【作者】 李涛；

【导师】 何怡刚；

【作者基本信息】 湖南大学 ， 电工理论新技术， 2013， 博士

【摘要】 随着以IC生产业为代表的电能质量敏感客户在国内的发展与壮大,暂态电能质量在供电部门及用户方都受到越来越多的重视。暂态电能质量问题通常是以频谱和暂态持续时间为特征,可以分为脉冲暂态扰动和振荡暂态扰动两大类。典型的暂态电能质量问题包括电压中断,电压跌落等,其信号通常具有非平稳、持续时间短、发生随机性强等特点。作为电能质量治理的基础,对暂态电能质量扰动的检测占有非常重要的地位。暂态电能质量扰动信号多是非平稳信号,具有持续时间短,干扰的随机性强等特点,使得暂态电能质量的检测难度很大,也是目前暂态电能质量分析和检测方法还不成熟的原因。如何快速、准确地从海量的电能质量扰动信号中提取特征并进行正确的分类和分析,为此广大研究人员对其进行了深入的研究,提出了小波变换、S变换、STFT等多种检测方法。本文在大量查阅和分析国内外相关文献之后,结合前人的研究,对暂态电能质量检测方法进行了改善。本文的主要研究和创新工作如下：第一,针对扰动种类多的特点,利用扰动信号的奇异性,提出了一种基于多重分形奇异谱和信息熵的叠加暂态扰动分析方法。MATLAB仿真结果表明：该方法可以有效检测出单独以及叠加的扰动信号,为叠加暂态扰动的检测提供了一种新思路。第二,针对暂态扰动检测的实时性要求,提出了一种基于小波分形的暂态电能质量扰动多分辨率分析方法。MATLAB仿真结果表明：该算法能很好地对暂态电能质量扰动信号进行识别,可以识别非重合小幅度的暂态扰动。提出的小波分形算法简单,可实现并行计算,便于硬件实现。通过该算法,可较好地分析暂态电能质量扰动。第三,针对暂态扰动检测的准确性要求,特别是暂态电能质量扰动信号含噪丰富的特点,提出了一种基于MAP小波的暂态电能质量扰动信号去噪分析方法,并与其他的几种常用的去噪算法进行了对比。MATLAB仿真结果表明：该算法对电能质量暂态扰动信号具有良好的去噪效果,可以有效滤除非常小幅度的信号噪声。提出的MAP小波去噪算法简单,便于硬件实现。通过该算法,可很好地滤除暂态电能质量扰动信号含有的噪声。第四,针对复小波含有丰富的幅值和相位信息以及提升算法计算简单的特点,提出了一种基于提升复小波的暂态电能质量扰动的检测与定位方法。MATLAB仿真结果表明：该方法利用提升复小波变换后得到的幅值和相位信息来定位扰动信号并计算扰动幅值,提出的算法实时性好、结构简单、精确度高。针对上述的各种方法进行了大量的仿真实验工作,并在此基础上,对电力系统的实际的电力故障信号进行了故障定位和电压下降幅度的分析,结果表明,所采取的方法定位准确,耗时短,幅度分析正确,效果良好。更多还原

【Abstract】 In modern power system, along with the widespread application of nonlinear loads and increasingly development of sensitive electronic equipment in the power grid, the pollution problems for transient power quality are becoming more and more serious. Transient power quality must be improved in order to ensure that all electrical equipments can operate properly and reliably. Therefore, it is very necessary that we depthly study the theory and the detection analysis techniques related to the transient power quality.Since the transient power quality disturbance signal are usually some non-stationary signals, and have with shorter duration and strong randomness of perturbation, the detection of transient power quality is very difficult, therefore, it is the premature reason of the analysis and detection method for transient power quality. How to acquire these features quickly and accurately from the massive amount of power quality disturbance signal and to carry out correct classification and analysis, more and more researcher have obtained some method and deepen relative study. In order to accurately detect the disturbances of transient power quality, and efficiently improve power quality, researchers at home and abroad have investigated depthly these problems and then obtained a lots of research results, which propose many kinds of detection methods, for examples, Wavelet Transform, S transform, Short Time Fourier Transform (STFT) have been constructed for transient power quality. Based on analysis of existing relevant literatures and previous studies, this thesis improves efficiently and feasibly some detection methods for transient power quality. The main research and innovation work are as follows:(1) Aim to the characteristic of many kinds of disturbance, we utilize singularity of the disturbance signal, and propose an information overlay method based on multi-fractal singularity spectrum analysis and information entropy of transient disturbances. Furthermore, MATLAB simulation results show that the method can be detect effectively the individual and superimposition disturbance signal, and the approach of superposition transient disturbance detection offers a new idea.(2) In order to meet the requirements of real-time detection, a multi-resolution analysis method for transient power quality is proposed based on wavelet fractal approach. Furthermore, MATLAB simulation results show that:the above algorithm has better adaptability and robustness of the identification of transient power quality disturbances, and can identify non-coincident minor power quality disturbances. Moreover, the proposed wavelet fractal algorithm can calculate simply and be executed concurrently, and then is beneficial to the hardware implementation. Therefore, the algorithm can be used to analyze the power quality transient disturbance.(3) In order to meet the requirements of the transient disturbance detection accuracy for power quality transient disturbance signal including noise with rich characteristics, this thesis proposes a MAP wavelet-based power quality transient disturbance signal denoising analytical methods, and compares detailedly to some other existing denoising algorithms. Furthermore, MATLAB simulation results show that:the above algorithm has a good denoising effect and capability for transient power quality disturbance signal, which can effectively filter out noise with very small amplitude from power quality signal. Moreover, the proposed MAP wavelet denoising algorithm is simple and then is conducive to the hardware implementation. In practice, this algorithm can effectively filter out noise from transient power quality disturbance signal.(4) Since that the complex wavelet contains rich amplitude and phase information and to enhance the lifting wavelet algorithm is simple, we construct a novel complex wavelet method that can be used to the detection and localization of transient power quality disturbances. Similarly, MATLAB simulation results show that:the method is to locate the disturbance signal and analyze the amplitude value by take fully advantage of the amplitude and phase information obtained from the complex wavelet transform based on lifting wavelet. Meanwhile, the proposed lifting-complex wavelet algorithm has good real-time performance, simpler structure, and higher accuracy.In addition, on the basis of the above algorithms/schemes, a large number of MATLAB simulation are given, and some real applications to the fault location and the voltage drop for the actual power of the power system fault signal are given to demonstrate that the present results carries out accurate width analysis with exact position and shorter time-consuming, and have good effect for the real-world power quality transient disturbance signal.更多还原