The GA -BP neural network and the orthogonal test design method were used to analyze the sensitivity of affecting factors of the bearing capacity of PHC piles in a area of Fujian province, in which the piles bearing behavior belongs to end bearing - friction type and the pile shaft teems with mucky soil and residual cohesive soil layers primarily. The main conclusions are as follows; ①according to priority,the sensitivity order of the influence factors on vertical bearing capacity of the single pile are pile diameter,the weighted average of pile shaft resistance values,resting stage,pile length, buried depth of pile, permeability coefficient and final pile static pressure; ②the contribution of pile diameter and pile shaft resistance value to vertical bearing capacity of the single pile is the largest, resting stage is in the next place which influence is also very notable,so the pile test time should not be in advance blindly, otherwise the pile test results would be affected objectivity.%利用GA - BP神经网络,结合正交试验设计方法,对以淤泥层和残积粘性土层为主的福建某地区桩侧土层,承载性状为端承摩擦型的PHC管桩实际工程,进行PHC管桩单桩承载力影响因素的敏感性分析.结果表明,影响单桩竖向极限承载力因素的敏感性排序为桩径、桩侧摩阻力加权平均值、休止期、桩长、入土深度、渗透系数和终止压桩力；桩径和桩侧摩阻力值对单桩竖向极限承载力的贡献最大,其次是休止期,影响也很显著,因此工程上不宜盲目提前试桩时间,否则将影响试桩结果.
展开▼
