论文部分内容阅读
清江隔河岩水电站自1993年蓄水发电至今已运行近20年,各项性态趋于稳定。根据清江隔河岩大坝近年来的大坝安全监测资料,主要针对拱冠梁15号坝段的位移,在基于物理推断分析的基础上,采用统计模型和反向传播(BP)神经网络模型进行了分析与研究。通过对2种模型分析的结果进行比较,应用BP神经网络模型进行分析具有更高的拟合精度和预测精度。分析结果表明:隔河岩大坝拱冠梁径向位移与上游水位呈正相关,与气温呈显著的负相关;拱冠梁切向位移与上游水位和气温无明显的关系,且径向位移非常小,拱冠基本呈对称状态,符合拱坝变形规律。
Qinghe Geheyan Hydropower Station has operated for nearly 20 years since its storage of water and electricity in 1993 and has stabilized in various aspects. According to dam safety monitoring data of Geheyan dam in Qingjiang River in recent years, the displacement of arch dam No.15 in Gongguan Beam is mainly analyzed. Based on the physical inference and analysis, statistical model and back propagation (BP) neural network model Carried out analysis and research. By comparing the results of two kinds of model analysis, the application of BP neural network model has higher fitting accuracy and prediction accuracy. The analysis results show that the radial displacement of the arch beam of Geheyan dam is positively correlated with the water level upstream and negatively correlated with the temperature. The tangential displacement of arch crown has no obvious relationship with the upstream water level and temperature, and the radial displacement is not Small, arch basically symmetrical state, in line with arch dam deformation laws.