Cutting Pattern Identification for Coal Mining Shearer through a Swarm Intelligence–Based Variable Translation Wavelet Neural Network

Sensors

Jing Xu¹  Xinhua Liu¹  Lei Si¹  Zhongbin Wang¹  Chao Tan¹ 

[1] School of Mechatronic Engineering, China University of Mining and Technology, No.1 Daxue Road, Xuzhou 221116, China;

关键词: cutting pattern identification;    sound signal;    variable translation wavelet neural network;    bat algorithm;    ensemble empirical mode decomposition;    disturbance coefficient;   

DOI  :  10.3390/s18020382

来源: DOAJ

【 摘 要 】

As a sound signal has the advantages of non-contacted measurement, compact structure, and low power consumption, it has resulted in much attention in many fields. In this paper, the sound signal of the coal mining shearer is analyzed to realize the accurate online cutting pattern identification and guarantee the safety quality of the working face. The original acoustic signal is first collected through an industrial microphone and decomposed by adaptive ensemble empirical mode decomposition (EEMD). A 13-dimensional set composed by the normalized energy of each level is extracted as the feature vector in the next step. Then, a swarm intelligence optimization algorithm inspired by bat foraging behavior is applied to determine key parameters of the traditional variable translation wavelet neural network (VTWNN). Moreover, a disturbance coefficient is introduced into the basic bat algorithm (BA) to overcome the disadvantage of easily falling into local extremum and limited exploration ability. The VTWNN optimized by the modified BA (VTWNN-MBA) is used as the cutting pattern recognizer. Finally, a simulation example, with an accuracy of 95.25%, and a series of comparisons are conducted to prove the effectiveness and superiority of the proposed method.

【 授权许可】

Unknown