A method for analyzing the influence of source-load power characteristics on transformer stability margin

Abstract

The present application belongs to the technical field of electric power, and provides a kind of source and load power characteristic influence analysis method to transformer stability margin, first in the metering automation system and SCADA system relevant data are collected, including wind power, photovoltaic power and load power change data with time and the actual voltage of transformer side, temperature rise, load current, winding compression force etc., and the data is preprocessed;Then, reference sequence and comparison sequence are established;Then, the unweighted correlation coefficient is obtained by improved difference quotient grey correlation analysis, and the influence degree and the influenced degree are obtained by applying improved Grey-DEMATEL method through comprehensive influence matrix, and then the centrality is calculated;Finally, the weight is calculated by centrality and the comprehensive grey correlation degree is obtained, and this comprehensive evaluation method can more accurately reflect the real situation of transformer stability margin.

Description

Technical Field

[0001] The present invention belongs to the field of electric power technology, and in particular relates to a method for analyzing the influence of source-load power characteristics on transformer stability margin. Background Art

[0002] With the massive influx of distributed generation (DGs) in power systems, and the increasing diversity and volatility of loads, source-load power characteristics are becoming increasingly complex. As a key component of power systems, the stability margin of transformers is directly related to the safe and stable operation of the power system. However, there is currently a lack of effective methods to accurately analyze the relationship between source-load power characteristics and transformer stability margins. This makes it difficult to identify key influencing factors and their relative importance, making it difficult to take targeted measures to ensure the stable operation of transformers. Therefore, a method is needed to analyze the impact of source-load power characteristics on transformer stability margins. Summary of the Invention

[0003] To solve the above technical problems, the present invention provides a method for analyzing the influence of source-load power characteristics on transformer stability margin to solve the problems in the prior art. The technical solution adopted by the present invention is:

[0004] A method for analyzing the influence of source-load power characteristics on transformer stability margin includes the following steps:

[0005] Step 1: Obtain raw data from the metering automation system and SCADA system;

[0006] Step 2: Preprocess the original data by removing abnormal data and normalizing it;

[0007] Step 3: Analyze the transformer stability margin from multiple angles and establish a reference sequence;

[0008] Step 4: Obtain the unweighted correlation coefficient through the improved difference quotient grey correlation analysis;

[0009] Step 5: Apply the improved Grey-DEMATEL method to obtain the index weights and calculate the comprehensive grey relational degree;

[0010] Furthermore, step 2 includes:

[0011] Step 2.1, missing data filling:

[0012] Interpolation method:

[0013]

[0014] where x t and x t+1 are the data at time t and time t+1, respectively, x t+k is the missing data at time t+k, and n is the total amount of data;

[0015] Multiple imputation method:

[0016] Assume that the variable to be interpolated is X, and its posterior distribution of observed data is given by the following formula:

[0017] P(X|X obs )=∫P(X|X obs ,X mis )P(Y mis |X obs )dX

[0018] where X obs represents the observed data, X mis Represents missing data, and the posterior predictive distribution P(X mis |X obs ) Perform m sets of conditional independent extraction to obtain repeated interpolation sets, and calculate the repeated interpolation value from each interpolated data set The final interpolated value of X:

[0019]

[0020] Step 2.2, duplicate handling:

[0021] Detect duplicate data by checking the timestamp. If duplicates are found, delete the redundant items and record the deleted duplicates.

[0022] Step 2.3, normalization, maps the original data to a specific interval so that they can be compared and analyzed on the same scale:

[0023]

[0024] where x i (β) is the data after normalization, is the original data indicator value, Indicates the average value of all load data.

[0025] Furthermore, in step 3, the transformer stability margin is analyzed from multiple angles, including voltage stability margin, thermal stability margin, short-circuit capacity margin, and winding compression force margin, where:

[0026] Voltage stability margin:

[0027]

[0028] V min ≤V a ≤V max

[0029]

[0030] Q g =Q c +Q l

[0031] V M Indicates the voltage stability margin, V a Indicates the actual voltage, V r Indicates rated voltage, V min Indicates the voltage lower limit, V max Indicates the voltage upper limit, Q g Indicates the reactive power generated by the reactive power source, Q c Indicates the reactive power consumed by the load, Q l represents reactive power loss, X represents system reactance;

[0032] Thermal stability margin:

[0033] θ M =θ al -θ a

[0034] θ a ≤θ max

[0035] P h =I 2 R

[0036]

[0037] θ M represents the thermal stability margin, θ a Indicates the actual temperature rise of the winding, θ al Indicates the allowable temperature rise, θ max Indicates the maximum temperature allowed by the insulation material, P h Indicates the heating power of the winding, I indicates the load current, R indicates the winding resistance, and C indicates the equivalent thermal capacity of the transformer;

[0038] Short circuit capacity margin:

[0039]

[0040] I sc ≤I sc,max

[0041] I sc ≤I t,sc And t sc ≤t t,sc

[0042] I sc ≤I br，sc

[0043] S MIndicates the short-circuit capacity margin, Z s Represents the system short-circuit impedance, Z t Represents the transformer short-circuit impedance, I sc Indicates short-circuit current, V r Indicates rated voltage, I sc,max Indicates the maximum allowable short-circuit current, I t,sc Indicates the transformer short-circuit withstand current, t sc Indicates the short circuit duration, t t,sc Indicates the short-circuit duration allowed by the transformer, I br，sc Indicates the rated short-circuit breaking current of the circuit breaker;

[0044] Winding compression force margin:

[0045]

[0046] P M is the winding compression force margin, P a is the winding compression force P min P is the lower limit of the normal range of winding compression force, max P is the upper limit of the normal range of winding compression force, o is the ideal compression force value;

[0047] The reference sequence established in step 3 includes: taking the transformer stability margin data sequence as the reference sequence, the voltage stability margin Thermal stability Short-circuit capacity margin Winding compression force margin

[0048] Furthermore, step 4 includes:

[0049] The source-load power characteristic data sequence is used as a comparison sequence, including wind power X1 = {x1(1), x1(2), ..., x1(n)}, photovoltaic power X2 = {x2(1), x2(2), ..., x2(n)}, and load power X3 = {x3(1), x3(2), ..., x3(n)};

[0050] Compute the discrepancy matrix and the quotient matrix:

[0051] Calculate reference sequence and compare sequence X i The difference between the index values ​​​​is obtained by the difference matrix:

[0052]

[0053] Where Δx i (j) represents the difference between the i-th comparison sequence and the reference sequence in the j-th index;

[0054] Calculate reference sequence and compare sequence X i The quotient values ​​of each indicator value are used to obtain the quotient matrix:

[0055]

[0056] Where Δx i (j) represents the quotient value of the i-th comparison sequence and the reference sequence at the j-th index;

[0057] Calculate the grey relational coefficient:

[0058] Grey relational coefficient of geometric similarity based on difference matrix:

[0059]

[0060] Grey relational coefficient based on the numerical proximity of the quotient matrix:

[0061]

[0062] Furthermore, step 5 includes:

[0063] Step 5.1: Form an expert evaluation team;

[0064] Step 5.2, expert evaluation and data collection;

[0065] Step 5.3, construct the direct grey relationship matrix:

[0066] The semantic variables evaluated by experts are converted into gray numbers according to the gray language scale in the table, and the direct impact matrix is ​​established:

[0067]

[0068] Where n is the number of indicators, is the gray number of the evaluation of the αth expert on the direct impact of indicator i on indicator j;

[0069] Step 5.4, calculate the clear relationship matrix:

[0070] Normalize the gray value:

[0071]

[0072] are the upper and lower limits of the normalized gray number;

[0073] Calculate the total normalized clarity value:

[0074]

[0075] Calculate the final clarity value:

[0076]

[0077] Step 5.5, calculate the direct impact matrix:

[0078]

[0079] Where f is the total number of experts;

[0080] Step 5.6, determine the comprehensive impact matrix:

[0081] Normalize the real direct influence matrix C by dividing each element value in the matrix by the maximum value of the direct influence matrix to obtain the normalized direct relationship matrix S, and then calculate the comprehensive influence matrix X:

[0082]

[0083] Step 5.7, calculate the influence, influence, centrality and cause:

[0084] Impact:

[0085]

[0086] Affected degree:

[0087]

[0088] Centrality:

[0089] M j =p j +q j

[0090] Cause degree:

[0091] N i =p j -q j

[0092] Step 5.8, calculate indicator weights:

[0093] Calculate the weight of each indicator based on the centrality:

[0094]

[0095] Step 5.9, calculate the comprehensive grey relational degree:

[0096]

[0097] Step 5.10: Determine the priority of influencing factors based on the comprehensive grey correlation degree.

[0098] The present invention has the following beneficial effects:

[0099] The present invention uses an improved difference-quotient grey correlation analysis method to calculate unweighted correlation coefficients. This method, by introducing a difference matrix and a quotient matrix, considers the relationship between data sequences from the perspectives of geometric similarity and numerical proximity, thereby improving the accuracy of transformer stability margin analysis. The improved Grey-DEMATEL method is used to obtain indicator weights, determining the relative importance of each indicator by calculating the influence and influence degree, and rationally allocating indicator weights, thus avoiding subjectivity in weight determination. This comprehensive evaluation method makes the evaluation results more scientific and reliable, and can more accurately reflect the true stability margin of the transformer under different operating conditions. BRIEF DESCRIPTION OF THE DRAWINGS

[0100] Figure 1 The figure is a flow chart of the overall method of the present invention. DETAILED DESCRIPTION

[0101] The following is a combination of the embodiments of the present invention Figure 1 , the technical solutions in the embodiments of the present invention are described clearly and completely. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments. Unless otherwise specified, the technical means used in the embodiments are conventional means well known to those skilled in the art.

[0102] The present invention first collects relevant data in the metering automation system and the SCADA system and preprocesses the data; then, a reference sequence and a comparison sequence are established; then, an unweighted correlation coefficient is obtained through an improved difference quotient grey correlation analysis, and an improved Grey-DEMATEL method is applied to obtain the influence degree and the influenced degree through a comprehensive influence matrix, thereby calculating the centrality; finally, the weight is calculated from the centrality to obtain the comprehensive grey correlation degree. This comprehensive evaluation method can more accurately reflect the true situation of the transformer stability margin.

[0103] The present invention proposes a method for analyzing the influence of source-load power characteristics on transformer stability margin, comprising the following steps:

[0104] Step 1: Obtain raw data from the metering automation system and SCADA system. This includes data on power and load changes over time for wind power, photovoltaic power, and other sources, as well as actual transformer voltage, temperature rise, load current, and winding compression force. Metering automation systems and SCADA systems are crucial technical tools for the operation, control, and management of modern power systems. Working together, they enable comprehensive monitoring and effective management of the power system.

[0105] Step 2: Preprocess the original data by removing abnormal data and normalizing it:

[0106] Step 2.1, missing data filling:

[0107] Interpolation method (a small number of discontinuous missing values):

[0108]

[0109] where x t and x t+1 are the data at time t and time t+1, respectively, x t+k is the missing data at time t+k, and n is the total amount of data.

[0110] Multiple imputation method (consecutive missing values):

[0111] Assume that the variable to be interpolated is X, and its posterior distribution of observed data is given by the following formula:

[0112] P(X|X obs )=∫P(X|X obs ,X mis )P(Y mis |X obs )dX

[0113] where X obs represents the observed data, X mis Represents missing data, and the posterior predictive distribution P(X mis |X obs ) Perform m sets of conditional independent extraction to obtain repeated interpolation sets, and calculate the repeated interpolation value from each interpolated data set The final interpolated value of X

[0114]

[0115] Step 2.2, duplicate handling:

[0116] Duplicate data is detected by checking the timestamp. If duplicates are confirmed, the redundant items are deleted and the information related to the deleted duplicates is recorded.

[0117] Step 2.3, normalization processing, since the collected data have different dimensions and magnitudes, the data are mapped to a specific interval so that they can be compared and analyzed on the same scale.

[0118]

[0119] where x i (β) is the data after normalization, is the original data indicator value, Indicates the average value of all load data;

[0120] Step 3: Analyze the transformer stability margin from multiple angles and establish a reference sequence:

[0121] Transformer stability margin is not a single indicator, but encompasses multiple aspects, including voltage stability, thermal stability, short-circuit capacity, and winding compression force. Different stability margin indicators have different significance and emphasis on the safe and stable operation of power systems.

[0122] Considering the voltage stability margin:

[0123]

[0124] V min ≤V a ≤V max

[0125]

[0126] Q g =Q c +Q l

[0127] V M Indicates the voltage stability margin, V a represents the actual voltage, V r Indicates rated voltage, V min Indicates the voltage lower limit, V max Indicates the voltage upper limit, Q g Indicates the reactive power generated by the reactive power source, Q c Indicates load consumption, Q l represents reactive power loss, and X represents system reactance.

[0128] Considering thermal stability margin:

[0129] θ M =θ al -θ a

[0130] θ a ≤θ max

[0131] P h =I 2 R

[0132]

[0133] θ M represents the thermal stability margin, θ a Indicates the actual temperature rise of the winding, θ al Indicates the allowable temperature rise, θ max Indicates the maximum temperature allowed by the insulation material, P h It represents the heating power of the winding, I represents the load current, R represents the winding resistance, and C represents the equivalent thermal capacity of the transformer.

[0134] Considering short-circuit capacity margin:

[0135]

[0136] I sc ≤I sc,max

[0137] I sc ≤I t,sc And t sc ≤t t,sc

[0138] I sc ≤I br，sc

[0139] S M Indicates the short-circuit capacity margin, Z s Represents the system short-circuit impedance, Z t Represents the transformer short-circuit impedance, I sc Indicates short-circuit current, V r Indicates rated voltage, I sc,max Indicates the maximum allowable short-circuit current, I t,sc Indicates the transformer short-circuit withstand current, t sc Indicates the short circuit duration, t t,sc Indicates the short-circuit duration allowed by the transformer, I br，sc Indicates the rated short-circuit breaking current of the circuit breaker; considering the winding compression force margin:

[0140]

[0141] P M is the winding compression force margin, P a is the winding compression force P min P is the lower limit of the normal range of winding compression force, max P is the upper limit of the normal range of winding compression force, o It is the ideal compression force value.

[0142] The reference sequence established in step 3 includes: taking the transformer stability margin data sequence as the reference sequence, the voltage stability margin Thermal stability margin Short-circuit capacity margin Winding compression force margin

[0143] Step 4: Obtain the unweighted correlation coefficient through the improved difference quotient grey relational analysis DQ-GRA:

[0144] The source-load power characteristic data sequence is used as a comparison sequence, including wind power X1 = {x1(1), x1(2), ..., x1(n)}, photovoltaic power X2 = {x2(1), x2(2), ..., x2(n)}, load power X3 = {x3(1), x3(2), ..., x3(n)}, etc.

[0145] Compute the discrepancy matrix and the quotient matrix:

[0146] Calculate reference sequence and compare sequence X i The difference between the index values ​​​​is obtained by the difference matrix:

[0147]

[0148] Where Δx i (j) represents the difference between the i-th comparison sequence and the reference sequence in the j-th index;

[0149] Calculate reference sequence and compare sequence X i The quotient values ​​of each indicator value are used to obtain the quotient matrix:

[0150]

[0151] Where Δx i (j) represents the quotient value of the i-th comparison sequence and the reference sequence at the j-th index;

[0152] Calculate the grey relational coefficient:

[0153] Grey relational coefficient of geometric similarity based on difference matrix:

[0154]

[0155] Grey relational coefficient based on the numerical proximity of the quotient matrix:

[0156]

[0157] Step 5: Apply the improved Grey-DEMATEL method to obtain the indicator weights and calculate the comprehensive grey relational degree:

[0158] Step 5.1, form an expert evaluation team: The evaluation team is composed of experts in power-related fields who can accurately judge the mutual influence relationship between various indicators.

[0159] Step 5.2, expert evaluation and data collection: using semantic variables, such as “very high impact VH”, “high impact H”, “low impact L”, “very low impact VL”, and “no impact NI”.

[0160] Step 5.3, construct the direct grey relationship matrix:

[0161] The semantic variables evaluated by experts are converted into gray numbers according to the gray language scale in the table, and a direct impact matrix is ​​established. Refer to the table below:

[0162] Exact value Semantic variables Gray Number 4 Very High Impact VH [0.75，1] 3 High Impact H [0.5，0.75] 2 Low Impact [0.25，0.5] 1 Very Low Impact VL [0，0.25] 0 No impact NI [0，0]

[0163] Direct impact matrix:

[0164] Where n is the number of indicators, is the gray number of the evaluation of the αth expert on the direct impact of indicator i on indicator j;

[0165] Step 5.4, calculate the clear relationship matrix: use the improved CFCS method to convert the gray value into the clear value to obtain the clear direct influence matrix.

[0166] First, normalize the gray value:

[0167]

[0168] are the upper and lower limits of the normalized gray number;

[0169] Calculate the total normalized clarity value:

[0170]

[0171] Finally calculate the final clarity value:

[0172]

[0173] Step 5.5, calculate the direct impact matrix:

[0174]

[0175] Where f is the total number of experts;

[0176] Step 5.6, determine the comprehensive influence matrix; normalize the real direct influence matrix C, that is, divide each element value in the matrix by the maximum value of the direct influence matrix to obtain the normalized direct relationship matrix S, and then calculate the comprehensive influence matrix X:

[0177]

[0178] Step 5.7, calculate the influence, influence, centrality and cause:

[0179] Influence degree indicates the comprehensive influence of indicator j on other indicators:

[0180]

[0181] The degree of influence indicates the comprehensive influence of indicator j on other indicators:

[0182]

[0183] Centrality reflects the position and importance of an indicator in the system. The greater the centrality, the more important the indicator:

[0184] M j =p j +q j

[0185] Cause degree reflects the net impact of an indicator on the system. The greater the cause degree, the greater the impact on other indicators:

[0186] N i =p j -q j

[0187] Step 5.8, calculate indicator weights:

[0188] Calculate the weight of each indicator based on the centrality:

[0189]

[0190] Step 5.9, calculate the comprehensive grey relational degree:

[0191]

[0192] Step 5.10: Determine the priority of influencing factors based on the comprehensive grey correlation degree.

[0193] The embodiments described above are merely descriptions of preferred embodiments of the present invention and are not intended to limit the scope of the present invention. Without departing from the spirit of the present invention, various deformations, modifications, and substitutions made to the technical solutions of the present invention by ordinary technicians in this field should fall within the scope of protection determined by the claims of the present invention.

Claims

1. A method for analyzing the influence of source-load power characteristics on transformer stability margin, characterized in that: The following steps are involved: Step 1: Obtain raw data from EMS and SCADA systems; Step 2: Preprocess the original data by removing abnormal data and normalizing it; Step 3: Analyze the transformer stability margin from multiple angles and establish a reference sequence; Step 4: Obtain the unweighted correlation coefficient through the improved difference quotient grey correlation analysis; Step 5: Apply the improved Grey-DEMATEL method to obtain the index weights and calculate the comprehensive grey relational degree; Step 5 includes: Step 5.1: Form an expert evaluation team; Step 5.2, expert evaluation and data collection; Step 5.3, construct the direct grey relationship matrix: The semantic variables evaluated by experts are converted into gray numbers according to the gray language scale in the table, and the direct impact matrix is ​​established: Where n is the number of indicators, is the gray number of the evaluation of the αth expert on the direct impact of indicator i on indicator j; Step 5.4, calculate the clear relationship matrix: Normalize the gray value: are the upper and lower limits of the normalized gray number; Calculate the total normalized clarity value: Calculate the final clarity value: Step 5.5, calculate the direct impact matrix: Where f is the total number of experts; Step 5.6, determine the comprehensive impact matrix: Normalize the real direct influence matrix C by dividing each element value in the matrix by the maximum value of the direct influence matrix to obtain the normalized direct relationship matrix S, and then calculate the comprehensive influence matrix X: Step 5.7, calculate the influence, influence, centrality and cause: Impact: Affected degree: Centrality: M j =p j +q j Cause degree: N i =p j -q j Step 5.8, calculate indicator weights: Calculate the weight of each indicator based on the centrality: Step 5.9, calculate the comprehensive grey relational degree: Step 5.10: Determine the priority of influencing factors based on the comprehensive grey correlation degree.

2. The method for analyzing the influence of source-load power characteristics on transformer stability margin according to claim 1, characterized in that: Step 2 includes: Step 2.1, missing data filling: Interpolation method: where x t and x t+1 are the data at time t and time t+1, respectively, x t+k is the missing data at time t+k, and n is the total amount of data; Multiple imputation method: Assume that the variable to be interpolated is X, and its posterior distribution of observed data is given by the following formula: P(X|X obs )=∫P(X|X obs ,X min )P(Y mis | Xobs ) d X where X obs represents the observed data, X mis Represents missing data, and the posterior predictive distribution P(X mis |X obs ) Perform m sets of conditional independent extraction to obtain repeated interpolation sets, and calculate the repeated interpolation value from each interpolated data set The final interpolated value of X: Step 2.2, duplicate handling: Detect duplicate data by checking the timestamp. If duplicates are found, delete the redundant items and record the deleted duplicates. Step 2.3, normalization, maps the original data to a specific interval so that they can be compared and analyzed on the same scale: where x i (β) is the data after normalization, is the original data indicator value, Indicates the average value of all load data.

3. The method for analyzing the influence of source-load power characteristics on transformer stability margin according to claim 1, characterized in that: In step 3, the transformer stability margin is analyzed from multiple angles, including voltage stability margin, thermal stability margin, short-circuit capacity margin, and winding compression force margin, where: Voltage stability margin: In min ≤V a ≤V max Q g =Q c +Q l V M Indicates the voltage stability margin, V a Indicates the actual voltage, V r Indicates rated voltage, V min Indicates the voltage lower limit, V max represents the voltage upper limit, Qg represents the reactive power generated by the reactive power source, Qc represents the reactive power consumed by the load, Ql represents the reactive power loss, and X represents the system reactance; Thermal stability margin: i M =θ al -θ a i a ≤θ max P h =I 2 R θ M represents the thermal stability margin, θ a Indicates the actual temperature rise of the winding, θ al Indicates the allowable temperature rise, θ max Indicates the maximum temperature allowed by the insulation material, P h Indicates the heating power of the winding, I indicates the load current, R indicates the winding resistance, and C indicates the equivalent thermal capacity of the transformer; Short circuit capacity margin: I sc ≤I sc,max I sc ≤I t,sc And t sc ≤t t,sc I sc ≤I br，sc S M Indicates the short-circuit capacity margin, Z s Represents the system short-circuit impedance, Z t Represents the transformer short-circuit impedance, I sc Indicates short-circuit current, V r Indicates rated voltage, I sc,max Indicates the maximum allowable short-circuit current, I t,sc Indicates the transformer short-circuit withstand current, t sc Indicates the short circuit duration, t t,sc Indicates the short-circuit duration allowed by the transformer, I br，sc Indicates the rated short-circuit breaking current of the circuit breaker; Winding compression force margin: P M is the winding compression force margin, P a is the winding compression force P min P is the lower limit of the normal range of winding compression force, max P is the upper limit of the normal range of winding compression force, o is the ideal compression force value; The reference sequence established in step 3 includes: taking the transformer stability margin data sequence as the reference sequence, the voltage stability margin Thermal stability Short-circuit capacity margin Winding compression force margin 4. The method for analyzing the influence of source-load power characteristics on transformer stability margin according to claim 3, characterized in that: Step 4 includes: The source-load power characteristic data sequence is used as a comparison sequence, including wind power X1 = {x1(1), x1(2), ..., x1(n)}, photovoltaic power X2 = {x2(1), x2(2), ..., x2(n)}, and load power X3 = {x3(1), x3(2), ..., x3(n)}; Compute the discrepancy matrix and the quotient matrix: Calculate reference sequence and compare sequence X i The difference between the index values ​​​​is obtained by the difference matrix: Where Δx i (j) represents the difference between the i-th comparison sequence and the reference sequence in the j-th index; Calculate reference sequence and compare sequence X i The quotient values ​​of each indicator value are used to obtain the quotient matrix: Where Δx i (j) represents the quotient value of the i-th comparison sequence and the reference sequence at the j-th index; Calculate the grey relational coefficient: Grey relational coefficient of geometric similarity based on difference matrix: Grey relational coefficient based on the numerical proximity of the quotient matrix:

Images