WO2018209913A1 - 一种考虑电压质量的配电网台区缺供电量预测方法 - Google Patents

一种考虑电压质量的配电网台区缺供电量预测方法 Download PDF

Info

Publication number
WO2018209913A1
WO2018209913A1 PCT/CN2017/111344 CN2017111344W WO2018209913A1 WO 2018209913 A1 WO2018209913 A1 WO 2018209913A1 CN 2017111344 W CN2017111344 W CN 2017111344W WO 2018209913 A1 WO2018209913 A1 WO 2018209913A1
Authority
WO
WIPO (PCT)
Prior art keywords
power
voltage
station
area
station area
Prior art date
Application number
PCT/CN2017/111344
Other languages
English (en)
French (fr)
Inventor
欧阳森
刘丽媛
陈丹伶
Original Assignee
华南理工大学
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 华南理工大学 filed Critical 华南理工大学
Publication of WO2018209913A1 publication Critical patent/WO2018209913A1/zh

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q50/00—Systems or methods specially adapted for specific business sectors, e.g. utilities or tourism
    • G06Q50/06—Electricity, gas or water supply

Definitions

  • the invention relates to the field of electric power technology, in particular to a method for predicting a shortage of power supply in a distribution network area considering voltage quality.
  • the user's load is interrupted or the load is reduced due to insufficient system power supply, power failure, etc., and the reduced power supply amount is the power supply shortage of the power system to the user.
  • the shortage of power supply in the existing distribution network only the situation of power limitation and power failure is considered.
  • the power shortage caused by the low supply voltage has not been included in the statistical range.
  • current voltage-sensitive devices such as PLCs, computers, AC contactors, and motors are widely used in industrial and agricultural production. Low voltages often cause user equipment to be out of service or unable to start, and it has become an important reason for users to passively reduce load.
  • the present invention provides a power distribution considering voltage quality.
  • the method for predicting the shortage of power supply in the network area is not limited to a power distribution considering voltage quality.
  • a method for predicting power shortage in a distribution network area considering voltage quality includes the following steps:
  • S1 obtains the characteristic indicators of each station in the distribution network during the set statistical time period, and the industry classification and power consumption of each user;
  • S2 calculates the total electricity consumption of each industry in the Taiwan area according to the industry classification of the user
  • S3 uses fuzzy C-means clustering algorithm to cluster all the stations in the distribution network according to the characteristics of the station area, and select the area closest to the center sample as the typical station area;
  • S4 conducts voltage monitoring on typical stations, calculates the voltage pass probability of each level, and represents the voltage quality of such stations with the low voltage probability of the typical station;
  • S5 counts the operation status of the main electrical equipment of users in each industry under the low voltage level, and determines the correlation table between the low voltage amplitude and the loss rate.
  • S6 calculates the hidden power supply caused by the low voltage according to the voltage qualification rate of each stage of the station and the total power consumption of each industry;
  • S7 combines the user's power outage and power limit record, and calculates the total power shortage of the station area after considering the hidden power shortage caused by the low voltage.
  • the characteristic indicators of each station in the distribution network include voltage qualification rate, annual maximum load rate, power supply radius and line power factor.
  • a fuzzy C-means clustering algorithm is used to cluster all the stations in the distribution network according to the characteristic indicators of the station, and multiple stations are obtained, and the station area closest to the center sample is selected as a typical station area, specifically Including the following steps:
  • the cluster validity refers to the ratio of the intra-class compactness to the inter-class separation degree, which is recorded as V xie , and the calculation formula is:
  • U is the membership matrix
  • V is the cluster center matrix
  • m is the number of samples
  • ie the number of stations
  • n is the number of clusters
  • ⁇ is the fuzzy factor
  • u ij is the element in the U matrix
  • v i is the V matrix The i-th row element, The corresponding n is the best cluster when the V xie calculated value is minimized;
  • each type of station area calculates the Euclidean distance between each station area and the sample center of this type. Select each type of station area and select the smallest station area as a typical station area.
  • the S4 performs voltage monitoring on a typical station area, calculates a voltage pass probability of each level, and represents a voltage quality of such a stage area with a low voltage probability of a typical station area;
  • S4.1 performs voltage detection on n typical stations obtained by S3, and records the power supply voltage of the station every T minutes;
  • S4.2 calculates the voltage pass rate VER k% of all the power supply voltages in the statistical time period.
  • the amplitude of the power supply voltage in the station is as low as K% ⁇ (K + 10)% of the rated voltage.
  • the expression is:
  • C K represents the number of time nodes whose monitoring voltage is greater than or equal to K%Ue and less than (K+10)% Ue; Ue represents the rated supply voltage of the station; and N represents the total number of days included in the set statistical period.
  • the loss rate is specifically as follows:
  • the calculation calculates a hidden power shortage caused by a low voltage, specifically:
  • the load loss rate of the jth industry user is A Kj
  • the ratio of the time when the power supply voltage of a certain area is low to K%Ue is the same as the total statistical time.
  • the K% voltage pass rate of the typical station area is VER k%
  • the calculation formula of the hidden power supply amount ⁇ Q caused by the low voltage in the station area is:
  • the S7 combines the user power outage and the power limit record to calculate the total power shortage of the station area after considering the hidden power shortage caused by the low voltage, specifically:
  • the formula is:
  • ⁇ k i represents the load rate difference of the transformer in the station area before or after the ith power failure or power limitation
  • t i represents the duration of the ith power failure or power limitation
  • S represents the transformer of the station area.
  • the invention quantitatively calculates the problem of hidden power shortage caused by voltage quality for the first time, enriches the meaning of power shortage, and makes the statistical work of power shortage more truly and comprehensively reflect the actual power consumption of the user; the method is practical and economical. Well, it can provide guidance for the reliability and power quality improvement work of power supply companies.
  • the calculation method for the power supply shortage in the distribution network area designed by the present invention takes into account the influence of voltage quality for the first time, broadens the statistical range of the power supply shortage, and provides a quantitative indicator to reflect that the user cannot use the voltage due to the low voltage.
  • the electric equipment is forced to reduce the load. Compared with the traditional power supply calculation method, it can more fully evaluate the power supply reliability and power quality level of the station area, and better reflect the actual power consumption experience of the user, so that the user is convinced;
  • the calculation method of the hidden power supply quantity considering the voltage quality in the design of the invention not only considers the influence of the low voltage amplitude on the user's load loss rate, but also takes into account the difference in voltage quality sensitivity characteristics of the main electrical equipment in different industries.
  • the prediction result of hidden power supply has good accuracy and credibility;
  • the typical platform selection method based on fuzzy C-means clustering algorithm designed by the present invention can effectively solve the problem that the distribution network has a large number of points and the overall monitoring and research cost is too high, and adopts the characteristic indicators related to voltage quality.
  • Clustering in place of a small number of typical station area voltage monitoring and user load characteristics tracking statistics to replace the full coverage monitoring statistics, greatly reducing equipment costs and statistical workload, with outstanding economic and efficiency.
  • Fig. 1 is a flow chart showing the operation of a method for predicting the shortage of power supply in a distribution network in consideration of voltage quality according to the present invention.
  • a method for predicting power shortage in a distribution network area considering voltage quality includes the following steps:
  • S1 obtains the characteristic indicators of each station in the distribution network during the set statistical time period, and the industry classification and power consumption of each user;
  • the characteristic index of each station in the distribution network is an indicator parameter related to the overall voltage quality level of the station, including voltage qualification rate, annual maximum load rate, power supply radius and line power factor.
  • S2 calculates the total electricity consumption of each industry in the Taiwan area according to the industry classification of the user
  • the total electricity consumption of the jth industry is the sum of the electricity consumption of the users in the j-sector industry in the statistical period. For Qj.
  • S3 uses fuzzy C-means clustering algorithm to cluster all the stations in the distribution network according to the characteristics of the station area, and select the area closest to the center sample as the typical station area;
  • the fuzzy C-means clustering algorithm is used to cluster all the stations in the distribution network according to the characteristics of the station area, and multiple stations are obtained.
  • the station area closest to the center sample is selected as the typical station area, which includes the following steps:
  • the cluster validity refers to the ratio of the intra-class compactness to the inter-class separation degree, which is recorded as V xie , and the calculation formula is:
  • U is the membership matrix
  • V is the cluster center matrix
  • m is the number of samples
  • ie the number of stations
  • n is the number of clusters
  • ⁇ is the fuzzy factor
  • u ij is the element in the U matrix
  • v i is the V matrix The i-th row element, The corresponding n is the best cluster when the V xie calculated value is minimized;
  • each type of station area calculates the Euclidean distance between each station area and the sample center of this type. Select each type of station area and select the smallest station area as a typical station area.
  • S4 conducts voltage monitoring on a typical station area, calculates the voltage pass probability of each level, and represents the voltage quality of such a station area with a low voltage probability of a typical station area, specifically;
  • S4.1 performs voltage detection on n typical stations obtained by S3, and records the power supply voltage of the station every T minutes;
  • S4.2 calculates the voltage pass rate VER k% of all the power supply voltages in the statistical time period.
  • the amplitude of the power supply voltage in the station is as low as K% ⁇ (K + 10)% of the rated voltage.
  • the expression is:
  • C K represents the number of time nodes whose monitoring voltage is greater than or equal to K%Ue and less than (K+10)% Ue; Ue represents the rated supply voltage of the station; and N represents the total number of days included in the set statistical period.
  • S5 counts the operation status of the main electrical equipment of users in each industry under the low voltage level, and determines the correlation table between the low voltage amplitude and the loss rate.
  • the loss rate is specifically as follows:
  • S6 calculates the hidden power supply caused by the low voltage according to the voltage qualification rate of each stage of the station and the total power consumption of each industry;
  • the load loss rate of the jth industry user is A Kj
  • the ratio of the time when the power supply voltage of a certain area is low to K%Ue is the same as the total statistical time.
  • the K% voltage pass rate of the typical station area is VER k%
  • the calculation formula of the hidden power supply amount ⁇ Q caused by the low voltage in the station area is:
  • S7 combines the user's power outage and power limit record, and calculates the total power shortage of the station area after considering the hidden power shortage caused by the low voltage.
  • the formula is:
  • ⁇ k i represents the load rate difference of the transformer in the station area before or after the ith power failure or power limitation
  • t i represents the duration of the ith power failure or power limitation
  • S represents the transformer of the station area.
  • a low-voltage distribution network has a total of 10 power supply stations, which are recorded as stations 1 to 10. Get all the power first
  • the characteristic index data of the Taiwan area in the past year namely the voltage pass rate of the station area, the annual maximum load rate, the power supply radius, and the line power factor, are shown in Table 1.
  • the industry to which the user belongs is divided into five categories: sensitive industry, general industry, sensitive commercial agriculture, general commercial agriculture, and residents, and the industry classification and power consumption of the customers in each station area are counted.
  • Table 2 is the statistical data of the electricity consumption of industry users in Taiwan.
  • the fuzzy C-means clustering algorithm is used to cluster all the stations in the distribution network according to the characteristics of the station area, and the station area closest to the center sample is selected as the typical station area.
  • the clustering validity index V xie is the minimum, and the optimal cluster number is determined to be 3.
  • 10 power supply stations are grouped into 3 categories according to the characteristic index data of Table 1, and various cluster centers are determined.
  • the index values of the three types of station clustering centers, the number of the station numbers included, and the typical station area are shown in Table 3.
  • step S4 voltage monitoring is performed on three typical stations (table area 1, station area 2 and station area 6), the power supply voltage of the station area is recorded every 15 minutes, and the voltage pass rate of each stage is calculated based on the one-year monitoring data.
  • the settlement results are shown in Table 4.
  • the hidden power shortage caused by the low voltage is calculated from the voltage qualification rate of each stage of the station and the total power consumption of each industry.
  • the voltage pass rate of each stage of the atypical station area the voltage pass rate of each stage of the typical station area classified by the station area can be approximated.
  • the hidden power shortage ⁇ Q of the station area 1 is determined by the industry power consumption of Table 2, the voltage qualification rate of each level of Table 4, and the load rejection rate of each industry under the different voltages of Table 5. Calculated.
  • step S7 according to the user power outage and power limit record, obtain the transformer load rate difference, the duration and the rated capacity of the transformer before and after each power outage or power limit, and calculate the stage after considering the hidden power shortage caused by the low voltage.
  • the total power supply shortage in the area is AENS.
  • the power factor ⁇ of station area 1 is 0.9
  • the rated capacity S of the station area transformer is 1250KVA.
  • a total of 8 power failure accidents occur during the statistical period, and 2 power cuts are recorded.
  • the duration of each power failure or power limit and the load ratio of the front and rear transformers are recorded. Poor, it is calculated that the traditional power shortage of the power failure or power limitation in the station area is 15.75MW ⁇ h, and the total power supply shortage is 446.664MW ⁇ h.
  • the hidden power supply caused by voltage quality is related to the low voltage level and the proportion of electricity consumption of users in sensitive industries in the Taiwan area.
  • the user's power equipment is not available due to voltage quality, especially low voltage, and the hidden power supply is large. Part of the lack of power supply can not be reflected in the traditional power supply statistics, which is not conducive to guiding the planning and reform of power supply enterprises. Work.
  • the embodiment further demonstrates that the method for predicting the power shortage of the power distribution network in the distribution network considering the voltage quality according to the present invention can quantitatively predict the hidden power shortage caused by the voltage quality, enriching the meaning of the power shortage.
  • the lack of power supply statistics work more realistically and comprehensively reflects the actual power consumption of users, and provides reference for the reliability and power quality improvement of power supply enterprises.

Abstract

一种考虑电压质量的配电网台区缺供电量预测方法，包括获取设定统计时间段内，配电网中各台区的特征指标，每个用户所属的行业分类和用电量(S1)，根据用户所属行业分类，计算台区内各行业的总用电量(S2)，采用模糊C均值聚类算法，根据台区特征指标对配电网中所有台区进行聚类，选择与中心样本最近的台区作为典型台区(S3)，对典型台区进行电压监测，计算各级电压合格概率(S4)，以典型台区的电压偏低概率代表此类台区的电压质量，统计每个电压偏低等级下各行业用户主要用电设备的运行情况，确定电压偏低幅值与失负荷率的关联关系表，根据台区的各级电压合格率和各行业总用电量，计算由电压偏低引起的隐性缺供电量(S6)进一步得到台区总缺供电量。

Description

一种考虑电压质量的配电网台区缺供电量预测方法 技术领域
本发明涉及电力技术领域，具体涉及一种考虑电压质量的配电网台区缺供电量预测方法。
背景技术
在供电可靠性领域中，由于系统电源不足、停电等原因造成用户负荷中断或负荷缩减，所减少的供电量就是电力系统对用户的缺供电量。在现有的配电网缺供电量计算中只考虑了限电和停电两种情况，对于供电电压偏低造成的缺供电量尚未纳入统计范围。事实上，目前PLC、计算机、交流接触器、电动机等电压敏感型设备广泛应用于工农业生产，电压偏低常常导致用户设备停运或无法启动，也成为了用户被动缩减负荷的一个重要原因。在高新企业和工业密集的城市中，电压偏低导致用户无法正常用电已经成为用户投诉的主要电能质量问题。随着电源和主网建设的不断完善，电网的限电和停电事件将大大减少，电压质量导致的隐性缺供电量的占比也将愈发突出。因此，电压质量问题及其导致的隐性缺供电量必将引起供电企业和用户越来越高的重视。
在供电企业开展配电网改造时，预测评估考虑电压质量影响的台区缺供电量，将有助于快速准确筛选提升需求较高的台区，一次性发现并解决供电可靠性与电压偏低两大问题。现有研究针对配电台区电压偏低问题开展了偏低概率预测和对设备运行的影响分析，但对电压偏低引起的隐性缺供电量暂未有成熟的研究成果。目前的缺供电量统计方法仅计及由停电和限电导致的用户用电量缩减，未考虑用户由于电压质量问题而被迫减少的用电量，难以反映用户的实际用电体验，不便于指导电压质量问题的改造提升工作，存在缺陷与不足。
发明内容
为了克服现有技术中现有台区缺供电量统计方法未能全面计及电压偏低带来的隐性缺供电量，存在局限性的技术问题，本发明提供一种考虑电压质量的配电网台区缺供电量预测方法。
本发明采用如下技术方案：
一种考虑电压质量的配电网台区缺供电量预测方法，包括如下步骤：
S1获取设定统计时间段内，配电网中各台区的特征指标，每个用户所属的行业分类和用电量；
S2根据用户所属行业分类，计算台区内各行业的总用电量；
S3采用模糊C均值聚类算法，根据台区特征指标对配电网中所有台区进行聚类，选择与中心样本最近的台区作为典型台区；
S4对典型台区进行电压监测，计算各级电压合格概率，以典型台区的电压偏低概率代表此类台区的电压质量；
S5在典型台区中，统计每个电压偏低等级下各行业用户主要用电设备的运行情况，确定电压偏低幅值与失负荷率的关联关系表；
S6根据台区的各级电压合格率和各行业总用电量，计算由电压偏低引起的隐性缺供电量；
S7结合用户停电和限电记录，计算考虑电压偏低引起的隐性缺供电量后的台区总缺供电量。
所述配电网中各台区的特征指标包括电压合格率、年最大负载率、供电半径及线路功率因数。
所述S2中，计算台区内各行业的总用电量，具体是指：假设一个供电台区所有用户分为j个行业，第j行业总用电量是台区中所有行业分类为第j行业的用户在统计时间段内的用电量的总和，记为Qj。
所述S3中，采用模糊C均值聚类算法，根据台区特征指标对配电网中所有台区进行聚类，得到多个台区，选择与中心样本最近的台区作为典型台区，具体包括如下步骤：
S3.1配电网内有m个台区，以聚类有效性为目标，确定最优的聚类数n，
其中所述聚类有效性是指类内紧凑度与类间分离度的比值，记为Vxie，计算公式为：
Figure PCTCN2017111344-appb-000001
其中，U是隶属矩阵，V是聚类中心矩阵，m是样本个数，即台区数，n是聚类数，α是模糊因子，uij是U矩阵中的元素，vi是V矩阵中的第i行元素，
Figure PCTCN2017111344-appb-000002
Vxie计算值取得最小时对应的n就是最佳聚类；
S3.2，采用模糊C均值聚类算法，根据电压合格率、年最大负载率、供电 半径及线路功率因数四项特征指标将m个台区分为n类，并计算每类的样本中心；
S3.3在每类台区中，计算各台区和该类样本中心的欧式距离，选择每类台区中，选择最小的台区为典型台区。
所述S4对典型台区进行电压监测，计算各级电压合格概率，以典型台区的电压偏低概率代表此类台区的电压质量；
S4.1对S3得到的n个典型台区进行电压检测，每T分钟记录一次台区的供电电压；
S4.2由统计时间段内的所有供电电压监测数据，计算各级电压合格率VERk％，其中，各级电压合格率VERk％，K＝90,80,70,…,10是指：统计时间内，台区供电电压幅值偏低至K％～(K+10)％额定电压的发生概率，表达式为：
Figure PCTCN2017111344-appb-000003
式中：CK表示监测电压大于等于K％Ue且小于(K+10)％Ue的时间节点数；Ue表示台区额定供电电压；N表示设定统计时段包含的总天数。
所述失负荷率，具体为：
台区供电电压偏低至K％Ue，K＝90,80,70,…,10时，台区内各行业用户因电压质量而无法正常使用的负荷占该行业总负荷的百分比，即失负荷率。
所述计算由电压偏低引起的隐性缺供电量，具体为：
假设台区供电电压偏低至K％Ue时，第j行业用户的失负荷率为AKj，某台区供电电压偏低至K％Ue的时间与总统计时间的比值近似等于其所属台区分类的典型台区的K％电压合格率VERk％，该台区由电压偏低引起的隐性缺供电量ΔQ的计算公式为：
Figure PCTCN2017111344-appb-000004
所述S7结合用户停电和限电记录，计算考虑电压偏低引起的隐性缺供电量后的台区总缺供电量，具体为：
根据用户停电和限电记录，获取每次停电或限电的前后变压器负载率差、持续时间和台区变压器额定容量，计算考虑电压偏低引起的隐性缺供电量后的台区总缺供电量AENS，计算公式为：
Figure PCTCN2017111344-appb-000005
其中，
Figure PCTCN2017111344-appb-000006
表示该台区的功率因数，Δki表示该台区变压器在第i次停电或限电前后的负载率差，ti表示第i次停电或限电的持续时间，S表示该台区变压器的额定容量。
本发明首次定量计算了电压质量造成的隐性缺供电量问题，丰富了缺供电量的含义，使缺供电量统计工作更真实全面地反映用户实际用电情况；该方法实用性强、经济性好，可为供电企业的可靠性和电能质量提升工作提供指导。
本发明的有益效果：
1、本发明设计的配电网台区缺供电量计算方法首次计及了电压质量的影响，拓宽了缺供电量的统计范围，提供了一个量化指标来反映用户由于电压偏低无法正常使用用电设备而被迫削减负荷的情况，与传统的缺供电量计算方法相比，能够更全面地评估台区的供电可靠性和电能质量水平，更能反映用户实际用电体验，使用户信服；
2、本发明设计的考虑电压质量的隐性缺供电量计算方法，不仅考虑了电压偏低幅值对用户失负荷率的影响，而且兼顾了不同行业主要用电设备的电压质量敏感特性差异，隐性缺供电量的预测结果具有较好的准确性和可信度；
3、本发明设计的基于模糊C均值聚类算法的典型台区选取方法，可以有效解决配电网台区点多量大、全面监测调研成本过高的问题，采用与电压质量相关的特征指标进行聚类，以对少量典型台区的电压监测和用户负荷特性跟踪统计来代替全覆盖的监测统计，大大减少了设备成本和统计工作量，具有突出的经济性和高效性。
附图说明
图1是本发明的一种考虑电压质量的配电网台区缺供电量预测方法的工作流程图。
具体实施方式
下面结合实施例及附图，对本发明作进一步地详细说明，但本发明的实施方式不限于此。
实施例
如图1所示，一种考虑电压质量的配电网台区缺供电量预测方法，包括如下步骤：
S1获取设定统计时间段内，配电网中各台区的特征指标，每个用户所属的行业分类和用电量；
所述配电网中各台区的特征指标是与台区整体电压质量水平相关的指标参数，包括电压合格率、年最大负载率、供电半径及线路功率因数。
S2根据用户所属行业分类，计算台区内各行业的总用电量；
具体是指：假设一个供电台区所有用户分为j个行业，第j行业总用电量是台区中所有行业分类为第j行业的用户在统计时间段内的用电量的总和，记为Qj。
S3采用模糊C均值聚类算法，根据台区特征指标对配电网中所有台区进行聚类，选择与中心样本最近的台区作为典型台区；
采用模糊C均值聚类算法，根据台区特征指标对配电网中所有台区进行聚类，得到多个台区，选择与中心样本最近的台区作为典型台区，具体包括如下步骤：
S3.1配电网内有m个台区，以聚类有效性为目标，确定最优的聚类数n，
其中所述聚类有效性是指类内紧凑度与类间分离度的比值，记为Vxie，计算公式为：
Figure PCTCN2017111344-appb-000007
其中，U是隶属矩阵，V是聚类中心矩阵，m是样本个数，即台区数，n是聚类数，α是模糊因子，uij是U矩阵中的元素，vi是V矩阵中的第i行元素，
Figure PCTCN2017111344-appb-000008
Vxie计算值取得最小时对应的n就是最佳聚类；
S3.2，采用模糊C均值聚类算法，根据电压合格率、年最大负载率、供电半径及线路功率因数四项特征指标将m个台区分为n类，并计算每类的样本中心；
S3.3在每类台区中，计算各台区和该类样本中心的欧式距离，选择每类台区中，选择最小的台区为典型台区。
S4对典型台区进行电压监测，计算各级电压合格概率，以典型台区的电压偏低概率代表此类台区的电压质量，具体为；
S4.1对S3得到的n个典型台区进行电压检测，每T分钟记录一次台区的供电电压；
S4.2由统计时间段内的所有供电电压监测数据，计算各级电压合格率VERk％，其中，各级电压合格率VERk％，K＝90,80,70,…,10是指：统计时间内，台 区供电电压幅值偏低至K％～(K+10)％额定电压的发生概率，表达式为：
Figure PCTCN2017111344-appb-000009
式中：CK表示监测电压大于等于K％Ue且小于(K+10)％Ue的时间节点数；Ue表示台区额定供电电压；N表示设定统计时段包含的总天数。
S5在典型台区中，统计每个电压偏低等级下各行业用户主要用电设备的运行情况，确定电压偏低幅值与失负荷率的关联关系表；
所述失负荷率，具体为：
台区供电电压偏低至K％Ue，K＝90,80,70,…,10时，台区内各行业用户因电压质量而无法正常使用的负荷占该行业总负荷的百分比，即失负荷率。
S6根据台区的各级电压合格率和各行业总用电量，计算由电压偏低引起的隐性缺供电量；
具体为：
假设台区供电电压偏低至K％Ue时，第j行业用户的失负荷率为AKj，某台区供电电压偏低至K％Ue的时间与总统计时间的比值近似等于其所属台区分类的典型台区的K％电压合格率VERk％，该台区由电压偏低引起的隐性缺供电量ΔQ的计算公式为：
Figure PCTCN2017111344-appb-000010
S7结合用户停电和限电记录，计算考虑电压偏低引起的隐性缺供电量后的台区总缺供电量。
具体为：
根据用户停电和限电记录，获取每次停电或限电的前后变压器负载率差、持续时间和台区变压器额定容量，计算考虑电压偏低引起的隐性缺供电量后的台区总缺供电量AENS，计算公式为：
Figure PCTCN2017111344-appb-000011
其中，
Figure PCTCN2017111344-appb-000012
表示该台区的功率因数，Δki表示该台区变压器在第i次停电或限电前后的负载率差，ti表示第i次停电或限电的持续时间，S表示该台区变压器的额定容量。
本实施例中，
某低压配电网共有10个供电台区，分别记为台区1～10。首先获取所有供电 台区在近一年的特征指标数据，即台区的电压合格率、年最大负载率、供电半径、线路功率因数，如表1所示。
表1 10个供电台区的特征指标数据
Figure PCTCN2017111344-appb-000013
在本实施例中将用户所属行业分为5类：敏感工业、一般工业、敏感商业农业、一般商业农业和居民，统计各个台区的客户所属的行业分类和用电量。
根据统计情况，计算台区内各行业用户的总用电量，表2为近一年台区1的行业用户用电量统计表。
表2 台区1的行业用户用电量统计表
Figure PCTCN2017111344-appb-000014
根据步骤S3，采用模糊C均值聚类算法，根据台区特征指标对配电网中所 有台区进行聚类，选择与中心样本最近的台区作为典型台区。首先，以聚类有效性指标Vxie最小为目标，确定最优聚类数为3。利用模糊C均值聚类算法，根据表1的特征指标数据将10个供电台区聚为3类，确定各类的聚类中心。3类台区聚类中心的指标数值、所包含的台区编号以及典型台区如表3所示。
表3 台区聚类结果及典型台区
Figure PCTCN2017111344-appb-000015
根据步骤S4，对3个典型台区(台区1、台区2和台区6)进行电压监测，每15分钟记录一次台区的供电电压，基于1年的监测数据计算各级电压合格率，结算结果如表4所示。
表4 3个典型台区的各级电压合格率统计结果
Figure PCTCN2017111344-appb-000016
根据3个台区监测结果统计出电压偏低幅值与失负荷率的关联关系，如表5所示。
表5 不同电压偏低幅值下各行业的失负荷率
Figure PCTCN2017111344-appb-000017
根据步骤S6，由台区的各级电压合格率和各行业总用电量，计算由电压偏低引起的隐性缺供电量。对于非典型台区的各级电压合格率，可以用其所属台区分类的典型台区的各级电压合格率近似表示。在本实施例中，台区1的隐性缺供电量ΔQ由表2的行业用电量、表4的各级电压合格率和表5的不同电压偏低幅值下各行业的失负荷率计算得出。
Figure PCTCN2017111344-appb-000018
根据步骤S7，根据用户停电和限电记录，获取每次停电或限电的前后变压器负载率差、持续时间和台区变压器额定容量，计算考虑电压偏低引起的隐性缺供电量后的台区总缺供电量AENS。台区1的功率因数φ为0.9，台区变压器额定容量S为1250KVA，在统计时段内共发生8次停电事故，2次限电，记录每次停电或限电的持续时间和前后变压器负载率差，计算得出台区1考虑停电或限电情况的传统缺供电量为15.75MW·h，总缺供电量为446.664MW·h。
Figure PCTCN2017111344-appb-000019
由隐性缺供电量和总缺供电量的计算过程可以看出，电压质量带来的隐性缺供电量与电压偏低程度、台区中敏感行业用户的用电量占比有关。在含有较多敏感行业用户且供电电压质量较差的台区或配电网中，由电压质量尤其是电压偏低问题造成的用户用电设备不可用情况突出，隐性缺供电量大，这部分缺供电量在传统缺供电量统计中无法体现，不利于指导供电企业的规划和改造工 作。
本实施例进一步表明了：本发明所述的考虑电压质量的配电网台区缺供电量预测方法能够定量预测了电压质量造成的隐性缺供电量问题，丰富了缺供电量的含义，使缺供电量统计工作更真实全面地反映用户实际用电情况，为供电企业的可靠性和电能质量提升工作提供参考。
上述实施例为本发明较佳的实施方式，但本发明的实施方式并不受所述实施例的限制，其他的任何未背离本发明的精神实质与原理下所作的改变、修饰、替代、组合、简化，均应为等效的置换方式，都包含在本发明的保护范围之内。

Claims (8)

  1. 一种考虑电压质量的配电网台区缺供电量预测方法，其特征在于，包括如下步骤：
    S1获取设定统计时间段内，配电网中各台区的特征指标，每个用户所属的行业分类和用电量；
    S2根据用户所属行业分类，计算台区内各行业的总用电量；
    S3采用模糊C均值聚类算法，根据台区特征指标对配电网中所有台区进行聚类，选择与中心样本最近的台区作为典型台区；
    S4对典型台区进行电压监测，计算各级电压合格概率，以典型台区的电压偏低概率代表此类台区的电压质量；
    S5在典型台区中，统计每个电压偏低等级下各行业用户主要用电设备的运行情况，确定电压偏低幅值与失负荷率的关联关系表；
    S6根据台区的各级电压合格率和各行业总用电量，计算由电压偏低引起的隐性缺供电量；
    S7结合用户停电和限电记录，计算考虑电压偏低引起的隐性缺供电量后的台区总缺供电量。
  2. 根据权利要求1所述的配电网台区缺供电量预测方法，其特征在于，所述配电网中各台区的特征指标包括电压合格率、年最大负载率、供电半径及线路功率因数。
  3. 根据权利要求1所述的配电网台区缺供电量预测方法，其特征在于，所述S2中，计算台区内各行业的总用电量，具体是指：假设一个供电台区所有用户分为j个行业，第j行业总用电量是台区中所有行业分类为第j行业的用户在统计时间段内的用电量的总和，记为Qj。
  4. 根据权利要求1所述的配电网台区缺供电量预测方法，其特征在于，所述S3中，采用模糊C均值聚类算法，根据台区特征指标对配电网中所有台区进行聚类，得到多个台区，选择与中心样本最近的台区作为典型台区，具体包括如下步骤：
    S3.1配电网内有m个台区，以聚类有效性为目标，确定最优的聚类数n，
    其中所述聚类有效性是指类内紧凑度与类间分离度的比值，记为Vxie，计算公式为：
    Figure PCTCN2017111344-appb-100001
    其中，U是隶属矩阵，V是聚类中心矩阵，m是样本个数，即台区数，n是聚类数，α是模糊因子，uij是U矩阵中的元素，vi是V矩阵中的第i行元素，
    Figure PCTCN2017111344-appb-100002
    Vxie计算值取得最小时对应的n就是最佳聚类；
    S3.2，采用模糊C均值聚类算法，根据电压合格率、年最大负载率、供电半径及线路功率因数四项特征指标将m个台区分为n类，并计算每类的样本中心；
    S3.3在每类台区中，计算各台区和该类样本中心的欧式距离，选择每类台区中，选择最小的台区为典型台区。
  5. 根据权利要求1所述的配电网台区缺供电量预测方法，其特征在于，所述S4对典型台区进行电压监测，计算各级电压合格概率，以典型台区的电压偏低概率代表此类台区的电压质量；
    S4.1对S3得到的n个典型台区进行电压检测，每T分钟记录一次台区的供电电压；
    S4.2由统计时间段内的所有供电电压监测数据，计算各级电压合格率VERk％，其中，各级电压合格率VERk％，K＝90,80,70,…,10是指：统计时间内，台区供电电压幅值偏低至K％～(K+10)％额定电压的发生概率，表达式为：
    Figure PCTCN2017111344-appb-100003
    式中：CK表示监测电压大于等于K％Ue且小于(K+10)％Ue的时间节点数；Ue表示台区额定供电电压；N表示设定统计时段包含的总天数。
  6. 根据权利要求1所述的配电网台区缺供电量预测方法，其特征在于，所述失负荷率，具体为：
    台区供电电压偏低至K％Ue，K＝90,80,70,…,10时，台区内各行业用户因电压质量而无法正常使用的负荷占该行业总负荷的百分比，即失负荷率。
  7. 根据权利要求6所述的配电网台区缺供电量预测方法，其特征在于，所述计算由电压偏低引起的隐性缺供电量，具体为：
    假设台区供电电压偏低至K％Ue时，第j行业用户的失负荷率为AKj，某台区供电电压偏低至K％Ue的时间与总统计时间的比值近似等于其所属台区分类的典型台区的K％电压合格率VERk％，该台区由电压偏低引起的隐性缺供电量ΔQ的计算公式为：
    Figure PCTCN2017111344-appb-100004
  8. 根据权利要求1所述的配电网台区缺供电量预测方法，其特征在于，所述S7结合用户停电和限电记录，计算考虑电压偏低引起的隐性缺供电量后的台区总缺供电量，具体为：
    根据用户停电和限电记录，获取每次停电或限电的前后变压器负载率差、持续时间和台区变压器额定容量，计算考虑电压偏低引起的隐性缺供电量后的台区总缺供电量AENS，计算公式为：
    Figure PCTCN2017111344-appb-100005
    其中，
    Figure PCTCN2017111344-appb-100006
    表示该台区的功率因数，Δki表示该台区变压器在第i次停电或限电前后的负载率差，ti表示第i次停电或限电的持续时间，S表示该台区变压器的额定容量。
PCT/CN2017/111344 2017-05-18 2017-11-16 一种考虑电压质量的配电网台区缺供电量预测方法 WO2018209913A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201710353190.9A CN107330540B (zh) 2017-05-18 2017-05-18 一种考虑电压质量的配电网台区缺供电量预测方法
CN201710353190.9 2017-05-18

Publications (1)

Publication Number Publication Date
WO2018209913A1 true WO2018209913A1 (zh) 2018-11-22

Family

ID=60193551

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2017/111344 WO2018209913A1 (zh) 2017-05-18 2017-11-16 一种考虑电压质量的配电网台区缺供电量预测方法

Country Status (2)

Country Link
CN (1) CN107330540B (zh)
WO (1) WO2018209913A1 (zh)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110264107A (zh) * 2019-07-01 2019-09-20 国网安徽省电力有限公司滁州供电公司 一种基于大数据技术的台区线损率异常诊断方法
CN113434690A (zh) * 2021-08-25 2021-09-24 广东电网有限责任公司惠州供电局 基于聚类算法的用电预测评估方法、装置、系统及介质

Families Citing this family (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107330540B (zh) * 2017-05-18 2019-03-05 华南理工大学 一种考虑电压质量的配电网台区缺供电量预测方法
CN108256724B (zh) * 2017-11-27 2021-12-07 国网浙江省电力公司嵊泗县供电公司 一种基于动态行业系数的配电网可开放容量规划方法
CN109615207B (zh) * 2018-12-04 2021-11-02 国网山东省电力公司无棣县供电公司 一种低压台区供电负荷调配方法及系统
CN110288215A (zh) * 2019-06-17 2019-09-27 国网甘肃省电力公司兰州供电公司 一种基于熵权法的电力客户服务态势感知评价方法
CN111191966B (zh) * 2020-02-18 2023-07-28 南京工程学院 一种基于时空特性的配电网电压不合格时段识别方法

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2007109156A2 (en) * 2006-03-16 2007-09-27 Leveler Llc Site analysis methodology and device for testing electric power
CN104573853A (zh) * 2014-12-19 2015-04-29 华南理工大学 低压台区配电网电压质量评估与优化的系统化解决方法
CN105389636A (zh) * 2015-12-11 2016-03-09 河海大学 一种低压台区kfcm-svr合理线损预测方法
CN105760997A (zh) * 2016-02-16 2016-07-13 国网山东省电力公司经济技术研究院 基于模糊评价的配电网异常电压评估方法
CN107330540A (zh) * 2017-05-18 2017-11-07 华南理工大学 一种考虑电压质量的配电网台区缺供电量预测方法

Family Cites Families (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103903189B (zh) * 2014-03-20 2017-11-07 华南理工大学 基于模糊聚类的低压配网台区聚类方法
CN104408549A (zh) * 2014-10-31 2015-03-11 国家电网公司 一种城市配电网运行状态评估方法

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2007109156A2 (en) * 2006-03-16 2007-09-27 Leveler Llc Site analysis methodology and device for testing electric power
CN104573853A (zh) * 2014-12-19 2015-04-29 华南理工大学 低压台区配电网电压质量评估与优化的系统化解决方法
CN105389636A (zh) * 2015-12-11 2016-03-09 河海大学 一种低压台区kfcm-svr合理线损预测方法
CN105760997A (zh) * 2016-02-16 2016-07-13 国网山东省电力公司经济技术研究院 基于模糊评价的配电网异常电压评估方法
CN107330540A (zh) * 2017-05-18 2017-11-07 华南理工大学 一种考虑电压质量的配电网台区缺供电量预测方法

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
ZHANG, YING ET AL.: "Prediction Model of Low Voltage Area Based on Self-organizing Competitive Neural Network", WATER RESOURCES AND POWER, vol. 30, no. 12, 25 December 2012 (2012-12-25), ISSN: 1000-7709 *

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110264107A (zh) * 2019-07-01 2019-09-20 国网安徽省电力有限公司滁州供电公司 一种基于大数据技术的台区线损率异常诊断方法
CN113434690A (zh) * 2021-08-25 2021-09-24 广东电网有限责任公司惠州供电局 基于聚类算法的用电预测评估方法、装置、系统及介质
CN113434690B (zh) * 2021-08-25 2022-02-08 广东电网有限责任公司惠州供电局 基于聚类算法的用电预测评估方法、装置、系统及介质

Also Published As

Publication number Publication date
CN107330540B (zh) 2019-03-05
CN107330540A (zh) 2017-11-07

Similar Documents

Publication Publication Date Title
WO2018209913A1 (zh) 一种考虑电压质量的配电网台区缺供电量预测方法
CN106815709B (zh) 一种服务快速响应中心支撑系统及方法
CN110110881B (zh) 电力客户需求预测分析方法及系统
CN113159339A (zh) 一种基于大数据的一台区一指标线损管理方法及系统
CN108805443A (zh) 一种电网性能综合评价方法和系统
CN110927654B (zh) 一种智能电能表批次运行状态评价方法
CN111291076A (zh) 基于大数据的异常用水监测报警系统及其构建方法
CN107666148B (zh) 基于配变停电信号的线路故障研判方法
CN111191966B (zh) 一种基于时空特性的配电网电压不合格时段识别方法
CN109634942A (zh) 一种用能数据异常判断方法及装置
CN108490285B (zh) 一种基于电压降落法的低压台区线损率计算方法
CN110378510B (zh) 一种基于时间序列和层次聚合的配网物资需求预测方法
CN114878934A (zh) 一种电能耗数据异常预警方法
CN106651093A (zh) 一种用于低压集抄系统的智能复核管控方法
CN111967684B (zh) 一种基于大数据分析的计量资产主动配送方法
CN111552686B (zh) 一种电力数据质量评估方法及其装置
CN111127186A (zh) 一种基于大数据技术的客户信用等级评价体系的使用方法
CN114676931B (zh) 一种基于数据中台技术的电量预测系统
CN113327006A (zh) 一种满足差异化可靠性需求的配电网供电恢复系统和方法
CN114399192A (zh) 一种在线稽查自动化rpa系统
CN108492033A (zh) 电网客户集中投诉智能预警方法
Zuhri et al. Inequality in Indonesia's electricity access
CN112001551B (zh) 一种基于大用户电量信息的地市电网售电量预测方法
CN115115470A (zh) 一种基于排放因子法的绿色数据中心碳排放管理方法
Ya’An Application of artificial intelligence in computer network technology in the era of big data

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 17909959

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 17909959

Country of ref document: EP

Kind code of ref document: A1

32PN Ep: public notification in the ep bulletin as address of the adressee cannot be established

Free format text: NOTING OF LOSS OF RIGHTS PURSUANT TO RULE 112(1) EPC (EPO FORM 1205A DATED 08/05/2020)

122 Ep: pct application non-entry in european phase

Ref document number: 17909959

Country of ref document: EP

Kind code of ref document: A1