With the rapid development of the transformation of the city construction, the demandfor the development of underground space and the intensity is increasing, large numbers ofthe deep foundation pit engineering have emerged. Both the depth of exploitation and theconstruction difficulty is increasing continuously, due to the complexity and the uncertaintyof engineering geological traits, improper excavation may lead to instability of the pit andslope, uneven subsidence or sinking of the earth’s surface, and other disasters. To realize thework of reliable prediction and control of soil settlement of the deep foundation pit, is animportant link of risk control in the underground engineering.According to the characteristics of settlement and deformation in the deepexcavation construction, this paper researched on the settlement mechanism and theimpact factors of deep foundation pit, based on the research of wavelet analysis andthe intelligent algorithm of RBF neural network, combined with the excellent thecharacteristics of them, this paper put forward the coupling algorithm of the"Wavelet-RBF", in which the wavelet analysis played as a preprocessing tool of theradial basis function neural network, structured the analysis and prediction modelof settlement of deep foundation pit based on intelligent algorithm (WRPM). Thepredicted results showed that, this WRPM settlement prediction model has thecharacteristics of high precision, strong generalization ability, which can providethe risk control of foundation pit engineering with high precision. |