Power distribution network source-load extreme scene extraction method under high-proportion distributed wind power access

Abstract

The invention relates to a power distribution network source-load extreme scene extraction method under high-proportion distributed wind power access, and belongs to the field of power distribution network operation and planning. The method comprises the following steps: S1, acquiring historical wind power output data and power distribution network load data from a power dispatching department, and establishing a source load uncertainty model, namely establishing a wind power output uncertainty model and a load uncertainty model respectively through a Kullback-Leibler divergence method and a confidence interval method; s2, randomly generating a large number of power distribution network source load data sets by adopting a Monte Carlo method according to the source load uncertainty model; and S3, extracting an extreme scene of the power distribution network by adopting a source-load balance entropy index. According to the method, training samples are not needed, outliers, edges and extreme scenes with small occurrence probability in the power distribution network are easily extracted, and the method can be applied to the fields of distributed wind storage system optimization configuration, power distribution network operation and planning and the like.

Description

Technical Field

[0001] The present invention belongs to the field of distribution network operation and planning, and relates to a method for extracting extreme source-load scenarios of a distribution network under high-proportion distributed wind power access. Background Art

[0002] In recent years, the proportion of distributed wind power in the distribution network has continued to increase, and its output has shown large fluctuations and strong randomness. At the same time, the electrification trend on the load side has also increased with the surge in electricity demand, increasing the uncertainty of source and load. The frequent occurrence of extreme events of "low probability-high loss" has posed a serious threat to the safe operation of the distribution network. Therefore, it is necessary to carry out research on the accurate and effective extraction of extreme scenarios, which has important engineering application value for the optimal configuration of distributed wind storage systems, distribution network operation and planning, etc.

[0003] At present, the scenario reduction methods based on data drive, including discrete-aggregation methods and multi-objective generative adversarial network methods, etc., are to extract extreme scenarios from time series scenarios by using the statistical results of source-load data. For example, the discrete-aggregation method projects each group of distributed power output-load data in the time series into a grid of a certain step length, and regards the data falling into the grid as a type of scenario, thereby realizing the extraction of extreme scenarios; the multi-objective generative adversarial network method considers the uncertainty of wind power to extract extreme scenarios, but it is necessary to ensure the accuracy and diversity of the generated scenarios through careful design. The implementation of the above methods requires a large amount of relatively comprehensive and long-term source-load operation scenario data support. Therefore, in order to effectively extract outliers, edges, and high-risk extreme scenarios with low probability of occurrence, considering the correlation of source-load data and then quantifying various operation scenarios will be an effective way to accurately extract extreme scenarios, so extracting valuable information from source-load data will be the key to quantifying operation scenarios. The present invention takes into account information entropy as a classic way to characterize information value, and measures the correlation of source-load data by constructing source-load balance entropy, thereby realizing the accurate extraction of source-load extreme scenarios of distribution networks with high proportion of distributed wind power access. Summary of the invention

[0004] In view of this, the purpose of the present invention is to provide a method for extracting extreme source-load scenarios of distribution networks under a high proportion of distributed wind power access. The method obtains historical wind power output data and distribution network load data from the power dispatching department, establishes a source-load uncertainty model, and adopts information entropy theory to establish a source-load balance entropy index, thereby realizing accurate extraction of extreme scenarios of the distribution network.

[0005] In order to achieve the above object, the present invention provides the following technical solutions:

[0006] The method for extracting extreme scenarios of source and load of distribution network under high proportion of distributed wind power access includes the following steps:

[0007] S1: Obtain historical wind power output data and distribution network load data from the power dispatching department and establish a source-load uncertainty model;

[0008] S11: Establish wind power output uncertainty model through KL divergence method;

[0009] S12: Establish load uncertainty model through confidence interval method;

[0010] S2: Based on the source-load uncertainty model, a large number of distribution network source-load data sets are randomly generated using the Monte Carlo method;

[0011] S3: Use the source-load balance entropy index to extract extreme scenarios of the distribution network.

[0012] Furthermore, in S11, the KL divergence method is used to establish a wind power output uncertainty model, and the specific steps are:

[0013] ① Reference probability distribution P0

[0014] There are N wind farm historical power data samples, where the xth data A x = {P WF,1 ,P WF,2 ,…,P WF,T}, each data sample has T sampling moments, and the power is divided into L power intervals with an interval of ε;

[0015] ε=P r / L (1)

[0016]

[0017] P0={ρ1,ρ2,...,ρ L} (4)

[0018] Among them, P r is the installed capacity of distributed wind farm; ρ l is the average probability of interval l corresponding to N data samples; is a 0-1 variable; A x The wind farm power at the next moment t;

[0019] ②Collection distance d KL

[0020]

[0021] in, is a chi-square distribution with L-1 degrees of freedom α * The upper quantile is used to ensure that the actual distribution of wind power output probability distribution is not less than α * probability;

[0022] ③ Wind power output probability distribution set G

[0023] KL divergence is used to describe the distance between distribution functions. The smaller the distance, the more similar the distributions are. KL distance D KL for:

[0024]

[0025] Where ξ is an uncertain variable; п is the sample space of ξ; f(ξ) and f0(ξ) are the probability density functions of P and P0 respectively; suppose the KL distance of the reference distribution P0 does not exceed the threshold d KL The set of all distribution functions of is G;

[0026] G={P|D KL (P||P0)≤d KL} (7)

[0027] Among them, when d KL >0, G contains infinite distribution functions; when d KL When it approaches 0, G=P0.

[0028] Furthermore, in S12, the load uncertainty is characterized by the influence of price-based demand response, and the demand price elasticity coefficient e is established according to the definition of price-based demand response. xy for:

[0029]

[0030] Among them, q x , Δq x are the power consumption and power consumption change in time period x respectively; Θ y , ΔΘ y are the electricity price and electricity price change in period y respectively;

[0031] When x=y, e xy is the self-elastic coefficient and e xy <0; when x≠y, e xy is the cross elastic coefficient and e xy >0; the elastic coefficients at each moment in the same period are the same; according to formula (8), the multi-period elastic coefficient matrix E is:

[0032]

[0033] According to the matrix E, the updated value of the load due to electricity price in each period is calculated;

[0034]

[0035] Among them, q t ′ is the electricity price load power consumption in time period t.

[0036] Furthermore, in S12, the confidence interval method is used to deal with the uncertainty of the load; there are multiple confidence intervals under the same confidence level, and infinite load distributions are obtained through calculation of historical load samples; if there is a set of load post-data q t ′ is in the range of [0,ξ2], when there exists P(ξ1≤q t When ′≤ξ2)=1-α, [ξ1,ξ2] is called the confidence interval corresponding to the confidence level 1-α, ξ1 and ξ2 are called the lower and upper confidence limits under the confidence level 1-α, respectively. At this time, the corresponding load is

[0037] Further, in S2, according to the source-load uncertainty model, k source-load operation scenarios are randomly generated through Monte Carlo simulation, and then the set distance d is determined. KL Then the probability distribution set G is obtained, and k distribution functions are randomly selected from it to generate the corresponding number of wind power output scenarios At the same time, k loads with confidence level 1-α are obtained

[0038] Further, in S3, the source-charge balance entropy H is used s-l , and calculate H for each scene s-l , where the largest H s-l The source-load scenario corresponding to the value is the extreme scenario.

[0039] Furthermore, in S3, the source-load balance entropy calculation process is as follows: Assume that the source-load scenario f includes f={f1, f2, …, f k}, the timing net load Normalized to Finally, the source-charge balance entropy is obtained The calculation formula is as follows:

[0040]

[0041]

[0042] in, is the net load at time t under scenario f; and are the load of node i with confidence level 1-α and the grid-connected power of the jth distributed wind farm at time t under scenario f; n is the total number of nodes in the distribution network; m is the number of distributed wind farms; let The probability distribution of {P f},Right now in

[0043] The beneficial effects of the present invention are as follows: the method for extracting extreme scenarios of source and load of distribution network under high proportion of distributed wind power access proposed in this scheme obtains historical wind power output data and distribution network load data from the power dispatching department, and establishes a source and load uncertainty model, that is, respectively establishes a wind power output uncertainty model and a load uncertainty model by using the KL (Kullback-Leibler) divergence method and the confidence interval method; then, according to the source and load uncertainty model, a large number of distribution network source and load data sets are randomly generated by using the Monte Carlo method; finally, the source and load balance entropy index is used to extract extreme scenarios of distribution network. The present invention can be used to extract outliers, edges, and extreme scenarios with low probability of occurrence in the distribution network, and can be applied to the fields of distributed wind storage system optimization configuration, distribution network operation and planning.

[0044] Other advantages, objectives and features of the present invention will be described in the following description to some extent, and to some extent, will be obvious to those skilled in the art based on the following examination and study, or can be taught from the practice of the present invention. The objectives and other advantages of the present invention can be realized and obtained through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] In order to make the purpose, technical solutions and advantages of the present invention more clear, the present invention will be described in detail below in conjunction with the accompanying drawings, wherein:

[0046] Figure 1 The flowchart of the method for extracting extreme scenarios of source and load in distribution network is provided;

[0047] Figure 2 It is the topology diagram of IEEE33 node distribution network including distributed wind farm;

[0048] Figure 3 The source-load balance entropy and voltage over-limit frequency of the distribution network;

[0049] Figure 4 It is the timing change of the net load of the scene;

[0050] Figure 5 is the daily node voltage distribution of the distribution network.

[0051] In the figure, 1, 2, 3, ... represent node numbers; ①, ②, ③, ... represent branch numbers. DETAILED DESCRIPTION

[0052] The following describes the embodiments of the present invention by specific examples, and those skilled in the art can easily understand other advantages and effects of the present invention from the contents disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and the details in this specification can also be modified or changed in various ways based on different viewpoints and applications without departing from the spirit of the present invention. It should be noted that the illustrations provided in the following embodiments only illustrate the basic concept of the present invention in a schematic manner, and the following embodiments and features in the embodiments can be combined with each other without conflict.

[0053] Among them, the drawings are only used for illustrative explanations, and they only represent schematic diagrams rather than actual pictures, and should not be understood as limitations on the present invention. In order to better illustrate the embodiments of the present invention, some parts of the drawings may be omitted, enlarged or reduced, and do not represent the size of actual products. For those skilled in the art, it is understandable that some well-known structures and their descriptions in the drawings may be omitted.

[0054] The same or similar numbers in the drawings of the embodiments of the present invention correspond to the same or similar parts; in the description of the present invention, it should be understood that if the terms "upper", "lower", "left", "right", "front", "rear", etc. indicate the orientation or position relationship, they are based on the orientation or position relationship shown in the drawings, which is only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operate in a specific orientation. Therefore, the terms describing the position relationship in the drawings are only used for illustrative purposes and cannot be understood as limiting the present invention. For ordinary technicians in this field, the specific meanings of the above terms can be understood according to specific circumstances.

[0055] See also Figure 1 to Figure 5 , which is a method for extracting extreme scenarios of source and load in distribution network under high proportion of distributed wind power access, mainly includes three steps: source and load uncertainty modeling, extreme scenario set generation and distribution network extreme scenario extraction.

[0056] To apply this invention, it is necessary to obtain historical wind power output data and distribution network load data from the power dispatching department, establish a source-load uncertainty model, and establish a source-load balance entropy index, so as to accurately extract extreme scenarios of the distribution network. Figure 1 The method and implementation steps for extracting extreme scenarios of source and load in distribution network under high proportion of distributed wind power access are described in detail. The effectiveness of the present invention is illustrated by taking IEEE33 distribution network as an example, where the distribution network contains 4 6MW distributed wind farms, and the location distribution is shown in Figure 2 As shown in the figure, 1, 2, 3, ... represent node numbers; ①, ②, ③, ... represent branch numbers.

[0057] Step 1: Source-load uncertainty modeling. Based on the historical wind power output data and load data provided by the power dispatching department, the wind power output uncertainty model and load uncertainty model are established by the KL divergence method and the confidence interval method respectively. Among them, the error of wind power output relative to the reference value follows a normal distribution with a mean of 0, and the standard deviation of the error is 20% of the reference value; according to formula (5), in α * = 0.95, select the set distance d KL = 0.0124. For the load, the confidence level is set to 1-α = 0.95.

[0058] Step 2: Generate extreme scenario sets. Use the Monte Carlo method to generate 5,000 corresponding source-load operation scenarios, that is, determine the probability distribution set G at a set distance of 0.0124, randomly extract 5,000 distribution functions from it to generate the corresponding number of wind power output scenarios, and at the same time obtain 5,000 load scenarios at a confidence level of 0.95.

[0059] Step 3: Extraction of extreme scenarios of distribution network. s-l When the normalized net load probability sampling interval is set to 10, the equal value interval is taken as 0.1pu, and the extreme scenario extraction results are shown in Figure 3-Figure 4 .from Figure 3 (a) H in visible scene C s-l At most, according to claim 6, scenario C is the extracted extreme scenario of the distribution network.

[0060] To illustrate the effectiveness of the above extreme scene extraction results, the following analysis is conducted. Figure 3 As shown in (b), the voltage crossing frequency of scenario C is significantly higher than that of other scenarios. Figure 5 As shown in the figure, the distribution network has upper and lower limits at different times, and the node voltage per unit value fluctuation range is [0.9075, 1.1240], which is far beyond the distribution network allowable voltage deviation range [0.93, 1.07]. However, compared with the node voltage per unit value fluctuation ranges of scenarios A, B and D, which are [0.9239, 1.1221], [0.9257, 1.1121], and [0.911, 1.1102] respectively, it can be seen that scenario C has the largest fluctuation range. Finally, combined with Figure 4 Source-load data analysis found that when operating scenario C occurs, there is a large and volatile wind speed, a wide load fluctuation range, and an extreme mismatch in source-load timing, which leads to a large fluctuation in the net load of the distribution network, and some reverse power is sent to the head end of the distribution network, which is easy to cause continuous overvoltage. In summary, it can be seen that the scenario C extracted by the present invention is an extreme scenario and is correct and effective.

[0061] Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solution of the present invention can be modified or replaced by equivalents without departing from the purpose and scope of the technical solution, which should be included in the scope of the claims of the present invention.

Claims

1. A method for extracting extreme scenarios of source and load in distribution network with high proportion of distributed wind power access, characterized by: The following steps are involved: S1: Obtain historical wind power output data and distribution network load data from the power dispatching department and establish a source-load uncertainty model; S11: Establish wind power output uncertainty model through KL divergence method; S12: Establish load uncertainty model through confidence interval method; S2: Based on the source-load uncertainty model, a large number of distribution network source-load data sets are randomly generated using the Monte Carlo method; S3: Use the source-load balance entropy index to extract extreme scenarios of the distribution network.

2. The method for extracting extreme scenarios of source and load in distribution network with high proportion of distributed wind power access according to claim 1 is characterized in that: In S11, the KL divergence method is used to establish a wind power output uncertainty model, and the specific steps are: ① Reference probability distribution P0 There are N wind farm historical power data samples, among which the xth data A x = {P WF,1 ,P WF,2 ,…,P WF,T }, each data sample has T sampling moments, and the power is divided into L power intervals with an interval of ε; e=P r / L (1) P0={ρ1,ρ2,...,ρ L } (4) Among them, P r is the installed capacity of distributed wind farm; ρ l is the average probability of interval l corresponding to N data samples; is a 0-1 variable; A x The wind farm power at the next moment t; ②Collection distance d KL in, is a chi-square distribution with L-1 degrees of freedom α * The upper quantile is used to ensure that the actual distribution of wind power output probability distribution is not less than α * The probability of ③ Wind power output probability distribution set G KL divergence is used to describe the distance between distribution functions. The smaller the distance, the more similar the distributions are. KL distance D KL for: Among them, ξ is an uncertain variable; п is the sample space of ξ; f(ξ) and f0(ξ) are the probability density functions of P and P0 respectively; Assume that the KL distance of the reference distribution P0 does not exceed the threshold d KL The set of all distribution functions of is G; G={P|D KL (P||P0)≤d KL } (7) Among them, when d KL >0, G contains infinite distribution functions; when d KL When it approaches 0, G=P0.

3. The method for extracting extreme scenarios of source and load in distribution network with high proportion of distributed wind power access according to claim 1 is characterized in that: In S12, the load uncertainty is characterized by the impact of price-based demand response, and the demand price elasticity coefficient e is established according to the definition of price-based demand response. xy for: Among them, q x , Δq x are the power consumption and power consumption change in time period x respectively; Θ y , ΔΘ y are the electricity price and electricity price change in period y respectively; When x=y, e xy is the self-elastic coefficient and e xy <0; when x≠y, e xy is the cross elastic coefficient and e xy >0; the elastic coefficients at each moment in the same period are the same; according to formula (8), the multi-period elastic coefficient matrix E is: According to the matrix E, the updated value of the load due to electricity price in each period is calculated; Among them, q t ′ is the electricity price load power consumption in time period t.

4. The method for extracting extreme scenarios of source and load in distribution network with high proportion of distributed wind power access according to claim 3 is characterized in that: In S12, the confidence interval method is used to deal with the uncertainty of the load; there are multiple confidence intervals under the same confidence level, and infinite load distributions are obtained through calculation of historical load samples; if there is a set of load post-data q t ′ is in the range of [0,ξ2], when there exists P(ξ1≤q t When ′≤ξ2)=1-α, [ξ1,ξ2] is called the confidence interval corresponding to the confidence level 1-α, ξ1 and ξ2 are called the lower and upper confidence limits under the confidence level 1-α, respectively. At this time, the corresponding load is 5. The method for extracting extreme scenarios of source and load in distribution network with high proportion of distributed wind power access according to any one of claims 2 to 4, characterized in that: In S2, according to the source-load uncertainty model, k source-load operation scenarios are randomly generated through Monte Carlo simulation, and then the set distance d is determined. KL Then the probability distribution set G is obtained, and k distribution functions are randomly selected from it to generate the corresponding number of wind power output scenarios At the same time, k loads with confidence level 1-α are obtained 6. The method for extracting extreme scenarios of source and load in distribution network with high proportion of distributed wind power access according to claim 1 is characterized in that: In S3, the source-charge balance entropy H is used s-l , and calculate H for each scene s-l , where the largest H s-l The source-load scenario corresponding to the value is the extreme scenario.

7. The method for extracting extreme scenarios of source and load in distribution network with high proportion of distributed wind power access according to claim 6 is characterized in that: In S3, the source-load balance entropy calculation process is as follows: Assume that the source-load scenario f includes f={f1, f2, ..., f k }, the timing net load Normalized to Finally, the source-charge balance entropy is obtained The calculation formula is as follows: in, is the net load at time t under scenario f; and are the load of node i with confidence level 1-α and the grid-connected power of the jth distributed wind farm at time t under scenario f; n is the total number of nodes in the distribution network; m is the number of distributed wind farms; let The probability distribution of {P f },Right now in

Images