基于粒子滤波的统一电能质量分析理论研究

项目名称： 基于粒子滤波的统一电能质量分析理论研究

项目编号： No.51277080

项目类型： 面上项目

立项/批准年度： 2013

项目学科： 电工技术

项目作者： 李开成

作者单位： 华中科技大学

项目金额： 83万元

中文摘要： 随着各种新能源的并网及负载的多样性，导致电能质量恶化，这严重危害电力系统的高效经济和稳定运行，影响人们的正常生活和产品质量，因此对电能质量的分析和治理显得十分紧迫和重要。电力系统的规模化、复杂化使得电能质量参数计算问题逐渐由单一事件的计算向混合事件的计算方向发展。IEC（国际电工委员会）推荐的方法主要针对各类单一电能质量事件的参数计算，且有些方法存在局限性。由于电能质量分析问题的重要性，使他成为人们关注的热点问题，许多新的计算方法被不断提出，但这些方法五花八门，缺乏统一性，且计算条件限制较多。本项目提出采用粒子滤波器的理论和方法来计算电能质量参数，通过对电能质量事件产生的机理和现象进行分析来建立合理的事件数学建模，对模型中的各个参数进行状态跟踪估计，并在建模过程中充分考虑模型的非线性、非高斯问题，以实现工业环境下电能质量参数的精确计算，由此建立一套统一的电能质量分析理论。

中文关键词： 电能质量；粒子滤波；s变换；贝叶斯网络；强跟踪滤波

英文摘要： Along with the wide application of new energy and versatile of loads, the power quality is becoming worse, which severely harms the efficient, safe, steady and economical operation of power system, lowers product quality and people's life, so analysis and treatment of power quality seems very urgent and important.The expansion of power system and increasing complication of power consumers leads a trend of the parameters estimation of mixed power quality events. In current, IEC proposed methods mainly focus on estimation of single event of power quality, and part of them have limitation. Due to the fact, parameters estimation of power quality remains popular, and many new algorithms are proposed. By the synthesis of the most interesting methods currently, this paper proposes a particle filter based method for this application. Based on the analysis of the mechanism and phenomenon of power quality events, rational mathematic model are founded. Taking the consideration of non-linear and non-Gaussian situation which can not be avoided in industry application, state tracing techniques are applied for the parameters estimation of power qualities. Finally, it is supposed to found a unified method for power quality evaluation.

英文关键词： power quality；particle filter；s-transform；Bayesian network；strong trace filter