论文部分内容阅读
鉴于大坝变形监测资料分析是大坝结构性态安全评价与预报的重要手段,针对单测点模型存在的缺点,建立了既考虑坝体不同方向的位移又考虑空间多个测点分布的多测点多方向位移模型,并利用BP神经网络较强的非线性映射能力,直接选取了对大坝变形有较大影响的自变量因子,解决了在建立大坝多测点多方向传统模型时自变量因子数众多、计算工作量大等问题。实例应用结果表明,多测点多方向BP网络模型可反映大坝变形的分布及变化规律,可见采用BP神经网络建立大坝多测点多方向变形监测模型具有可行性和有效性。
In view of dam deformation monitoring data analysis is an important means of structural state safety evaluation and prediction of dam, aiming at the shortcomings of single measurement point model, this paper established a method that not only considers the displacement in different directions of dam but also considers the distribution of multiple measuring points in space Multi-directional displacement model of the measuring point, and by using the strong nonlinear mapping ability of BP neural network, the independent variable factors that greatly affect the deformation of the dam are directly selected, thus solving the problems that when the traditional multi-directional measuring model of the dam is established, There are many variables such as the number of independent variables and the workload of calculation. The results of practical application show that the multi-measuring point multi-directional BP network model can reflect the distribution and variation of dam deformation. It is feasible and effective to use BP neural network to build multi-measuring dam multi-directional deformation monitoring model.