【摘要】 通过声波信号在线分析反应器内物料特征具有快速、无损、安全、不侵入流场的优点。讨论了利用声发射信号在线分析搅拌釜内物料浓度与粒度的方法。先通过小波变换多尺度分解原始声信号,然后以分解后各细节信号小波系数幅值绝对值加和为特征变量,釜内物料浓度或颗粒平均粒度和声发射信号的关系再用判别分析方法加以区分,考察了信号长度、分解尺度以及不同小波对判别准确性的影响,结果表明,当物料的浓度一定时,基于多尺度分析的声测量方法可以准确地对釜内物料的粒度进行分类。更多还原

【Abstract】 It is advantage to analysize material’s character in a stirrer by acoustic emission signals for the sake of celerity,lossless, safe and non-invasive.Method on how to measure the concentration and the average particle size on-line in a stirrer by acoustic emission (AE) signals had been detailedly discussed.Original AE signals were firstly multi-scale decomposed by the wavelet.After that, the absolute wavelet coefficients of the approximation and the detail signals were summed.The summations(energy of signals) that had relationship with the material concentration and the average particle size were used as characteristic variables of discriminant analysis. Effects of the signal length,decomposition level and wavelet species on classification accuracy had been discussed.The results show the concentration and the particle size in a stirrer could be clasisfied with high accuracy by discrimination analysis employed with multi-scale analysis.更多还原