摘要
针对量子粒子群优化(QPSO)算法对越界粒子处理方式的不足,提出了一种基于边界控制的改进方法,并将其应用在有限长脉冲响应(FIR)滤波器的频率采样设计法中,给出了算法的具体实施步骤。对FIR低通和带通滤波器的仿真结果表明,相对于查表法及标准QPSO算法,改进后的QPSO算法能够快速、有效地求得频率过渡带样本值的最优解,同时通带波动变小,最小阻带衰减变大,从而对FIR滤波器的设计进行了进一步的优化,验证了改进算法的有效性。
Considering the premature convergence problem in the conventional quantum particle swarm optimization （QPSO） algorithm,an improved method based on boundary control was proposed and was applied in finite impulse response （FIR）filter design with frequency sampling method.The specific implementation steps of the improved algorithm were presented.In contrast with look-up table method and QPSO algorithm,the simulation results of FIR low pass and band pass filter verified that the improved method could rapidly and effectively find the optimal sample value in the frequency sampling,and the pass band ripple became small and the minimum stop band attenuation became large.The efficiency of the improved QPSO algorithm was verified.
出处
《辽宁石油化工大学学报》
CAS
2014年第6期67-70,共4页
Journal of Liaoning Petrochemical University
基金
辽宁省高校杰出青年学者成长计划项目(LJQ2011032)
辽宁省科技攻关项目(2011216011)
关键词
量子粒子群优化
边界控制
FIR滤波器
频率采样法
最优解
Quantum particle swarm optimization
Boundary control
FIR filter
Frequency sampling method
Optimal solution
