论文部分内容阅读
针对大坝工作条件复杂,影响因素繁多,致使现有监控模型预报精度偏差过大问题,基于递阶对角神经网络能够逼近任意非线性函数的特点,使用串并联模型辨识器,采用动态BP学习算法,以水压、温度和时效因子为输入量,坝体位移为输出量,结合工程实例提出了大坝变形监测的递阶对角神经网络模型,并将该模型用于坝体变形数据的拟合分析及其预测预报.研究表明,该网络不仅收敛速度快,提高了算法的效率,而且对实测数据具有较好的拟合效果,提高了预报精度,在大坝安全预测分析中具有有效性和优越性.
In view of the complex working conditions of dams and the various influencing factors, the existing problems of large deviation of forecasting accuracy of monitoring model are caused. Based on the fact that the hierarchical diagonal neural network can approximate the characteristics of any nonlinear function, the parallel-series model identifier is used and dynamic BP learning The algorithm takes the hydraulic pressure, the temperature and the aging factor as the input volume and the dam displacement as the output volume. Combined with engineering examples, a hierarchical diagonal neural network model of dam deformation monitoring is proposed. The model is applied to dam deformation data Fitting analysis and forecasting of the dam.The results show that this network not only has fast convergence and improves the efficiency of the algorithm but also has good fitting effect on the measured data and improves the forecasting accuracy and is effective in dam safety prediction and analysis Sex and superiority.