软土固结系数Cv是岩土工程结构设计中的重要参数，测试固结系数需要花费大量的时间和成本。基于生物地理学优化算法（biogeographical optimization algorithm，BBO）来改进人工神经网络（artificial neural network，ANN）模型，使用生物地理优化的人工神经网络（ANN-BBO）模型对软土固结系数进行计算。基于连淮高速公路改扩建项目的路基软土数据对计算模型进行了训练和测试。通过相关系数矩阵和主成分分析对软土的11个物理力学参数进行了统计分析，确定了7个参数作为计算模型的输入参数，并对计算模型进行了训练和测试；模型通过相关系数、均方根误差和方差比进行检验，并对误差进行了统计分析；采用蒙特卡罗模拟对计算模型的鲁棒性进行了分析。结果表明：ANN-BBO模型可以用于估计软土的固结系数，相关系数R2 = 0.947 1，均方根误差RMSE = 0.165 7×10−3 cm2/s，方差比VAF= 94.54%；ANN-BBO模型预测精度明显优于ANN模型；ANN-BBO模型的鲁棒性比ANN模型更好。The consolidation coefficient Cv of soft soil is an important parameter ingeotechnical engineering. The artificial neural network (ANN) model is improvedwith the biogeography-based optimization algorithm (BBO). And the artificial neuralnetwork with biogeography-based optimization (ANN-BBO) model

trained and tested byusing the subgrade soft soil data of the reconstruction and expansion projectof Lianyungang–Huai’an expressway

has been adopted to calculate the soft soilconsolidation coefficient. Using the correlation coefficient matrix andprincipal component analysis

eleven physical and mechanical parameters werestatistically analyzed

seven of which were identified as input parameters ofcalculation model which was then trained and tested. The model was tested bycorrelation coefficient

root mean square error

and variance ratio

whoserobustness was analyzed by using Monte Carlo simulation. The results show thatthe ANN-BBO model can be used to calculate theconsolidation coefficient of soft soil

the correlation coefficient R 2 = 0.947 1

the root mean square error RMSE= 0.165 7×10−3 cm2/s

and the variance ratio VAF = 94.54%. TheANN–BBO model has significantly higher predictionaccuracy and better robustness

compared with the ANN model.