Abstract
To solve the problem that motor data does not have intuitive semantics, and it is hard to get complete fault information as the signal kinds monitored is more and its complexity is higher than before, this paper proposes a fault diagnosis method based on Ontology and Particle Swarm-Immune Optimization algorithm. It first creates an ontology library using the expert knowledge. Secondly, it extends the fault data and creates a fault diagnosis trainer by the particle swarm optimization (PSO) and immune algorithms. At last, it will obtain an effective fault diagnosis trainer, which could improve the final fault diagnosis' accuracy and validity. Experiment results prove that the new fault diagnosis algorithm is an effective method, which effectively completes the fault information database and improves the fault diagnosis' accuracy and validity.