【摘要】 利用小波变换在时频域具有一定时间和频率多分辨率的特性,设计了小波阈值优化方法.根据信号与随机噪音在小波变换下各尺度不同的性质,同时减小拒真概率和虚报概率,在有效抑制随机噪音基础上,很好地保留了信号的细节成分.峰值误差为0.7%～1.0%,峰位误差为0.1%～0.3%.提高了光谱信号分析的准确性及系统的判停准确度.更多还原

【Abstract】 Using the time and frequency multi-resolution properties of wavelet transform,adaptive threshold adjustment method was designed.According to the different characteristics of signal and the random noise processed by wavelet transform in different scales,both rejecting true probability and false declaration probability were reduced.On the basis of reserving signal detail,the random noise was suppressed effectively.All the details were retained well.Peak error is from 0.7% to 1.0%,and peak position error is from 0.1% to 0.3%.And the analysis accuracy of spectrum signal and the precision of systematic decision pause were improved.更多还原