基于PSO与LS-SVM的飞机总体设计的综合论证

Study on Intelligent Evaluation for Airplane Design Based on PSO and LS-SVM

研究小样本数据对飞机武器系统的设计和改型方案是航空系统工程的重要内容。针对提高设计的进度和质量问题,利用粒子群优化算法的群体智能优化理论与最小二乘回归支持向量机的回归思想,提出了一种基于粒子群算法与最小二乘回归支持向量机的飞机设计综合智能论证模型。提出应用粒子群算法对支持向量机核函数参数进行寻优,再利用优化的核函数参数支持向量机回归模型,建立映射模型来对飞机的作战效能进行预测。仿真实例验证了方法的适用性和结果的可靠性。

The evaluation on design or approach of aerial weapon system is one of important research field for aerial system engineering. Based on the idea of swarm optimization of the particle swarm optimization （PSO） and regression of the least squares support vector machine （ LS - SVM ）, a kind of evaluation model of airplane design based on PSO and LS - SVM has been presented. The parameters in LS - SVM are optimized by PSO, so the best regression model of LS - SVM can be determined. Afterwards the non - linearity mapping model between the mostly factors of airplane intelligent evaluation and LS -SVM is established and the cost forecasting is performed. The simulation ex- ample shows that the model is applicable and reliable.