简述了数据挖掘理论和相空间重构理论的原理和算法，并通过实例讨论了其在大坝裂缝开度预测中的应用。将混沌理论和数据挖掘模型相结合，对揭示大坝裂缝复杂的非线性结构是很有效的。经实例研究，并与用单一的混沌相空间重构理论模型预测的计算结果相对比，表明该模型应用在大坝裂缝开度预测中是可行的、合理的，有很好的预报精度和应用价值。

The thesis expounds the basic principle and method of data mining and phase - space reconstruction theory, and discusses its application in the prediction of dam crack. The combined method of chaos reconstruction and data mining model is very effective for discovering the complex nonlinear structure of dam crack. It is proved that the application of the model above is practicable and reasonable on the basis of practical examples researching and in comparison with the predicted results obtained by unitary chaos reconstruction theory. And it has high predicting accuracy and much applicable significance.