摘要
对单自由度刚度可调非线性隔振模型进行数学建模仿真和实验 .对于待训练的神经网络模型 ,将数值仿真结果作为学习样本 ,将实验结果作为检验样本 ,训练成功的神经网络模型具有较好的内延能力和一定的外推能力 ,在设定非线性隔振装置要求的隔振效能参数后 ,可以给出隔振装置对应的刚度和阻尼参数 。
In this paper the mathematical simulating and actual test are carried out for a single degree of freedom stiffness adjustable nonlinear antivibration device. With the numerical simulating results as learning samples and the test results as examining samples, a BP neural network is trained and obtained. After setting the antivibration efficiency parameters of nonlinear antivibration device, the BP neural network will give the values of stiffness and damping for the nonlinear antivibration device.
出处
《实验力学》
CSCD
北大核心
2002年第3期340-344,共5页
Journal of Experimental Mechanics