摘要
本文将B样条神经网络用于有源压力脉动衰减系统。该系统与基于传统的滤波 XLMS(FLMS)算法的有源脉动衰减系统相比 ,可匹配系统中的非线性 ,初值稳定性好、收敛速度快、在所试验的 2 0～ 70Hz范围内 ,衰减压力脉动 8.7～
In this paper,B spline neural network is introduced into active attenuator for pressure pulsation.Compared with the attenuator using “Fllitered X”LMS algolition,the attenuatol using D splino noural notwork can match the nonlinearity in the system,the stability of initial value is good,and the convergence rate is quick.Within the experimental limits of 20～70Hz,the attenuation for pressure pulsation is 8.7～11dB.
出处
《液压气动与密封》
2001年第2期2-4,共3页
Hydraulics Pneumatics & Seals
