基于遗传算法和神经网络的结构损伤识别研究

应用遗传算法来优化神经网络的连接权值,在此基础上提出了基于遗传算法和神经网络的结构损伤诊断方法,并以一个矩形截面简支钢梁的损伤识别为例,进行了实例应用研究。该方法弥补了传统的BP神经网络收敛速度慢,易陷入局部极小点等缺陷。研究结果表明遗传算法和神经网络结合后兼有神经网络广泛的映射能力和遗传算法快速的全局收敛性能。

Research on Structural Damage Identification Based on Genetic Algorithm and Neural Networks

The article puts forward a new method using genetic algorithm to optimize neural networks' weights, and a structural damage identification method based on genetic algorithm and neural networks is given on the foundation. Furthermore, a simple supported rectangle steel beam model is presented as a calculated example for applying the above-mentioned method. The method remedies the defects of the traditional BP neural networks such as restraining on calculating speed slowly and falling into the partly extreme minimum value easily and so on. The research results show that the extensive mapping ability of neural networks and the rapid and global convergence of genetic algorithm can be acquired simultaneously by combining them together.