技术领域technical field
本申请涉及智能电网技术领域，尤其涉及一种基于主成分分析与EM距离的低压用户相别识别方法。The present application relates to the technical field of smart grid, in particular to a low-voltage user identification method based on principal component analysis and EM distance.
背景技术Background technique
配电网中低压用户大多采用单相电，随着用户的增加、维护、恢复、重新配置以及用户消费模式的变化，三相电压可能由平衡状态转为不平衡状态。负载不平衡会产生电能质量差，功率损耗增加、变压器使用寿命缩短等问题。准确识别用户相别可调整部分用户相别，尽可能达到三相负载平衡，降低配电线路损耗及低压供电线路耗电量，提高电能质量，提高配变寿命。Most low-voltage users in the distribution network use single-phase power. With the increase, maintenance, restoration, reconfiguration of users and changes in user consumption patterns, the three-phase voltage may change from a balanced state to an unbalanced state. Unbalanced loads can lead to problems such as poor power quality, increased power loss, and shortened transformer life. Accurate identification of user phase difference can adjust part of user phase difference to achieve three-phase load balance as much as possible, reduce distribution line loss and low-voltage power supply line power consumption, improve power quality, and increase distribution transformer life.
传统相位识别通过手动或主动注入信号实现，两种方式成本高、劳动密集、耗时且容易出错。错误的相位识别会导致拓扑检测、状态估计和故障定位检测的错误。目前，电力公司将配网中手动读数模拟仪表升级到智能仪表，提高了配电网络的可观察性和可控性。但是，现有智能电表大多只具备采集电压及电量功能，若只利用电压，判别结果准确率难以保证，导致现有的智能电表对低压用户相别识别准确性和选择性不高。Traditional phase identification is achieved by manually or actively injecting signals, both of which are costly, labor-intensive, time-consuming, and error-prone. Incorrect phase identification can lead to errors in topology detection, state estimation, and fault localization detection. At present, power companies have upgraded the manual reading analog meters in the distribution network to smart meters, which has improved the observability and controllability of the distribution network. However, most of the existing smart meters only have the function of collecting voltage and power. If only the voltage is used, the accuracy of the discrimination results cannot be guaranteed. As a result, the accuracy and selectivity of the existing smart meters for identifying low-voltage users are not high.
发明内容Contents of the invention
本申请提供了一种基于主成分分析与EM距离的低压用户相别识别方法，以解决现有的智能电表对低压用户相别识别准确性和选择性不高的问题。The present application provides a low-voltage subscriber identification method based on principal component analysis and EM distance to solve the problem of low accuracy and selectivity of existing smart meters for identifying low-voltage subscribers.
Earth Mover's Distance为推土机距离，简称EMD，也叫Wasserstein距离，用来表示两个分布的相似程度。Earth Mover's Distance is bulldozer distance, referred to as EMD, also known as Wasserstein distance, which is used to indicate the similarity of two distributions.
本申请提供的一种基于主成分分析与EM距离的低压用户相别识别方法，包括以下步骤：The application provides a low-voltage user identification method based on principal component analysis and EM distance, which includes the following steps:
采集关口表三相电压及低压用户电压；Collect the three-phase voltage of the gateway meter and the low-voltage user voltage;
对所述低压用户电压中的缺失数据采用牛顿法插值处理，得到高维数据；Using Newton's method to interpolate the missing data in the low-voltage user voltage to obtain high-dimensional data;
利用主成分分析将所述高维数据映射到低维空间，将所有低压用户电压形成矩阵，将矩阵中的所有特征去均值，得到协方差矩阵C；Using principal component analysis to map the high-dimensional data to a low-dimensional space, form a matrix of all low-voltage user voltages, and remove the mean value of all features in the matrix to obtain a covariance matrix C;
计算所述矩阵C的特征值和特征向量；Calculate the eigenvalues and eigenvectors of the matrix C;
根据所述特征值和特征向量计算得到主成分PC1、PC2并进行聚类，得到聚类后的各相用户的曲线；Calculate and obtain the principal components PC1 and PC2 according to the eigenvalues and eigenvectors, and perform clustering to obtain the clustered user curves of each phase;
根据所述曲线计算聚类中值；calculating the cluster median from the curve;
将所述聚类中值与关口表各相电压的EM距离进行归类分析得到识别结果。Classify and analyze the cluster median and the EM distance of each phase voltage of the gateway table to obtain the identification result.
可选的，所述低压用户电压中的缺失数据采用牛顿法插值处理，得到高维数据步骤包括：Optionally, the missing data in the low-voltage user voltage is interpolated using Newton's method, and the step of obtaining high-dimensional data includes:
设样本包含n个点{(x1，(x1)),…,(xn，f(xn))}，其中缺失点为(xi，f(xi))，则插值多项式f(xi)为：Suppose the sample contains n points {(x1 , (x1 )),…,(xn , f(xn ))}, where the missing point is (xi , f(xi )), then the interpolation polynomial f (xi ) is:
其中，f(xi)为牛顿插值所得函数值，根据所述插值多项式可以计算缺失数据。Wherein, f(xi ) is the function value obtained by Newton interpolation, and the missing data can be calculated according to the interpolation polynomial.
可选的，所述利用主成分分析将所述高维数据映射到低维空间，将所有低压用户电压形成矩阵，将矩阵中的所有特征去均值，得到协方差矩阵C的步骤包括：Optionally, the step of using principal component analysis to map the high-dimensional data to a low-dimensional space, forming a matrix of all low-voltage user voltages, removing the mean value of all features in the matrix, and obtaining the covariance matrix C includes:
利用主成分分析将高维数据映射到低维空间，设离散时域信号X包含M个样本{X1,X2,...,XM}，每个样本有N个特征，即各特征xj均有各自特征值；Using principal component analysis to map high-dimensional data to low-dimensional space, suppose the discrete time-domain signal X contains M samples {X1 ,X2 ,...,XM }, each sample has N features, namely Each feature xj has its own eigenvalue;
对所有特征去均值，将所有样本同一特征求均值，并将自身特征减去均值得到协方差矩阵C；Remove the mean value of all features, calculate the mean value of the same feature of all samples, and subtract the mean value from its own features to obtain the covariance matrix C;
其中，求均值公式如下：Among them, the average formula is as follows:
协方差矩阵C的表达式如下：The expression of the covariance matrix C is as follows:
可选的，所述计算所述矩阵C的特征值和特征向量的步骤包括：Optionally, the step of calculating the eigenvalues and eigenvectors of the matrix C includes:
根据协方差求解公式计算协方差，所述矩阵C对角线为方差，非对角线为各自协方差；其中，协方差求解公式为：Calculate the covariance according to the covariance solution formula, the diagonal of the matrix C is the variance, and the off-diagonal lines are the respective covariances; wherein, the covariance solution formula is:
计算所述矩阵C的特征值及对应的特征向量，其中计算公式如下：Calculate the eigenvalues and corresponding eigenvectors of the matrix C, wherein the calculation formula is as follows:
Cμ＝λμ。Cμ=λμ.
可选的，根据所述特征值和特征向量计算得到主成分PC1、PC2并进行聚类，得到聚类后的各相用户的曲线的步骤包括：Optionally, according to the eigenvalues and eigenvectors, the principal components PC1 and PC2 are calculated and clustered, and the steps of obtaining the clustered user curves of each phase include:
将所述特征值从大到小排列得到{λ1,λ2,...,λN}，对应的特征向量为{μ1,μ2,...,μN}，最大特征值λ1对应的特征向量μ1为第1主成分，记为PC1；特征值λ2对应的特征向量μ2为第2主成分，记为PC2，将PC1与PC2作为表征X的主要信息，利用PC1与PC2进行不同相别用户的聚类，得到各相用户的曲线。Arrange the eigenvalues from large to small to get {λ1 ,λ2 ,...,λN }, the corresponding eigenvectors are {μ1 ,μ2 ,...,μN }, the largest eigenvalue λ The eigenvector μ1 corresponding to1 is the first principal component, denoted as PC1 ; the eigenvector μ2 corresponding to the eigenvalue λ2 is the second principal component, denoted as PC2 , and PC1 and PC2 are used as the main components representing X information, using PC1 and PC2 to cluster users of different phases to obtain the curves of users of each phase.
可选的，将所述聚类中值与关口表各相电压的EM距离进行归类分析得到识别结果的步骤包括：Optionally, the step of classifying and analyzing the cluster median and the EM distance of each phase voltage of the gateway table to obtain the identification result includes:
将聚类后所有曲线计算平均值得到的聚类中值与关口各相电压结合计算EM距离；Calculate the EM distance by combining the cluster median value obtained by calculating the average value of all curves after clustering with the voltage of each phase at the gate;
将所述EM距离进行归类分析以及相关性比较，得到各聚类类别归属以及识别结果。Classification analysis and correlation comparison are performed on the EM distances to obtain the classification and identification results of each cluster category.
可选的，所述EM距离通过下式计算：Optionally, the EM distance is calculated by the following formula:
式中p1，p2为两个概率分布，Π(p1,p2)为p1与p2组合所得的联合分布集合；样本x与y的联合分布用γ表示；||x-y||为样本间距离，计算联合分布γ下，样本对距离期望值表示为E(x,y):γ[||x-y||]。In the formula, p1 and p2 are two probability distributions, Π(p1 , p2 ) is the joint distribution set obtained by combining p1 and p2 ; the joint distribution of sample x and y is represented by γ; ||xy|| is the distance between samples, and under the calculation of the joint distribution γ, the expected value of the distance between samples is expressed as E(x,y):γ [||xy||].
本申请提供的一种基于主成分分析与EM距离的低压用户相别识别方法包括：采集关口表三相电压及低压用户电压；对所述低压用户电压中的缺失数据采用牛顿法插值处理，得到高维数据；利用主成分分析将所述高维数据映射到低维空间，将所有低压用户电压形成矩阵，将矩阵中的所有特征去均值，得到协方差矩阵C；计算所述矩阵C的特征值和特征向量；根据所述特征值和特征向量计算得到主成分PC1、PC2并进行聚类，得到聚类后的各相用户的曲线；根据所述曲线计算聚类中值；将所述聚类中值与关口表各相电压的EM距离进行归类分析得到识别结果。A low-voltage user identification method based on principal component analysis and EM distance provided by the present application includes: collecting the three-phase voltage of the gateway meter and the low-voltage user voltage; using Newton's method interpolation for the missing data in the low-voltage user voltage to obtain High-dimensional data; use principal component analysis to map the high-dimensional data to a low-dimensional space, form a matrix of all low-voltage user voltages, remove the mean value of all features in the matrix, and obtain a covariance matrix C; calculate the characteristics of the matrix C value and eigenvector; calculate the principal components PC1 and PC2 according to the eigenvalue and eigenvector, and perform clustering to obtain the curves of users of each phase after clustering; calculate the cluster median value according to the curve; The cluster median and the EM distance of each phase voltage of the gateway table are classified and analyzed to obtain the identification results.
本申请的有益效果为：首先利用主成分分析提取低压用户电压主成分进行聚类，然后计算聚类中值与关口各相电压EM距离进行相别归属，实测数据证明该方法对相别识别准确性较高，且较于传统相似度算法在判别相别时具有明显区分度，解决现有的智能电表对低压用户相别识别准确性和选择性不高的问题。The beneficial effect of this application is: firstly, the principle component analysis is used to extract the main components of the voltage of low-voltage users for clustering, and then the median value of the cluster and the EM distance of each phase voltage at the gate are calculated for phase identification. The measured data proves that this method is accurate for phase identification Compared with the traditional similarity algorithm, it has a clear degree of discrimination when distinguishing phases, which solves the problem of low accuracy and selectivity of the existing smart meters for low-voltage user phase identification.
附图说明Description of drawings
为了更清楚地说明本申请的技术方案，下面将对实施例中所需要使用的附图作简单地介绍，显而易见地，对于本领域普通技术人员而言，在不付出创造性劳动的前提下，还可以根据这些附图获得其他的附图。In order to illustrate the technical solution of the present application more clearly, the accompanying drawings that need to be used in the embodiments will be briefly introduced below. Obviously, for those of ordinary skill in the art, on the premise of not paying creative work, there are also Additional figures can be derived from these figures.
图1为本申请实施例提供的一种基于主成分分析与EM距离的低压用户相别识别方法流程图；Fig. 1 is a flow chart of a low-voltage user identification method based on principal component analysis and EM distance provided by the embodiment of the present application;
图2为本申请实施例1中数据预处理差值前后对比图；Fig. 2 is the comparison chart before and after the data preprocessing difference in the embodiment 1 of the present application;
图3为本申请实施例1中低压用户电压曲线簇在PCA空间聚类点簇图；Fig. 3 is the point cluster diagram of low-voltage user voltage curve cluster in PCA space clustering in embodiment 1 of the present application;
图4为本申请实施例1中低压用户电压曲线簇聚类结果图；Fig. 4 is the graph of clustering results of voltage curves of low-voltage users in Example 1 of the present application;
图5为本申请实施例1中关口电压ABC三相电压图。FIG. 5 is a three-phase voltage diagram of gate voltage ABC in Embodiment 1 of the present application.
具体实施方式Detailed ways
下面将详细地对实施例进行说明，其示例表示在附图中。下面的描述涉及附图时，除非另有表示，不同附图中的相同数字表示相同或相似的要素。以下实施例中描述的实施方式并不代表与本申请相一致的所有实施方式。仅是与权利要求书中所详述的、本申请的一些方面相一致的系统和方法的示例。The embodiments will be described in detail hereinafter, examples of which are illustrated in the accompanying drawings. When the following description refers to the accompanying drawings, the same numerals in different drawings refer to the same or similar elements unless otherwise indicated. The implementations described in the following examples do not represent all implementations consistent with this application. These are merely examples of systems and methods consistent with aspects of the present application as recited in the claims.
参见图1，为本申请实施例提供的一种基于主成分分析与EM距离的低压用户相别识别方法流程图。Referring to FIG. 1 , it is a flow chart of a method for separately identifying low-voltage users based on principal component analysis and EM distance provided by an embodiment of the present application.
本申请实施例提供的一种基于主成分分析与EM距离的低压用户相别识别方法，包括以下步骤：An embodiment of the present application provides a low-voltage user identification method based on principal component analysis and EM distance, including the following steps:
1、采集关口表三相电压及低压用户电压1. Collect the three-phase voltage of the gateway meter and the low-voltage user voltage
2、对缺失数据采用牛顿法差值2. Use Newton's method difference for missing data
收集低压用户电压曲线时可能由于测量装置的异常导致数据不完备，需对数据进行插值处理，设样本包含n个点{(x1,f(x1)),…,(xn,f(xn))}，其中缺失点为(xi,f(xi))，则插值多项式f(xi)可写为如下形式：When collecting the low-voltage user voltage curve, the data may be incomplete due to the abnormality of the measuring device, and the data needs to be interpolated. Suppose the sample contains n points {(x1 ,f(x1 )),…,(xn ,f( xn ))}, where the missing point is (xi ,f(xi )), then the interpolation polynomial f(xi ) can be written as follows:
式(1)中，f(xi)为牛顿插值所得函数值。In formula (1), f(xi) is the function value obtained by Newton interpolation.
3、低压用户电压形成矩阵，所有特征去均值，求协方差矩阵C3. Low-voltage user voltage forms a matrix, removes the mean value of all features, and finds the covariance matrix C
利用主成分分析将高维数据映射到低维空间，设离散时域信号X包含M个样本{X1,X2,...,XM}，每个样本有N为特征，即各特征xj均有各自特征值。Use principal component analysis to map high-dimensional data to low-dimensional space. Suppose the discrete time-domain signal X contains M samples {X1 ,X2 ,...,XM }, and each sample has N features, namely Each feature xj has its own eigenvalue.
对所有特征去均值，将所有样本同一特征求均值，并将自身特征减去均值。To remove the mean of all features, take the mean of the same feature of all samples, and subtract the mean from its own features.
求协方差矩阵C，协方差矩阵的表达形式如下：Find the covariance matrix C, the expression of the covariance matrix is as follows:
4、计算所有低压用户电压的PC1、PC2进行聚类4. Calculate PC1 and PC2 of all low-voltage user voltages for clustering
求C的特征值及对应的特征向量。Find the eigenvalues and corresponding eigenvectors of C.
Cμ＝λμ (4)Cμ=λμ (4)
将特征值从大到小排列得{λ1,λ2,...,λN}，对应特征向量{μ1,μ2,...,μN}，最大特征值λ1对应的特征向量μ1称第1主成分，记为PC1；特征值λ2对应的特征向量μ2称第2主成分，记为PC2，以此类推。一般情况下PC1和PC2包含样本90％以上的信息，因此可将PC1与PC2两个主元作为表征X的主要信息，利用PC1与PC2可实现不同相别用户的聚类。Arrange the eigenvalues from large to small to {λ1 ,λ2 ,...,λN }, corresponding to the eigenvectors {μ1 ,μ2 ,...,μN }, and the feature corresponding to the largest eigenvalue λ1 The vector μ1 is called the first principal component, denoted as PC1 ; the eigenvector μ2 corresponding to the eigenvalue λ2 is called the second principal component, denoted as PC2 , and so on. Generally, PC1 and PC2 contain more than 90% of the information of the sample, so the two principal components of PC1 and PC2 can be used as the main information representing X, and the clustering of different users can be realized by using PC1 and PC2 .
5、计算聚类后的各相用户电压中值5. Calculate the median value of the user voltage of each phase after clustering
6、计算聚类中值与关口各相电压的Wasserstein距离进行归类6. Calculate the cluster median and the Wasserstein distance of each phase voltage at the gate for classification
将聚类后所有曲线计算平均值得聚类中值，再与关口各相进行相关性比较，可得各聚类结果类别归属。Wasserstein(EM)距离定义如下：Calculate the average value of all the curves after clustering, and then compare the correlation with each phase of the gate to obtain the classification of each clustering result. The Wasserstein (EM) distance is defined as follows:
式中p1，p2为两个概率分布，Π(p1,p2)为p1与p2组合所得的联合分布集合。样本x与y的联合分布用γ表示。||x-y||为样本间距离，计算联合分布γ下，样本对距离期望值表示为E(x,y):γ[||x-y||]，所有可能的联合分布中取期望值下界即为p1与p2的Wasserstein距离。简单来讲，就是将概率分布p1搬到概率分布p2消耗的最小能量。In the formula, p1 and p2 are two probability distributions, and Π(p1 , p2 ) is the joint distribution set obtained by combining p1 and p2 . The joint distribution of samples x and y is denoted by γ. ||xy|| is the distance between samples. When calculating the joint distribution γ, the expected value of the sample pair distance is expressed as E(x, y): γ [||xy||], and the lower bound of the expected value in all possible joint distributions is p1 to the Wasserstein distance of p2 . Simply put, it is the minimum energy consumed by moving the probability distribution p1 to the probability distribution p2 .
在实际应用中，会计算出多种曲线簇聚类中值，通过与三相电压的Wasserstein距离比较，能够准确的分辨出所属类别，因为各类聚类中值与其他相别距离均相差较大，最终能够识别聚类结果以及相别类属。In practical applications, the median value of a variety of curve clusters will be calculated. By comparing it with the Wasserstein distance of the three-phase voltage, the category to which it belongs can be accurately distinguished, because the median value of each type of cluster has a large difference from other phase distances. , and finally be able to identify clustering results and similar categories.
7、输出识别结果7. Output recognition results
在实际应用中，若要识别用户相，需要对数据进行聚类，为充分反映各相别之间的相似性，兼顾算法运算效率，本申请利用主成分分析(Principal Component Analysis，PCA)进行特征降维再聚类，计算聚类中值与关口电压Wasserstein距离以识别聚类结果相别类属。In practical applications, if you want to identify user phases, you need to cluster the data. In order to fully reflect the similarity between the phases and take into account the efficiency of algorithm operations, this application uses Principal Component Analysis (PCA) to perform feature analysis. Dimensionality reduction and re-clustering, calculate the distance between the cluster median and the gate voltage Wasserstein to identify the different categories of the clustering results.
下面结合具体实测数据对本申请提供的一种基于主成分分析与EM距离的低压用户相别识别方法说明。The following describes a low-voltage user identification method based on principal component analysis and EM distance provided by the present application in combination with specific measured data.
实施例1：以云南某地区关口三相电压及关口下70个低压用户电压实测数据。Example 1: Based on the actual measurement data of the three-phase voltage at a gateway in a certain area of Yunnan and the voltage of 70 low-voltage users at the gateway.
具体步骤如图1所示：The specific steps are shown in Figure 1:
首先，对缺失数据采用牛顿法进行插值处理，插值前后24点低压用户电压曲线如图2所示。First, Newton's method is used to interpolate the missing data, and the 24-point low-voltage user voltage curve before and after interpolation is shown in Figure 2.
其次，对70个低压用户电压进行PCA聚类，计算各成分贡献度及贡献率，第一个成分贡献率为77.26％，第二成分贡献率为14.94％，PC1与PC2的总贡献率为92.2％，超过90％，PC1与PC2已包含了原数据大部分信息，所以选择前两个主成分进行聚类，70个低压用户电压曲线簇在PCA空间聚类结果如图3所示，70个低压用户电压曲线簇聚类结果如图4所示，关口电压ABC三相电压如图5所示。Secondly, carry out PCA clustering on 70 low-voltage user voltages, and calculate the contribution degree and contribution rate of each component. The contribution rate of the first component is 77.26%, the contribution rate of the second component is 14.94%, and the total contribution rate of PC1 and PC2 is 92.2%, more than 90%, PC1 and PC2 already contain most of the information of the original data, so the first two principal components are selected for clustering, and the clustering results of 70 low-voltage user voltage curve clusters in PCA space are shown in Figure 3 Figure 4 shows the clustering results of the voltage curves of 70 low-voltage users, and Figure 5 shows the gate voltage ABC three-phase voltage.
最后，根据公式(5)对各聚类结果计算中值，并与关口各相电压计算Wasserstein距离。第I类曲线簇聚类中值与A相电压Wasserstein距离最小，最小距离为0.206；第II类曲线簇聚类中值与B相电压Wasserstein距离最小，距离为0.149；第III类曲线簇聚类中值与C相电压Wasserstein距离最小，最小距离为0.197，且各类聚类中值与其他相别距离均相差较大。Finally, calculate the median value of each clustering result according to formula (5), and calculate the Wasserstein distance with each phase voltage of the gate. The distance between the cluster median of type I curves and the Wasserstein voltage of phase A is the smallest, and the minimum distance is 0.206; the distance between the cluster median of type II curves and the Wasserstein voltage of phase B is the smallest, and the distance is 0.149; the cluster cluster of type III curves has the smallest distance The distance between the median value and C-phase voltage Wasserstein is the smallest, and the minimum distance is 0.197, and the median value of each cluster has a large difference from other phase distances.
本申请首先利用主成分分析提取低压用户电压主成分进行聚类，然后计算聚类中值与关口各相电压EM距离进行相别归属，实测数据证明该方法对相别识别准确性较高，且与Pearson相关系数，Kendall相关系数和Spearman相关系数三种相关性分析算法进行比较，较于传统相似度算法在判别相别时具有明显区分度，解决现有的智能电表对低压用户相别识别准确性和选择性不高的问题。This application first uses principal component analysis to extract the main components of low-voltage user voltage for clustering, and then calculates the cluster median and the EM distance of each phase voltage at the gate for phase identification. The measured data proves that this method has a high accuracy for phase identification, and Compared with the three correlation analysis algorithms of Pearson correlation coefficient, Kendall correlation coefficient and Spearman correlation coefficient, compared with the traditional similarity algorithm, it has a clear degree of discrimination when distinguishing phases, and solves the problem of identifying low-voltage users for existing smart meters. Problems with low accuracy and selectivity.
本申请提供的实施例之间的相似部分相互参见即可，以上提供的具体实施方式只是本申请总的构思下的几个示例，并不构成本申请保护范围的限定。对于本领域的技术人员而言，在不付出创造性劳动的前提下依据本申请方案所扩展出的任何其他实施方式都属于本申请的保护范围。The similar parts between the embodiments provided in the present application can be referred to each other, and the specific implementations provided above are only a few examples under the general concept of the present application, and do not constitute a limitation of the protection scope of the present application. For those skilled in the art, any other implementations expanded based on the proposal of the present application without creative work shall fall within the scope of protection of the present application.
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110854226.8ACN113505852B (en) | 2021-07-28 | 2021-07-28 | Low-voltage user phase identification method based on principal component analysis and EM distance |
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110854226.8ACN113505852B (en) | 2021-07-28 | 2021-07-28 | Low-voltage user phase identification method based on principal component analysis and EM distance |
| Publication Number | Publication Date |
|---|---|
| CN113505852A CN113505852A (en) | 2021-10-15 |
| CN113505852Btrue CN113505852B (en) | 2023-07-28 |
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110854226.8AActiveCN113505852B (en) | 2021-07-28 | 2021-07-28 | Low-voltage user phase identification method based on principal component analysis and EM distance |
| Country | Link |
|---|---|
| CN (1) | CN113505852B (en) |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114626436A (en)* | 2022-02-10 | 2022-06-14 | 珠海格力电器股份有限公司 | User classification method and device, electronic equipment and storage medium |
| CN119915333B (en)* | 2025-04-03 | 2025-06-06 | 四川中测仪器科技有限公司 | Humidity sensor fault detection method based on PCA and Jaccard distance |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH095362A (en)* | 1995-06-16 | 1997-01-10 | Koyo Denki Kk | Waveform detection method and apparatus |
| CN110389267A (en)* | 2019-07-17 | 2019-10-29 | 国网陕西省电力公司电力科学研究院 | A method for identifying the relationship between households in low-voltage station areas based on data collected by smart watt-hour meters |
| CN111244979A (en)* | 2020-01-19 | 2020-06-05 | 国网冀北电力有限公司电力科学研究院 | Three-phase unbalanced load adjustment method based on low-voltage distribution network power and electricity consumption |
| CN111612056A (en)* | 2020-05-16 | 2020-09-01 | 青岛鼎信通讯股份有限公司 | A low-voltage household variable relationship identification method based on fuzzy clustering and zero-crossing offset |
| CN111654392A (en)* | 2020-05-11 | 2020-09-11 | 国网浙江省电力有限公司电力科学研究院 | Method and system for topology identification of low-voltage distribution network based on mutual information |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8947258B2 (en)* | 2008-08-08 | 2015-02-03 | Powermax Global Llc | Reliable, long-haul data communications over power lines for meter reading and other communications services |
| CN103745232B (en)* | 2014-01-23 | 2017-01-18 | 西安电子科技大学 | Band migration-based hyperspectral image clustering method |
| CN109740641A (en)* | 2018-12-18 | 2019-05-10 | 清华四川能源互联网研究院 | A method for user identification in Taiwan area based on principal component analysis and k-means clustering |
| CN109683029A (en)* | 2018-12-29 | 2019-04-26 | 云南电网有限责任公司电力科学研究院 | A kind of outlier identification method of metering automation system |
| US11657405B2 (en)* | 2019-09-17 | 2023-05-23 | University Of Florida Research Foundation, Incorporated | Detection of recycled integrated circuits and system-on-chips based on degradation of power supply rejection ratio |
| CN113159105B (en)* | 2021-02-26 | 2023-08-08 | 北京科技大学 | Driving behavior unsupervised mode identification method and data acquisition monitoring system |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH095362A (en)* | 1995-06-16 | 1997-01-10 | Koyo Denki Kk | Waveform detection method and apparatus |
| CN110389267A (en)* | 2019-07-17 | 2019-10-29 | 国网陕西省电力公司电力科学研究院 | A method for identifying the relationship between households in low-voltage station areas based on data collected by smart watt-hour meters |
| CN111244979A (en)* | 2020-01-19 | 2020-06-05 | 国网冀北电力有限公司电力科学研究院 | Three-phase unbalanced load adjustment method based on low-voltage distribution network power and electricity consumption |
| CN111654392A (en)* | 2020-05-11 | 2020-09-11 | 国网浙江省电力有限公司电力科学研究院 | Method and system for topology identification of low-voltage distribution network based on mutual information |
| CN111612056A (en)* | 2020-05-16 | 2020-09-01 | 青岛鼎信通讯股份有限公司 | A low-voltage household variable relationship identification method based on fuzzy clustering and zero-crossing offset |
| Title |
|---|
| Method for Single-phase Electric Meter Phase Discrimination Based on Multiple Linear Regression;Liqiang Zhang等;《2019 IEEE 8th International Conference on Advanced Power System Automation and Protection ( APAP)》;177-181* |
| 基于智能电表数据的供电网络拓扑识别方法研究;张丽强;《中国优秀硕士学位论文全文数据库 (基础科学辑)》(第2020（10）期);A002-55* |
| Publication number | Publication date |
|---|---|
| CN113505852A (en) | 2021-10-15 |
| Publication | Publication Date | Title |
|---|---|---|
| CN111080105A (en) | Recognition method and system of household change relationship in station area based on voltage time series data | |
| CN113505852B (en) | Low-voltage user phase identification method based on principal component analysis and EM distance | |
| CN109740641A (en) | A method for user identification in Taiwan area based on principal component analysis and k-means clustering | |
| Feng et al. | Data mining for abnormal power consumption pattern detection based on local matrix reconstruction | |
| CN109828184B (en) | Voltage sag source identification method based on mutual approximate entropy | |
| CN111505433A (en) | A method for error correction and phase identification of household-change relationship in low-voltage station area | |
| CN109816033A (en) | A method for user identification in Taiwan area based on optimized supervised learning | |
| CN108562821B (en) | Method and system for determining single-phase earth fault line selection of power distribution network based on Softmax | |
| CN105160416A (en) | Transformer area reasonable line loss prediction method based on principal component analysis and neural network | |
| CN112819299A (en) | Differential K-means load clustering method based on center optimization | |
| CN110687361A (en) | A method and system for identifying phase sequence of low-voltage distribution network station area | |
| CN110110907B (en) | Method for extracting characteristic parameters of low-voltage transformer area | |
| CN111709668A (en) | Method and device for risk identification of power grid equipment parameters based on data mining technology | |
| CN118336720B (en) | Low-voltage distribution network topology identification method based on interval 2-type fuzzy clustering algorithm | |
| CN115578217A (en) | Non-intrusive load monitoring method and monitoring system based on decision tree model | |
| CN114722098A (en) | Typical load curve identification method based on normal cloud model and density clustering algorithm | |
| CN112819649B (en) | Method and device for determining user change relation of station area | |
| Grigoraş et al. | Load estimation for distribution systems using clustering techniques | |
| CN114676783B (en) | A load identification method based on single classification combined with fuzzy width learning | |
| CN114611869B (en) | A method for household identification in low-voltage station area | |
| CN114266457A (en) | Method for detecting different loss inducement of distribution line | |
| CN115659553A (en) | A low-voltage power supply network topology identification method and system | |
| CN106291239A (en) | A kind of direct current transmission line fault recognition methods using filter branches electric current and principal component analytical method | |
| CN117454203A (en) | A topology identification method based on autoencoder dimensionality reduction and BIRCH clustering | |
| Khumchoo et al. | Cluster analysis for primary feeder identification using metering data |
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |