Study On Fault Diagnosis For Gear-box Based On The Neural Network Of Particle Swarm Optimization

Particle swarm optimization (PSO) is a kind of optimization algorithm based on thetheory of swarm intelligence. It instructs optimization searching by competition andcooperation of each particle in race swarm. The kind of intelligence algorithm can be used tosolve various optimization problems and shows great potential in practice. Now, it has beenwidely applied in many other areas, such as artificial neural network and fuzzy systemcontrol.The paper introduces the basal theory of particle swarm optimization, and studies on thesocial behavior by the analysis of evolvable equation of particle's velocity. Then, the particleswarm optimization is thought of as the evolvement of dynamic system. By analysis methodof linear discrete time system, the paper analyses the certain behavior of particle swarmoptimization, and educes the convergence condition of sample particle swarm optimization.The neural network of particle swarm optimization is programmed by c++ based on analysisneural network.The gear-box fault vibration theory is analyzed deeply, the usual fault forms and causesare studied, signal eigenvalues are chosen. The gear-box(JZQ250) is thought of as researchfulobject, the faults of gear-box are diagnosed by trained neural network, experimental resultsshow all right.