Analysis And Prediction Of Subsidence Time Series

There are many method to forecast subsidence, the traditional forecasting methods mainly dynamics and mathematical statistical methods, the expected results by the commonly used formula of mining subsidence is expected to have greater error. The common characteristics of these methods establish a sequence of subjective data model firstly, and then based on subjective model to calculate and predict the surface subsidence or deformation. These methods are very effective to linear question, but if the system being to the chaotic state, these methods will be inappropriate, or even have big errors. Researching surface subsidence with chaotic dynamics is less, so there will be significance meaning to carry out this thesis.The subsidence system is a complex nonlinear dynamic system under the influence of various factors, causing the predicting of subsidence accurately difficultly. The chaos theory and artificial neural networks theory are introduced into the model of subsidence forecasting and a new method of subsidence prediction is put forward in this thesis. The time series of subsidence displacement is used for research in this thesis because it is obtained easily in practice project. The main content is as follows:Based on the research of basic chaos theory, the reconstruction of phase space suited to subsidence time series is put forward. The embedding dimension and delay time of phase space are found. And the subsidence time series is reconstructed using this method. Then the chaos characters of subsidence are researched in the reconstructed phase space. In this research, the correlation dimension and the maximum lyapunov exponent of subsidence are computed, which prove subsidence system has the chaos characteristic.Based on the study of network training and network design of BP artificial neural networks, a new global predicting model by BP network is built, which fits for the time series of subsidence. The phase points in the reconstructed phase space of subsidence are fitted and the subsidence are predicted by this model, and the predicted result is better.In a word, the model based on chaos theory and artificial neural networks theory is able to reflect and predict the characteristic of subsidence well.