论文部分内容阅读
施工期混凝土坝的温度场是复杂的不稳定温度场,显式温度统计模型有时难以合理反映不同温控因素和混凝土坝温度之间的非线性关系.首先从施工期温度影响各因素的理论解析解入手,挖掘并选取施工期混凝土坝温度场的影响因子,以挖掘出的温度影响因子作为输入矢量,实测温度为输出矢量,建立了施工期混凝土坝温度的神经网络隐式时空分布模型.结合西南某在建混凝土坝工程的分布式光纤测温资料,分别建立了显式温度时空分布模型以及神经网络隐式时空分布模型.分析表明,相对于显式温度时空分布模型,建立的神经网络隐式温度时空分布模型重构的温度场精度更高,可准确反映不同时刻施工期混凝土坝的温度状态.
The temperature field of concrete dam during construction is complicated and unstable temperature field, and the explicit temperature statistic model is sometimes difficult to reasonably reflect the nonlinear relationship between different temperature control factors and concrete dam temperature.Firstly, from the theoretical analysis of various factors that influence the temperature during construction period To solve this problem, the influencing factors of temperature field of concrete dams during excavation were selected, and the temperature influence factors of excavation were taken as input vector and the measured temperature was output vector. The neural network implicit space-time distribution model of concrete dams during construction was established. The distributed optical fiber temperature measurement data of a concrete dam project under construction in Southwest China are respectively established the temporal and spatial distribution model of explicit temperature and the implicit spatial and temporal distribution model of neural network.The analysis shows that compared with the explicit spatial and temporal distribution model of temperature, The temperature field reconstructed by spatio-temporal distribution model has higher accuracy and can accurately reflect the temperature state of concrete dams at different times.