【摘要】 依据天然河冰垂直温度剖面数据,构造了以拟合函数的系数为参变量的分布参数系统优化辨识模型,并分别用指数函数和线性函数拟合了河冰导温系数。对目标函数的优化采用Newton-Raphson算法,对热传导方程和灵敏度方程采用半隐式差分格式迭代求解。利用太原汾河二库天然冰垂直温度剖面数据进行了参数辨识和数值模拟,模拟结果与实测数据相吻合,表明所构造的参数辨识模型和优化算法是有效的。更多还原

【Abstract】 Based on the vertical temperature profile data of natural river ice,this paper establishes optimal identification model about distributed parameter system where parameters are the coefficients of fitted functions,and fits coefficients of thermal diffusivity of river ice by exponential function and linear function respectively.Furthermore,this paper employs Newton-Raphson algorithm to optimize objective function and use half-implicit difference scheme to obtain the solutions of heat-conduction equation and sensitivity equation.Using the vertical temperature profile data of natural ice in second reservoir of Taiyuan fen-river,the parametric identification and numerical simulation are processed.The results show that simulation results are corresponding with practical measured data.Hence,the identification model and optimal algorithm are effective and feasible.更多还原