基于遗传算法的混凝土湿度场特征参数求解及反分析

Solution of Concrete Moisture Diffusion Parameters Based on Genetic Algorithms and Back Analysis

为了求解干燥环境中混凝土内部湿度分布,提出了利用遗传算法来求解混凝土湿度场特征参数.根据室内实验实测的湿度值,运用遗传算法对混凝土试块湿度场进行反演计算,得到反映混凝土湿度扩散性能的参数.利用这些参数和三维有限元仿真计算程序对湿度场进行分析,将计算值和实测值进行对比得知二者变化规律一致,各组数据相关系数大于0.98,说明利用反演的参数仿真计算是完全合理的,可以指导施工,同时说明遗传算法作为一种优化算法,具有收敛速度快、效果好等特点.

To obtain moisture distribution of concrete exposed to dry air,genetic algorithms were proposed to calcu-late moisture diffusion parameters.Based on the measured relative humidity,relevant parameters were obtained by inverse analysis.With those parameters and 3-D FEM,the long-term distribution of relative humidity was gained.From the result,it is known that the calculated values agree well with the experimental values.The correlation coeffi-cient R of all statistics is higher than 0.98,which indicates that simulation analysis using inversion parameters is quite reasonable and can be applied to guiding construction.Meanwhile,as one of optimization methods,genetic algo-rithms are simple and highly efficient.