CN110163507A - A kind of processing method of Line Loss of Distribution Network System - Google Patents

Abstract

The present invention provides a kind of processing methods of Line Loss of Distribution Network System, which comprises determines the feeder line sample in the power distribution network, acquires the feeder line index parameter of the feeder line sample；From the feeder line index parameter, clustering target parameter is determined；Cluster calculation is carried out to the feeder line sample using the clustering target parameter, obtains the benchmark feeder line of each cluster；Calculate the limit line loss per unit of the benchmark feeder line of each cluster；For the limit line loss per unit of the benchmark feeder line of each cluster, the limit line loss per unit of the power distribution network is calculated；Calculation amount of the invention is smaller, and the limit line loss per unit of resulting power distribution network can reflect that the line loss of current power distribution network route is horizontal to a certain extent, determines that basic preparation is carried out in practical decreasing loss space for work about electric power personnel.

Description

Method for processing line loss of power distribution network

Technical Field

The invention relates to the technical field of power distribution networks, in particular to a method for processing line loss of a power distribution network.

Background

The line loss and the line loss rate comprehensively reflect the economic operation level of a power grid and the management level of a power supply enterprise. The line loss and the line loss rate are closely related to energy conservation and emission reduction, and how to reduce the line loss has economic value and energy conservation significance. The current relatively mature line loss calculation methods include a root mean square current method, an average current method, a maximum current method, a loss factor method, a voltage loss method, an equivalent resistance method and a power flow algorithm. The method is widely applied by a root mean square current method, an equivalent resistance method and a power flow algorithm. The new line loss calculation method comprises a genetic algorithm, an interval algorithm, a clustering algorithm and an artificial neural network method. The algorithms are improved compared with the traditional line loss calculation method, and are gradually applied to the calculation of line loss.

At present, the research on the clustering of lines and the calculation of limit line loss is not enough, but the research on applying line clustering to the calculation of line loss is less. In the prior art, a method for calculating line loss by applying line clustering to a neural network exists, but the established RBFNN model is relatively complex, and meanwhile, a load flow algorithm is adopted to calculate the line loss rate, so that the method is not convenient to popularize in actual engineering; the method only considers the line loss rate and the digital characteristics thereof for clustering, does not calculate the line loss rate, and has low practicability.

Disclosure of Invention

The invention provides a method for processing the line loss of a power distribution network, which aims to overcome the technical problem.

In order to solve the above problems, the present invention discloses a method for processing the line loss of a power distribution network, wherein the method comprises:

determining a feeder line sample in the power distribution network, and acquiring a feeder line index parameter of the feeder line sample;

determining a clustering index parameter from the feeder index parameters;

clustering calculation is carried out on the feeder line samples by utilizing the clustering index parameters to obtain a reference feeder line of each cluster;

calculating the limit line loss rate of the reference feeder line of each cluster;

and calculating the limit line loss rate of the power distribution network aiming at the limit line loss rate of the reference feeder line of each cluster.

Further, before the feeder indicator parameters of the feeder samples are collected, the method includes:

and preprocessing the feeder line sample.

Further, the step of determining a cluster index parameter from the feeder index parameters includes:

calculating a correlation coefficient rho between the feeder line index parameters; wherein, the calculation formula of the correlation coefficient rho is as follows:

(1) wherein ρ is a correlation coefficient between the feeder index parameter X and the feeder index parameter Y, and e (X), e (Y), and e (xy) are X, Y and a mathematical expectation of a product of X and Y, respectively; d (X), D (Y) are the variance of the index X, Y;

and determining the clustering index parameter according to the correlation coefficient rho.

Further, the sub-step of determining the clustering index parameter according to the correlation coefficient ρ further includes:

selecting a target feeder line index parameter with the correlation coefficient rho larger than a preset coefficient value;

judging whether the target feeder index parameter meets a preset condition or not;

and when the target feeder line index parameter meets a preset condition, determining the target feeder line index parameter as the clustering index parameter.

Further, the preset conditions include:

the target feeder index parameter reflects a basic attribute of the feeder sample; wherein the fundamental properties are used to distinguish different lines of the feeder samples;

the target feeder index parameter reflects a normal load level of the feeder sample.

Further, the step of performing cluster calculation on the feeder line samples by using the cluster index parameter to obtain a reference feeder line for each cluster includes:

calculating the optimal clustering number K of the feeder line samples by using the clustering index parameters;

wherein,

(2) in the formula, XB (-) represents a certain balance point between intra-class compactness and inter-class separability of the clustering index parameter, and the smaller the calculation result of XB (-) is, the better the clustering effect is, so as to determine the optimal clustering number K; the feeder samples are { x1, x2, …, xn }, where n represents the total number of feeder samples; m represents an ambiguity parameter, d (x)_j,v_i) Point x representing a certain said feeder line sample_jTo the center v_iThe Euclidean distance of;

performing validity fuzzy clustering calculation on the feeder line samples by using a target function and a constraint condition according to the optimal clustering number K; wherein:

the objective function is:

the constraint condition is as follows:

0≤u_ij≤1,1≤i≤K,1≤j≤n (4)；

(3) in the formula (6), U is a membership matrix, and V is a clustering center matrix of p multiplied by k; mu.s_ijRefers to the jth sample feeder x_jMembership belonging to class i;

initializing a cluster center v_iContinuously updating the membership degree matrix U and the clustering center matrix V through iterative calculation when the objective function J is_m(U, V) stopping iterative computation when the difference ξ between the objective function values of the two iterations is within a preset range, and outputting a clustering result;

wherein the initialization cluster center v_iThe calculation formula of (2) is as follows:

degree of membership mu_ijAnd from the center v_iThe update formula of (2) is:

and according to the clustering result, selecting a feeder corresponding to the highest membership degree in the final membership degree matrix in the clustering, and determining the feeder as a reference feeder of the clustering.

Further, before calculating the limit line loss rate of the reference feeder of each cluster, the method includes:

replacing all feeder sections of the feeder sample with maximum section type wires specified by a power supply company;

and/or replacing the transformer of the feeder line sample with an optimal type transformer specified by a power supply company;

and/or when the limit line loss rate of the reference feeder line of each cluster is calculated, assuming that the calculation condition is three-phase balance.

Further, the step of calculating the limit line loss rate of the reference feeder of each cluster includes:

defining a calculation condition; wherein the calculation condition includes:

the load is proportional to the rated capacity of the transformer;

the power factors of the load points are the same, the node voltages are the same, and the voltage drop is not considered;

under the calculation conditions, for each cluster, an equivalent resistance R is used_eqCalculating the limit line loss rate of the clustered reference feeder lines;

wherein,

(10) in the formula, R_eqHas the unit of omega; assuming that the line of the reference feeder of each cluster is divided into g segments, where S_Ni、S_N∑The rated capacity of the distribution transformer on the i-th section branch line and the rated capacity of the total distribution of the line are respectively, and the unit is kVA; r_iThe equivalent resistance of the transformer on the ith branch line is in omega; r is₀Means the unit length of the i-th section of feeder line is omega/km；l_iThe unit of the length of the ith feeder line is km;

(11) in the formula, delta_lIs the limit line loss rate, I, of the clustered reference feeder_jfIs root mean square current, with unit a; t is the running time and the unit is h; e_lThe unit of the power supply for the feeder is kW/h.

Further, the step of calculating the limit line loss rate of the power distribution network for the limit line loss rate of the reference feeder of each cluster includes:

(12) in the formula, delta is the limit line loss rate of the power distribution network, f is the total number of the classified feeders, and delta_lsThe limit line loss rate of the s-th type reference feeder line; e_lsThe unit is the sum of the power supply of the s-th feeder line and is ten thousand kW/h.

Further, the method further comprises:

clustering calculation is carried out on the feeder line samples by utilizing the clustering index parameters to obtain non-reference feeder lines of each cluster;

and respectively calculating the limit line loss rate of the non-reference feeder of each cluster for the non-reference feeders of each cluster.

Compared with the prior art, the invention has the following advantages:

according to the method, relevant indexes influencing line loss are considered, and the limit line loss rate of the power distribution network is calculated by using a clustering algorithm, so that the limit line loss rate of each line is not required to be calculated, the calculation amount of the limit line loss rate of the power distribution network is greatly reduced, and the practice guidance effect of the line loss rate is further improved;

according to the method, the feeder line samples are preprocessed, so that data redundancy can be reduced, indexes with high influence on line loss are screened out, and the line loss calculation accuracy is improved.

Drawings

Fig. 1 is a flowchart illustrating steps of a method for processing line loss of a power distribution network according to an embodiment of the present invention;

FIG. 2 is a flowchart illustrating steps of a method for calculating a limit line loss rate of a non-reference feeder according to an embodiment of the present invention;

FIG. 3 is a clustering result diagram of the present example;

fig. 4 is a limit line loss topological diagram of the class B reference feeder line under the condition of the limit parameter in the present example.

Detailed Description

In order to make the aforementioned objects, features and advantages of the present invention comprehensible, embodiments accompanied with figures are described in further detail below.

The noun explains:

limiting line loss rate: for a distribution network line with a determined structure, under the current load characteristics and load distribution, the line loss after all the loss reduction measures which can be adopted under the limit condition is called limit line loss, and the line loss rate at the moment is called limit line loss rate.

Clustering: the fuzzy clustering calculation is carried out according to the clustering index, and then the data is divided into a plurality of groups. The clustering is to find potential similar modes existing in data without the guidance of any prior information, classify the data after fuzzy clustering calculation according to clustering indexes, and enable the data classification to meet the principle that the similarity in the classes is as large as possible and the difference between the different classes is as large as possible.

To solve the technical problem of the present invention, referring to fig. 1, a flowchart illustrating steps of a method for processing a line loss of a power distribution network according to an embodiment of the present invention is shown, where the method specifically includes the following steps:

step S101: determining a feeder line sample in the power distribution network, and acquiring a feeder line index parameter of the feeder line sample;

step S102: determining a clustering index parameter from the feeder index parameters;

step S103: clustering calculation is carried out on the feeder line samples by utilizing the clustering index parameters to obtain a reference feeder line of each cluster;

step S104: calculating the limit line loss rate of the reference feeder line of each cluster;

step S105: and calculating the limit line loss rate of the power distribution network aiming at the limit line loss rate of the reference feeder line of each cluster.

The number of distribution network lines is large, and the number of single line branch lines and elements is also large. Therefore, a feeder sample in the power distribution network is determined first, and the feeder sample in the embodiment of the present invention may be a line from which a lossless line, a full-loss line (power line), and the like have been removed and which do not meet requirements.

Next, considering that the feeder line sample has a huge data volume and a large redundancy, before acquiring the feeder line index parameter of the feeder line sample, the embodiment of the present invention further includes the following steps:

and preprocessing the feeder line sample.

By preprocessing the feeder line samples, the data redundancy is reduced, the indexes with high influence on the line loss are screened out, and the data are subjected to dimensionality reduction by utilizing the correlation principle, so that the clustering calculation amount is reduced.

Preferably, the feeder index parameters screened by the embodiment of the present invention include a power supply radius, a cabling rate, a distribution transformation capacity, an average power amount, a maximum current, and a maximum load rate.

The supply radius is defined as the length of the line from the power supply point to the farthest load point. The length of the power supply radius directly influences the equivalent resistance of the feeder line, and the influence degree on the limit line loss is high.

The cabling rate is the percentage of the cable length over the total length of the line. Since the eddy current generates a loss in the cable, the loss of the cable is larger than that of the overhead wire, and therefore, the higher the cabling ratio, the larger the limit line loss.

The sum of the capacities is the sum of the rated capacities of all the distribution transformers on the line. Generally, the larger the sum of the capacities is, the more the number of transformers is, and the equivalent resistances of the same capacity and different models are different. The iron loss is different, and the economic operation interval is slightly different.

The average electric quantity is the average value of the annual active power supply quantity, the load of the power distribution network is reflected, and the size of the load directly influences the size of the limit line loss.

The maximum current is the maximum working current which appears in a feeder of the power distribution network in a short time under the condition of not influencing equipment safety. The maximum current is an effective index of a load curve, and meanwhile, the line loss is proportional to the square of the current, so that the influence degree of the maximum current on the limit line loss is high.

The maximum load rate reflects the magnitude of the ultimate loss to some extent. The transformer loss comprises copper loss and iron loss, and the iron loss is the same when the load rates of transformers with the same capacity are different; as the load factor increases, the copper loss increases.

After the feeder line index parameters are determined, the clustering index parameters need to be determined from the feeder line index parameters. The clustering index parameter selection should have correlation with line loss, can represent line attributes, and simultaneously needs to reduce data scale and dimensionality as much as possible; meanwhile, the influence degree of the clustering index parameters on the line loss is also considered, so that the clustering index is selected by considering the linear correlation coefficient among the line indexes and the influence degree of the indexes on the line loss.

In an optional embodiment of the present invention, it is shown that the specific implementation method of the step S102 includes the following sub-steps:

substep 1: calculating a correlation coefficient rho between the feeder line index parameters; wherein, the calculation formula of the correlation coefficient rho is as follows:

(1) wherein ρ is a correlation coefficient between the feeder index parameter X and the feeder index parameter Y, and e (X), e (Y), and e (xy) are X, Y and a mathematical expectation of a product of X and Y, respectively; d (X), D (Y) are the variance of the index X, Y;

substep 2: and determining the clustering index parameter according to the correlation coefficient rho.

Preferably, the substep 2 further may comprise:

selecting a target feeder line index parameter with the correlation coefficient rho larger than a preset coefficient value;

judging whether the target feeder index parameter meets a preset condition or not;

and when the target feeder line index parameter meets a preset condition, determining the target feeder line index parameter as the clustering index parameter.

The inventor researches and finds that when the correlation coefficient rho is larger than 0.7, the two indexes can be considered to have high correlation, so that the feeder line index parameter with the correlation coefficient rho larger than 0.7 is preferably used as the primary screening condition of the clustering index parameter.

And then, screening target feeder line index parameters with the correlation coefficient rho larger than 0.7 again through preset conditions to determine the influence degree of line loss and ensure that indexes related in a clustering index system have clear meanings and have large influence on the line loss rate. Wherein the preset condition may include:

the target feeder index parameter reflects a basic attribute of the feeder sample; wherein the fundamental properties are used to distinguish different lines of the feeder samples;

the target feeder index parameter reflects a normal load level of the feeder sample.

Then, after the clustering index parameter is determined, clustering calculation needs to be performed on the feeder line samples by using the clustering index parameter to obtain a reference feeder line of each cluster;

in a preferred embodiment of the present invention, it is shown that the step S103 may specifically include the following sub-steps:

substep 3: calculating the optimal clustering number K of the feeder line samples by using the clustering index parameters;

(2) in the formula, XB^(-)A certain balance point between intra-class compactness and inter-class separability representing a search for said clustering index parameter, said XB^(-)The smaller the calculation result of (2), the better the clustering effect, and the best clustering number K is determined; the feeder samples are { x1, x2, …, xn }, where n represents the total number of feeder samples; m represents an ambiguity parameter, d (x)_j,v_i) Point x representing a certain said feeder line sample_jTo the center v_iThe Euclidean distance of;

substep 4: performing validity fuzzy clustering calculation on the feeder line samples by using a target function and a constraint condition according to the optimal clustering number K; wherein:

the objective function is:

the constraint condition is as follows:

0≤u_ij≤1,1≤i≤K,1≤j≤n (4)；

(3) in the formula (6), U is a membership matrix, and V is a clustering center matrix of p multiplied by k; mu.s_ijRefers to the jth sample feeder x_jMembership belonging to class i;

substep 5: initializing a cluster center v_iContinuously updating the membership degree matrix U and the clustering center matrix V through iterative calculation when the objective function J is_m(U, V) stopping iterative computation when the difference ξ between the objective function values of the two iterations is within a preset range, and outputting a clustering result;

wherein the initialization cluster center v_iThe calculation formula of (2) is as follows:

degree of membership mu_ijAnd from the center v_iThe update formula of (2) is:

substep 6: and according to the clustering result, selecting a feeder corresponding to the highest membership degree in the final membership degree matrix in the clustering, and determining the feeder as a reference feeder of the clustering.

From formulas (2) to (9), the feeder lines in the power distribution network are clustered by using the characteristics of the FCM algorithm in the embodiment of the present invention. The FCM algorithm is a clustering algorithm based on data partitioning, and its objective is to maximize the similarity between objects partitioned into the same class and minimize the similarity between different classes. The partitioning of data by FCM is a fuzzy, flexible partitioning.

In the embodiment of the invention, after FCM algorithm clustering calculation is performed on the feeder line samples by using the step shown in step S103, various types of reference feeder lines are obtained, and then the limit line loss rate of each clustered reference feeder line can be calculated.

For the calculation of the feeder line limit line loss rate, the line parameters of the reference feeder line can be calculated in a limit mode according to the definition of the limit line loss. Before calculation, the limit parameters of the feeder line samples are constrained first, and a specific constraint method may include:

replacing all feeder sections of the feeder sample with maximum section type wires specified by a power supply company;

and/or replacing the transformer of the feeder line sample with an optimal type transformer specified by a power supply company;

and/or when the limit line loss rate of the reference feeder line of each cluster is calculated, assuming that the calculation condition is three-phase balance.

Further, calculating the limit line loss rate of the reference feeder line of each cluster;

in a preferred embodiment of the present invention, it is shown that the step S104 may specifically include the following sub-steps:

substep 7: defining a calculation condition; wherein the calculation condition includes:

the load is proportional to the rated capacity of the transformer;

the power factors of the load points are the same, the node voltages are the same, and the voltage drop is not considered;

substep 8: under the calculation conditions, for each cluster, an equivalent resistance R is used_eqCalculating the limit line loss rate of the clustered reference feeder lines;

wherein,

(10) in the formula, R_eqHas the unit of omega; assuming that the line of the reference feeder of each cluster is divided into g segments, where S_Ni、S_N∑The rated capacity of the distribution transformer on the i-th section branch line and the rated capacity of the total distribution of the line are respectively, and the unit is kVA; r_iThe equivalent resistance of the transformer on the ith branch line is in omega; r is₀The unit length of the unit of the i-th section of feeder line is the alternating current resistance, and the unit is omega/km; l_iThe unit of the length of the ith feeder line is km;

(11) in the formula, delta_lIs the limit line loss rate, I, of the clustered reference feeder_jfIs root mean square current, with unit a; t is the running time and the unit is h; e_lThe unit of the power supply for the feeder line is kW.h.

Next, the limit line loss rate of the power distribution network is calculated for the limit line loss rate of the reference feeder of each cluster, and the limit line loss rate of the power distribution network is calculated by using a formula (12).

Specifically, the implementation method of step S105 is as follows:

(12) in the formula, delta is the limit line loss rate of the power distribution network, f is the total number of the classified feeders, and delta_lsThe limit line loss rate of the s-th type reference feeder line; e_lsFor the supply of class-s feederAnd, in units of ten thousand kW/h.

Because the line loss under different load curves is changed, the limit line loss rate is also different, the load curve of a typical day is generally selected for calculating the line loss rate, the calculation amount of one line is not large, but the calculation amount of the whole distribution network becomes huge. Therefore, the embodiment of the invention calculates the limit line loss rate of the power distribution network by using the clustering algorithm, so that the limit line loss rate of each line does not need to be calculated, and the calculation amount of the limit line loss rate of the power distribution network is greatly reduced.

The following further verifies with a specific example for the present invention.

Feeder line sample: 87 lines of a certain power distribution network in a city are actual data of 10kV lines, and the 87 lines are data obtained after non-conforming requirements such as lossless lines and full-loss lines (power lines) are removed. And according to the collected data characteristics, the calculated limit line loss rate is the average limit line loss rate.

The collected feeder indicator parameters are as follows: power supply radius (X1), average power (X2), maximum current (X3), cabling rate (X4), capacity sum (X5), and maximum load rate (X6).

And calculating a correlation coefficient according to the feeder line index parameter, wherein the calculation result is shown in table 1.

TABLE 1

According to the correlation coefficient calculation results in the above table, it can be seen that the line loss indicators with correlation coefficients greater than 0.7 include X1, X2, X3, X4, X5, X6, X2 and X6, and X3 and X6, that is, the indicators have strong correlation, so one should be selected as a clustering indicator parameter.

And then, analyzing the line loss influence degree, namely judging whether the target feeder index parameter meets a preset condition, wherein the analysis is as follows:

1) the power supply radius can represent the basic attribute of the line, the distinguishing requirement on the line can be met, and the influence degree on the generated line loss is large. Therefore, the power supply radius (X1) is selected as the clustering index parameter.

2) The average electric quantity can objectively reflect the normal operation state of the line through the average calculation of the monthly electric quantity, and the extreme value parameters cannot objectively reflect the operation condition of the line, so that instability exists. Meanwhile, according to the calculation result of the correlation coefficient, the correlation among the average electric quantity (X2), the maximum current (X3) and the maximum load rate (X6) is strong, and in order to reduce the data dimension, the average electric quantity (X2), the maximum current (X3) and the maximum load rate (X6) are not suitable to be selected as the clustering index parameters.

3) In urban networks, the cabling rate (X4) is higher and higher, and the cabling rate of a plurality of lines reaches 100%, and the index cannot effectively distinguish and classify the lines, so that the cabling rate (X4) is not suitable as a clustering parameter of the lines.

4) The sum of the capacities is the sum of the capacities of all transformers on the line. Generally, the load rate of the transformer fluctuates at any time, the acquired load rate data is the load rate at a certain moment, the load condition cannot be objectively reflected, meanwhile, the number of the transformers in a 10V distribution network is large, and the load rate data is difficult to receive. Therefore, the capacity sum (X5) is not suitable for selection as the clustering index parameter.

Therefore, the power supply radius (X1) can be finally determined as the clustering index parameter of the line.

The present example uses Matlab2016 for fuzzy clustering calculations. First, adopt XB^(-)And calculating the optimal clustering number, and determining the optimal clustering number to be 6. And then, performing validity fuzzy clustering by adopting an FCM algorithm. The maximum iteration number is 100, the clustering number k is 6, the ambiguity m is 2, and the variation of the objective function value of the iteration termination condition is less than 1 × e^-5. The clustering result graph of the present example refers to fig. 3, in which the abscissa is the power supply radius/km and the ordinate is the average electric quantity/kWh. Each point represents a 10kV line. As can be seen from the figure, the classification result satisfies the separation between classes, within classThe criteria of compactness. The statistical results of the clustering data are shown in Table 2.

TABLE 2

As can be seen from Table 2, the cluster classification is clear, and the results have practical significance. The power supply radius and the average electric quantity of the data of the A-type feeder line are both extremely small, and the power supply radius and the average electric quantity of the data of the second E-type feeder line are both extremely large. And the influence of the power supply radius and the average electric quantity on the line loss rate is great, so that the line loss rate distribution of various feeder lines is relatively concentrated.

Meanwhile, as can be seen from fig. 3, the cluster center usually does not fall on a certain data point, and is mostly in the vicinity of a certain data point. And selecting the feeder corresponding to the maximum membership degree as a reference feeder of each type of feeder according to the membership degree matrix of 6 multiplied by 87 in the process, and calculating the limit line loss rate. The membership degree of each type of reference feeder reaches over 0.94, and the specific data of the maximum membership degree of each clustering reference feeder is shown in table 3.

TABLE 3

Cluster numbering	Degree of membership	Line name
			A	0.9987	F9
B	0.9925	F31
			C	0.9887	F46
D	0.9834	F55
			E	0.9451	F76
F	0.9809	F81

From table 3, it can be seen that the membership degree of the E-type reference feeder is 0.9451, and the membership degrees of the other types of feeder are all above 0.98. Although the various clustering centers do not fall on the feeder data points, the reference feeders are close to the clustering centers, and the feeders can be selected as various reference feeders.

And after a reference feeder line is selected, drawing a network topological graph according to actual data, and replacing line parameters to limit parameters. Taking the type-B reference feeder F28 as an example, the 3 x 240 cables outgoing from the original line are replaced by the 3 x 400 cables, the trunk lines all adopt JKLVJ-3 x 240 overhead lines, the branch lines adopt JKLVJ-3 x 150 overhead lines, and the distribution transformers all adopt S15 series. The limit line loss topological graph of the class B reference feeder line under the condition of the limit parameter of the present example refers to fig. 4.

Next, the limit line loss rate of the reference line was calculated to be 1.275% by using the equivalent resistance method. As shown in table 4, the limit line loss rate of the reference feeder for each cluster.

TABLE 4

Cluster numbering	Limit line loss rate of each reference feeder line%
		A	0.949
B	1.275
		C	1.435
D	2.477
		E	1.585
F	1.847

And calculating the average limit line loss rate of the A-F by using a formula (12) aiming at the limit line loss rate of the reference feeder of the clustering feeders with clustering numbers of A-F. And then comparing the calculated average limit line loss rate with the existing line loss rate data, according to the requirement that the comprehensive line loss rate (deducting lossless lines) of the lines of 10kV and below in the city is less than or equal to 5%, and if the comprehensive line loss rate is less than 5.5% of the current statistical line loss rate and 3.2% of the theoretical line loss rate, considering that the value approaches to the ideal limit line loss rate to a certain extent, and using the ideal limit line loss rate as the limit line loss rate of the current power distribution network. The obtained limit line loss rate of the power distribution network can reflect the line loss level of the current power distribution network line to a certain extent, and basic preparation is made for determining the actual loss reduction space for power workers.

In addition, other feeder lines, collectively referred to as non-reference feeders, are included in the same cluster of the distribution grid. The embodiment of the present invention further calculates the limit line loss rate of each clustered non-reference feeder, and referring to fig. 2, a flowchart of the steps of the method for calculating the limit line loss rate of the non-reference feeder according to the embodiment of the present invention is shown, where the method specifically includes the following steps:

step S201: clustering calculation is carried out on the feeder line samples by utilizing the clustering index parameters to obtain non-reference feeder lines of each cluster;

step S202: and respectively calculating the limit line loss rate of the non-reference feeder of each cluster for the non-reference feeders of each cluster.

In a preferred embodiment of the present invention, the step S202 further includes the following sub-steps:

defining an adjustment factor P between non-reference and reference feeders of said cluster_ij；

Wherein, the

(13) Wherein q is the total number of the feeder index parameters, X_ijIs the ith index parameter value, X, of the jth feeder line in the cluster_i0And the ith index parameter value of the clustered reference feeder line is used as the index parameter value of the clustered reference feeder line.

Calculating the limit line loss rate of the clustered non-reference feeder lines according to the adjusting coefficient;

wherein,

δ_j＝(1+P_ij)δ₀(14)；

(14) in the formula, delta_jThe jth limit line loss of the clustered non-reference feeder lineRate, delta_lAnd the limit line loss rate of the clustered reference feeder line.

Then, in the embodiment of the present invention, the limit line loss rate of the non-reference feeder line may be calculated with reference to step S105, and then the limit line loss rate of the power distribution network may be calculated by using formula (12), so that the limit line loss rate data of the power distribution network may be more complete. It should be noted that, in the embodiment of the present invention, for the calculation of the line loss of other feeders in the same cluster, the calculation of the limit line loss rate is performed by using the index parameter difference between the calculated line loss and the reference feeder, and the calculation amount of the formula (10) and the formula (11) can be omitted.

In a specific example of the present invention, the limit line loss rate of the non-reference feeder of each cluster is also calculated. When the limit line loss rates of other feeders in each cluster are calculated, the influence of other parameters on the line loss rates is considered, difference coefficient calculation is carried out through feeder line loss rate parameters such as the average electric quantity and the power supply radius of the clustered lines, and the limit line loss rates of all the lines are obtained and are shown in table 5.

TABLE 5

The embodiments in the present specification are described in a progressive manner, each embodiment focuses on differences from other embodiments, and the same and similar parts among the embodiments are referred to each other.

The above method for processing the line loss of the power distribution network provided by the invention is described in detail, and a specific example is applied in the text to explain the principle and the implementation of the invention, and the description of the above embodiment is only used to help understand the method and the core idea of the invention; meanwhile, for a person skilled in the art, according to the idea of the present invention, there may be variations in the specific embodiments and the application scope, and in summary, the content of the present specification should not be construed as a limitation to the present invention.

Claims (10)

1. A method for processing line loss of a power distribution network is characterized by comprising the following steps:

determining a feeder line sample in the power distribution network, and acquiring a feeder line index parameter of the feeder line sample;

determining a clustering index parameter from the feeder index parameters;

clustering calculation is carried out on the feeder line samples by utilizing the clustering index parameters to obtain a reference feeder line of each cluster;

calculating the limit line loss rate of the reference feeder line of each cluster;

and calculating the limit line loss rate of the power distribution network aiming at the limit line loss rate of the reference feeder line of each cluster.

2. The method of claim 1, wherein prior to collecting the feeder indicator parameters for the feeder samples, the method comprises:

and preprocessing the feeder line sample.

3. The method of claim 1, wherein the step of determining a cluster metric parameter from the feeder metric parameters comprises:

calculating a correlation coefficient rho between the feeder line index parameters; wherein, the calculation formula of the correlation coefficient rho is as follows:

(1) wherein ρ is a correlation coefficient between the feeder index parameter X and the feeder index parameter Y, and e (X), e (Y), and e (xy) are X, Y and a mathematical expectation of a product of X and Y, respectively; d (X), D (Y) are the variance of the index X, Y;

and determining the clustering index parameter according to the correlation coefficient rho.

4. The method according to claim 3, characterized in that the sub-step of determining the cluster indicator parameter according to the correlation coefficient p further comprises:

selecting a target feeder line index parameter with the correlation coefficient rho larger than a preset coefficient value;

judging whether the target feeder index parameter meets a preset condition or not;

and when the target feeder line index parameter meets a preset condition, determining the target feeder line index parameter as the clustering index parameter.

5. The method according to claim 4, wherein the preset conditions include:

the target feeder index parameter reflects a basic attribute of the feeder sample; wherein the fundamental properties are used to distinguish different lines of the feeder samples;

the target feeder index parameter reflects a normal load level of the feeder sample.

6. The method of claim 1, wherein the step of performing cluster computation on the feeder samples by using the cluster index parameter to obtain a reference feeder for each cluster comprises:

calculating the optimal clustering number K of the feeder line samples by using the clustering index parameters;

wherein,

(2) in the formula, XB^(-)A certain balance point between intra-class compactness and inter-class separability representing a search for said clustering index parameter, said XB^(-)The smaller the calculation result of (2), the better the clustering effect, and the best clustering number K is determined; the feeder samples are { x1, x2, …, xn }, where n represents the total number of feeder samples; m represents an ambiguity parameter, d (x)_j,v_i) Point x representing a certain said feeder line sample_jTo the center v_iThe Euclidean distance of;

performing validity fuzzy clustering calculation on the feeder line samples by using a target function and a constraint condition according to the optimal clustering number K; wherein:

the objective function is:

the constraint condition is as follows:

0≤u_ij≤1,1≤i≤K,1≤j≤n (4)；

(3) in the formula (6), U is a membership matrix, and V is a clustering center matrix of p multiplied by k; mu.s_ijRefers to the jth sample feeder x_jMembership belonging to class i;

initializing a cluster center v_iContinuously updating the membership degree matrix U and the clustering center matrix V through iterative calculation when the objective function J is_m(U, V) stopping iterative computation when the difference ξ between the objective function values of the two iterations is within a preset range, and outputting a clustering result;

wherein the initialization cluster center v_iThe calculation formula of (2) is as follows:

degree of membership mu_ijAnd from the center v_iThe update formula of (2) is:

and according to the clustering result, selecting a feeder corresponding to the highest membership degree in the final membership degree matrix in the clustering, and determining the feeder as a reference feeder of the clustering.

7. The method according to claim 1, wherein before calculating the limit line loss rate of the reference feeders of each cluster, the method comprises:

replacing all feeder sections of the feeder sample with maximum section type wires specified by a power supply company;

and/or replacing the transformer of the feeder line sample with an optimal type transformer specified by a power supply company;

and/or when the limit line loss rate of the reference feeder line of each cluster is calculated, assuming that the calculation condition is three-phase balance.

8. The method of claim 7, wherein the step of calculating the limit line loss rate of the reference feeder of each cluster comprises:

defining a calculation condition; wherein the calculation condition includes:

the load is proportional to the rated capacity of the transformer;

the power factors of the load points are the same, the node voltages are the same, and the voltage drop is not considered;

under the calculation conditions, for each cluster, an equivalent resistance R is used_eqCalculating the limit line loss rate of the clustered reference feeder lines;

wherein,

(10) in the formula, R_eqHas the unit of omega; assuming that the line of the reference feeder of each cluster is divided into g segments, where S_Ni、S_N∑The rated capacity of the distribution transformer on the i-th section branch line and the rated capacity of the total distribution of the line are respectively, and the unit is kVA; r_iThe equivalent resistance of the transformer on the ith branch line is in omega; r is₀The unit length of the unit of the i-th section of feeder line is the alternating current resistance, and the unit is omega/km; l_iThe unit of the length of the ith feeder line is km;

(11) in the formula, delta_lIs the limit line loss rate, I, of the clustered reference feeder_jfIs root mean square current, with unit a; t is the running time and the unit is h; e_lTo be fedThe power supply of the wire is kW/h.

9. The method of claim 1, wherein the step of calculating the limit line loss rate of the distribution network for the limit line loss rate of the reference feeder of each cluster comprises:

(12) in the formula, delta is the limit line loss rate of the power distribution network, f is the total number of the classified feeders, and delta_lsThe limit line loss rate of the s-th type reference feeder line; e_lsThe unit is the sum of the power supply of the s-th feeder line and is ten thousand kW/h.

10. The method of claim 1, further comprising:

clustering calculation is carried out on the feeder line samples by utilizing the clustering index parameters to obtain non-reference feeder lines of each cluster;

and respectively calculating the limit line loss rate of the non-reference feeder of each cluster for the non-reference feeders of each cluster.