Abstract: Traditional regression model predicts badly when the independent variable has a complex relation with dependent variable. Intelligent model develops fast due to his visibility, networking and easy availability. Based on the measured data of a concrete dam, extreme learning machine was used in dam safety monitoring and a comparison with the operation process and the prediction results by BP neural network and support vector machine for regression was carried out. Moreover, their actual performances and application occasion were carefully analyzed.

Key words: safety monitoring of dam, intelligent model, extreme learning machine, BP neural network, support vector machine for regression

摘要： 当效应量和自变量关系复杂时，传统回归模型预测效果较差，智能模型以可视化、网络化、易于实现等特征发展迅速。基于某混凝土坝实测资料，将极限学习机模型运用于大坝安全监控中，并与BP 神经网络、支持向量机模型的训练、预测结果对比，分析了各个模型的实际运行性能和适用场合。

关键词: 大坝安全监控, 智能模型, 极限学习机, BP 神经网络, 支持向量机拟合