Scholarlink

<p>The paper shows that large and interrelated data processing is a challenge in the power industry because traditional RDBMSs are not supported by real-time analytics. It suggests a graph database solution with tasks divided between OLTP in real time tasks and OLAP in general analyses. The system is fast and efficient with the use of Graph Neural Networks and Temporal Graph Networks. Findings indicate a 35% response time (120ms to 78ms) and 40% reduction in complexity of OLAP (0.45s to 0.27s). The predictive maintenance is 25% more accurate (MAP 92% vs. 67%), and redundancy is decreased by 20%, reducing the storage costs and increasing scalability.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Graph Database Design for Power Industry Applications Integrating OLTP and OLAP Workflows

  • 期刊论文

  • 发表日期：

    2026年6月17日

查看全文
  • Lei Wei,
  • Honghua Xu,
  • Zijian Hu,
  • Ye Ji,
  • Qilong Qina,
  • Shuai Reng

摘要

The paper shows that large and interrelated data processing is a challenge in the power industry because traditional RDBMSs are not supported by real-time analytics. It suggests a graph database solution with tasks divided between OLTP in real time tasks and OLAP in general analyses. The system is fast and efficient with the use of Graph Neural Networks and Temporal Graph Networks. Findings indicate a 35% response time (120ms to 78ms) and 40% reduction in complexity of OLAP (0.45s to 0.27s). The predictive maintenance is 25% more accurate (MAP 92% vs. 67%), and redundancy is decreased by 20%, reducing the storage costs and increasing scalability.