Distribution networks nontechnical power loss estimation: A hybrid data-driven physics model-based framework
Distribution networks nontechnical power loss estimation: A hybrid data-driven physics model-based framework
复制标题
配电网非技术功率损耗估计：基于混合数据驱动的物理模型的框架
DOI：
10.1016/j.epsr.2020.106397
复制
发表时间：
2020
影响因子：
3.9
通讯作者：
Bretas, Newton G.
中科院分区：
文献类型：
--
作者：
Bretas, Arturo S.;Rossoni, Aquiles;Trevizan, Rodrigo D.;Bretas, Newton G.
提出了一种基于混合数据驱动物理模型的配电网非技术线损估算框架。非技术性电力损失被定义为输送给用户但不由公用事业公司计费的能量。与技术损失不同，这些损失并不是电力运输过程所固有的。最先进的非技术线损估算解决方案要么是数据驱动的，要么是基于物理模型的。然而，由于非技术功率损耗的演变特性，数据驱动的解决方案本身是不够的。基于物理模型的解析解，否则，考虑准静态系统模型，完全依赖于物理现象的观察，然而，它几乎不可能模拟所有的网格动态。在这种情况下，基于数据驱动的物理模型分析模型的联系使问题的解决成为可能。混合框架由三个相互依存的进程组成。首先，执行不平衡潮流分析以获得操作系统状态的初始估计。第二，应用数据驱动的消费者分类方法。第三，考虑到测量的创新性和关键测量的元组，建立了综合测量，旨在改进粗差分析。以IEEE4母线、13母线和123母线不平衡试验馈线为例进行了方案验证。对比测试结果突出了非技术线损估计误差的减少。建立在经典加权最小二乘状态估计器基础上的简单实现，以及容易获得的参数，表明了现实应用的潜在方面。
This paper presents a hybrid data-driven physics model-based framework for distribution networks nontechnical power loss estimation. Nontechnical power loss is defined as energy delivered to the consumers but not billed by the utility. These losses, unlike technical losses, are not inherent to the process of transportation of electricity. State-of-the-art solutions for nontechnical power loss estimation are either data-driven or physics model-based. However, due to the evolving nature of nontechnical power losses, data-driven solutions by themselves are not sufficient. Physics model-based analytical solutions, otherwise, which consider a quasi-static system model, rely solely on physics phenomena observation, however it is virtually impossible to model all grid dynamics. In this case, the nexus of data-driven physics model-based analytic models enable the solution of the problem. The hybrid framework is composed of three interdependent processes. First, an unbalanced load flow analysis is performed to obtain an initial estimate of the operating system state. Second, a data-driven method for consumer classification is applied. Third, synthetic measurements are created considering the measurement's innovation andn-tuple of critical measurements aiming to improve gross error analysis. Solution validation is made considering the IEEE 4-bus, 13-bus and 123-bus unbalance test feeders. Comparative test results highlight decreased nontechnical power loss estimation errors. Simplicity of implementation, with easy-to-obtain parameters, built on the classical weighted least squares state estimator, indicate potential aspects for real-life applications.
登录
查看更多内容
影响因子：
9.6
作者：
J. Leite;J. Mantovani
通讯作者：
J. Leite;J. Mantovani
DOI：
10.1109/tcsii.2018.2796938
发表时间：
2018-01
期刊：
IEEE Transactions on Circuits and Systems II: Express Briefs
影响因子：
--
作者：
N. Bretas;A. Bretas
通讯作者：
N. Bretas;A. Bretas
配电系统技术和非技术损失估计的概率方法
影响因子：
3.9
作者：
E. A. C. Neto;J. Coelho
通讯作者：
J. Coelho
影响因子：
9.6
作者：
Marcelo Zanetti;Edgard Jamhour;M. Pellenz;M. Penna;V. Zambenedetti;Ivan Chueiri
通讯作者：
Marcelo Zanetti;Edgard Jamhour;M. Pellenz;M. Penna;V. Zambenedetti;Ivan Chueiri
DOI：
10.1201/9781420041736
发表时间：
2001-08
期刊：
Electric Power Generation, Transmission, and Distribution: The Electric Power Engineering Handbook
影响因子：
--
作者：
W. Kersting
通讯作者：
W. Kersting