基于BP-Adaboost算法的输电塔-线结构整体可靠性分析

Global Reliability Analysis of Transmission Tower-line Structures Based on BP-Adaboost Algorithm

针对传统结构可靠性分析方法的应用局限性和计算成本较大等不足,引入BP-Adaboost集成学习算法,提出一种准确高效、简单易行的输电塔线结构整体可靠性分析方法。首先,建立输电塔线结构有限元分析模型,分析结构的失效准则并推导相应的极限状态功能函数;然后,将BP-Adaboost集成学习算法与蒙特卡洛模拟方法相结合,建立高精度预测模型来替代分析过程复杂的有限元模型,提出基于BP-Adaboost模型的输电塔线结构整体可靠性分析方法;最后,通过数值算例对所提方法进行应用与验证,证明所提方法具有更高的精度和效率,并进一步将其应用于实际输电塔线结构工程,分析多失效模式下输电塔线结构的整体可靠性。

Considering application limitations of and large calculation cost in traditional structural reliability methods,this paper introduces the BP-Adaboost ensemble learning algorithm and proposes a simple and efficient global reliability analysis method of transmission tower-line structures.Firstly,it establishes a finite element analysis model of the transmission tower-line structures,analyzes the failure criteria of the structure and derives corresponding limit state functions.Secondly,combing the BP-Adaboost ensemble learning algorithm and Monte Carlo simulation method,the paper builds a high accuracy prediction model to replace the finite element model with complicated analysis process.It also develops a global reliability analysis method of the transmission tower-line structure based on BP-Adaboost algorithm.Finally,the proposed method is applied and verified by a numerical example,which demonstrates that the proposed method has higher accuracy and efficiency.Additionally,it is further applied to the practical engineering,enabling the analysis of global reliability of transmission tower-line structure under multiple failure modes.