Converter station optical current transformer state evaluation method
Through the subjective and objective combination of empowerment method and weighted rank sum ratio method, a state evaluation method for optical current transformers is constructed, which solves the problem of difficulty in evaluating the health status of the equipment in the prior art, and achieves rapid and accurate evaluation in extreme environments, improving the reliability and safety of the equipment.
Patent Information
- Application Number
- CN202411792693.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-12-07
- Publication Date
- 2025-03-28
AI Technical Summary
The existing technology lacks standardized optical current transformer state evaluation methods, which makes it difficult to quickly and accurately evaluate the health status of the equipment in extreme environments, increasing failure rate and safety risks.
The subjective and objective combination empowerment method is used to assign weights to characteristic parameters, and the weighted rank sum ratio method is used to evaluate health status. By constructing the rank matrix and weighted rank sum ratio estimates, a rapid and accurate evaluation of the state of the optical current transformer is achieved.
It realizes a rapid and accurate assessment of the health status of optical current transformers in extreme environments, improves the reliability and safety of equipment operation, and reduces the failure rate and maintenance costs.
Smart Images
Figure CN119719604A_ABST
Abstract
Description
Technical Field
[0001] The present invention belongs to the field of power equipment operation and maintenance technology, and specifically relates to a method for evaluating the state of an optical current transformer in a converter station. Background Art
[0002] Optical Current Transducer (OCT) has attracted wide attention in power systems due to its advantages such as no magnetic saturation, good insulation performance and strong anti-interference ability. Optical current transformers have been increasingly used in smart substations, and their operating reliability and stability directly or indirectly affect the safe operation of smart substations.
[0003] At present, there are relatively few studies on the state assessment methods of optical current sensors at home and abroad, and there is no standardized assessment method. The optical current transformer based on optical principles is an optomechanical and electrical integrated device, and the factors affecting its operating status are wide-ranging and complex. During operation, the optical current transformer is subjected to the long-term effects of various factors such as electricity, heat, and mechanics, and the failure rate will gradually increase. It is necessary to find and replace the current transformer with hidden dangers in time. In order to ensure production safety and efficiency, it is necessary to establish a standardized state assessment method for optical current transformers. Summary of the invention
[0004] In view of the shortcomings of the existing methods, the present invention proposes a method for evaluating the status of optical current transformers in converter stations. The subjective and objective combined weighting method is used to assign weights to characteristic parameters. In addition, the weighted rank sum ratio method with statistical characteristics is applied to the evaluation of the health status of OCT. The advantages of the weighted rank sum ratio method, such as simple model principle and stable performance, are utilized to achieve rapid and accurate evaluation of the health status of OCT under extreme environments.
[0005] In order to solve the above technical problems, the technical solution adopted by the present invention is:
[0006] A method for evaluating the state of an optical current transformer in a converter station comprises the following steps:
[0007] Step 1: Obtain the state characteristic parameters of the evaluation object and construct a rank matrix based on the merits and demerits of the single state characteristic parameters;
[0008] Step 2: Calculate the rank sum ratio estimate of the evaluation object or the weighted rank sum ratio estimate of the evaluation object according to whether the comprehensive weights of each state characteristic parameter are the same;
[0009] Step 3: Group the evaluation objects according to the estimated values, and Calculate the probability unit of the evaluation object;
[0010] Step 4: Construct a regression equation between the rank sum ratio estimate or the weighted rank sum ratio estimate and the probability unit of the evaluation object;
[0011] Step 5: Sorting and evaluation.
[0012] Furthermore, the state characteristic parameters include: LED current, semiconductor cooler temperature, modulator drive voltage, input voltage peak value, input voltage average deviation and second harmonic voltage.
[0013] Furthermore, the method of constructing the rank matrix is as follows: the evaluation matrix formed by constructing all state characteristic parameters is as follows:
[0014] A=(a ij ) m×n (1)
[0015] Where i is the number of the device to be evaluated, and j is the number of the state characteristic parameter;
[0016] The characteristic parameters of the same state of each device to be evaluated are sorted from best to worst. After the ranking is completed, the new rank matrix is as follows:
[0017] R=(r ij ) m×n (2)
[0018] Furthermore, the comprehensive weight of the state characteristic parameters is:
[0019]
[0020] Where W i is the comprehensive weight value of the state characteristic parameter i, ω is the subjective weight, is the objective weight;
[0021] If the comprehensive weight of each state characteristic parameter is the same, the rank sum ratio estimate δ of the evaluation object is directly calculated RSRi , otherwise calculate the weighted rank sum ratio estimate δ of the evaluation object WRSRi , and the calculation formulas are:
[0022]
[0023] Where, i = 1, 2, ..., m; R ij is the rank of the i-th row and j-th column, ω j is the comprehensive weight of the j-th state characteristic parameter, ∑ω j =1.
[0024] Furthermore, the rank sum ratio estimates or weighted rank sum ratio estimates are arranged in ascending order, and the same values are grouped together. The frequency f and cumulative frequency f of each group of data are counted. i , calculate the rank range S and average rank Get the cumulative frequency p of each group i :
[0025]
[0026] The last group p in the formula i The result is 1. To correct it, the p of the last group is specified. i =1-1 / 4m;
[0027] Then we get the probability unit P of each evaluation object. robiti :
[0028] P robiti =u(p i )+5 (7)
[0029] Among them, u(p i ) is p i The standard normal deviation obtained by converting according to the probability unit correspondence table.
[0030] Furthermore, with δ RSRi or δ WRSRi is the dependent variable, P robiti The linear relationship constructed for the independent variables is as follows:
[0031] δ RSRi (or δ WRSRi )=a+b×P robiti (8)
[0032] Wherein, a and b are both constants.
[0033] Further, the bin sorting and evaluation steps include:
[0034] According to p i Size classification evaluates equipment grade;
[0035] The P of each evaluation device robiti Substitute the linear regression equation into the calculation, and archive the calculation results according to the divided classification table to achieve the status assessment of the assessment equipment.
[0036] Compared with the prior art, the present invention has the following beneficial effects:
[0037] The present invention uses a subjective and objective combined weighting method to assign weights to characteristic parameters. In addition, a weighted rank sum ratio method with statistical characteristics is applied to the assessment of OCT health status. The advantages of the model's simple principle and stable performance are utilized to realize the assessment of OCT health status under extreme environments. BRIEF DESCRIPTION OF THE DRAWINGS
[0038] Figure 1The figure is a schematic diagram of the steps of a method for evaluating the state of an optical current transformer in a converter station according to an embodiment of the present invention. DETAILED DESCRIPTION
[0039] 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. 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 making creative work are within the scope of protection of the present invention.
[0040] Example
[0041] See also Figure 1 The method for evaluating the state of the optical current transformer in the converter station specifically includes the following steps:
[0042] 1) Obtain state characteristic parameters and construct rank matrix.
[0043] There are two ranking methods: the integer rank sum ratio method and the non-integer rank sum ratio method, which are not limited in the present invention. This embodiment adopts the integer rank sum ratio method.
[0044] The weighted rank-sum ratio method (WRSR) can be used for four-cell table evaluation and row × column table data evaluation. For each evaluation object to be evaluated and each evaluation indicator of each evaluation device, its specific parameters are listed as a matrix with m rows and n columns. Each parameter in the matrix can obtain a corresponding weighted sum ratio estimate after rank conversion. Then, each weighted sum ratio estimate is sorted according to the positive or negative impact of the weighted sum ratio estimate on the evaluation device, so as to obtain a comprehensive evaluation of each evaluation object. The larger the calculated weighted sum ratio estimate, the better the evaluation result.
[0045] The WRSR state assessment method is usually used to assess multiple assessment objects, while the WRSR-based OCT state assessment only assesses a single device, so OCT devices at different times are selected as multiple assessment objects. After determining the state feature parameters for assessment and classifying the state features, the OCT state assessment can be performed according to the WRSR assessment steps.
[0046] Specifically, determine the number of devices m that need to be evaluated and n state characteristic parameters based on which the evaluation is based. The state characteristic parameters are: LED current, semiconductor cooler temperature, modulator drive voltage, input voltage peak value, input voltage average deviation and second harmonic voltage. The evaluation matrix formed by all state characteristic parameters is as follows:
[0047] A=(a ij ) m×n(1)
[0048] Where i is the number of the device to be evaluated, and j is the number of the state characteristic parameter.
[0049] The data in the evaluation matrix are ranked according to specific rules, that is, the data is sorted, and its sequence number is used as the rank. The specific rules are: all data of the same state characteristic parameter are sorted from best to worst. For example, the larger the benefit index, the better it is, so it is ranked from large to small, and the cost index is the opposite. Then, according to the sorting result, the corresponding equipment and other state characteristic parameters are sorted.
[0050] The new matrix obtained after completing the ranking becomes the rank matrix, as follows:
[0051] R=(r ij ) m×n (2)
[0052] 2) Calculate the rank sum ratio estimate or the weighted rank sum ratio estimate.
[0053] Specifically, this embodiment adopts a combination method to perform weight analysis and calculate the final comprehensive weight, and the calculation method is as follows:
[0054]
[0055] Where W i is the comprehensive weight value of the state characteristic parameter i, ω is the subjective weight, The objective weights are not limited in the present invention. The subjective weights and the calculation method of the objective weights are preferably calculated using the Analytic Hierarchy Process (AHP) after manual assignment, and the objective weights are calculated using the Coefficient of Variation (CoV) method.
[0056] According to the actual situation of the evaluation equipment, there are two possibilities. If the comprehensive weight of each state characteristic parameter is the same, the rank sum ratio estimate δ of the evaluation object is directly calculated. RSRi That is, if the comprehensive weight of each state characteristic parameter is different, the weighted rank sum ratio estimate δ of the evaluation object is calculated WRSRi .
[0057] The calculation formulas are:
[0058]
[0059] Where, i = 1, 2, ..., m; R ij is the rank of the i-th row and j-th column, ω j is the comprehensive weight of the j-th state characteristic parameter, ∑ω j=1.
[0060] By comparing δ RSRi or δ WRSRi After that, we can sort them by size to achieve comparative evaluation of multiple evaluation objects, but this cannot confirm the specific operating status of each evaluation object, so we need to further classify the evaluation objects and analyze the estimated value δ RSRi or δ WRSRi The distribution of .
[0061] 3) Calculate probability units.
[0062] Specifically, the estimated value δ RSRi or δ WRSRi Rearrange in ascending order, group the same size data into the same group, and count the frequency f and cumulative frequency f of each group of data. i , to form a frequency distribution table, it is also necessary to calculate the rank range S and average rank of each group Thus, the cumulative frequency p of each group is obtained i :
[0063]
[0064] But according to the above method, p of the last group is calculated i When the result is 1, in order to correct it, the p of the last group is specified i =1-1 / 4m.
[0065] Finally, calculate the probability unit P of each evaluation object robiti , and then complete the establishment of the subsequent regression equation:
[0066] P robiti =u(p i )+5 (7)
[0067] In the formula, u(p i ) is p i The standard normal deviation obtained by converting according to the probability unit correspondence table.
[0068] 4) Calculate the regression equation.
[0069] Using the univariate linear regression equation to establish δ RSRi or δ WRSRi With P robiti The linear relationship between them is as follows, where δ RSRi or δ WRSRi is the dependent variable, P robiti For the independent variable:
[0070] δ RSRi (or δ WRSRi )=a+b×P robiti (8)
[0071] Wherein, a and b are both constants.
[0072] 5) Sorting and evaluation.
[0073] According to p i The size of the evaluation equipment is divided into grades. The number of grades depends on the specific situation, usually 3-5 grades. In this embodiment, it is divided into 4 grades. The variance between each grade is required to be consistent. The probability unit critical value under the commonly used graded conditions is shown in Table 1. After the grades are divided, the P of each evaluation equipment is robiti Substitute the linear regression equation into the calculation, archive the calculation results according to the divided classification table, and finally achieve the status assessment of the assessment equipment.
[0074] Table 1 Probability unit critical division values
[0075]
[0076] Table 1 Probability unit critical division values
[0077]
[0078] The six indicators determined in step 1) can reflect the degradation of the OCT equipment to varying degrees. RSRi or δ WRSRi The values are divided into four states, namely normal, caution, abnormal and severe. According to the four states, three critical values 3.5, 5 and 6 are obtained by looking up Table 1, and finally the OCT evaluation system can be established. The definitions of the four states are shown in Table 2.
[0079] Table 2 OCT operation status classification
[0080]
[0081] 6) Inspection and evaluation.
[0082] In order to verify the OCT status assessment method introduced in this paper, an OCT device was selected from a converter station and some of its operating data in August 2021 were collected, as shown in Table 3.
[0083] Table 3 Daily average values of OCT operation data
[0084]
[0085] Table 3 Daily average values of OCT operation data
[0086]
[0087] Table 3 Daily average values of OCT operation data
[0088]
[0089] According to the analytic hierarchy process introduced above, subjective weights are assigned to the six characteristic state quantities. Four experts assign subjective weights to each parameter individually, and the subjective weights ω of the six characteristic state quantities are calculated. i =[0.141, 0.185, 0.099, 0.068, 0.215, 0.292]; Calculate the objective weights of the six characteristic state quantities according to the coefficient of variation method =[0.204, 0.133, 0.129, 0.13, 0.174, 0.23]; Finally, the subjective and objective weights are combined to obtain the comprehensive weight W of the six characteristic state quantities i =[0.165, 0.135, 0.066, 0.046, 0.239, 0.35].
[0090] The OCT operation data is listed as a matrix and then ranked. Since the four indicators of LED current, semiconductor cooler temperature, modulator drive voltage, and input voltage peak are neither cost indicators nor benefit indicators, their average deviation is taken and converted into cost indicators to unify the evaluation method of all feature quantities, which is convenient for the final unified calculation and ranking. The rank matrix R is shown in formula (9):
[0091]
[0092] According to the calculation formula (5) of the weighted rank sum ratio estimate, the δ of each day is calculated as WRSRi , and calculated several other parameters one by one according to the formula to form the corresponding frequency distribution table as shown in Table 4.
[0093] Table 4 Frequency distribution table
[0094]
[0095] Table 4 Frequency distribution table
[0096]
[0097] Table 4 Frequency distribution table
[0098]
[0099] Taking P in Table 4 as robiti As the independent variable, δ WRSRi As the dependent variable, the univariate linear regression equation is constructed as follows:
[0100] δ WRSRi =-0.3171+0.164P robiti(10)
[0101] The daily P in Table 4 robiti Substitute the value of into the established univariate linear regression equation, and calculate the weighted rank sum ratio of all dates: [0.713, 0.688, 0.666, 0.884, 0.811, 0.624, 0.647, 0.580, 0.608, 0.511, 0.566, 0.536, 0.318, 0.293, 0.340, 0.426, 0.549, 0.495, 0.594, 0.359, 0.412, 0.482, 0.382, 0.397, 0.440, 0.470, 0.194, 0.248, 0.457, 0.524, 0.758]. In order to correspond to the four OCT operating states (normal, caution, abnormal, and severe) divided in the previous article, the weighted rank sum ratio calculated in the end needs to be divided into four levels. From Table 1, it can be determined that the critical values corresponding to the four levels are 3.5, 5, and 6, respectively. Substituting the critical values into the univariate linear regression equation, the critical values of the weighted rank sum ratio are 0.257, 0.501, and 0.667, respectively. Thus, the OCT operating state classification table can be established as shown in Table 5.
[0102] Table 5 OCT operation status classification table
[0103]
[0104] In this paper, OCT devices on different dates are regarded as different evaluation objects, and finally the δ calculated on all dates is WRSRi The average value is calculated to get δ WRSRi =0.515, according to the OCT operating status classification table, the result is "Caution".
[0105] Finally, it should be noted that the above embodiments are only preferred embodiments of the present invention, which are only used to illustrate the technical solution of the present invention rather than to limit it. Any modification, improvement, other modification or equivalent replacement made to the technical solution of the present invention by ordinary technicians in this field, as long as they do not depart from the spirit and scope of the technical solution of the present invention, should be included in the protection scope of the present invention.
Claims
1. A method for evaluating the state of an optical current transformer in a converter station, characterized in that: The steps include: Step 1: Obtain the state characteristic parameters of the evaluation object and construct a rank matrix based on the merits and demerits of the single state characteristic parameters; Step 2: Calculate the rank sum ratio estimate of the evaluation object or the weighted rank sum ratio estimate of the evaluation object according to whether the comprehensive weights of each state characteristic parameter are the same; Step 3: Group the evaluation objects according to the estimated values, and Calculate the probability unit of the evaluation object; Step 4: Construct a regression equation between the rank sum ratio estimate or the weighted rank sum ratio estimate and the probability unit of the evaluation object; Step 5: Sorting and evaluation.
2. The method for evaluating the state of an optical current transformer in a converter station according to claim 1, characterized in that: The state characteristic parameters include: LED current, semiconductor cooler temperature, modulator drive voltage, input voltage peak value, input voltage average deviation and second harmonic voltage.
3. The method for evaluating the state of an optical current transformer in a converter station according to claim 1, characterized in that: The method to construct the rank matrix is: The evaluation matrix formed by constructing all state characteristic parameters is as follows: A=(a ij ) m×n (1) Where i is the number of the device to be evaluated, and j is the number of the state characteristic parameter; The characteristic parameters of the same state of each device to be evaluated are sorted from best to worst. After the ranking is completed, the new rank matrix is as follows: R＝(r ij ) m×n 。 (2) 4. The method for evaluating the state of an optical current transformer in a converter station according to claim 3, characterized in that: The comprehensive weight of the state characteristic parameters is: Where W i is the comprehensive weight value of the state characteristic parameter i, ω is the subjective weight, is the objective weight; If the comprehensive weight of each state characteristic parameter is the same, the rank sum ratio estimate δ of the evaluation object is directly calculated RSRi , otherwise calculate the weighted rank sum ratio estimate δ of the evaluation object WRSRi , and the calculation formulas are: Where, i = 1, 2, ..., m; R ij is the rank of the i-th row and j-th column, ω j is the comprehensive weight of the j-th state characteristic parameter, ∑ω j =1.
5. The method for evaluating the state of an optical current transformer in a converter station according to claim 4, characterized in that: The rank sum ratio estimates or weighted rank sum ratio estimates are arranged in ascending order. The values with the same value are grouped together. The frequency f and cumulative frequency f of each group of data are counted. i , calculate the rank range R and the average rank Get the cumulative frequency p of each group i : The last group of p in the formula i The result is 1. To correct it, the p of the last group is specified. i =1-1 / 4m; Then we get the probability unit P of each evaluation object. robiti : Q robiti =u(p i )+5 (7) Among them, u(p i ) is p i The standard normal deviation obtained by converting according to the probability unit correspondence table.
6. The method for evaluating the state of an optical current transformer in a converter station according to claim 5, characterized in that: By δ RSRi or δ WRSRi is the dependent variable, P robiti The linear relationship constructed for the independent variables is as follows: d RSRi (or WRSRi )=a+b×P robiti (8) Where a and b are both constants.
7. The method for evaluating the state of an optical current transformer in a converter station according to claim 6, characterized in that: The bin sorting and evaluation steps include: According to p i Size classification evaluates equipment grade; The P of each evaluation device robiti Substitute the linear regression equation into the calculation, and archive the calculation results according to the divided classification table to achieve the status assessment of the assessment equipment.