Multichannel Vibration Fault Diagnosis for Rolling Bearings Based on QPCA and SVM

Abstract

A new method had been proposed in this paper of fault diagnosis for rolling bearings based on multichannel vibration signals and QPCA-SVM-based method. The vibration signals were obtained by some multi-sensors with three channels X, Y, Z, that were orthogonal axes. The three orthogonal axes signals were constructed a pure quaternion sequences as samples for processing. The pure quaternion sequences data set was processed by quaternion principle components analysis (QPCA) for feature extraction, and then combined with pattern recognition tools support vector machine (SVM) for classifying some faults patterns. The experimental results indicated its efficiency, and it provided a method for fault diagnosis on multichannel vibration signals.

The Research on Fault Diagnosis of Diesel Engine Based on EMD

“…As classified by the Libsvm when it comes to choosing the penalty parameter C and kernel function parameters of G, using the cross validation (CV, Cross, Validation) method can obtain the optimal parameters in some sense, avoid excessive learning and less learning state, is more effective in classification than the random parameter SVM model training. In addition, heuristic algorithm to optimize parameters, such as genetic algorithm (GA, Genetic and Algorithm) and particle swarm optimization (PSO, Particle parameter optimization Swarm Optimization) algorithm parameters optimization of [6][7][8]. In this paper, by three kinds of parameter optimization method for parameter selection, results in table 2.…”

State of charge estimation for lithium‐ion battery pack based on real vehicle data and optimized backpropagation method by adaptive cross mutation sparrow search algorithm

“…Directly using SSA to optimize the model parameters of BP neural networks may not achieve the expected results. 25,30 Cloud-based remote estimation of SOC is an emerging trend in BMS, providing a fundamental backbone for big data-driven SOC estimation. Unfortunately, existing machine learning methods have yet to gain widespread application in this domain, while the precision of existing optimized BP network techniques falls short of meeting practical requirements.…”

“…However, in the SSA algorithm, the number of producers and scroungers remains constant, and there is no mutation mechanism, which can quickly reduce the diversity of the population and lead to trapping in local optima, thereby reducing the search accuracy. Directly using SSA to optimize the model parameters of BP neural networks may not achieve the expected results 25,30 …”