遗传神经网络在大坝安全评价中的应用

Application of genetic neural network to dam safety evaluation

利用神经网络方法评价大坝安全具有一定的优势,但传统大坝安全评价方法不能为神经网络模型提供合适的学习样本。文中引入安全度值的概念,为神经网络提供可量化的学习样本,并针对BP神经网络收敛速度慢、稳定性差、易陷入局部极小等问题,利用遗传算法进行改进,提出基于遗传神经网络的大坝安全评价方法。工程实例表明,评价方法合理、可行。

Application of neural network to the dam safety evaluation has some advantages, but the traditional methods of dam safety evaluation can not provide the appropriate learning samples for the neural network model. The concept of safety degree value is introduced to provide quantifiable learning samples for neural network. And aiming at the problems of slow convergent rate, poor stability and local minimum of the BP neural network, an improved dam safety evaluation method based on genetic neural network is proposed. The example shows that the method is reasonable and feasible.