摘要
对农电负荷进行非侵入式负荷辨识,有助于掌握用户负荷信息,合理开展电网经济调度,达到降低网损、调节峰谷差等目的。本文针对农电多类型负荷同时运行时的负荷特征,以混合特征和智能优化算法为基础,提出1种基于频率加权因子遗传算法(Frequency Weighting Factor Genetic Algorithm,FWF-GA)的异种负荷同时识别方法。该方法以时域信号的有功功率、无功功率以及频域信号的幅频特性建立混合特征模型,以有功功率和无功功率构建同时识别的优化模型,并以混合特征模型构建遗传算法异种负荷同时识别的适应度函数。通过农村居民5种用电设备的负荷识别对所提方法进行验证。采用本文所提方法对5000组混合负荷进行识别,识别结果表明5种电器的单个识别准确率以及整体识别的平均准确率均在90%以上;采用3种不同适应度函数的识别方法对8组混合负荷进行识别,识别结果表明本文所提方法的识别效率最高。实例分析的结果表明,基于FWF-GA的异种负荷同时识别方法具有较好的识别效率和精度。
The non-invasive load identification of rural power load is helpful to master the user load information and to reasonably carry out the economic dispatching of power grid,and thus to achieve the reduction of network loss and the adjustment of peak valley difference.This study proposed a simultaneous identification method on rural power load with multiple appliances.The method adopted the frequency weighting factor genetic algorithm(FWF-GA)based on the load hybrid feature and intelligent optimization algorithm.In the proposed method,a hybrid feature model which was composed of the active power,reactive power and amplitude-frequency characteristic was firstly established,and then the active and reactive power was used to construct the simultaneous identification and optimization model.On the basis of the constructed optimization model and genetic algorithm,the FWF-GA was developed to solve the simultaneous identification and optimization problem,in which a new fitness function was established based on the hybrid feature model to efficiently evaluate the fitness of individuals.Finally,the effectiveness of the proposed method was verified by load identification of five kinds of electrical equipment for rural residents.In the validation example,five thousand groups of combined loads were identified by the proposed method,and the identification results showed that the single identification accuracy of the five electrical appliances as well as the average accuracy of overall identification was more than 90%.Moreover,three identification methods with different fitness functions were used to identify eight groups of mixed loads,and the comparison results showed that the proposed method was more efficiency than the other two identification methods.The example analysis results indicated that the proposed load identification method can effectively identify the simultaneously existed loads with satisfied efficiency and accuracy.
作者
易姝慧
王健
刘俊杰
李强
欧阳含熠
YI Shuhui;WANG Jian;LIU Junjie;LI Qiang;OU YANG Hanyi(Department of Metrology,China Electric Power Research Institute,Wuhan 430070,China;College of Information Science and Technology,Hebei Agricultural University,Baoding 071001,China)
出处
《河北农业大学学报》
CAS
CSCD
北大核心
2024年第3期121-129,共9页
Journal of Hebei Agricultural University
基金
国家电网公司总部科技项目(5600-202024168A-0-0-00,5700-202319273A-1-1-ZN).
关键词
农电负荷
遗传算法
频率加权因子
负荷识别
混合特征
rural power load
genetic algorithm
frequency weighting factor
load identification
hybrid feature
