中文摘要：

针对目前软基沉降预测中最常用的生长曲线模型以及人工神经网络模型的不足,提出将自适应神经模糊推理系统（ANFIS）应用于软基沉降预测。ANFIS将专家的模糊推理过程蕴含于神经网络结构中,使神经网络的结点和权值具有明确的物理意义,避免了传统神经网络工作过程的＂黑盒＂性。同时该系统可以采用最小二乘法和梯度下降法相结合的混合算法,既具有神经网络的自适应性和学习能力,又克服了它的局部极小值等缺点,预测精度也远高于生长曲线模型。最后用工程实例与生长模型和神经网络模型进行了对比,结果表明：ANFIS模型优于这两种模型,特别是在模拟多输入变量、高维数下软基沉降预测问题时有着独特的优势,具有一定的推广应用价值。

英文摘要：

To compensate the defect of the growth model and the Artificial Neural Network model,which are the most commonly used method to settlement prediction of soft clay roadbed,this paper proposes the application of the adaptive neuro-fuzzy inference system（ANFIS）to soft clay roadbed settlement prediction.In the ANFIS the fuzzy reasoning process of the expert inherent is put in the neural network structure.The neural network nodes and the weights have a clear physical meaning,and the ＂black box＂ through the working process of Neural Network is avoided.The system not only uses the least square method and gradient descent algorithm for the combination of mixed with both the self-adaptive and learning ability of a neural network,but also overcomes its disadvantages such as the local minimum.The prediction accuracy is also much higher than that with the growth curve model.The methods are applied to the calculation the examples.The result show that the ANFIS model is of great theoretic significance and has practical value in preventing roadbed sink and guaranteeing the road behavior in practice.