运用改进的遗传算法进行框架结构损伤检测

Application study of an improved genetic algorithm for frame structure damage detection

运用一种改进的遗传算法来进行结构的损伤检测研究。该方法在传统遗传算法的变异算子里引入了一种被称为零变异率因子的参数,使得种群中时刻保持一定数量的零值元素,即相当于用结构的损伤只是发生在局部这个信息约束了传统的遗传算法,从而使得检测的结果更加准确。通过对某一框架模型试验数据的损伤检测研究,证明了该方法应用于实际工程结构的可行性,同时阐明了实际应用中的若干重要问题。

Although general genetic algorithm （GGA） has a powerful when applied to large and complex structures because of its blindness global-searching capacity, there exist some difficulties existed in searching. In order to overcome its shortcomings, an improved genetic algorithm （IGA）, for which a zero mutation ratio factor （ZMRF） is added to the mutation operator of general genetic algorithm （GGA）, is adopted in this paper for the damage detection. In order to obtain more accurate and reliable results, a certain number of zero elements in population of each generation are kept in this method, which is in fact a kind of GGA under a constraint condition presented as an auxiliary information such that damage always occurs locally. A laboratory frame structure with bolted joints is adopted as an example to demonstrate the feasibility of this method, and some related important issues existed in real applications are also illuminated at the same time.