粒子群优化算法用于缺陷的红外识别研究

Research on PSO Algorithm Applied in Defect Identification Using Thermal Imager

讨论了利用一种较新的仿生优化算法———粒子群优化算法(PS0)进行缺陷的红外识别的一条途径。PSO算法可以不用计算梯度,因此可以和通用的有限元软件结合起来,对比较复杂的缺陷识别问题都可以采用同一手法进行求解,并使得优化算法和有限元编程实现了有效的隔离。最后给出了PSO算法在泛圆台类缺陷红外识别中一个简单的应用例子。

A novel biological optimization algorithm, particle swarm optimization （PSO） algorithm, is applied in defect identification in the paper. Diversified universal FEM software such as ANSYS or NASTRAN can be combined with the algorithm, since the hard-wan calculation of gradient is not required, and the programming of FEM and optimization algorithm can be isolated effectively, which makes many sophisticated cases solved easily. A simple defect identification case is also discussed in the paper.