多源非平稳信号频率跟踪的改进方法研究

Study of Method Improvement for Racking Frequency of Multi-Source Non-Stationary Signal

摘要: 智能微电网的应用越来越受到重视正逐渐成为研究的热点之一，其发电系统中存在多个发电源和负载，其系统电流也是由多个频率分量而组成。由于智能微电网的组成复杂性，其发电系统更容易受到干扰和影响，因此系统电流会表现为非平稳特性。本文针对此类情况，提出了基于短时傅里叶变换,然后采用了相位建模的改进频率估计方法，对智能微电网的系统信号进行分析处理，有效实现了多分量分平稳信号的分离和频率跟踪，并且适用于能量相差悬殊的多分量信号，同时可以准确提取各频率分量的初始相位估计。

Abstract: The applications of Smart MicroGrid are placed more and more attention, and it is becoming one of the hot issues. There are multiple generation sources and loads in the power generation system, and its system current consists of a number of frequency components. In addition, because of the complexity of Smart MicroGrid, the power system is vulnerable to interference and influence, and the current have characteristics of non-stationary. In such cases, this paper proposes the improved method base on short-time Fourier transform and phase modeling. It can effectively realize the separation and frequency tracking of multi-source non-stationary signal, suitable for the signal that energy of each component is very different. At the same time it can accurately extract initial phase estimationof of each frequency com- ponent of signal.