A new framework for intelligent simultaneous-fault diagnosis of rotating machinery using pairwise-coupled sparse Bayesian extreme learning committee machine

• DOI: 10.1177/0954406216632022
• Publication Date: 2016-02-18T01:53:30Z
• Authors: Pak Kin Wong; Jian-Hua Zhong; Zhi-Xin Yang; Chi Man Vong
• Keywords: Extreme Learning Machine

This paper proposes a new diagnostic framework, namely, probabilistic committee machine, to diagnose simultaneous-fault in the rotating machinery. The framework combines feature extraction method with ensemble empirical mode decomposition and singular value decomposition, multiple pairwise-coupled sparse Bayesian extreme learning machines (PCSBELM), parameter optimization algorithm create an intelligent framework. is employed find features of single faults pattern. Multiple PCSBELM networks are built as different signal members, each member trained using vibration or sound signals respectively. individual result from fault detection then combined by method, which can improve overall accuracy increase number detectable compared classifier acting alone. effectiveness proposed verified case study on gearbox detection. Experimental results show superior existing classifier. Moreover, system both single- simultaneous-faults for machinery while single-fault patterns only.