数理统计中在处理回归的问题时,常用的传统参数估计方法存在着一些严重不足之处。为解决此问题,提出了将基于量子行为的微粒群优化(QPSO)算法应用于复杂函数的参数估计中。通过仿真实验,表明了该算法不仅可以准确地估计出复杂函数的参数,并且具有计算简便、收敛速度快等特点。通过与传统微粒群(PSO)算法的比较,证明了QPSO算法的优越性。
The quantum-behaved particle swarm optimization (QPSO) algorithm is developed for some serious disadvantages of traditional parameter estimation methods of complex functions in statistics. It indicates that QPSO algorithm can not only estimate parameters of complex functions correctly, but also can calculate simply and constringe fast. By comparing results with traditional particle swarm optimization (PSO) algorithm, QPSO algorithm is demonstrated superior to PSO algorithm.