【作者】 李春林；

【导师】 潘文霞；

【作者基本信息】 河海大学 ， 电力系统及其自动化， 2005， 硕士

【摘要】 随着经济的发展，新型用电设备使用，越来越多的电力部门和用户关注的电能质量问题已经不仅仅是传统意义上的供电可靠性、稳定性以及电压、频率、波形等参量偏离标称值的问题，而且包含了诸如振荡和脉冲之类的暂态或者瞬态特征。 采取合理的措施是提高电能质量的必要保证，首要的是应对其正确的检测和识别。基于这一点，本文对基于小波变换的电能质量扰动检测和识别方法进行了研究，在总结前人经验的基础上提出了自己对算法的改进，并且利用MATLAB软件对改进算法进行仿真分析，取得了令人满意的效果，验证了所用方法的可行性。 首先，为了压缩电能质量扰动数据存储空间，并且减少传输时间，本文针对电能质量扰动的频率具有非常广泛的特点，提出将基于最佳小波包基的数据压缩方法用于电能质量扰动数据的压缩，结果是对于暂态和瞬态扰动的压缩效果要优于基于小波变换的数据压缩方法。接着，本文利用小波变换模极大值原理对电能质量扰动中的电压暂降、暂时、中断，脉冲，振荡的起止时间等特征参数进行了提取，仿真结果表明该方法的有效性。然后在提取电能质量扰动的小波变换的各尺度处的特征量后，对其进初步分为三类，并且提出利用自组织特征映射神经网络(SOFM)的改进算法对其每一大类进一步的分类，经过仿真分析取得了较其他分类方法较好的分类结果。最后针对电力系统中实际情况下各种扰动有可能同时发生的情况，对某些电能质量多重扰动的情况也进行了仿真分析，取得了一定的成果。更多还原

【Abstract】 As a result of the development of economy, the power quality to which is paid attention by more and more users and electric power department is not only the deviation of voltage and frequency from the rating value, but also include the phenomenon of oscillation and pulse.So, it is necessarily to detect and identify the power quality. Based on this, this paper studied the methods of the detection and identification of power quality.Firstly, in order to save the storage space and the transmission time of data, it is necessarily to compress the data of power quality. According to characteristic of the transient disturbance data of power quality which frequency is very high, this paper put forward that the method based on optional wavelet packet apply to the compression of transient disturbance data of power quality. Finally, by comparing with the method based on the wavelet, we can draw the conclusion that the distortion rate of former is lower when the compression rate of the both is approximately same. Secondly, this paper used the theory of wavelet transform modulus maxima to realize the detection of power quality. Finally, the improved method of self-Organizing Feature Map (SOFM) was put forward to identify the kind of power quality. By the simulation of MATLAB. these methods are verified to be feasible.更多还原