0120] Where K is the overall conflict coefficient of n evidence theories. The closer its value is to 1, the greater the conflict. ij is the conflict coefficient between evidence source i and evidence source j.

[0121] b. Definition is the credibility of the evidence, where is the comprehensive average of each pair of evidence sets in n pieces of evidence, reflecting the conflict between the two pieces of evidence. It is different from k in the classical evidence theory, which reflects the overall conflict degree. When k is large, But not necessarily big.

[0122] c. The new synthesis rules are as follows:

[0123]

[0124] m(X)=p(X)+K×ε×q(X)+K(1-ε)

[0125]

[0126] m(A) can also be written as follows:

[0127]

[0128] The first term in the above formula It is the DS synthesis formula. When K = 0, it is equivalent to the DS synthesis formula. When K → 1, the highly conflicting evidence is mainly determined by q(A) and ε. Among them, q(A) is the average support of evidence for A, and ε is the credibility of evidence. This synthesis formula effectively solves the problem of excessive conflict between different evidence sources, making the fusion result more reasonable.

[0129] S3. Transformer daily, monthly and annual assessment data sources.

[0130] In this embodiment, the daily, monthly, and annual evaluation of the transformer is a comprehensive reflection of the equipment evaluation in different periods. It is necessary to fully integrate the data in different periods. Therefore, it is necessary to make a selection based on the actual situation of each indicator during the evaluation. This embodiment uses the DS evidence theory to fuse the membership vector matrix of the real-time evaluation results to obtain the daily evaluation results, and the fusion result is also the probability distribution size of the five states. After normalization, it is used as the basic evidence source and then the evidence theory is used to obtain the monthly equipment evaluation results. The above process is also used for the monthly and annual evaluations. In the evaluation of different periods, it is necessary not only to consider the probability distribution vector of the results in the previous period, but also to correct the results according to the actual operating environment and preventive test data. Only in this way can the evaluation results be more reasonable. The data sources that need to be integrated for each period evaluation are as follows: Figure 1 As shown, the specific process is:

[0131] S31. Daily, monthly and annual assessment of transformers, including:

[0132] S311. Daily assessment of transformers.

[0133] In this embodiment, the real-time evaluation of the transformer has strong real-time and rapidity, and can reflect the real-time changes of the transformer. However, data collection is performed more than once in a day, and the real-time evaluation is also performed multiple times. Using only one evaluation result to characterize the status of the transformer on a certain day has uncertainty and large errors. Figure 2 As shown in the figure, in order to make up for this defect, the above-mentioned evidence theory algorithm is introduced. The transformer daily evaluation steps based on the improved DS evidence theory are as follows:

[0134] a. Arrange the membership vectors of the monitoring data obtained by the real-time evaluation of the day to obtain the membership vector matrix B;

[0135] In this embodiment, the calculation process of the membership vector matrix B is as follows: first, quantitative data preprocessing is performed on each monitoring data to serve as the cloud model input value; second, the transformer health index is calculated according to the equipment operating years and the number of overhauls to correct the cloud model digital eigenvalues; then, the health index is used to convert the five evaluation results into cloud model digital features (Ex, En, He); finally, the indicator membership discrimination method based on the combination of trapezoidal cloud and normal cloud is adopted to obtain the membership vector matrix B corresponding to the five states of each indicator of the monitoring data.

[0136] Furthermore, quantitative data refers to the value of each indicator reflected by a specific numerical value. The main indicators include monitoring data, preventive tests, etc. Since the units and dimensions of each indicator are different, and there are two types of data (the bigger the better and the smaller the better), the quantitative data should be preprocessed first, and the indicator state quantity should be normalized to obtain the indicator score. The formula for normalizing the indicator is as follows:

[0137]

[0138] In the formula, x n is the index degradation score, when x n >1, let x n =1; when x n <1, let x n =0; w n is the measured value of the characteristic quantity of the indicator. When the indicator is larger, the better, the value w′=1.3w a ; When the indicator is the smaller the better, pay attention to the value w′=w a / 1.3,w f is the factory value of this indicator, w aThe value of refers to the "Test Procedure for Condition Inspection and Maintenance of Power Transmission and Transformation Equipment" and the transformer equipment manufacturer.

[0139] b. Normalize the membership vector matrix B according to the order of real-time evaluation time to obtain vector B0 as the evidence source of evidence theory;

[0140] c. Adopt the improved DS evidence theory to fuse the daily real-time evaluation results in chronological order to obtain the transformer daily evaluation probability vector C. C1 is the monitoring data state discrimination matrix; a is the expected score matrix of each state, a=[95,75,55,35,15], i=1,2,3,4,5 represents the state is excellent, good, caution, abnormal, serious.

[0141] Furthermore, the specific process of obtaining the monitoring data state discrimination matrix is ​​as follows: first, the subjective weight WA1 of the monitoring data indicators is determined by using the improved hierarchical analysis method; secondly, the objective weight WA2 of each indicator is calculated by the CRITIC method based on the historical data of the transformer; finally, based on the maximum entropy principle, the subjective and objective weights are integrated to obtain the optimal weight matrix WA3 of the monitoring data indicators, and the monitoring data state discrimination matrix C1=WA3×B is calculated, where B is the membership vector matrix.

[0142] d. According to the formula:

[0143]

[0144] In the formula, Scorej is the monitoring data score.

[0145] e. According to the evaluation score score_day after integration, the environmental factor score factor S is assigned according to the actual operating environment of the day. 11 , get the transformer's final daily evaluation score Score_day:

[0146] Score_day=score_day×S 11 ;

[0147] f. Based on the final daily assessment score of the transformer and combined with Table 4, the daily comprehensive status assessment level is obtained.

[0148] Table 4 Status level distribution interval

[0149]

[0150] S312. Monthly evaluation of transformers.

[0151] In this embodiment, in the daily operation of the transformer, the preventive test is the main means of routine detection of the transformer in operation, which is generally required to be performed once a year or even once every six months. Therefore, in the monthly evaluation of the transformer, it is necessary not only to integrate the daily evaluation results of the transformer, but also to consider the results of the preventive test, such as Figure 3 As shown in the figure, the process of transformer monthly evaluation is:

[0152] a. Normalize the probability vector C obtained from the daily comprehensive evaluation results obtained through evidence theory within a month, such as 30 days, to obtain a probability vector as the basic evidence source D1 of the monthly evaluation model;

[0153] b. Select the most recent preventive trial data as supplementary evidence based on the assessment time;

[0154] c. The normalized data obtained in step b is used as the cloud model input value, and the membership vector matrix B2 of each indicator is obtained through the normal cloud + trapezoidal cloud model membership discrimination; at this time, the digital characteristics of the cloud model are based on Table 3, and the membership matrix B2 is weighted and summed in an equal weighted manner to obtain the supplementary evidence source D2;

[0155] d. The basic evidence source D1 in step a is fused in chronological order to obtain the basic support vector D3, and finally the evidence sources D2 and D3 are fused based on the improved DS evidence theory to obtain the monthly evaluation probability vector D;

[0156] e. Obtain the post-fusion daily evaluation score score_month according to the monitoring data scoring formula in step S311;

[0157] f. According to the final monthly score of the transformer and Table 4, the comprehensive status assessment grade within the month is obtained.

[0158] S313. Annual assessment of transformers.

[0159] In this embodiment, the annual comprehensive evaluation of the transformer is mainly used to characterize the health of the equipment in that year. Compared with daily and monthly evaluations, its evaluation cycle is longer, with more data and lower timeliness. It is mainly used to evaluate the health of the equipment. At the same time, the annual evaluation results provide data basis for equipment life prediction. Figure 4 As shown, the steps of the transformer status assessment method proposed in this embodiment are as follows:

[0160] a. Normalize the monthly evaluation probability vectors D for each of the 12 months in the year to obtain the basic evidence source vector E1;

[0161] b. The annual evaluation probability vector E is obtained by fusion in chronological order through the DS evidence theory with improved fusion rules;

[0162] c. Obtain the fusion annual evaluation score score_year according to the monitoring data scoring formula in S311 above;

[0163] d. According to the final annual assessment score of the transformer and Table 4, the daily comprehensive status assessment grade is obtained.

[0164] The application of the transformer multi-time scale state assessment method based on the improved DS evidence theory of the present invention is described in detail below through specific embodiments.

[0165] The object of analysis in this embodiment is a 220kV oil-immersed transformer on a certain offshore oil and gas platform group, with a model of SZ20-630000 / 220-NX2, a capacity of 150 / 90 / 70MVA, a rated voltage of 220±8×1.25% / 115 / 35, a connection group of YN, yn0, d11, and a cooling method of natural oil circulation (ONAN). The transformer is located on an offshore operating platform, and there are two identical main transformers on the operating station.

[0166] Based on the above settings, the transformer multi-time scale state assessment method based on the improved DS evidence theory in this embodiment has the following specific process:

[0167] 1. Daily evaluation and analysis of transformers.

[0168] According to the real-time evaluation results of the transformer on a certain day in January 2023, the specific time evaluation results and the evaluation result membership vector matrix are shown in Table 5 below.

[0169] Table 5

[0170]

[0171]

[0172] The weather conditions on that day are shown in Table 6 below.

[0173] Table 6 Weather conditions on that day

[0174]

[0175] According to the weather conditions on the day in the table above, on the day of assessment, the environmental factor score factor S 11 =0.98, the weather conditions are good.

[0176] According to the transformer daily evaluation steps, the evaluation results of each time node in Table 5 are fused using the DS evidence theory: first, the five-level membership of each time point is normalized to obtain the basic evidence source, and then the DS evidence theory with improved fusion rules is used to fuse the basic evidence sources. This can effectively avoid errors caused by strong conflicts between the evidence sources, and finally obtain the daily evaluation result vector C:

[0177] C=[0.54390.24450.13140.05580.0238];

[0178] The calculation results in score_month=79.97.

[0179] Combined with the environmental factor score factor, the comprehensive score Score_day of the transformer on that day is calculated as:

[0180] Score_day=79.97×0.98=78.38.

[0181] The final comprehensive evaluation score of the transformer on that day was 78.38. Combined with the evaluation level division interval in Table 4, the equipment is currently in good condition and does not need maintenance. However, the comprehensive score on that day is only 78.38, which is below the good condition. It is recommended to increase equipment inspections to find problems in time. The transformer status comments and descriptions are shown in Table 7:

[0182] Table 7

[0183]

[0184] Comparison of daily evaluation results and real-time evaluation results: Table 5 shows that the equipment's status level changes from excellent in the early stage to attention in the middle stage, and finally stabilizes in the excellent state, with a large level change. Because another device on the platform is in a power outage and maintenance state, only one transformer on the platform is working, resulting in excessive load rate during work, which causes the equipment's status level to change continuously. However, in the real-time evaluation of the day, the equipment score range is [64.61, 92.16], and the daily comprehensive score is 78.38. From the data, the daily comprehensive score is between the two, which is more reasonable; from the perspective of equipment operation status, the daily comprehensive evaluation level is good, which is also between the real-time evaluation level results. By retaining the membership of the real-time evaluation results, the influence of the monitoring data of various indicators of the transformer on the results is retained. By improving the evidence theory to integrate the evaluation results at different time points, a daily comprehensive evaluation is achieved, which can accurately reflect the equipment's operating status on the day, timely discover potential faults, and provide corresponding theoretical basis for inspection personnel.

[0185] 2. Monthly evaluation and analysis of transformers.

[0186] On the basis of the daily comprehensive evaluation, the real-time monitoring data of the equipment in the month is counted, and the daily comprehensive evaluation results are obtained by calculation as shown in Table 8 below.

[0187] Table 8 Comprehensive evaluation results for each day of the month

[0188]

[0189]

[0190] By consulting the equipment test records, the most recent preventive test record of the transformer as well as the factory values ​​of the indicators and the attention values ​​are shown in Table 9 below.

[0191] Table 9 Test values ​​of preventive test indicators

[0192]

[0193] Firstly, the normalized values ​​of the prevention test indicators were calculated, and the normalized scores of each indicator were 0.886, 0.7273, 0.8892, 0.7592, and 0.4387.

[0194] Substitute the normalized scores of each indicator into the cloud model membership discrimination algorithm. When calculating the membership of the preventive test indicators, the digital features of the cloud model refer to the above table to obtain the membership vector matrix B2:

[0195]

[0196] Therefore, we obtain the supplementary evidence source of preventive trials D2 = {0.327, 0.499, 0.119, 0.114, 0.001}. According to the improved DS evidence theory fusion rule, we fuse the daily evaluation results within 30 days of the month to obtain D3.

[0197] D3=[0.5430.3320.1000.0210.005];

[0198] Finally, the basic support vector D3 is fused with the supplementary evidence source D2 to obtain the monthly evaluation result vector D.

[0199] D=[0.4960.4630.0320.0060.001];

[0200] The final monthly evaluation score of the transformer is: 81.53, and the status level is good. In the month of the evaluation, the equipment has been in normal working condition, but the level change has also experienced three status levels: excellent, good, and attention. However, most of the time, the transformer is still operating in a good or above state. The final monthly evaluation result is good and the score is above 80, indicating that the equipment is in normal operating condition. Combined with the preventive test, the winding insulation dielectric loss index score is low, the maximum membership is at the abnormal level, and the final supplementary evidence source is the maximum membership of the good level. It is recommended to pay attention to whether there is a fault in the winding part. The monthly evaluation of the transformer integrates the comprehensive evaluation results of the equipment every day in the current month, which can reflect the operation of the equipment from a medium- and long-term perspective; at the same time, the monthly evaluation results can help formulate corresponding maintenance plans and promote the development of "planned maintenance" to "status maintenance".

[0201] 3. Annual evaluation and analysis of transformers.

[0202] The transformer is equipped with a monitoring system, which monitors the transformer operating parameters in real time and uploads them to the server center. Therefore, according to the server center data, the real-time monitoring data of the transformer in the past year is cleaned and calculated to obtain the monthly evaluation results, as shown in Table 10 below.

[0203] Table 10 Comprehensive evaluation results for the month within the year

[0204]

[0205]

[0206] According to the above annual evaluation steps, the fused probability vector E = {0.387, 0.453, 0.139, 0.018, 0.003} is obtained, and the final annual comprehensive score of the equipment is: 79.12. According to the monthly evaluation results of the transformer within one year, the equipment scores are mostly above 75 points, and the comprehensive status is mostly good. The equipment can operate normally and inspect according to the plan. However, the annual comprehensive evaluation score is 79.12 points, and the status is good. According to the inspection records of this year, the equipment has been in normal operation. During the daily inspection, it was found that only the winding had slight deformation and overload operation occurred at some time. After comparing the actual evaluation results with the actual operation status, the annual comprehensive evaluation reflects the health status of the equipment from the comprehensive operation status of the equipment within one year. It has a certain objectivity. It not only integrates the actual operation monitoring data of the equipment within one year, but also retains the digital characteristics of the previous evaluation, which can characterize the comprehensive status of the equipment within one year. At the same time, the annual comprehensive evaluation can provide data support for the later life prediction.

[0207] Embodiment 2: The above-mentioned embodiment 1 provides a transformer multi-time scale state assessment method based on the improved DS evidence theory. Correspondingly, this embodiment provides a transformer multi-time scale state assessment device based on the improved DS evidence theory. The device provided in this embodiment can implement the transformer multi-time scale state assessment method based on the improved DS evidence theory of embodiment 1, and the device can be implemented by software, hardware, or a combination of software and hardware. For the convenience of description, the description of this embodiment is divided into various units and described separately according to their functions. Of course, the functions of each unit can be implemented in the same or more software and / or hardware during implementation. For example, the device may include integrated or separate functional modules or functional units to perform the corresponding steps in each method of embodiment 1. Since the device of this embodiment is basically similar to the method embodiment, the description process of this embodiment is relatively simple, and the relevant parts can refer to the partial description of embodiment 1. The embodiment of the transformer multi-time scale state assessment device based on the improved DS evidence theory provided by the present invention is only illustrative.

[0208] Specifically, the transformer multi-time scale state assessment device based on the improved DS evidence theory provided by the present invention comprises:

[0209] The transformer daily evaluation unit is configured to perform transformer daily evaluation based on the improved DS evidence theory according to the membership vector matrix of the real-time monitoring data of the day and the transformer operating environment;

[0210] The transformer monthly evaluation unit is configured to perform monthly transformer evaluation based on the improved DS evidence theory according to the transformer daily evaluation results and preventive test results;

[0211] The transformer annual evaluation unit is configured to perform transformer annual evaluation based on the improved DS evidence theory according to the transformer monthly evaluation result.

[0212] Embodiment 3: This embodiment provides an electronic device corresponding to the transformer multi-time scale state assessment method based on the improved DS evidence theory provided in this embodiment 1. The electronic device can be an electronic device for a client, such as a mobile phone, a laptop computer, a tablet computer, a desktop computer, etc., to execute the method of embodiment 1.

[0213] The electronic device includes a processor, a memory, a communication interface and a bus, and the processor, the memory and the communication interface are connected through the bus to complete mutual communication. The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Component (EISA) bus, etc. The memory stores a computer program that can be run on the processor, and the processor executes the method of the first embodiment when running the computer program. Its implementation principle and technical effect are similar to those of the first embodiment, and will not be repeated here.

[0214] In a preferred embodiment, the logic instructions in the above-mentioned memory can be implemented in the form of a software functional unit and can be stored in a computer-readable storage medium when it is sold or used as an independent product. Based on this understanding, the technical solution of the present application can be essentially or partly embodied in the form of a software product that contributes to the prior art. The computer software product is stored in a storage medium, including several instructions to enable a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), optical disk and other media that can store program codes.

[0215] In a preferred embodiment, the processor may be a central processing unit (CPU), a digital signal processor (DSP) or other general-purpose processors of various types, which are not limited here.

[0216] Embodiment 4: This embodiment provides a computer-readable storage medium storing one or more programs, wherein the one or more programs include computer instructions. When the computer instructions are executed by a computer, the computer executes the method provided in the above-mentioned embodiment 1.

[0217] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (apparatus), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0218] These computer program instructions may also be stored in a computer readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture including an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0219] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process in the computer or other programmable device. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0220] Each embodiment in this specification is described in a progressive manner, and the same and similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In the description of this specification, the description of reference terms "a preferred embodiment", "further", "specifically", "in the present embodiment", etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the embodiment of this specification. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art can combine and combine the different embodiments or examples described in this specification and the features of the different embodiments or examples without contradiction.

[0221] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A transformer multi-time scale state assessment method based on improved DS evidence theory, characterized in that: include: According to the membership vector matrix of the real-time monitoring data of the day and the transformer operating environment, the transformer daily evaluation is carried out based on the improved DS evidence theory; According to the daily transformer evaluation results and preventive test results, monthly transformer evaluation is carried out based on the improved DS evidence theory; According to the monthly evaluation results of transformers, annual evaluation of transformers is carried out based on the improved DS evidence theory.

2. The transformer multi-time scale state assessment method based on improved DS evidence theory according to claim 1 is characterized in that: According to the membership vector matrix of the real-time monitoring data of the day and the transformer operating environment, the transformer daily evaluation is carried out based on the improved DS evidence theory, including: The membership vectors of the monitoring data obtained by the real-time evaluation of the day are sorted to obtain the membership vector matrix B; Normalize the membership vector matrix B according to the time sequence of real-time evaluation to obtain vector B0 as the evidence source of evidence theory; The improved DS evidence theory is used to fuse the daily real-time evaluation results in chronological order to obtain the transformer daily evaluation probability vector C, where: C1 is the monitoring data state discrimination matrix; a is the expected score matrix of each state, a=[95,75,55,35,15], i=1,2,3,4,5 represents the state is excellent, good, caution, abnormal, severe; Based on the transformer daily evaluation probability vector C, the fused daily evaluation score score_day is obtained: Assign environmental factor score factor S according to the actual operating environment on that day 11 And the fusion daily evaluation score score_day, get the final daily evaluation score of the transformer: Score_day=score_day×S 11 ; The daily comprehensive status assessment grade is obtained based on the final daily assessment score of the transformer and the assessment grade division interval.

3. The transformer multi-time scale state assessment method based on improved DS evidence theory according to claim 2 is characterized in that: The calculation process of the membership vector matrix B is: Firstly, quantitative data preprocessing is performed on each monitoring data to serve as cloud model input value; Secondly, the transformer health index is calculated according to the equipment operating years and the number of overhauls to correct the digital eigenvalues ​​of the cloud model; Then, the five evaluation results are converted into digital features of the cloud model using the health index; Finally, the indicator membership discrimination method based on the combination of trapezoidal cloud and normal cloud is adopted to obtain the membership vector matrix B corresponding to the five states of each indicator of the monitoring data. Among them, the membership vector matrix B divides the real-time evaluation indicators into five qualitative concepts: excellent, good, attention, abnormal and serious.

4. The transformer multi-time scale state assessment method based on improved DS evidence theory according to claim 3 is characterized in that: The calculation formula for the fused daily evaluation score based on the transformer daily evaluation probability vector C is: Where Scorej is the monitoring data score.

5. The transformer multi-time scale state assessment method based on improved DS evidence theory according to claim 2 is characterized in that: Based on the transformer daily assessment results and preventive test results, monthly transformer assessment is carried out, including: The probability vector C obtained from the daily comprehensive evaluation results within the month is normalized to obtain a probability vector as the basic evidence source D1 of the monthly evaluation model; Data from preventive trials with set times were selected as supplementary evidence based on the evaluation period; The normalized data of the supplementary evidence is used as the input value of the cloud model membership discrimination algorithm. The membership vector matrix B2 of each indicator is obtained through the normal cloud + trapezoidal cloud model membership discrimination, and the membership vector matrix B2 is weighted and summed in an equal weighted manner to obtain the supplementary evidence source D2; The basic evidence source D1 is fused in chronological order based on the improved DS evidence theory fusion rule to obtain the basic support vector D3; The basic support vector D3 and the supplementary evidence source D2 are fused based on the improved DS evidence theory to obtain the monthly evaluation probability vector D; Based on the monthly evaluation probability vector D, the fused monthly evaluation score score_month is obtained; The monthly comprehensive status assessment grade is obtained based on the final score of the transformer and the assessment grade division interval.

6. The transformer multi-time scale state assessment method based on improved DS evidence theory according to claim 5 is characterized in that: The test values ​​of preventive test indicators include insulation resistance absorption ratio, polarization coefficient, volume resistivity, winding DC resistance difference and / or winding insulation dielectric loss.

7. The transformer multi-time scale state assessment method based on improved DS evidence theory according to claim 5 is characterized in that: According to the monthly transformer assessment results, the annual transformer assessment is carried out based on the improved DS evidence theory, including: Normalize the monthly evaluation probability vectors D of the 12 months in the year respectively to obtain the basic evidence source vector E1; The annual evaluation probability vector E is obtained by fusing in chronological order based on the improved DS evidence theory; Based on the annual evaluation probability vector E, the fused annual evaluation score score_year is obtained; The comprehensive status assessment grade within the year is obtained based on the transformer annual assessment score and the assessment grade division interval.

8. A transformer multi-time scale state assessment device based on improved DS evidence theory, characterized in that: include: The transformer daily evaluation unit is configured to perform transformer daily evaluation based on the improved DS evidence theory according to the membership vector matrix of the real-time monitoring data of the day and the transformer operating environment; The transformer monthly evaluation unit is configured to perform monthly transformer evaluation based on the improved DS evidence theory according to the transformer daily evaluation results and preventive test results; The transformer annual evaluation unit is configured to perform transformer annual evaluation based on the improved DS evidence theory according to the transformer monthly evaluation result.

9. An electronic device, characterized in that: include: at least one processor; and a memory communicatively connected to the processor; wherein the memory stores instructions executable by the processor, and the instructions are executed by the processor so that the processor can execute the method according to any one of claims 1-7.

10. A computer-readable storage medium storing one or more programs, characterized in that: The one or more programs include computer instructions for causing a computer to execute the method according to any one of claims 1-7.

Images