基于贝叶斯与属性识别模型的电能质量综合评估方法  

A synthetic power quality evaluation method based on Bayesian algorithm and attribute recognition model

作　　者：张乙[1] 潘天红[1] 李正明[1] 

机构地区：[1]江苏大学电气信息学院,江苏镇江212013

出　　处：《电力系统保护与控制》2013年第7期41-45,共5页Power System Protection and Control

基　　金：国家自然科学基金资助(61273142);江苏省科技支撑计划(BE2011143);江苏省高校优势学科建设工程资助项目PAPD;江苏省自然科学基金(BK2011466)~~

摘　　要：电能质量的合理评估是实现电能监管的基础,也是提高能源使用效率的有力保证。为提高电能质量综合评估的客观性和准确性,提出一种新的综合评估算法。首先,利用贝叶斯算法修正各等级下的主观权重,得到各等级下的权重优化值;其次,借助于属性识别模型,对监测点的电能质量的各项指标进行分级评估,再结合权重的优化值,计算得到监测点的综合属性测度(亦即电能质量整体的等级数);最后通过实例验证了所提算法的实用性与客观性。A good power quality evaluation method can not only realize the power management, but also improve the energy efficiency. In order to make the power quality evaluation more objective and accurate, a new synthetic assessment algorithm is proposed. Firstly, the optimal weights for all power quality levels are calculated by using the Bayesian algorithm under each level. Then, all kinds of power indexes of monitoring point are evaluated based on the attribute recognition model. Combined with the optimized weight, the degree of attribute measurement can be calculated and the power quality evaluation can be achieved. Finally, a case study demonstrates the effectiveness and objectivity of the proposed method.

关 键 词：电能质量 贝叶斯算法 权重 属性识别模型 电能评估 

分 类 号：TM73[电气工程—电力系统及自动化]