Study on prediction model of deep pit deformation

Elman neural network (ENN) is one of the well-known dynamic recurrent neural networks. A new self-adaptive particle swarm optimization (SPSO) was proposed to improve Elman in order to solve problems of dynamic prediction in this paper. SPSO combines ENN and form SPSONN hybrid algorithm. Based on the algorithm, a nonlinear time-varying model was established to prediction deformation of deep foundation pit. The results of an engineering case indicate that the intelligent prediction model is efficiencies in the complex underground structures.