Reliability analysis of functionally graded beam structure using support vector machine
In practical engineering applications,the reliability analysis is implemented for obtaining the probability of an event that leads to the failure of the concerned systems.Due to the complexity of the engineering structures,the limit state functions(LSFs)are not guaranteed to be available explicitly.In this context,surrogate models can be adopted to approximate the relationship between the input system parameters and critical structural responses.Recently,a statistical learning based technique,the support vector machine(SVM),has been employed as an efficient and effective surrogate method for structurally reliability analysis.Assisted by the design of experiments,the closed-form estimated expression of the LSF can be constructed by solving a quadratic programming problem.Thus,the LSF is approximated as an expression of the input random variables and the then failure probability can be calculated accordingly with less computational effort.In this paper,the reliability of the functionally graded beam structure is studied by using a SVM approach.The performance of the adopted SVM-based reliability approach is demonstrated to satisfy the engineering requirement by comparing with the results of the conventional sampling method.
Structural reliability Support vector machine Functional graded material Finite element analysis Surrogate model
Yuan.Feng Qingya.Li Wei.Gao
Centre for Infrastructure Engineering and Safety(CIES),School of Civil and Environmental Engineering,The University of New South Wales Sydney,NSW 2052,Australia
国际会议
The 7th World Conference on Structural Control and Monitoring(7WCSCM)(第七届结构控制与监测世界大会)
青岛
英文
86-92
2018-07-22（万方平台首次上网日期，不代表论文的发表时间）