Abstract: Three kinds of programming operators involving combination operator, harmony operator and abruption operator have been defined to perform community dynamic distribution in particle swarm optimization. Thus a new random optimization tool has been developed, i.e. community dynamic assignation-based particle swarm optimization (CDAPSO). Via changing organization structure and distribution character of particle swarm to increase varieties of eliciting information during optimization, this algorithm notably improves global searching abilities and effectively avoid problems of premature and convergence.