摘要
电力系统以及部分用户均安装有监测装置,收集了大量的扰动数据,因此有必要研究出实用简单的、自动的扰动分类器,从而为进一步研究奠定基础。提出了电能质量扰动的专家概率分类器模型,用于常见电能质量扰动的分类。概率分类器基于数理统计规律,概念清楚、运算简单,引入专家反馈环节可以提高分类的准确性、鲁棒性,使得分类器具备一定的自适应能力。根据同一监测地点检测到的电能质量扰动样本,构建并测试了该分类器的可行性及性能,结果令人满意。
Power quality is a growing concern on the electric power systems. High power quality is required as a result of increase of sensitive electronic and computer-controlled load. When disturbances detected and collected, it is necessary to classify them. An expert based probability classification model is employed to investigate generic power quality disturbances. It is based on the statistical principle with the character of simple and low calculation cost. Improved accurateness and robustness are achieved by adding expert interference feedback to the model. The model has certain adaptive ability with expert feedback. The model is created and its feasibility and performance are evaluated through the field collected disturbance sample of the same site. The result shows much satisfying.
出处
《电力系统自动化》
EI
CSCD
北大核心
2004年第8期45-49,56,共6页
Automation of Electric Power Systems
关键词
电能质量扰动分类
概率分类器
专家反馈
power quality disturbances classification
probability classification
expert feedback
