针对包含认知不确定性的结构可靠性问题,提出了一种基于证据理论的可靠性优化设计方法.首先,在优化过程中基于代理模型对迭代点处各约束进行可靠性分析,计算其似真度;其次,通过证据变量均匀化方法构造了一种证据理论可靠性的近似梯度计算方法,从而可采用基于梯度的常规优化算法进行优化求解;最后,数值算例和工程算例表明了本方法的有效性. Aiming at the problem of structural reliability including cognitive uncertainty, a reliability-based optimal design method based on evidence theory is proposed.Firstly, the reliability of each constraint at iteration point is analyzed based on agent model in the optimization process, Fidelity. Secondly, an approximate gradient calculation method for the reliability of evidence theory is constructed by the method of homogeneity of evidence variables, so that a conventional gradient-based optimization algorithm can be used to optimize the solution. Finally, numerical examples and engineering examples show that this method The effectiveness of the method.