Power supply station comprehensive line loss rate evaluation method based on dynamic calibration mechanism

Abstract

The invention relates to the technical field of power supply, in particular to a power supply station comprehensive line loss rate evaluation method based on a dynamic calibration mechanism. According to the technical scheme, the method for evaluating the comprehensive line loss rate of the power supply station based on the dynamic calibration mechanism comprises the following steps: S1, acquiring historical line loss data of a distribution line in a power supply station area in the first 12 months; s2, performing two rounds of screening on the data, including eliminating line data of which the commissioning time is less than or equal to 2 months, and screening line loss rate abnormal data through a dynamic threshold method; s3, constructing a dynamic threshold interval by adopting a modified Z-score algorithm, performing day-level verification on the abnormal line loss rate data, and identifying suspected electric quantity abnormal points; and S4, performing electric quantity reduction calculation on the abnormal data, wherein the electric quantity reduction calculation comprises meter bottom integrity verification, rate consistency verification and electric quantity rationality correction. By introducing a dynamic calibration mechanism and constructing a comprehensive evaluation index system, the accuracy, practicability and management efficiency of the evaluation method for the comprehensive line loss rate of the power supply station are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of power supply, and in particular to a method for evaluating the comprehensive line loss rate of a power supply station based on a dynamic calibration mechanism. Background Art

[0002] In the existing power grid line loss management technology system, there is a structural problem that needs to be solved urgently: in the process of pursuing the optimization of specific performance indicators, various professional departments often adopt local optimization strategies, resulting in a significant trade-off effect between line loss management indicators.

[0003] From a technical perspective, the current line-loss assessment system relies on a "four-part line-loss" management model, which implements step-by-step control based on zone, voltage, line, and substation. This assessment model lacks a multi-dimensional, comprehensive evaluation mechanism for line and substation loss indicators, making it difficult to achieve global optimization and coordinated control of power grid line-loss management. This localized optimization strategy not only compromises the scientific nature of line-loss management decisions but also reduces the economic efficiency of the power grid.

[0004] In the process of calculating the comprehensive line loss rate, meticulous data cleaning and preprocessing are necessary to accurately reflect the actual line loss situation. Specifically, in the daily operation of power lines, load reversals often occur due to temporary repairs and planned maintenance. This can cause a discrepancy between the power consumption statistics at the power supply gateway and the actual power supply to the line, thus affecting the accuracy of the line loss rate.

[0005] Secondly, we also need to consider the impact of sudden changes in electricity sales on line loss rates. To accurately identify and analyze such changes, we need to comprehensively consider multiple key factors, including changes in user profile information, the integrity of meter readings, the accuracy of meter replacement records, and seasonal load fluctuations. This helps determine whether there is a sudden change in electricity sales and how to correct the sudden change.

[0006] Patent publication number CN119848605A discloses a method, device, equipment, and medium for managing line loss in a substation. The method includes determining the total power supply and total power consumption of a target substation, and determining the actual line loss in the target substation based on the difference between the total power supply and total power consumption. If the actual line loss is greater than the theoretical line loss, determining whether there is an abnormality in power consumption based on the power consumption information of users in the target substation. The theoretical line loss is determined based on the power supply radius, line length, and number of meters in the target substation. If there is an abnormality in power consumption, generating an early warning based on the abnormal power consumption information and pushing the warning information to the target management terminal for processing. While the method primarily addresses the inefficiency of manual inspections, it still has shortcomings in terms of line loss data processing. Summary of the Invention

[0007] The present invention proposes a comprehensive line loss rate evaluation method for power supply stations based on a dynamic calibration mechanism, which solves the problem in the prior art that the power statistics at the power supply gateway deviate from the actual power supplied to the line, thereby affecting the accuracy of the line loss rate.

[0008] In order to solve the above technical problems, the technical solution adopted by the present invention is:

[0009] The comprehensive line loss rate evaluation method of the power supply station based on the dynamic calibration mechanism includes the following steps:

[0010] Step S1: Obtain historical line loss data of the power distribution lines in the power supply station area for the previous 12 months;

[0011] Step S2: Perform two rounds of data screening, including excluding line data with an operation time of ≤ 2 months, and screening abnormal line loss rate data using a dynamic threshold method;

[0012] Step S3: Use the modified Z-score algorithm to construct a dynamic threshold interval, perform daily verification on abnormal line loss rate data, and identify suspected power anomalies;

[0013] Step S4: Perform power restoration calculation on abnormal data, including meter bottom integrity check, rate consistency check and power rationality correction;

[0014] Step S5: Restore and calculate the power consumption of the power supply gateway of the cross-unit packaging line to correct the input / output deviation;

[0015] Step S6: Calculate the comprehensive line loss rate of each sub-line and the comprehensive line loss rate of each sub-station.

[0016] Furthermore, the two rounds of data screening in step S2 include:

[0017] In the first round of screening, we eliminated lines that were put into operation within two months and had abnormal data due to low load, large fluctuations, or the influence of system file parameter transmission and model configuration;

[0018] In the second round of screening, abnormal line data with line loss rate > 7% and power loss > 5000kWh are eliminated, and those that meet the whitelist rules are retained.

[0019] Furthermore, in step S3, a modified Z-score algorithm is used to construct a dynamic threshold, and the formula is:

[0020]

[0021] Where M is the median, MAD is the median absolute deviation, and k = 0.6745;

[0022] The dynamic threshold confidence level is set to 99.7%, corresponding to Z = 3, and the threshold range is calculated as [M-Δ, M+Δ], where Δ = 3 × k × MAD;

[0023] Line loss data that exceeds the threshold is marked as abnormal, and daily analysis of the line loss rate is triggered.

[0024] Furthermore, after the first and second rounds of screening, the abnormal monthly line loss data is screened and restored. The process is as follows:

[0025] For lines without packaging marks, if the monthly line loss rate exceeds the dynamic threshold range, the daily line loss rate will be checked daily;

[0026] For the packaged lines within the same power supply station, verify whether the total line loss rate after packaging is within the threshold range of the participating lines; when it is necessary to calculate the comprehensive line loss rate of each line, it is also necessary to verify whether the daily line loss rate of each line is within its respective threshold range.

[0027] For inter-unit packaged lines, directly verify whether the daily line loss rate of each line is within the respective threshold range.

[0028] Furthermore, 5 days before and after the packaging are selected as the reference period. The daily line loss rate of each day selected should be within the threshold range. If it does not meet the threshold, another day will be selected in the forward or backward direction to calculate the average input power, output power and power sales;

[0029] Based on the average line loss rate η in the base period A , restore the input power of the packaging day:

[0030]

[0031] Among them, E 售电，A(t) It refers to the actual electricity sales of the line on the package day, that is, the sum of the forward electricity of high-voltage users and the forward electricity of the public transformer area;

[0032] The monthly threshold power consumption is corrected based on the difference between the restored value and the system statistical value.

[0033] Furthermore, the verification method in step S4 includes:

[0034] Check the integrity of the bottom of the table: check if the bottom of the table is missing or reversed;

[0035] Ratio consistency check: verify the ratio change before and after the meter is replaced;

[0036] Power rationality check: abnormalities are determined based on the historical daily average power value ±3 times the standard deviation and the user capacity limit.

[0037] Furthermore, the abnormal power correction method obtained by verification includes:

[0038] Correction of historical daily average power consumption based on sliding window (window length n = 7 days);

[0039] Over / under capacity correction based on user capacity limits;

[0040] Estimate revisions based on the rate of change of the table base;

[0041] The final correction value is the average of the above three options.

[0042] Furthermore, in step S6, in the calculation of the line loss rate of the branch line:

[0043] Input power statistics include forward power at the line gateway, high-voltage distributed reverse power, and low-voltage distributed reverse power;

[0044] Output power statistics include reverse power at the line junction, forward power for high-voltage office use, and forward power for low-voltage office use.

[0045] Furthermore, in the calculation of the comprehensive line loss rate of the branch line:

[0046] Input power statistics include forward power at the line gateway, high-voltage distributed reverse power, and low-voltage distributed reverse power;

[0047] Output power statistics include reverse power at the line junction, forward power for high-voltage office use, and forward power for low-voltage office use.

[0048] The electricity sales statistics include the forward electricity sales to high-voltage users and the electricity sales to low-voltage users on the line;

[0049] Calculation of comprehensive line loss rate of the affiliated power supply station:

[0050] Input power statistics include the forward power of all line gateways within a power supply station and all high-voltage and low-voltage distributed on-grid power on the corresponding lines;

[0051] Output power statistics include reverse power at all line junctions within a power supply station and all high-voltage and low-voltage office power consumption on the corresponding lines;

[0052] The electricity sales statistics include the forward electricity sales to all high-voltage users and the electricity sales to all low-voltage users corresponding to all lines within a power supply station.

[0053] Furthermore, it also includes:

[0054] Output power supply station comprehensive line loss rate evaluation report, including line / substation anomaly location, loss reduction strategy recommendations and resource allocation optimization plan;

[0055] The historical data threshold is updated through a dynamic calibration mechanism to form a closed-loop feedback.

[0056] The positive effects of the present invention are: by introducing a dynamic calibration mechanism and building a comprehensive evaluation index system, the accuracy, practicality and management efficiency of the comprehensive line loss rate evaluation method of the power supply station are significantly improved. The specific technical effects are reflected in the following aspects:

[0057] First, data accuracy is improved: through meticulous data cleaning and preprocessing steps, especially the power return processing for situations such as load reversal and sudden changes in power sales, the accuracy and reliability of line loss rate calculation are significantly improved.

[0058] The second is the dynamic calibration mechanism: the introduction of dynamic threshold construction and reasonable interval verification enables the line loss rate evaluation to adapt to changes in different seasons, policy environments and electricity loads, enhancing the flexibility and adaptability of the evaluation.

[0059] The third is the improvement of management efficiency: the calculation and evaluation of the comprehensive line loss rate provides the power supply station with comprehensive and intuitive line loss management indicators, which helps to quickly locate problem lines and improve the efficiency and quality of grassroots line loss management.

[0060] Fourth, loss reduction strategy optimization: This method can inspire power supply stations to actively seek loss reduction strategies, implement precise policies through data analysis and mining, effectively reduce line losses, and improve the economic operation efficiency of the power grid.

[0061] Fifth, resource allocation optimization: comprehensive line loss assessment provides a scientific basis for the power supply station to rationally allocate resources. Through data analysis and mining, precise policy implementation can effectively reduce line and substation losses and improve the economic operation efficiency of the power grid.

[0062] Sixth, strengthening decision-making support: providing decision makers with more accurate and comprehensive data support, which will help to scientifically plan, build and operate power systems, and promote the development of power grid management towards intelligence and refinement. DETAILED DESCRIPTION

[0063] The following will be combined with the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the embodiments described 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 any creative efforts are within the scope of protection of the present invention.

[0064] Example 1

[0065] The comprehensive line loss rate evaluation method of the power supply station based on the dynamic calibration mechanism includes the following steps:

[0066] Step S1: Obtain historical line loss data of the power distribution lines in the power supply station area for the previous 12 months;

[0067] Step S2: Perform two rounds of data screening, including excluding line data with an operation time of ≤ 2 months, and screening abnormal line loss rate data using a dynamic threshold method;

[0068] Step S3: Use the modified Z-score algorithm to construct a dynamic threshold interval, perform daily verification on abnormal line loss rate data, and identify suspected power anomalies;

[0069] Step S4: Perform power restoration calculation on abnormal data, including meter bottom integrity check, rate consistency check and power rationality correction;

[0070] Step S5: Restore and calculate the power consumption of the power supply gateway of the cross-unit packaging line to correct the input / output deviation;

[0071] Step S6: Calculate the comprehensive line loss rate of each sub-line and the comprehensive line loss rate of each sub-station.

[0072] In this example, the previous 12 months of historical line loss data for all lines within the power grid zone are obtained. This includes the relationship between power stations, lines, substations, and users, as well as line loss rates, power consumption, meter readings, and packaging records. Data cleaning at the power supply gateway of distribution lines primarily involves cleaning and authenticating line loss data for abnormal monthly losses and lines marked as inter-station packaging, thereby improving data quality.

[0073] The two rounds of data screening in step S2 include:

[0074] In the first round of screening, we eliminated lines that were put into operation within two months and had abnormal data due to low load, large fluctuations, or the influence of system file parameter transmission and model configuration;

[0075] In the second round of screening, abnormal line data with line loss rate > 7% and power loss > 5000kWh are eliminated, and those that meet the whitelist rules are retained.

[0076] 1. Considering that the line loss rate will vary according to the changes in electricity load in different seasons and policy environments, and the seasonal load change trends are similar, the line loss rate data of the past 12 months are selected to obtain the original line loss data.

[0077] 2. Due to the large amount of line loss rate data for lines and substations, and the fact that line loss-related data may be abnormal due to data transmission, human intervention, special operating modes, etc., it is necessary to clean the line loss rate data for the selected 12 months.

[0078] First-round screening criteria: When a line is in operation for two months or less, anomalies are prone to occur due to low load, large fluctuations, or factors such as system file parameter transmission and model configuration. Therefore, such lines are not included in the calculation of the comprehensive line loss rate of the power supply station distribution network. In other words, the 12-month line information of abnormal lines is directly eliminated to obtain the historical line loss rate of each line after the first round of screening.

[0079] Second round of screening conditions: In this embodiment, based on the actual line loss management situation of the sub-area, the line loss rate generally does not exceed 7%, and the power loss does not exceed 5000kWh. Since the impact of metering accuracy error on line loss is limited, the line loss rate of the line will generally not be less than -1%. That is, when the line loss rate of the line is greater than 7% and the power loss is greater than 5000kWh, it is determined that the line loss rate fluctuates abnormally, and such data needs to be eliminated. In special cases, if the line loss meets the evaluation criteria of the State Grid Corporation's line whitelist, the line loss rate is determined to be normal and does not need to be eliminated. After this round of cleaning, the historical line loss rate of each line after the second round of screening is obtained.

[0080] Example 2

[0081] On the basis of Example 1, in step S3, a modified Z-score algorithm is used to construct a dynamic threshold, and the formula is:

[0082]

[0083] Where M is the median, MAD is the median absolute deviation, and k = 0.6745;

[0084] The dynamic threshold confidence level is set to 99.7%, corresponding to Z = 3, and the threshold range is calculated as [M-Δ, M+Δ], where Δ = 3 × k × MAD;

[0085] Line loss data that exceeds the threshold is marked as abnormal, and daily analysis of the line loss rate is triggered.

[0086] Calculate the reasonable threshold of line loss for each line after the second round of screening. The calculation method is as follows:

[0087] (1) Traditional Z-score standardization formula

[0088]

[0089] Where: X is the original data point, μ is the data mean, and σ is the standard deviation.

[0090] (2) Modify the Z-score improvement points

[0091] Use median (M) and absolute deviation (MAD) instead of mean and standard deviation to enhance noise resistance:

[0092] M=median(X1,X2,...,X n )

[0093] MAD=median(|X1-M|,|X2-M|,...,|X n -M|)

[0094] (3) Modify the Z-score calculation formula

[0095]

[0096] Where k is a correction factor, typically set to 0.6745, to maintain probability consistency with the traditional Z-score. In practice, the k value can be adjusted based on specific scenarios to balance sensitivity and stability.

[0097] (4) Reasonable interval construction

[0098] Set the confidence level (usually 99.7% corresponding to Z = 3), set the dynamic threshold to [M-Δ, M+Δ], and calculate the dynamic threshold:

[0099] Δ=3×k×MAD

[0100] For new data x new Make a judgment:

[0101]

[0102] When the monthly line loss rate is normal, it means that the electricity calculation data is reliable, and the data can be directly extracted and prepared for participation in the comprehensive line loss rate calculation.

[0103] When there is an abnormality in the monthly line loss rate, it is necessary to identify the abnormality one by one on a daily basis, perform power restoration calculations, and use the restored daily cumulative power instead of the original monthly power to participate in the comprehensive line loss rate calculation.

[0104] After the first and second rounds of screening, the abnormal monthly line loss data is screened and restored. The process is as follows:

[0105] For lines without packaging marks, if the monthly line loss rate exceeds the dynamic threshold range, the daily line loss rate will be checked daily;

[0106] For the packaged lines within the same power supply station, verify whether the total line loss rate after packaging is within the threshold range of the participating lines; when it is necessary to calculate the comprehensive line loss rate of each line, it is also necessary to verify whether the daily line loss rate of each line is within its respective threshold range.

[0107] For inter-unit packaged lines, directly verify whether the daily line loss rate of each line is within the respective threshold range.

[0108] Line bundling refers to a temporary load shedding measure implemented when a sudden on-site failure or maintenance prevents normal power supply to users. In this scenario, the original line topology model is not adjusted, and electricity sales statistics retain the existing line-to-line relationship. This can result in uneven power supply on different lines, leading to fluctuations in line loss rates, which in turn impact the statistical analysis of line loss rates. The following analysis will focus on three factors: bundling, cross-unit (across power stations), and monthly line loss rates.

[0109] (1) No package line processing

[0110] Case 1: If the line has no packaging mark and its monthly line loss rate meets x new ∈[M-Δ,M+Δ], the line loss rate of the line is determined to be normal and no further analysis is required.

[0111] Case 2: If the line has no packaging mark, but its monthly line loss rate It is necessary to conduct a rational analysis of the daily loss rate one by one:

[0112] When the line loss rate satisfies x new ∈[M-Δ,M+Δ], then the corresponding daily line loss rate of the line is determined to be normal.

[0113] Daily line loss rate The corresponding daily line loss rate of the line will be marked as suspected power abnormality.

[0114] (2) Packing line processing within the same power supply station

[0115] Case 3: Lines within the same power supply station are bundled, and the total monthly line loss rate of the bundled lines meets x new ∈[min(M1-Δ1,M2-Δ2,…),max(M1+Δ1,M2+Δ2,…)], the corresponding daily line loss rate is considered normal and can be directly included in the comprehensive line loss rate calculation of the substation. If it participates in the comprehensive line loss rate calculation of the substation, the line loss of each participating line needs to be restored using the same method as "Cross-unit Line Packaging Processing".

[0116] Case 4: Lines within the same power supply station are bundled, and the total monthly line loss rate of the bundled lines is It is necessary to conduct a rational analysis on the daily line loss rate of the lines involved in the package one by one:

[0117] During the packaging period: when the total daily line loss rate of the line after packaging meets x new∈[min(M1-Δ1,M2-Δ2,…),max(M1+Δ1,M2+Δ2,…)], the corresponding daily line loss rate of the line participating in the package is determined to be normal. If it is not satisfied, the corresponding daily line loss rate of the line participating in the package is marked as suspected abnormal power.

[0118] During the non-packaging period: the daily line loss rate of the lines participating in the packaging meets x new ∈[M-Δ,M+Δ], the line loss rate of the corresponding daily line is considered normal. If it is not satisfied, it is marked as a suspected power anomaly.

[0119] (3) Cross-unit package line processing

[0120] Case 5: Directly assess whether the daily line loss rates of the participating lines are within the reasonable threshold of their respective line loss rates:

[0121] During the packaging period: when the total daily line loss rate of the line after packaging meets x new ∈[min(M1-Δ1,M2-Δ2,…),max(M1+Δ1,M2+Δ2,…)], the corresponding daily line loss rate of the lines participating in the package is determined to be normal, and the line gateway power needs to be restored and calculated; if it is not satisfied, the corresponding daily line loss rate of the lines participating in the package is marked as suspected power abnormality.

[0122] During the non-packaging period: the daily line loss rate of the lines participating in the packaging meets x new ∈[M-Δ,M+Δ], the line loss rate of the corresponding daily line is considered normal. If it is not satisfied, it is marked as a suspected power anomaly.

[0123] In the above five situations, when the daily and monthly line loss rates meet the State Grid Corporation's line whitelist rules, there is no need to verify according to the reasonable range and the line loss rate can be directly determined to be normal.

[0124] In summary, there are two types of anomalies that require further restoration and processing. One is suspected abnormal daily power loss on packaged or non-packaged lines, and the other is cross-unit packaging that requires restoration and calculation of the power supply gateway power of the involved lines.

[0125] Step 3: Verify and restore suspected abnormal power loss on daily lines

[0126] Verification of suspected abnormal power consumption in line daily loss involves verifying the integrity and power fluctuation of four types of meters: power supply gateway, distributed grid meter, high-voltage user meter, and substation meter. The abnormal power consumption is then restored and calculated. The specific steps are as follows:

[0127] (1) Abnormal power verification rules

[0128] When there is no meter replacement, daily power consumption calculation:

[0129] E计算,t =(B down,t -B up,t )×R t

[0130] When the meter is replaced, the daily power calculation is: Assume that the multiplication rate before the meter is replaced is R 旧 , after changing the table, the magnification is R 新 .

[0131] E 计算,t =(B down,t旧 -B up,t旧 )×R 旧 +(B down,t新 -B up,t新 )×R 新

[0132] Verification 1: Integrity of bottom of table

[0133] Suppose the upper base of a meter on abnormal day t is B up,t , the bottom of the table below is B down,t , the verification conditions are:

[0134]

[0135] If the bottom of the watch is missing or inverted, it is marked as abnormal bottom of the watch.

[0136] Verification 2: Ratio consistency

[0137] Assume that the abnormal daily rate is R t , the rate at the beginning of the month (1st) is R 月初 , the monitoring is not later than the abnormal day t after the meter replacement rate R 新 , the verification conditions are:

[0138] On or before the tth day of the month, there is no record of changing the meter or the magnification rate. t =R 月初 ;

[0139] On or before day t of the month, there is a record of changing the meter or the magnification, R t =R 新 ;

[0140] If neither of the above two conditions is met, it will be marked as a magnification abnormality.

[0141] Verification 3: Power rationality verification

[0142]

[0143] Where: E 历史,t is the average daily electricity consumption in the previous 12 months, σ 历史 is the standard deviation of the average daily electricity consumption in the previous 12 months, S 用户is the user's transformer capacity (kVA), and cos is the power factor (the default value is 0.9).

[0144] When the electricity quantity of a meter does not meet one of the above three rationality verification conditions, it is determined that the corresponding equipment measurement is abnormal and abnormal electricity recovery calculation is required.

[0145] (2) Abnormal power recovery calculation

[0146] Solution 1: Correction method based on historical average daily electricity consumption

[0147] For abnormal daily electricity E 异常,t , the correction value is calculated using the sliding window method:

[0148]

[0149] Where n is the window length (recommended to be 7 days, covering a weekly period). α is the weight coefficient, generally set to 0.3, indicating a greater reliance on historical data. When the abnormal day is a holiday or special event, the weight of historical data needs to be reduced, and a value of 0.1 is recommended.

[0150] Solution 2: Correction method based on user capacity limit

[0151] Excess capacity correction: If E 计算,t >S 用户 ×24×cos, then the correction is:

[0152] E 修正,t =min(E 计算,t , S 用户 ×24×cos)

[0153] Low battery correction: If E 计算,t <0.1×E 历史,t , then it is corrected to:

[0154]

[0155] Solution 3: Power correction based on meter bottom change rate

[0156] According to the historical table bottom change rate r 表底 Estimate:

[0157] B down,t '=B up,t ×(1+r 表底 )

[0158]

[0159] Since the lower bottom of the chart on day t-1 is the upper bottom of the chart on day t, when the upper bottom is lost, the upper bottom of the chart on the day before the abnormal day t is used for calculation, and so on.

[0160] E 修正,t =(B down,t '-B up,t )*R t

[0161] The final corrected power is the average of the correction results of Scheme 1, Scheme 2, and Scheme 3.

[0162] Select 5 days before and after packaging as the reference period. The daily line loss rate of each day selected should be within the threshold range. If it does not meet the threshold, select another day in the previous or next day to replace it and calculate the average input power, output power and power sales;

[0163] Based on the average line loss rate η in the base period A , restore the input power of the packaging day:

[0164]

[0165] Among them, E 售电，A(t) It refers to the actual electricity sales of the line on the package day, that is, the sum of the positive electricity of high-voltage users and the positive electricity of the public transformer area.

[0166] The monthly threshold power consumption is corrected based on the difference between the restored value and the system statistical value.

[0167] The verification method in step S4 includes:

[0168] Check the integrity of the bottom of the table: check if the bottom of the table is missing or reversed;

[0169] Ratio consistency check: verify the ratio change before and after the meter is replaced;

[0170] Power rationality check: abnormalities are determined based on the historical daily average power value ±3 times the standard deviation and the user capacity limit.

[0171] The correction method for abnormal power obtained by verification includes:

[0172] Correction of historical daily average power consumption based on sliding window (window length n = 7 days);

[0173] Over / under capacity correction based on user capacity limits;

[0174] Estimate revisions based on the rate of change of the table base;

[0175] The final correction value is the average of the above three options.

[0176] When lines are bundled, the electricity sales model remains on the original line, but the line input electricity may be supplied to other line power equipment or supplied by other lines, affecting the input and output electricity statistics between power supply stations. Taking the bundling of two lines from two power supply stations as an example, the restoration calculation is as follows:

[0177] First, define the basic data

[0178] Line packaging period: date interval is T 打包 =[t start , t end ].

[0179] Benchmark data interval: Select T 5 days before packaging 前 =[t start -5,t start -1] and 5 days after unpacking T 后 =[t end +1,t end +5] a total of 10 days as the base period.

[0180] Base period statistics: average input power Average output power Average electricity sales They are the average values ​​of 10 input power, output power and sales power within the benchmark data interval.

[0181] Then, calculate the average line loss rate as follows,

[0182] Average power consumption during the base period:

[0183]

[0184] Average line loss rate during the base period:

[0185]

[0186] Average line loss rate during the base period:

[0187]

[0188] Then, the input power reduction value of the packaging day t is calculated

[0189] Enter the power recovery formula:

[0190]

[0191] Where: E 售电，A(t) It refers to the actual electricity sales of the line on the package day, that is, the sum of the positive electricity of high-voltage users and the positive electricity of the public transformer area.

[0192] Then, the power correction of the package day t

[0193] Input side correction:

[0194] ΔE 输入,A(t) =E 输入,还原,A (t)-E 输入,系统,A (t)

[0195] E 输入,系统,A (t) is the actual input power value of the system during the same period;

[0196] If ΔE 输入,A(t) >0: This means that the load of this line is transferred to other lines during the packaging period. It is necessary to add ΔE to the monthly forward power statistics. 输入,A(t) ;

[0197] If ΔE 输入,A(t) <0: This line has assumed the load of other lines during the packaging period, and ΔE needs to be subtracted when calculating the forward power statistics at the monthly threshold. 输入,A(t) .

[0198] Output side correction:

[0199] E 输出,还原,A (t) = E 办公,A (t)+E 反向,A (t)

[0200] ΔE 输出,A(t) =E 输出,还原,A (t)-E 输出,系统,A (t)

[0201] E 输出,系统,A (t) is the actual output power value of the system during the same period;

[0202] Among them E 办公,A (t) is the positive power consumption of high-voltage office on the line, which can be directly obtained from historical data, E 反向,A (t) is the reverse charge at the checkpoint.

[0203] If ΔE 输出,A(t) >0: ΔE needs to be added to the monthly reverse power statistics 输出,A(t) ;

[0204] If ΔE 输出,A(t) <0: ΔE needs to be subtracted when calculating the reverse electricity consumption at the monthly threshold 输出,A(t) .

[0205] Example 3

[0206] On the basis of Example 2: In step S6, in the calculation of the comprehensive line loss rate of the sub-lines:

[0207] Input power statistics include forward power at the line gateway, high-voltage distributed reverse power, and low-voltage distributed reverse power;

[0208] Output power statistics include reverse power at the line junction, forward power for high-voltage office use, and forward power for low-voltage office use.

[0209] The electricity sales statistics include the forward electricity sales to high-voltage users and the electricity sales to low-voltage users on the line;

[0210] Calculation of comprehensive line loss rate of the affiliated power supply station:

[0211] Input power statistics include the forward power of all line gateways within a power supply station and all high-voltage and low-voltage distributed on-grid power on the corresponding lines;

[0212] Output power statistics include reverse power at all line junctions within a power supply station and all high-voltage and low-voltage office power consumption on the corresponding lines;

[0213] The electricity sales statistics include the forward electricity sales to all high-voltage users and the electricity sales to all low-voltage users corresponding to all lines within a power supply station.

[0214] In the existing technology, the four-dimension calculation currently carried out by State Grid Corporation of China includes four line loss dimensions: zone, voltage, line, and substation, as shown below:

[0215] 1. Line loss rate statistics for each line

[0216] Line input power:

[0217] E 输入,线路 =E 关口正向,线路 +∑(E 高压分布式,反向 +E 台区总表,反向 )

[0218] Includes the forward power at the line gateway, the reverse power of all high-voltage distributed users under the line, and the reverse power of all substation total meters;

[0219] Line output power:

[0220] E 输出,线路 =E 关口反向,线路 +∑E 高压办公,正向

[0221] Includes the reverse direction of the line gateway and the forward direction of all high-voltage office electricity under the line;

[0222] Line electricity sales:

[0223] E 售,线路 =∑E 高压用户,正向 +∑E 台区总表,正向

[0224] Contains the forward electricity of all high-voltage users under the line and the forward electricity of the total meter in the substation area;

[0225] Line loss power:

[0226] E 损失,线路 =E 输入,线路 -E 输出,线路 -E售,线路

[0227] Line loss rate:

[0228]

[0229] 2. Line loss rate statistics in substation areas

[0230] Input power of the area:

[0231]

[0232] Contains the forward power of the total meter of the substation and the reverse power of all low-voltage distributed users under the substation; Substation output power:

[0233] E 输出,台区 =E 台区总表,反向 +∑E 低压办公,正向

[0234] Includes the reverse output of the total meter in the substation and the forward power consumption of all low-voltage office electricity in the substation; Substation electricity sales:

[0235] E 售,台区 =∑E 低压用户,正向

[0236] Contains the sum of electricity sales of all low-voltage non-office users in this substation;

[0237] Line loss electricity in the substation area:

[0238] E 损失,台区 =E 输入,台区 -E 输出,台区 -E 售,台区

[0239] Line loss rate in the substation area:

[0240]

[0241] Based on data from the State Grid Corporation of China's concurrent power and line loss management system, this paper proposes a comprehensive line loss rate for distribution networks based on power stations. This method can calculate the comprehensive line loss rate for a single line to evaluate the overall line loss performance of that line. It can also calculate the comprehensive line loss rate for all 10kV lines managed by a power station to evaluate the overall line loss performance of a particular station. In this embodiment, the improved comprehensive line loss rate for a power station is calculated as follows:

[0242] (1) Calculation of comprehensive line loss rate of a single line

[0243] Comprehensive input power:

[0244] E 输入,综合 =E 关口正向,线路 +∑E 高压分布式,反向+∑E 低压分布式,反向

[0245] Includes the forward power at the line gateway and all high-voltage and low-voltage distributed grid-connected power on the line;

[0246] Comprehensive output power:

[0247] E 输出,综合 =E 关口反向,线路 +∑E 高压办公,正向 +∑E 低压办公,正向

[0248] Including reverse power at the line junction, and all high-voltage and low-voltage office power consumption on the line;

[0249] Comprehensive electricity sales:

[0250] E 售,综合 =∑E 高压用户,正向 +∑E 低压用户,正向

[0251] Includes the forward electricity sales of all high-voltage users and the electricity sales of all low-voltage users on this line;

[0252] Comprehensive power consumption:

[0253] E 损失,综合 =E 输入,综合 -E 输出,综合 -E 售,综合

[0254] Comprehensive line loss rate of a single line:

[0255]

[0256] (2) Calculation of comprehensive line loss rate of power supply station

[0257] Comprehensive input power of power supply station:

[0258] E 输入,所综合 =∑(E 关口正向,线路 +E 高压分布式,反向 +E 低压分布式,反向 )

[0259] Includes the forward power of all line gateways in a power supply station and all high-voltage and low-voltage distributed grid-connected power on the corresponding lines;

[0260] Comprehensive output power of power supply station:

[0261] E 输出,所综合 =∑(E 关口反向,线路 +E 高压办公,正向 +E 低压办公,正向 )

[0262] Includes reverse power at all line junctions within a power supply station, and all high-voltage and low-voltage office power consumption on the corresponding lines;

[0263] Comprehensive electricity sales of power supply stations:

[0264] E 售,所综合 =∑E 高压用户,正向 +∑E 低压用户,正向

[0265] Includes all forward electricity sales to high-voltage users and all electricity sales to low-voltage users corresponding to all lines within a power supply station;

[0266] Comprehensive power consumption of power supply station:

[0267] E 损失,所综合 =E 输入,所综合 -E 输出,所综合 -E 售,所综合

[0268] Comprehensive line loss rate of power supply station:

[0269]

[0270] The comprehensive line loss rate of the power supply station calculated by the above formula can fully reflect the overall line loss level of the power supply station.

[0271] Based on the line loss rate obtained above, a comprehensive line loss rate evaluation report for the power supply station is output, including line / substation anomaly location, loss reduction strategy recommendations, and resource allocation optimization solutions;

[0272] The historical data threshold is updated through a dynamic calibration mechanism to form a closed-loop feedback.

[0273] The above-mentioned embodiments are described in a relatively detailed and specific manner, expressing preferred embodiments of the present invention. They are only used to illustrate the technical ideas and features of the present invention. Their purpose is to enable those skilled in the art to understand the contents of the present invention and implement them accordingly. However, they are not limited to the present invention alone, and the patent scope of the present invention cannot be limited solely by these embodiments. That is, any equivalent changes or modifications made to the spirit disclosed by the present invention, for researchers or technicians in this field, without departing from the structure of the present invention, local improvements within the system and changes and conversions between subsystems, etc., are still within the patent scope of the present invention.

Claims

1. A method for evaluating the comprehensive line loss rate of a power supply station based on a dynamic calibration mechanism, characterized in that: The following steps are involved: Step S1: Obtain historical line loss data of the power distribution lines in the power supply station area for the previous 12 months; Step S2: Perform two rounds of data screening, including excluding line data with an operation time of ≤ 2 months, and screening abnormal line loss rate data using a dynamic threshold method; Step S3: Use the modified Z-score algorithm to construct a dynamic threshold interval, perform daily verification on abnormal line loss rate data, and identify suspected power anomalies; Step S4: Perform power restoration calculation on abnormal data, including meter bottom integrity check, rate consistency check and power rationality correction; Step S5: Restore and calculate the power consumption of the power supply gateway of the cross-unit packaging line to correct the input / output deviation; Step S6: Calculate the comprehensive line loss rate of each sub-line and the comprehensive line loss rate of each sub-station.

2. The method for evaluating the comprehensive line loss rate of a power supply station based on a dynamic calibration mechanism according to claim 1 is characterized in that: The two rounds of data screening in step S2 include: In the first round of screening, we eliminated lines that were put into operation within two months and had abnormal data due to low load, large fluctuations, or the influence of system file parameter transmission and model configuration; In the second round of screening, abnormal line data with line loss rate > 7% and power loss > 5000kWh are eliminated, and those that meet the whitelist rules are retained.

3. The method for evaluating the comprehensive line loss rate of a power supply station based on a dynamic calibration mechanism according to claim 1 is characterized in that: In step S3, the modified Z-score algorithm is used to construct a dynamic threshold, and the formula is: Where M is the median, MAD is the median absolute deviation, and k = 0.6745; The dynamic threshold confidence level is set to 99.7%, corresponding to Z = 3, and the threshold range is calculated as [M-Δ, M+Δ], where Δ = 3 × k × MAD; Line loss data that exceeds the threshold is marked as abnormal, and daily analysis of the line loss rate is triggered.

4. The method for evaluating the comprehensive line loss rate of a power supply station based on a dynamic calibration mechanism according to claim 2 is characterized in that: After the first and second rounds of screening, the abnormal monthly line loss data is screened and restored. The process is as follows: For lines without packaging marks, if the monthly line loss rate exceeds the dynamic threshold range, the daily line loss rate will be checked daily; For the bundled lines within the same power supply station, verify whether the total line loss rate after bundling is within the threshold range of the participating lines. When calculating the comprehensive line loss rate of each line, it is also necessary to verify whether the daily line loss rate of each line is within its respective threshold range; For inter-unit packaged lines, directly verify whether the daily line loss rate of each line is within the respective threshold range.

5. The method for evaluating the comprehensive line loss rate of a power supply station based on a dynamic calibration mechanism according to claim 4 is characterized in that: Select 5 days before and after packaging as the reference period. The daily line loss rate of each day selected should be within the threshold range. If it does not meet the threshold, select another day in the previous or next day to replace it and calculate the average input power, output power and power sales; Based on the average line loss rate η in the base period A , restore the input power of the packaging day: Among them, E 售电，A(t) It refers to the actual electricity sales of the line on the package day, that is, the sum of the forward electricity of high-voltage users and the forward electricity of the public transformer area; The monthly threshold power consumption is corrected based on the difference between the restored value and the system statistical value.

6. The method for evaluating the comprehensive line loss rate of a power supply station based on a dynamic calibration mechanism according to claim 1 is characterized in that: The verification method in step S4 includes: Check the integrity of the bottom of the table: check if the bottom of the table is missing or reversed; Ratio consistency check: verify the ratio change before and after the meter is replaced; Power rationality check: abnormalities are determined based on the historical daily average power value ±3 times the standard deviation and the user capacity limit.

7. The method for evaluating the comprehensive line loss rate of a power supply station based on a dynamic calibration mechanism according to claim 6 is characterized in that: The correction method for abnormal power obtained by verification includes: Correction of historical daily average power consumption based on a sliding window with a window length of n = 7 days; Over / under capacity correction based on user capacity limits; Estimate revisions based on the rate of change of the table base; The final correction value is the average of the above three options.

8. The method for evaluating the comprehensive line loss rate of a power supply station based on a dynamic calibration mechanism according to claim 1 is characterized in that: In step S6, the line loss rate of the branch line is calculated: Input power statistics include forward power at the line gateway, high-voltage distributed reverse power, and low-voltage distributed reverse power; Output power statistics include reverse power at the line junction, forward power for high-voltage office use, and forward power for low-voltage office use.

9. The method for evaluating the comprehensive line loss rate of a power supply station based on a dynamic calibration mechanism according to claim 1, characterized in that: In the calculation of the comprehensive line loss rate of the branch line: Input power statistics include forward power at the line gateway, high-voltage distributed reverse power, and low-voltage distributed reverse power; Output power statistics include reverse power at the line gateway, forward power at high-voltage office power, and forward power at low-voltage office power. The electricity sales statistics include the forward electricity sales to high-voltage users and the electricity sales to low-voltage users on the line; Calculation of comprehensive line loss rate of the affiliated power supply station: Input power statistics include the forward power of all line gateways within a power supply station and all high-voltage and low-voltage distributed on-grid power on the corresponding lines; Output power statistics include reverse power at all line junctions within a power supply station and all high-voltage and low-voltage office power consumption on the corresponding lines; The electricity sales statistics include the forward electricity sales to all high-voltage users and the electricity sales to all low-voltage users corresponding to all lines within a power supply station.

10. The method for evaluating the comprehensive line loss rate of a power supply station based on a dynamic calibration mechanism according to claim 1, characterized in that: Also includes: Output power supply station comprehensive line loss rate evaluation report, including line / substation anomaly location, loss reduction strategy recommendations and resource allocation optimization plan; The historical data threshold is updated through a dynamic calibration mechanism to form a closed-loop feedback.

Images