摘要
针对传统方法获得路面功率谱密度的不足,提出了一种基于新型样条权函数(SWF)神经网络的路面功率谱密度识别方法。建立了4自由度汽车的振动系统模型,推导出了汽车振动系统的输出响应谱密度与输入激励谱密度之间的非线性关系,采用样条权函数神经网络对这种非线性映射关系进行了仿真。结果表明,基于样条权函数神经网络的路面功率谱密度识别方法比较成功地克服了传统算法的缺点,具有更高的识别精度。
Considering the shortage of traditional method for identifying the road surface power spectrum density（PSD）,a new method based on spline-weigh-function neural network（SWFNN） is introduced.A vibration model of four-freedom vehicle is set up and the nonlinear relation between response＇PSD and excitation＇PSD of the vibration system is established.Using the SWFNN,the nonlinear relation is simulated.The simulation results show that the proposed method succeedes in overcoming the shortcomings of conventional algorithmic and has higher identification accuracy.
出处
《潍坊学院学报》
2010年第2期31-34,44,共5页
Journal of Weifang University
关键词
路面不平度
功率谱密度
样条权函数
神经网络
road roughness
power spectrum density
spline weigh function
neural network
