【摘要】 为了有效利用结构健康诊断中的多源不确定数据,提高损伤识别的准确率,通过改进D-S证据理论(Dempster提出由Shafer完善形成的一种推理理论)中的基本概率赋值函数和组合规则,提出了一种基于改进D-S证据理论的结构损伤识别新方法。该方法利用灰色关联理论和熵权理论处理信息源的基本概率赋值问题,利用改进的D-S组合规则处理信息源间的相关、冲突问题。通过钢管焊接结构的试验研究验证所提方法的有效性,结果表明所提出的方法优于传统的D-S损伤识别方法;相对于单一损伤信息的损伤识别而言,基于多源损伤信息能够有效降低损伤误判的可能性,获得更准确的损伤识别结果。更多还原

【Abstract】 In order to make full use of multi-resource and uncertain data and to improve the damage identification accuracy of the objective from structural health diagnosis,a new damage identification method based on improved DS evidence theory are proposed in this paper via improving the basic belief assignment functions and compound rule.This method deals the basic probability assignment by the entropy weight coefficient and grey incidence analysis,and solves the problem of related or conflicting evidences by improved D-S compound rules.A steel tube welded structure is utilized to validate this proposed method.Results show that the method presented in this paper is superior to the classical D-S fusion theory.And in comparison to the single damage assessment method,more reliable identification results can be obtained by the proposed algorithm.更多还原