摘要
本文通过对线性结构体系的动力学方程进行正则变换，在理论上分析了主动控制力对残余振型的影响，提出了解决振型遗忘问题的矩阵奇异值辅助分析的新算法，并采用人工神经网络的BP算法对主动控制力进行了动态仿真。数值计算的结果表明：采用本文提出的算法可以方便、有效地减弱振型遗忘所引起的不利影响。
in this paper the effect of active controlling forces to the residual modes was analysed theoretically with canonical transformation for the dynamical equations of structural system. A new algorithm of Using SVD to solve the spill-over effect was presented. The BP algorithm of Artificial neural networks was used to simulate the dynamical controlling forces.The numerical results indicate the new algorithm which was given in this Paper is more convenient and efficient to eliminate the effect of spill-over.
出处
《齐齐哈尔大学学报（自然科学版）》
1998年第2期15-19,共5页
Journal of Qiqihar University(Natural Science Edition)
关键词
振型遗忘
人工神经网络
振动分析
结构振动
Spill-over effct, Canonical transformation, Arfificial neural networds
