小波包在双进双出磨煤机料位检测中的应用

Application of wavelet packet in material level measurement of double in/out ball mill

针对双进双出磨煤机料位难于检测的问题,研究了基于小波包技术噪声法的磨煤机料位检测技术.在噪声信号分析中运用小波包变换技术,构造出了噪声特征向量,并对特征向量进行优化.经过仿真分析,找到了反映料位信息的敏感频率段,进而降低了向量的维数.采用BP人工神经网络,通过该网络的学习和训练,实现了噪声特征向量和料位之间的非线性映射.通过仿真实验,证明了该算法的有效性,具有较高的检测精度.

Aimed at the difficulty in detecting the material level of double in/out ball mill, a technique for restraining noise based on wavelet packet was developed. Wavelet packet transform technique was adopted in the analysis of noise signal. The noise characteristic vector was constructed and optimized. By simulation, the sensitive frequency range reflecting the material level was founded, and thus the dimension of the vector was decreased. The BP neural network was employed in the measurement approach. Through learning and training of the network, the nonlinear mapping between noise characteristic vector and material level was obtained. The experiments show that the present method is available and has high detecting precise.