基于模态刚度的预应力混凝土梁损伤识别方法研究

阐明了模态刚度在损伤识别研究中的重要意义,并对11根多级损伤状态的预应力混凝土梁进行动力试验研究。通过对梁模态分析发现,由于噪音污染等多种因素的影响,仅凭各梁实测模态刚度数值的直观分析很难对梁的多级损伤状态进行有效的识别。为此,提出了以模态刚度变化率为损伤指标的BP神经网络和PNN神经网络的损伤识别方法,并利用实测数据验证所提方法的实用性。研究表明,两种神经网络分类器识别方法均能够有效应用于实际中,且具有很高的损伤识别精度,为结构损伤识别方法研究提供了新思路。

DAMAGE IDENTIFICATION OF PRESTRESSED CONCRETE BEAMS BASED ON MODAL STIFFNESS

This paper expounded that modal stiffness are of great significance for damage identification. It was proposed BP and PNN neural network for damage identification based on this modal parameter. The result,via experiment analysis on 11 prestressed concrete beams with multistage damage status,shows that the theoretical method based on modal stiffness is hardly practical in engineering application due to many factors such as noise and boundary condition etc,while the neural network can be effectively applied to identify damages with high precision.Moreover,which can offer a new idea for damage identification researches.