Particle swarm optimisation-based support vector machine for intelligent fault diagnosis

In this paper, we present an application of support vector machines (SVMs) and particle swarm optimisation (PSO) to fault diagnosis. SVMs have been successfully employed to solve regression problem of nonlinearity and small sample. However, the practicability of SVM is affected due to the difficulty of selecting appropriate SVM parameters. PSO is a new optimisation method, which is motivated by social behaviour of organisms such as bird flocking and fish schooling. The method not only has strong global search capability but also is very easy to implement. Thus, the proposed PSO-SVM model is applied to diagnosis operation of rolling bearing failure in this paper, in which PSO is used to determine free parameters of SVM. The experimental results also indicate that the SVM method can achieve greater accuracy than grey model, artificial neural network under the condition of availability of small training data.