Requirements of modern industryial development rapidly and reliably achieve the fault diagnosis. Against particle swarm algorithm for the reduction and other issues so easy to fall into local optimum problem,this paper aims to present the MIQPSO Algorithm. The quantum particle swarm algorithm for clustering by the MIQPSO Algorithm, and through vaccination, to guide the direction of the particle evolution towards more optimized, improve the convergence rates and optimization searching ability of the quantum particle swarm. The use of UCI data sets, and by Hu algorithm, particle swarm optimization, quantum particle swarm optimization, multi-species quantum particle swarm algorithm for rough set attribute reduction verification, the results show that the algorithm based on the quantum particle swarm optimization has good reduction effect on the reduction.