Scholarlink

The transition towards smart and sustainable cities is emerging as a fundamental aspect in the future of our society. In this context, this article proposes a design for a low-cost smart energy meter that allows recording of electrical parameters for user energy monitoring. Through data training, data are validated in real-time for the execution of predictive models. The methodology is based on a measurement architecture and real-time data recording, processing, and validation. The implementation of Wide Neural Networks results in optimal demand forecasting. It is analyzed in a case study with real-life profiles at specific measurement points. The results show that the implementation of these measurements and real-time forecasting enables monitoring and supervision of energy systems from different locations.

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

Smart Meter Based on Demand Forecasting for Real-Time Architecture

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Dario Benavides,
  • Paul Arévalo,
  • Alberto Ríos,
  • Leonardo Torres,
  • Danny Ochoa-Correa

摘要

The transition towards smart and sustainable cities is emerging as a fundamental aspect in the future of our society. In this context, this article proposes a design for a low-cost smart energy meter that allows recording of electrical parameters for user energy monitoring. Through data training, data are validated in real-time for the execution of predictive models. The methodology is based on a measurement architecture and real-time data recording, processing, and validation. The implementation of Wide Neural Networks results in optimal demand forecasting. It is analyzed in a case study with real-life profiles at specific measurement points. The results show that the implementation of these measurements and real-time forecasting enables monitoring and supervision of energy systems from different locations.