基于响应面法的带喷流激波针参数优化研究

Optimization research on combination of spike and forward-facing jet using response surface methodology

对基于响应面法的带喷流激波针参数优化方法进行了研究,优化目标是最佳减阻效果和最小喷流流量,优化参数是激波针长度、喷流出口总压和喷流出口直径,利用响应面模型对设计参数与响应目标的关系进行建模,样本点设计采用了Ⅳ-最优方法,样本点的气动响应通过数值计算得到,最后用期望函数法进行多目标寻优。研究表明:激波针长度、喷流总压和喷流出口直径与阻力呈现2阶或3阶非线性关系,且激波针长度和喷流出口直径耦合效应较强;响应面模型给出了阻力与各设计参数关系的数学表达式,仅用较少的样本点就获得了设计空间内任意参数组合的阻力预测值和置信区间,效率较高;通过响应面法获得了最优参数组合,其阻力预测值与校验结果相比精度较高;响应面法应用于带喷流激波针这类多参数、多目标优化设计问题中,有计算量小、结果可信、实用性强的特点。

Both spike and forward-facing jet can reduce the shock wave drag acted on a blunt head hypersonic vehicle significantly. However, they are not be adopted widely because of some shortages. The combination of spike and forward-facing jet takes advantages of both of them, while more complex aerodynamical phenomena and parameter optimization issue are involved. In this paper, the parametric optimization of combination of spike and forward-facing jet is carried out by response surface methodology. The objective of the optimization is to minimize the drag and the jet flux. The optimized parameters are the length of the spike, total pressure of the jet and the diameter of the jet out let. The response surface model is involved to feedback the relationship between the response and the designing parameters. IV-optimal is used to design the sample points, of which the response values are achieved through numerical simulation. The desirability value is used to estimate the benefits. The principal conclusions are as following： The function of drag and design parameters is supposed to be 2na or 3ra order non-linear polynomial, and the coupling effects among design parameters are evident, especially between the diameter of jet outlet and the total pressure of jet. Response surface model speculated from sample points simulations reveals the relationship between the drag and the design parameters, from which the drag prediction interval of arbitrary parameters combination in design space can be obtained. The optimal parameters of minimum drag and jet mass flow are acquired, where the drag is predicted in a good precision. The study shows that response surface methodology in parameters optimization has high reliability and practicality with less computation requirement, and has good prospect of application in aircraft design.