Search

Purposes A load feature extraction method based on a combination of event monitoring and Gaussian mixture model clustering is proposed to explore the potential of energy saving and emission reduction at the customer side, to finely analyze and manage the customers’ electricity consumption behavior, and to improve the efficiency of electricity utilization. Methods First, the active power fluctuation of each appliance during a single operation is extracted by the event monitoring algorithm based on sliding window, and the start-up time, number of times, and operation duration of the appliance can be obtained by the event monitoring algorithm. Second, to address the problem that the same appliance often has similar power but inconsistent operation status, the Gaussian mixture model clustering algorithm with the advantages of “soft classification” and flexible class clusters is adopted to finely classify the load operating status and form a load status feature library that is consistent with the actual operation of power-using equipment. Finally, with the public data set AMPds2 as the research object, the method proposed in this paper is applied to study the energy consumption habits of residential customers, and the validation analysis is carried out. Findings The results show that the proposed method can extract load features better than other models.

Clustering is a branch of unsupervised machine learning algorithms that has widespread applications in the information age. However, diverse research on clustering algorithms often faces issues such as the need to specify a fixed number of neighbors for density calculation, the requirement of predefining the number of clusters, and the necessity for multiple iterations to update information aggregation. These problems can lead to the loss of data features and increase computational complexity, resulting in higher time complexity of the models. To address these challenges,this study was inspired by the luminescence and light information transmission of fireflies and proposed a clustering algorithm called FLINCA（Firefly Luminescent Information Navigation Clustering Algorithm). FLINCA consists of two main modules: "Growing Fireflies" and "Merging Fireflies Trees. " Firstly, data points are treated as fireflies, and their brightness is determined using an adaptive number of neighbors to achieve preliminary clustering. Then, cluster fusion is performed based on the firefly trees, resulting in the final clustering outcome. Experimental results demonstrate that FLINCA exhibits favorable clustering performance on four benchmark clustering datasets and three real-world multidimensional datasets compared to twelve different algorithms. This confirms the extensive applicability of FLINCA, which is based on firefly luminescence and light information transmission, in addressing the limitations of traditional clustering algorithms and improving clustering accuracy. [ABSTRACT FROM AUTHOR]

Copyright of Natural Gas Industry is the property of Natural Gas Industry Journal Agency and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Copyright of Special Oil & Gas Reservoirs is the property of Special Oil & Gas Reservoirs Editorial Office and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Copyright of Journal of Computer Engineering & Applications is the property of Beijing Journal of Computer Engineering & Applications Journal Co Ltd. and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Fault diagnosis is one of the most challenging tasks in power communication.The fault diagnosis based on rules can no longer meet the demand of massive alarms processing.The existing approaches based on the supervised learning need large sets of the labeled data and sufficient time to train models for processing continuous data instead of alarms, which are far behind the feasibility of deployment.As for alarm correlation and fault pattern discovery, a self-learning algorithm based on the density-based clustering and frequent subgraph mining was proposed.A novel approach for automatic fault diagnosis and dispatch were also introduced, which provided the scalable and self-renewing ability and had been deployed to the automatic fault dispatch system.Experiments in the real-world datasets authorized the effectiveness for timely fault discovery and targeted fault dispatch.

Copyright of Journal of Jilin University (Science Edition) / Jilin Daxue Xuebao (Lixue Ban) is the property of Zhongguo Xue shu qi Kan (Guang Pan Ban) Dian zi Za zhi She and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Copyright of Journal Of Sichuan University (Natural Sciences Division) / Sichuan Daxue Xuebao-Ziran Kexueban is the property of Editorial Department of Journal of Sichuan University Natural Science Edition and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Copyright of Telecommunications Science is the property of Beijing Xintong Media Co., Ltd. and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Because the computational amount of the traditional active learning method increases exponentially with the increase of problem size, it is difficult to apply to the large-scale multi-category data classification tasks. To solve this problem, a subsampling-based active learning (SBAL) algorithm is designed. This algorithm integrates unsupervised clustering algorithm with traditional active learning method, and adds subsampling operation between them. This operation can significantly reduce the time complexity of the algorithm, reduce the experimental time-consuming on the basis of ensuring the accuracy of the experiment, so as to deal with the classification problem of large-scale data sets more efficiently. The experimental results show that the experimental performance of the SBAL algorithm is better than that of the traditional active learning algorithm, which proves that the proposed method can break through the limitation that the traditional active learning method can not deal with multi-category classification of large-scale data sets. [ABSTRACT FROM AUTHOR]

Copyright of Journal of Data Acquisition & Processing / Shu Ju Cai Ji Yu Chu Li is the property of Editorial Department of Journal of Nanjing University of Aeronautics & Astronautics and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

The existing vector clustering algorithm need to learn a lot of complex data in order to get a good performance for clustering, and it does not have good performance for big data. This paper proposed a quantization error and fractal theory based high computation efficiency unsupervised clustering algorithm to solve that problem. Firstly, it constructed a parametric modeling of the quantization error for data set, got the rate-distortion curve based on the space structure of the data set. Then, it computed the efficient dimensionality of the data set by estimation of the rate distortion curve. Lastly, it obtained the optimal clustering number of the target data set by fractal theory. Experiments result shows that the proposed quantization error modeling can estimate the quantization error very well and the proposed algorithm has better performance in search the best clustering number and computation efficiency than the existing vector clustering algorithm. [ABSTRACT FROM AUTHOR]