The recent widespread deployment of smart meters on a global scale has created an immense amount of fine-grained smart meter data, which requires effective and real-time analysis. Although the cloud center has powerful data processing capabilities, it is insufficient for real-time analysis, especially in the case of huge and distributed data volumes. Correspondingly, intelligent edge computing is merged with smart meters in this work to create an Internet-of-Things-based architecture for an edge-intelligence-enabled smart meter (EI-smart meter) system. To achieve its potential, we also propose two (one offline and one online) ultra-low-latency cloud-edge collaboration schemes regarding real-time data analytics. Unlike the existing work, we integrate a deep neural network (DNN) into the cloud-edge collaboration scheme in a bid to reduce execution time and improve the adaptability. Finally, numerical results are presented to validate the performance of our proposed system.

中文翻译：

---

基于物联网的智能计量系统中实时数据分析的边缘智能


近年来，智能电表在全球范围内的广泛部署产生了大量细粒度的智能电表数据，需要有效、实时的分析。云中心虽然拥有强大的数据处理能力，但对于实时分析的能力不足，尤其是在数据量庞大且分布的情况下。相应地，在这项工作中，智能边缘计算与智能电表相结合，为边缘智能智能电表（EI智能电表）系统创建基于物联网的架构。为了发挥其潜力，我们还提出了两种（一种离线、一种在线）关于实时数据分析的超低延迟云边协作方案。与现有工作不同，我们将深度神经网络（DNN）集成到云边协作方案中，以减少执行时间并提高适应性。最后，给出了数值结果来验证我们提出的系统的性能。