By Analysing speech signal and conventional denoising method of threshold, a new mult-threshold based on wavelet transform method of speech signal denoising is proposed.
研究了基于小波变换的语音去噪问题，在对传统阈值法去噪和语音特性分析基础上，提出一种改进的多阈值法语音信号去噪方法。
Prediction parameters preprocessing method is proposed which is based on time series analysis with wavelet transform method. And in this way, the prediction accuracy is increased.
提出了基于时间序列分析和小波变换方法的实测参数预处理方法，提高了预测精度。
This paper also does some examples to show that a curve matching method based on wavelet deal with false value better and have higher accuracy than modulus maxima of wavelet transform method.
通过算例验证，结果表明该方法与传统的小波模极大值法相比，较好地解决了行波波形的伪极值点问题，提高了故障定位精度。
This paper also does some examples to show that a curve matching method based on wavelet deal with false value better and have higher accuracy than modulus maxima of wavelet transform method.
通过算例验证，结果表明该方法与传统的小波模极大值法相比，较好地解决了行波波形的伪极值点问题，提高了故障定位精度。
应用推荐