摘要
用神经网络方法对系统状态转移矩阵进行识别的研究，首先建立结构动力学运动方程的状态转移矩阵的识别与神经网络的权值矩阵识别间的相互等价关系，其次利用神经网络功能强大的并行运算能力和丰富的学习功能以单层线性神经网络来识别线性时不变系统的广义状态转移矩阵以获取结构的状态参数。最后通过识别获得的神经网络仿真系统结构在任意激励下的响应，可用以抗震性能评估或振动控制。
This paper deals with the research on identification of generalized state transfer matrix of linear time - invariant (LTI) system by use of neural networks based on LM (Levenberg - Marquardt) method. The relationship between the identification of state transfer matrix of structural dynamics and the identification of the weight Inarix of neural network has been estabished in theery. A single layer neural network is adopted to obtain the structural state pmters as a power tool which has parallel distributed pmeessing ability and the property of adaptation or learning. The constaint condition of weight matrix of the neural network is deduced so that the learming and training of the designed neural network can be more effective. The identified neural network is used to simulate the structund response excited by any othe signals . In order to cope with its further application in practical problems, some noise (5% and 10% here) is expected to be present in the output measumnts. Results from computer simulation studies show that this method is valid and feasible in predicting structural response.
出处
《土木工程学报》
CSCD
北大核心
2000年第2期40-45,共6页
China Civil Engineering Journal
基金
陕西省建设厅资助
关键词
神经网络
结构动力学
参数识别
抗震
neural networks
structural dynamics
parameter identificaion
earthguake engineering
