论文部分内容阅读
本文探索了一种能多变量综合优化的方法,即对喷管进行参数化设计后,用均匀试验设计(UED)将试验样本均匀散布在设计区间内,求出各性能参数后,利用径向基神经网络(RBF)对试验样本进行拟合,再用粒子群算法(PSO)对训练好的神经网络进行寻优,找出了更好的双喉道气动矢量喷管设计参数组合。数值模拟结果显示,优化后的双喉道气动矢量喷管的矢量角有了明显提高。试验表明这种优化方法具有很好的优化能力,可以用来对喷管几何外形进行参数优化。
This paper explores a multi-variable comprehensive optimization method, that is, the nozzle after the parametric design, the uniform test design (UED) will be uniformly distributed in the design of the test sample within the design interval, the performance parameters obtained, the use of radial The RBF neural network (RBF) is used to fit the test sample, then the particle swarm optimization algorithm (PSO) is used to optimize the trained neural network, and a better design parameter combination of double throat pneumatic vector nozzle is found out. The numerical simulation results show that the vector angle of the optimized double throat pneumatic vector nozzle has been significantly improved. Experiments show that this optimization method has good optimization ability and can be used to optimize the nozzle geometry.