混凝土疲劳损伤累积神经网络模型

Fatigue damage cumulative model of concrete based on artificial neural network

针对混凝土疲劳累积损伤的发展受荷载、材料性质等多种因素影响、很难用显式的数学函数表达的难题,基于神经网络方法提出了一个考虑变幅疲劳的级数和加载顺序影响的混凝土疲劳损伤累积模型,采用神经网络技术描述了其与材料性能、荷载应力间复杂的非线性映射关系,准确反映了混凝土疲劳损伤演变的真实过程,并运用离散系统仿真原理构造了混凝土变幅疲劳寿命预测的智能化仿真系统.该系统可用于结构抗疲劳设计和疲劳可靠性分析、有损结构损伤评估及维修决策.

The accumulation of fatigue damage of concrete related to the loadings, properties of materials and other unknown factors is difficult to be described by mathematical function. The cumulative damage theory based on the artificial neural network approach of concrete that can model the magnitude and sequence of variable-amplitude fatigue loadings is proposed. The artificial neural network approach is used to describe the nonlinear relationship between the properties of materials and the loading, which can reflect the accumulation of damage of concrete accurately. An intelligent simulation system based on the principle of discrete system simulation is also proposed to predict the residual fatigue life of concrete under variable-amplitude loadings. The system is valuable to structural design in fatigue resistance, analysis of reliability on fatigue, evaluation of damaged structure and making decision for maintenance.