,
State Grid Fujian Electric Power Co., Ltd. Marketing Service Center , Fuzhou, Fujian , China
,
Xi'an Jiaotong University , Xi'an, Shaanxi , China
,
State Grid Fujian Electric Power Co., Ltd. Marketing Service Center , Fuzhou, Fujian , China
,
State Grid Fujian Electric Power Co., Ltd. Marketing Service Center , Fuzhou, Fujian , China
State Grid Fujian Electric Power Co., Ltd. Marketing Service Center , Fuzhou, Fujian , China
To solve the problems of traditional Apriori algorithm in power marketing big data processing, such as candidate item set redundancy, low single-machine computing efficiency, and difficulty in adapting to multi-dimensional time series data, this study proposes an improved Apriori algorithm that integrates Resilient Distributed Dataset (RDD) distributed architecture. This study takes two public data sets as the research object. It first uses RDD distributed architecture to complete data cleaning, missing value filling, outlier elimination and feature conversion. Then, it optimizes the pruning strategy and parallel support statistical method to address the shortcomings of insufficient pruning and redundant support calculation of traditional algorithms. The experimental results show that when the improved algorithm processes 1 million pieces of electricity marketing data, the running time is reduced from 486.5s to 183.4s compared to native Apriori. When processing 5 million pieces of real electricity marketing data, the speedup ratio of the improved algorithm reaches 3.75 at five nodes, and the expansion rate remains at 79%. A total of 12 core association rules for power marketing were discovered. Among them, typical rules such as "industrial users → high load from 9:00 to 18:00 on weekdays" and "high temperature >35°C+residential users → surge in air conditioning load" have an average support degree of 0.71, an average confidence level of 0.83, and an improvement degree greater than 1.2. The research conclusion confirms that the integration solution of the improved algorithm and RDD model can efficiently process power marketing big data, and the mined association rules have actual business value. This research provides data support and technical reference for power companies to formulate peak-shifting electricity price policies, optimize regional power supply planning, and provide precise marketing services. This is of great significance in promoting the transformation of electric power marketing to intelligence and refinement.
This is an open access article distributed under the Creative Commons Attribution Non-Commercial License (CC BY-NC) License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
0
The statements, opinions and data contained in the journal are solely those of the individual authors and contributors and not of the publisher and the editor(s). We stay neutral with regard to jurisdictional claims in published maps and institutional affiliations.