Analyzing harmonic disturbance using the conception of time-frequency distribution

• Main Author: 蔡佳憲
• Other Authors: 黎傳樹
• Format: Others
• Language: zh-TW
• Published: 2005
• Online Access: http://ndltd.ncl.edu.tw/handle/46243009781138130989

碩士 === 明新科技大學 === 電機工程研究所 === 93 === The major purpose of this thesis is to investigate the theories of various time-frequency analysis (TFA), and it can apply crucial analysis of non-stationary signals to overcome electromagnetic compatibility (EMC) problems which caused by harmonic interference. The conventional spectrum analysis technique utilizing Fourier theory cannot provide us with effective answers enough to solve the harmonics problems, since it cannot provide us localized time domain information. A method involving windowing of waveform to ameliorate this drawback is known as short-time Fourier transform. Its capabilities in dynamic harmonic analysis have been researched widely. Unfortunately, restricted to time and frequency resolution inherent in STFT lead to develop quadratic (analytic) time-frequency signal representations such as Wigner distribution. In general, quadratic time-frequency distribution have many good properties such as instantaneous frequency、instantaneous bandwidth、group delay, at the same time, these methods also exists cross-term in the spectrogram of a multi-component signal and zero value to obscure joint time-frequency atoms. The fundamental mathematical backgrounds of the several commonly used linear and quadratic time-frequency distributions are mentioned in this thesis. Further, the advantages and disadvantages about Wigner distribution and Wavelet transform base on the inference are discussed and develop harmonic analysis tools. Multi-resolution analysis also address in his thesis. MATLAB tool has been used to simulate the harmonic distortion of analyzed result.