The large number, scattered distribution, and complex working environment of idlers makes their faults challenging to detect. In this paper, a fault analysis method of belt conveyor idlers based on sound and thermal infrared image (TII) features is proposed. According to 18 classes of idler sound and TII data, the time-domain (TD) features of sound signals are analysed using statistical methods, the frequency-domain (FD) and time-frequency-domain (TFD) features of sound signals are analysed with the quantization and dimension reduction method based on Fisher’s linear discriminant, and the TII features are analysed using statistical methods. The analysis results show that final and catastrophic faults can be detected by using FD features of idler sound and TII temperature rise of the idler outer load area and shaft end, and TFD features of idler sound signals can be used to detect typical bearing defects, which features high reliability, low cost, and easy implementation.

托辊数量多、分布分散、工作环境复杂，使得其故障难以检测。本文提出了一种基于声热红外图像(TII)特征的带式输送机托辊故障分析方法。根据18类闲音和TII数据，采用统计方法分析声音信号的时域（TD）特征，分析声音信号的频域（FD）和时域（TFD）特征采用基于Fisher线性判别式的量化降维方法，并使用统计方法分析TII特征。分析结果表明，利用惰轮声音和惰轮外载荷区和轴端的 TII 温升的 FD 特征可以检测到最终和灾难性故障，