用于岩土工程位移预测的软科学方法研究

SOFT SCIENCE METHODS FOR DISPLACEMENT PREDICTION IN GEOTECHNICAL ENGINEERING

位移预测对岩土工程稳定性具有重要的实用意义,近来软科学方法得到了广泛重视及应用。对常用软科学方法——灰色系统、时序分析及神经网络进行了系统分析研究,指出了它们的优缺点及适用范围,并提出了一种新型软科学预测方法——进化神经网络法。最后,通过工程实例验证了理论分析的结果。

It is very important to study displacement prediction method in geotechnical engineering. Recently, the prediction methods based on soft science have been applied widely. Here, the generally used soft science methods-Grey System, Time Series Analysis and Neural Network are analyzed and studied comprehensively, and their merits and demerits are also revealed. Then, a new prediction method based on soft science-Evolutionary Neural Network is proposed. At last, the results of the theory analysis are verified b...