结构损伤的遗传神经网络检测方法

A Neural Network Based on Genetic Algorithm for Structural Damage Detection

将遗传算法与神经网络相结合,用遗传算法完成神经网络的学习过程,建立了结构损伤的遗传神经网络检测方法,并对遗传算法进行了改进。研究结果表明,用改进的遗传算法进化神经网络可以有效地避免BP算法有可能陷入局部极小的缺点,而且运算速度大大加快,精度提高。

Using genetic algorithm for neural network, completing the learning of neural network by genetic algorithm. Building a neural network base on genetic algorithm for structural damage detection, and improving the genetic algorithm. The research results prove that the improved genetic algorithm can get over the insufficiency of BP algorithm such as liable to get into local minimum. It can raise computing speed greatly, and improve precision。