Discrete Particle Swarm Optimization Algorithm for Gearbox Fault Symptom Selection

Particle swarm optimization (PSO) has been developed rapidly and has been applied widely since it is easy to understand and realize.Aimed at mechanical fault symptom selection,the paper proposed a discrete Particle Swarm Optimization(PSO) algorithm.In order to improve the ability to escape from the local optimum,the mutation operator was added to the algorithm.The discrete PSO algorithm was applied to helicopter gearbox gear fault symptom selection.Symptom set extracted by wavelet packet transform method was presented by binary-coded particle.The accuracy evaluated by k-fold cross-validation of the SVM classifiers was used as the particle fitness.Experimental results indicate that the discrete PSO can get the same optimal symptom subset as that from genetic algorithm,and achieve the better diagnosis accuracy,but the compute time consumed is only 0.4 time that of genetic algorithm,so the discrete PSO is a more effective method for mechanical fault symptom selection problem.