为了优化煤岩压裂声发射信号的降噪效果,采用小波变换与小波包变换相结合的方法,利用MATLAB软件的小波包降噪函数,分别对3种不同脆性矿物含量的煤岩压裂声发射信号进行了降噪处理.提出在运用小波包降噪算法的前提下,在常见小波基中选取最优基,可提高分析效率且降噪效果良好;选取测试时间相同的数据进行分析,是正确选择降噪层数的必要条件;结合多种评价指标进行综合分析,进而确定最佳的降噪层数,可有效避免降噪不足或降噪过度.实验结果表明,对脆性矿物含量为30％、50％、70％的煤岩压裂声发射信号进行小波包降噪处理的最优基分别为Sym2、Sym6、Sym5,且最佳降噪层数均为3层.SymN小波系最适合煤岩压裂声发射信号的降噪,N值的选取与煤岩的脆性矿物含量有关,且在DbN小波系、CoifN小波系中,各自的最优基均具有较小的N值.随着煤岩脆性矿物含量的增加,SymN小波系相对于其他2种小波系在小波包降噪效果上表现出的优势越加明显.%In order to optimize the de-noise effect,the paper adopt the method of wavelet transform to combine with wavelet packet transform,and use the wavelet packet de-noise function of MATLAB software to take de-noise processing about the acoustic emission experiment signal,which the signal is from coal rock fracturing of three kinds of brittle mineral content.And the paper puts forward that on the premise of applying wavelet packet de-noise algorithmic,we improve to analyze efficiency and obtained good de-noise effect;choosing and analyzing the data at the same testing time is the requirement of determining the best number of plies;combining with various evaluation index and generalized analysis and then confirm the best number of plies could effectively avoid de-noising deficiency or de-nosing excessive.The experiment results indicate that the signal of three kinds of coal rock samples those mineral content are 30％,50％,70％ and respectively the best wavelet basis is Sym2,Sym6,Sym5,and the best number of these plies is three.SymN wavelet is best for denosing of the coal rock fracturing acoustic emission signal,which the value of N relates to mineral content of rock.In the DbN and CoifN wavelet,each wavelet has smaller value of N.With the increasing of brittle mineral content,the SymN has more obvious advantage of denoise effect relative to other two kinds of wavelet have.
展开▼
