摘要：【中文】本发明公开了一种基于深度卷积神经网络的外表箱类型识别方法，涉及电能计量装置外表箱识别方法领域。随着人工智能技术的深度应用，电力行业出现了智能巡查的趋势，在智能巡查过程中，如何识别外表箱类型是一项基本的工作。本方法如下：首先采用深度卷积神经网络模型检测外表箱及其部件区域的位置和大小；再对外表箱的视窗进行检测；对检测到视窗进行判断，如果是单视窗，则直接识别为单表箱，如果是多视窗，则进行视窗布局的网格构建；通过构建好的视窗网格与已知的类型的视窗布局进行匹配；识别出外表箱类型。本方法能快速识别和定位外表箱及其视窗和其他的部件，实现智能巡查过程中电能计量装置外表箱的类型有效和快速准确的识别。 【EN】The invention discloses an appearance box type identification method based on a deep convolutional neural network, and relates to the field of appearance box identification methods of electric energy metering devices. With the deep application of artificial intelligence technology, the power industry has a tendency of intelligent inspection, and how to identify the type of an outer meter box is a basic task in the intelligent inspection process. The method comprises the following steps: firstly, detecting the positions and the sizes of an outer meter box and a component area thereof by adopting a deep convolutional neural network model; then detecting a window of the outer meter box; judging the detected window, if the detected window is a single window, directly identifying the detected window as a single table box, and if the detected window is a multi-window, constructing a grid of window layout; matching the constructed window grid with a window layout of a known type; the exterior meter box type is identified. The method can quickly identify and position the outer meter box, the window and other components of the outer meter box, and realizes effective, quick and accurate identification of the type of the outer meter box of the electric energy metering device in the intelligent patrol process.

摘要：【中文】本发明公开了一种基于极限学习机和密度聚类的海量报文掉线状态分析方法，涉及报文分析方法。目前，海量报文掉线分析方法适应性差，运算速度慢。本发明包括以下步骤将预处理之后的且带有标签的数据导入极限学习机中；通过极限学习机得出密度聚类的阈值；通过密度聚类模型的密度聚类选出掉线数据。本技术方案结合了机器学习方法的优势与聚类算法的优势，利用极限学习机来给出聚类算法中关键的阈值，可以拓宽聚类算法的应用范围和提高密度聚类算法的准确度。在面对海量报文数据时，聚类相对于神经网络有较块的响应速度，更适合应用于像报文这类需要较快知道是否掉线的问题。 【EN】The invention discloses a massive message offline state analysis method based on an extreme learning machine and density clustering, and relates to a message analysis method. At present, the method for analyzing the offline of massive messages has poor adaptability and low operation speed. The method comprises the following steps of importing preprocessed data with labels into an extreme learning machine; obtaining a threshold value of density clustering through an extreme learning machine; and selecting out-line data through density clustering of the density clustering model. The technical scheme combines the advantages of a machine learning method and the advantages of a clustering algorithm, utilizes an extreme learning machine to provide a key threshold value in the clustering algorithm, and can widen the application range of the clustering algorithm and improve the accuracy of the density clustering algorithm. When massive message data is faced, clustering has a faster response speed compared with a neural network, and is more suitable for the problem that whether a line is disconnected or not needs to be known quickly like a message.

摘要：【中文】本发明公开了一种面向电力采集设备的二级标签生成方法，涉及电力信息采集领域。现有电力业务系统中，标签画像通常是单维度，各标签的重要性一样，对标签的理解主要在展示，缺乏对标签的多维度、多场景化应用。本方法包括步骤：1）定位标签主体对象；2）主体对象建立二级标签目录；3）给主体对象标记标签，形成主体对象标签作为一级标签，标签标记时，需要根据标签等级分级标记；4）在主体对象标签下建立二级标签，二级标签包括静态标签和动态标签，动态标签可根据对象实际情况显示不同内容。从多维度和多场景角度出发全面反映采集设备状态信息，方便判断采集设备异常的影响情况，提升对采集设备维护和工作状态的判断能力。 【EN】The invention discloses a secondary label generation method for power acquisition equipment, and relates to the field of power information acquisition. In the existing power business system, a label portrait is usually in a single dimension, the importance of each label is the same, the understanding of the label is mainly displayed, and the multi-dimension and multi-scenario application of the label is lacked. The method comprises the following steps: 1) positioning a tag body object; 2) the main object establishes a secondary label directory; 3) labeling the main object with a label to form the main object label as a first-level label, wherein the label is labeled according to the label level in a grading way; 4) and establishing a secondary label under the main object label, wherein the secondary label comprises a static label and a dynamic label, and the dynamic label can display different contents according to the actual condition of the object. The state information of the acquisition equipment is comprehensively reflected from the multi-dimensional and multi-scene angles, the abnormal influence condition of the acquisition equipment is conveniently judged, and the judgment capability of the maintenance and working state of the acquisition equipment is improved.

摘要：【中文】本发明公开了一种基于PCA降维和K‑Means聚类的低压台区用户接入点识别方法,涉及低压台区用户接入识别方法。目前，排查都需要人工上门排查，且无法事先进行预测，只能逐户进行排查，耗费大量人力物力。本发明对标准化处理的数据进行主成分分析法PCA降维处理，经主成分分析法PCA降维处理的数据在保持各维数据维度内方差最大的前提下，通过寻找新的向量基，将原有高维数据投影在低维空间，剔除方差较小的噪声，保留信息量最大的主成分；聚类分析；对所分析台区进行现场排查，验证分析结果的准确性。本技术方案事先进行预测，不需要逐户进行排查，减少大量人力物力，不影响台区下其他用户的正常用电，且不需要投入载波通信设备，无需增加电力企业的运营成本。 【EN】The invention discloses a low-voltage transformer area user access point identification method based on PCA dimension reduction and K-Means clustering, and relates to a low-voltage transformer area user access identification method. At present, the investigation needs manual on-door investigation, prediction cannot be carried out in advance, only the investigation is carried out one by one, and a large amount of manpower and material resources are consumed. The invention carries out Principal Component Analysis (PCA) dimension reduction processing on the data subjected to the standardization processing, and under the premise of keeping the maximum variance in each dimension of the data, the data subjected to the PCA dimension reduction processing by the principal component analysis method projects the original high-dimensional data in a low-dimensional space by searching a new vector base, removes the noise with smaller variance, and keeps the principal component with the maximum information content; clustering analysis; and (4) carrying out on-site investigation on the analyzed region, and verifying the accuracy of the analysis result. According to the technical scheme, prediction is carried out in advance, troubleshooting is not needed, a large amount of manpower and material resources are reduced, normal power utilization of other users in a distribution area is not affected, carrier communication equipment does not need to be input, and the operation cost of a power enterprise does not need to be increased.

摘要：【中文】本发明公开了一种基于决策树模型的电力用能数据存储优化方法，涉及一种电力数据存储领域，对于高频采集、低实时性业务场景的数据，已有的高效存储方法不能满足要求。本发明基于适应于HBase多存储格式进行特点分析，将采集实际应用场景划分，通过决策树模型选择合适的数据存储格式，选出最优的HBase存储方式，以此来优化高频采集数据存储空间。本技术方案采用多种格式的数据存储方式，根据不同的情况择优进行存储，弥补目前对于高频采集数据存储空间优化的不足，节省电力用能数据在大数据平台存储资源空间，提高HBase读写性；解决电力用能数据存储问题，为电力用能提供一个规范的HBase存储优化方法，具有易实现的、高效的特点。 【EN】The invention discloses a decision tree model-based power energy consumption data storage optimization method, relates to the field of power data storage, and aims to solve the problem that the existing high-efficiency storage method cannot meet the requirements of data of high-frequency acquisition and low-real-time service scenes. The method carries out characteristic analysis based on multiple HBase storage formats, divides the practical acquisition application scene, selects a proper data storage format through a decision tree model, and selects an optimal HBase storage mode so as to optimize the high-frequency acquisition data storage space. According to the technical scheme, a data storage mode with various formats is adopted, and the data is preferentially stored according to different conditions, so that the defect of optimizing the storage space of high-frequency acquired data at present is overcome, the storage resource space of the power energy data on a large data platform is saved, and the reading and writing performance of HBase is improved; the method solves the problem of data storage of the power energy consumption, provides a standard HBase storage optimization method for the power energy consumption, and has the characteristics of easy implementation and high efficiency.

摘要：【中文】本发明公开了一种低压配电网三相单线图初始布局自动生成方法。本发明根据低压三相配电网单馈线的拓扑关系以及负荷节点的等效单相负荷，计算出abc单相按照120度的扇面布局，最终构成平面均匀布置的三相配电网单馈线单线图。本发明可以在一个平面上对低压配电网接线图分相进行其走线自动计算各节点的坐标，并根据节点间的连接关系绘制线段，并可通过系数l的选择，避免线段交叉，最终形成分布均匀且无交叉的低压配电网分相接线图；为下一步的图形优化计算提供优良的初始单线图，也可直接用于低压配电网接线图最终图形，在单线图上进行基于智能电表量测数据以及潮流计算等计算值进行各类专题的渲染，形成低压配电网准实时态势监视用途的效果更佳。 【EN】The invention discloses an automatic generation method for an initial layout of a three-phase single line diagram of a low-voltage distribution network. According to the topological relation of the single feeder of the low-voltage three-phase power distribution network and the equivalent single-phase load of the load node, the abc single-phase is calculated according to the sector layout of 120 degrees, and finally the three-phase power distribution network single feeder single line diagram with uniformly arranged planes is formed. According to the invention, the wiring of the low-voltage distribution network wiring diagram can be divided on one plane, the coordinates of each node can be automatically calculated, the line segments are drawn according to the connection relation between the nodes, the line segment crossing can be avoided through the selection of the coefficient l, and finally the low-voltage distribution network divided-phase wiring diagram which is uniformly distributed and has no crossing is formed; the method provides an excellent initial single line diagram for next graph optimization calculation, can also be directly used for final graphs of low-voltage distribution network wiring diagrams, and performs various special renderings on the single line diagram based on calculated values such as intelligent electric meter measurement data and load flow calculation, so that the effect of the low-voltage distribution network quasi-real-time situation monitoring purpose is better.