SWARM INTELLIGENT MODEL FOR DEFORMATION PREDICTING OF SLOPE

Deformation is an important index for the structure of geotechnical engineering;and the deformation monitoring is a key issue for the design and construction of geotechnical engineering. Technical staffs are interested in how to use the monitoring data to guide construction. Up to now,various approximate methods have been used for deformation predications. According to the method of time series analysis,a new method—particle swarm optimization(PSO) was presented and applied to the deformation predication of slope. It takes the deformation of slope as a time series variable and combines the time series analysis with PSO. The proposed method,having good characteristics of global optimization and better performance in practical engineering,can identify the structure and the relative parameters of predicted model. The method was successfully applied to predict the deformation of high slope of the permanent shiplock in the Three Gorges Project;and the achieved results show considerable satisfactoriness in the engineering application. The proposed method is quick and precise;and it can provide a good approach to the information construction and management in geotechnical engineering.