节点文献

基于高斯过程回归的零能耗建筑供配电系统参数优化

Parameter optimization of power supply and distribution system of zero energy consumption building based on Gaussian process regression

  • 推荐 CAJ下载
  • PDF下载
  • 不支持迅雷等下载工具,请取消加速工具后下载。

【作者】 曲锐陈雪刘成郑瑞春王焕忠

【Author】 QU Rui;CHEN Xue;LIU Cheng;ZHENG Ruichun;WANG Huanzhong;Foshan Power Supply Bureau of Guangdong Power Grid Co.,Ltd.;Guangdong Nanhai Electric Power Design Institute Engineering Co.,Ltd.;

【机构】 广东电网有限责任公司佛山供电局广东南海电力设计院工程有限公司

【摘要】 为解决建筑供配电系统电能利用率过低的问题,实现对建筑能耗量的精准控制,提出基于高斯过程回归的零能耗建筑供配电系统参数优化方案。以单调谐滤波器作为核心谐波输出元件,按需计算供配电网络的潮流量与目标约束条件,完成对零能耗建筑供配电系统的谐波特性分析。在此基础上,对各项供配电数据进行预处理,通过选择核函数的方式,确定必要的高斯回归参数,实现基于高斯过程回归的零能耗建筑供配电系统参数优化与处理。与传统供配电系统参数处理手段相比,优化后方案能承载的电能传输量更大,而所需消耗的能量却相对较少,可有效解决配电系统电能利用率过低的实际应用问题。

【Abstract】 In order to solve the problem of low power utilization rate of building power supply and distribution system and realize the accurate control of building energy consumption,the parameter optimization scheme of zero energy building power supply and distribution system based on Gaussian process regression is proposed. Taking single tuned filter as the core harmonic output component,the tidal flow and objective constraints of power supply and distribution network are calculated on demand,and the harmonic characteristics of zero energy consumption building power supply and distribution system are analyzed. On this basis,the power supply and distribution data are preprocessed,and the necessary Gaussian regression parameters are determined by selecting the kernel function,so as to realize the parameter optimization and processing of zero energy consumption building power supply and distribution system based on Gaussian process regression. Compared with the traditional power supply and distribution system parameter processing means,the optimized scheme can carry more electric energy transmission capacity,but the energy consumption is relatively less,which can effectively solve the practical application problem of low power utilization rate of distribution system.

  • 【文献出处】 电子设计工程 ,Electronic Design Engineering , 编辑部邮箱 ,2021年23期
  • 【分类号】TU852
  • 【被引频次】1
  • 【下载频次】270
节点文献中: 

本文链接的文献网络图示:

本文的引文网络