压电智能结构荷载识别方法的研究

采用压电智能结构实测荷载的输出响应，基于BP神经网络与有限元逆分析提出一种识别荷 载位置及大小的方法. 首先在结构的不同位置施加单位荷载由有限元方法计算得到网络的学 习样本，经网络作逆分析识别荷载位置，继而通过有限元逆逼近方法确定荷载大小的最小二 乘解. 数值算例表明，该方法计算速度快、精度高，不受结构几何形状和边界条件的限制， 用于识别实际压电智能结构不确定荷载的位置及大小是可行的.

On load identification for piezoelectric smart structures

Based on BP neural network and finite element inverse analysis, a method is proposed to identify the location and magnitute of loads by measuring the piezoelectric responsive charge on the piezoelectric smart structure. Firstly, the unit loads are acted on several different locations on the structure and the leafing stylebook of net is calculated by the finite element method, by which the location of loads may be finded. Then, the magnitute of loads is determined by finite element inverse analysis and the least square method. The calculation example shows that the method has high precise and rapid calculation velocity, which is suit for the piezoelectric smart structures with complex shape and boundary conditions and may find its uses in the loads identification of the applied smart structures.