Forecasting Sedimentation of constructions Based on BP Network

If sedimentation of constructions exceeds the prescribed limits, it would give rise to huge losses for community and people, so it is significant to establish the effective and practical deformation forecasting model for the safe operation and economic development. With the unique non-linear, non-convexity, non-locality, non-steadiness, adaptability and powerful ability of calculation and information process, BP (Back Propagation) neural network can adapt to the complicated and changeable dynamic characteristics of buildings,that has broad application foreground in deformation prediction. In this paper, on the basis of deformation observation data, a basic algorithm about establishing BP neural network model in sedimentation prediction is presented.at the same time, analysis of examples are given so that the application of neural network for deformation forecasting is studied comprehensively and systemically, the results show that neural network is very effective for Sedimentation prediction and can serve society and people in the future.