在冻土路基融沉变形极限状态方程的基础上,从可靠度指标的几何涵义出发,提出了基于遗传算法的冻土路基融沉可靠度指标和失效概率的计算方法。该法是一种全局优化算法,能有效克服传统搜索算法易陷入局部极小值的缺点,不需要对功能函数进行直接转换,避免了功能函数比较复杂时所带来的求解困难,特别适合求解非线性规划问题。以青藏铁路冻土路基的具体实例,验证了本方法的准确性和有效性。
On the basis of limit equilibrium equation for thawing settlement of permafrost subgrade, a global optimal algorithm is put forward to calculate the reliability index and failure probability based on genetic algorithm. It can overcome the drawbacks of local optimal wildly existing in classic searching method. By this method, difficulty brought by the transformation of complex performance function can be avoided and nonlinear programming problem can be solved efficiently. Engineering example of Qinghai-Tibet Railway shows that this method is accurate and reliable.