基于改进支持向量机回归的非线性飞机结构载荷模型建模

为进行飞机结构载荷安全监控并为飞机结构疲劳寿命评估积累相关数据,需建立与飞行参数相关的 飞机结构载荷模型。针对飞机结构载荷与飞行参数之间的非线性关系,采用改进停机准则的 SMO 算法及粒 子群模型参数优化算法对支持向量机回归方法进行改进,并通过飞行动力学理论分析结合皮尔逊相关系数的 方法对参与建模的飞行参数进行选取。以飞机跨声速俯仰机动为例,建立机翼某一测载剖面结构剪力模型,并 对该建模方法进行仿真验证。结果表明:采用改进支持向量机回归方法所建立模型精度优于原始支持向量机回归方法建立的模型,即采用改进支持向量机回归方法可提高建模精度及泛化能力。

Nonlinear Aircraft Structure Load Model Based on Improved Support Vector Machine Regression

In order to carry out aircraft structural load safety monitoring and accumulate relevant structural load data for aircraft fatigue life assessment, it is necessary to establish aircraft structural load model related to flight parameters. For the nonlinear relationship between aircraft structural loads and flight parameters,the SMO algorithm with improved stopping criterion and the particle swarm optimization algorithm were used to improve the support vector machine regression method,and the flight parameters involved in the modeling were selected by the method of flight dynamics analysis combined with the Pearson correlation coefficient. The aircraft"s transonic pitch maneuver was taken as an example , and a shear force model of a certain load-bearing section structure of the wing was established. The results show that the accuracy of improved support vector machine regression method is better than the original method. It is concluded that the improved support vector machine regression method can improve the accuracy and generalization ability of the established model.