大坝变形预测的ANFIS模型

大坝变形预报时，存在影响因素多且各因素之间的相互关系复杂，常规的变形预测方法难以满足大坝安全监控的要求。自适应神经模糊系统(ANFIS)兼备神经网络的自学习、自适应能力，以及模糊系统良好的知识表达性能。在系统分析大坝变形主要影响因素的基础上，以水库库水位、温度及时间效应为影响因子，建立基于自适应模糊神经系统的大坝变形预测模型，并以三峡二期围堰为例进行实证分析。研究表明，该模型计算简便，适用性强，精度高，为大坝变形预报提供了新的思路。

ANFIS model for predicting deformation of dam

In forecasting deformation of dam, many influential factors exist as well as interrelation of factors was complex. Hence, general simulation functions are difficultly to meet the needs of safety supervising of dam. As we know, adaptive neural-fuzzy inference system(ANFIS) has good capability of fuzzy system and artificial neural network. Influential factors of deformation of diaphragm wall analyzed systemically as well as effects of temperature, time, and water level of reservoir considered, predicting model of deformation of dam is established based on ANFIS. Taking the second stage cofferdam of Three Gorges Project for example, the studies are carried out with this model. The results show that the model is simple in calculation and good in applicability and precision, so as to provide a new method for forecasting deformation of dam.