人工神经网络在引水隧洞变形监测中的应用

Application of Artificial Neural Network to Deformation Monitoring of a Power Tunnel

针对引水隧洞围岩变形受地质、施工等因素综合影响的特点,引入人工神经网络理论,建立了隧洞收敛测值终值预测的BP网络模型。该模型将隧洞围岩的5个物性参数容重、弹性模量、泊松比、粘结力、内摩擦角作为BP网络的输入向量,将最终收敛位移作为网络输出向量。应用Matlab进行BP网络设计和程序编制。实例分析表明,利用所建立的BP网络模型进行围岩收敛变形预测,结果合理可行。

The BP network model was established for forecasting of the measured tunnel convergence final value by using the artificial neural network theory according to the characteristic of that surrounding rock deformation of a power tunnel is influenced by geological and construction conditions. Bulk density, elastic modulus, Poisson ratio, cohesion and internal friction angle five physical property parameters of tunnel surrounding rock were taken as input vectors of the BP network while the final convergence displacement was taken as an output vector of the network. Matlab was employed to carry out design of the BP network and programming. The analysis of examples indicates that the application of the established BP network model to forecasting of surrounding rock convergence deformation provides desirable and practicable results.