基于BP神经网络的拱坝温度监测数据预测模型

Forecast Model for Temperature Monitoring Data of Arch Dams Based on MATLAB Neural Network

实时监控拱坝的温度对工程进度和坝体安全具有重要意义。以白莲崖碾压混凝土拱坝温度监测数据为研究对象,建立基于MATLAB的拱坝温度监测反向传播(BP)神经网络预测模型,用原型观测数据对其进行校核和检验,并引入灰色理论中的GM(1,1)模型、混沌模型(最大Lyapunov指数法)与预测结果进行比较。结果证明,用人工神经网络建立坝体变形的神经网络模型对大坝变形能够进行较高精度的预测,具有良好的应用前景。

Temperature monitoring of arch dams is very important in their construction.Based on the BP neural network theory,a forecast model for temperature monitoring of arch dams is established by means of MATLAB.The temperature monitoring data of Bailianya Arch Dam are studied.The observed data are employed to validate and check the proposed model.By introducing GM（1,1） model in the gray theory and fuzzy model（the largest Lyapunov exponent method）,a comparison of the predicted results with the observed data shows that the proposed model can accurately predict dam deformation and is practical.