【摘要】 针对隧道初期支护结构可靠度几何优化模型搜索区域大而可行区域小且约束为隐式形式的特点,提出了基于改进遗传算法的可靠度计算方法。该法将初始种群分成可靠种群和失效种群2组,并对它们重复使用二分法来产生可行解,然后运用遗传算法实现全局最优的快速求解。通过算例验证了所提方法具有较高的精度和效率,且该方法对计算其它线弹性结构的可靠度也具有广泛的适用性。更多还原

【Abstract】 It is one of main directions of research on reliability of tunnel structure that research on efficient and precision meeting for engineering requirement method for reliability index calculation.To large search space and small feasible space of the tunnel primary support optimization model and the implicit of the limit state function of carrying capacity,the method for reliability calculation based on improved Genetic Algorithms is putted in the thesis.The initial population is divided into two parts,which one part lies in the safe domain and another part lies in the failure domain.A feasible population is produced by applying bisection method repeatedly.Then optimal solution in the whole space is solved by Genetic Algorithms quickly.Accuracy and efficiency of the method are tested by example,and the method is broadly applied for other linear elastic structure.更多还原