中文摘要：

利用声发射检测技术,根据固体和液体碰撞壁面产生不同频率的声发射信号的机理,结合小波分析,建立了搅拌釜中浆液浓度的预测模型,模型可以定量描述特征频段的声能量值随浆液浓度和搅拌转速的变化规律.以水-沙子体系为例,分别考察了浆液浓度为0～0.4 g·ml^-1、搅拌转速为350～550 r·min^-1条件下的声发射信号.实验发现,在完全悬浮的条件下,将声发射信号在0～30 kHz频率范围内作8尺度小波分解,第4、5、6等3个频段的声发射信号的能量值之和随浆液浓度的增加线性增加,随转速的增加呈三次方关系增加.而浆液浓度预测模型的计算值与实验值之间的平均相对误差为7.62%.研究表明,声发射检测技术能快速、准确地实现浆液浓度的在线检测.

英文摘要：

A model of the slurry concentration in a stirred vessel was proposed to explore the quantitative relationship between AE （acoustic emission） energy, stirring rate, and slurry concentration, by using wavelet transform. It was based on the mechanism that AE signals with different frequencies were emitted when solid and liquid impacted the wall. The system of water and sands was used as an example and was investigated in the slurry concentration range from 0 to 0.4 g · ml^-1 and at the stirring rates from 350 to 550 r · min^-1 in a stirred vessel. AE signals were analyzed by using 8 scales wavelet transform when the stirring rate was greater than 350 r · min^-1. It was found that the sum of the AE energy in the 4th, 5th, 6th frequency ranges was proportional to the slurry concentration, and to the 3rd power of the stirring rate. Based on these experimental data, the model parameters were estimated. The average absolute relative error between the calculated values and the actual values of the slurry concentration was 7.62 % . Therefore, this model could predict the slurry concentration with fairly good accuracy.