研究了特征值反问题求解的几种神经网络模型：直接逆模型，间接逆模型，优化方法模型，指出了各种方法的应用范围。
In this paper, several artificial neural network based models of solving inverse eigenvalue problem and their characters are studied. These models are direct, indirect and optimization inverse models.
从偏微分方程反问题的角度解释了结构拓扑优化中棋盘格式、网格依赖性等数值不稳定现象的本质。
Form the point view of inverse PDE, the essence of numerical instabilities in the topology optimization is given, including checkerboard pattern and mesh-dependence.
反问题的求解常常需要转化为非线性优化问题，其目标函数定义为观测数据与模型数据之间的残差平方和。
The solution of inverse problem usually requires nonlinear optimization of an objective function describing the difference between measured and simulated data.
因此许多学者提出了各种求解反问题的方法，比如脉冲谱方法，最佳摄动量法，蒙特卡罗方法，各种优化方法和正则化方法等。
So various methods are proposed by scholars to solve these problems, such as pulse spectrum method, the best perturbation method, Monte Carlo method, optimized and regularization method.
因此许多学者提出了各种求解反问题的方法，比如脉冲谱方法，最佳摄动量法，蒙特卡罗方法，各种优化方法和正则化方法等。
So various methods are proposed by scholars to solve these problems, such as pulse spectrum method, the best perturbation method, Monte Carlo method, optimized and regularization method.
应用推荐