High cut slopes have been widely formed due to excavation activities during the period of immigrant relocation in the reservoir area of the Three Gorges, China. Effective reinforcement meas-ures must be taken to guarantee the stability of the slopes and the safety of residents. This article pre-sents a comprehensive method for integrating particle swarm optimization （PSO） and support vector machines （SVMs）, combined with numerical analysis, to handle the determination of appropriate rein-forcement parameters, which guarantee both slope stability and lower construction costs. The relation-ship between reinforcement parameters and slope factor of safety （FOS） and construction costs is in-vestigated by numerical analysis and SVMs, PSO is adopted to determine the best SVM performance resulting in the lowest construction costs for a given FOS. This methodology is demonstrated by a prac-tical reservoir high cut slope stabilised with anti-sliding piles, which is located at the Xingshan （兴山） County of Hubei （湖北） Province, China. The determination process of reinforcement parameters is discussed profoundly, and the pile spacing, length, and section dimension are obtained. The results pro-vide a satisfactory reinforcement design, making it possible a signficant reduction in construction costs.
A new method integrating support vector machine (SVM),particle swarm optimization (PSO) and chaotic mapping (CPSO-SVM) was proposed to predict the deformation of tunnel surrounding rock mass.Since chaotic mapping was featured by certainty,ergodicity and stochastic property,it was employed to improve the convergence rate and resulting precision of PSO.The chaotic PSO was adopted in the optimization of the appropriate SVM parameters,such as kernel function and training parameters,improving substantially the generalization ability of SVM.And finally,the integrating method was applied to predict the convergence deformation of the Xiakeng tunnel in China.The results indicate that the proposed method can describe the relationship of deformation time series well and is proved to be more efficient.
