基于改进PSO的有源噪声控制算法关键参数在线获取

Online Acquisition of Key Parameters of Active Noise Control AlgorithmBased on Modified PSO

针对有源噪声控制(Active Noise Control, ANC)系统中自适应算法关键参数难以最优设置,导致降噪系统性能不理想的问题,提出了一种改进粒子群(PSO)算法用于在线获取ANC系统关键参数。该算法在ANC系统的实时采样过程中,随机选择粒子作为ANC算法的关键参数,并以数据块为单位迭代更新粒子群。算法采用动态的惯性权重和非线性的适应度函数,解决了传统PSO算法寻优能力不足和鲁棒性弱的问题。实验结果表明,改进PSO算法在ANC关键参数在线寻优过程中,仅需20次迭代即能找到最优参数值,收敛速度快,并且算法获得的稳态误差值最低,稳定值为-35.6 dB,降噪表现优异。

A modified Particle Swarm Optimization(PSO)algorithm is proposed in this paper for online acquisition of ANC(Active Noise Control)parameters in response to the problem of difficult optimal setting of key parameters in adaptive algorithms in Active Noise Control(ANC)systems,resulting in unsatisfactory performance of noise reduction systems.The algorithm randomly selects particles as key parameters of the ANC algorithm during real-time sampling of the ANC system,and iteratively updates the particle swarm on a data block basis.The algorithm uses dynamic inertia weights and non-linear fitness functions to solve the problems of insufficient optimization capability and weak robustness of the traditional PSO algorithm.The experimental results show that the improved PSO algorithm can find the optimal parameter value in the online optimization process of ANC key parameters with only 20 iterations.Moreover,the steady-state error obtained by algorithm is the lowest,with a stable value of-35.6 dB and excellent noise reduction performance.