INTELLIGENT METHOD OF COMBINATORIAL OPTIMIZATION OF EXCAVATION SEQUENCE AND SUPPORT PARAMETERS FOR LARGE UNDERGROUND CAVERNS UNDER CONDITION OF HIGH GEOSTRESS
Abstract:Aiming at distinctness of deformation and failure of rockmass under high geostress and that optimization of excavation schemes and support schemes for large caverns is a complicated problem having a large search space and large scale of numerical calculation,a new intelligent optimization integrated method is proposed for optimization of excavation sequence and support parameters for large underground caverns under condition of high geostress. The method,which takes the integration optimization indexes including elastic release energy,plastic zone volume,displacement around caverns,support cost as fitness,integrates the 3D numerical method based on a new constitutive model which performs excellently under condition of high geostress,and the intelligent technique including particle swarm optimization(PSO) and support vector machine(SVM). In detail,learning samples are established by numerical analysis for some typical construction schemes firstly. Then,using SVM trained by learning samples,the nonlinear mapped relationship of excavation sequence and support parameters with integrated optimization indexes are established. Finally,the globally optimum excavation sequence and support parameters are achieved by PSO search technique. Using the method mentioned above,the excavation sequence and support parameters of the large caverns of Laxiwa Hydropower Station located on the Yellow River in China are optimized. The result proves that the proposed method is feasible.