| 引用本文： |
-
于红德,章辉,邢晨欣,等.融合随机网络演算的光伏电力物联网数字孪生流量边界构建[J].电讯技术，2026，（4）：565 - 574. [点击复制]
- YU Hongde,ZHANG Hui,XING Chenxin,et al.Construction of Digital Twin Flow Boundaries for Photovoltaic Power Internet of Things Integrating Stochastic Network Calculus(SNC)[J].，2026，（4）：565 - 574. [点击复制]
|
|
| 摘要: |
| 在新型电力系统中，由于光伏等新能源存在波动性和不确定性，这使得在构建光伏电力物联网的数字孪生系统时，依靠传统仿真或数据驱动方法易造成模型鲁棒性差、预测精度低等问题。为此，提出了一种基于随机网络演算（Stochastic Network Calculus，SNC）的光伏电力物联网流量预测理论框架，将业务流建模为周期监测业务与泊松事件的复合到达过程，利用矩母函数（Moment Generating Function，MGF）表征其统计特性。采用Gilbert信道模型构建马尔可夫服务过程模型，并依照MGF-SNC理论推导系统积压与时延的概率上界。仿真结果表明，不同参数下的性能边界能为光伏电力物联网数字孪生流量边界的参数配置、资源调度提供可量化的依据。所提方法计算效率优于传统深度学习算法，可实现95.6%的预测覆盖率。 |
| 关键词: 电力物联网 新型电力系统 数字孪生 随机网络演算(SNC) |
| DOI：10.20079/j.issn.1001-893x.260116001 |
|
| 基金项目:教育部“长江学者奖励计划（讲席学者项目）”（21069000222） |
|
| Construction of Digital Twin Flow Boundaries for Photovoltaic Power Internet of Things Integrating Stochastic Network Calculus(SNC) |
| YU Hongde,ZHANG Hui,XING Chenxin,CHEN Hongming |
| (1.College of Electronic Information and Optical Engineering，Nankai University，Tianjin 300350，China;2.School of Information Engineering，Zhejiang Ocean university，Zhoushan 316000，China) |
| Abstract: |
| In new power systems，the volatility and uncertainty of new energy sources such as photovoltaics make traditional simulation or data-driven methods prone to poor model robustness and low prediction accuracy when constructing digital twin systems for photovoltaic power Internet of Things(IoT).Therefore，a theoretical framework for traffic prediction in photovoltaic power IoT based on stochastic network calculus(SNC) is proposed，in which the traffic flow is modeled as a composite arrival process of periodic monitoring traffic and Poisson event traffic，and its statistical characteristics are characterized by the moment generating function(MGF).Based on this，a Markov service process model is constructed using the Gilbert channel model，and the probabilistic upper bounds of system backlog and delay are derived according to MGF-SNC theory.Simulation results show that the proposed method provides effective traffic boundary prediction for photovoltaics power IoT digital twin systems，offering quantitative guidance for parameter configuration and resource scheduling.Compared with deep learning algorithms，the proposed method achieves superior computational efficiency and a prediction coverage of 95.6%. |
| Key words: power Internet of Things new power systems digital twins stochastic network calculus(SNC) |
《电讯技术》编辑部 