Risk analysis on long inclined-shaft construction in coalmine by TBM techniques based on multiple variables chaotic time series

Multi-variable chaotic time series are used to predict the long inclined-shaft construction in coalmine construction by TBM techniques, and principal component analysis (PCA) is used to determine the main factors that impact risk (shield) of the long inclined-shaft construction in coalmine by TBM techniques. Phase space of risk time series for construction by TBM are reconstructed; time delay and embedding dimension are determined. Maximum Lyapunov indexes of risk are obtained by using small data quantity method; it is found that the time series have characteristics of chaos. Prediction model is established using the combination of first-order local method and double hidden layer neural network. Simulation experiments show that the combined model has a strong ability of prediction and achieves better effect. As a result, it provides a new way for long inclined-shaft construction in coalmine by TBM techniques.