791翼型水动力性能的多目标遗传优化

首先采用贝塞尔曲线参数化翼型背面型线,控制翼型的厚度变化规律;其次,通过均匀试验设计方法和人工神经网络建立背面型线控制点与翼型性能参数之间的近似模型;最后通过多目标遗传算法优化791翼型的背面型线,获得具有较好绕流特性下的厚度变化规律.以791翼型的升阻比F最大化和阻力F_D最小化为优化目标,以贝塞尔曲线的控制点为优化变量进行计算.优化结果表明:单纯地以翼型升阻比和阻力为优化目标时,优化后翼型阻力最大减小了4.17%,升阻比最大增加了15.33%,“失速”冲角增大;优化后翼型的最大厚度的位置更加靠近来流方向,出现在弦长约0.390处,最大厚度增加,约为弦长的0.090;当改变优化目标时,该多目标优化策略可也以较好地优化翼型的结构,提高翼型的绕流性能,可为其他流体机械的优化方法研究提供一定的参考.

Study on multi-objective genetic optimization of hydrodynamic performance of airfoil 791

Firstly, Bezier curve is used to parameterize the suction side of an airfoil to control its thickness profile. Secondly, a surrogate model is established to represent an approximate relationship between the control nodes distributed on the suction side and the performance parameters of airfoil by using uniform experiment design and artificial neural network method. Finally, the optimal thickness profile is obtained based on the multi-objective genetic algorithm(MOGA). Lift-drag ratio F and drag F_D of airfoil 791 are served as the optimization objective functions and the control nodes of the Bezier curve are chosen as optimization variables. The result indicates that MOGA is feasible. As a result, the drag is decreased by 4.17%, the lift-drag ratio is increased by 15.33%, and the angle of attack at which a stall appears becomes larger. The optimal location of maximum thickness is more close to the leading edge as near as 0.390 the chord length and the maximum thickness is increased as thick as 0.090 the chord length. When other the objective functions are changed, an optimum airfoil still can be obtained by means of the optimization strategy. Such that the optimization strategy proposed can provide a reference for a study on optimization methods applied in other fluid machines.