基于遗传算法的二维气动优化-流体力学专业论文.docxVIP

南京航空航天大学硕士学位论文 南京航空航天大学硕士学位论文 基于遗传算法的二维气动优化 基于遗传算法的二维气动优化 i i ii ii 摘 要 本文针对翼型气动优化问题，就翼型参数化、流场求解器和遗传算法等相关问题进行了研 究。基于 CST 参数化技术构建翼型参数化模型，使参数寻优空间得到了有效地控制；遗传算法 适应度评估涉及的流场求解器则采用基于 Jameson 有限体积法的 Euler 方程解算程序。通过引 入隐式残值光顺、多重网格法和弹簧网格处理等措施，改进了求解器的计算效率；优化方法基 于基本遗传算法，通过引入精英保留策略和 OpenMP 技术的并行化处理，提高了算法的收敛性 能和计算效率。文中首先给出了亚音速条件下 NACA0012 翼型表面压力分布重构算例，重构的 压强分布能与目标分布吻合，展示出本文发展的方法可用于翼型气动优化。 接着，本文进行了翼型减阻凸包外形的优化设计，即在翼型上表面放置优化的凸包外形降 低激波强度，达到减阻效果。文中以 RAE2822 翼型为例，给出了这一减阻凸包外形的优化设计， 展示出减阻的显著效果(本算例波阻减少达 40%)。最后本文进行了基于翼型面积和后缘角约束 的翼型气动优化设计，文中给出了以 RAE2822 翼型为基准翼型的带约束升力系数极大化设计算 例，达到了明显增升效果。上述算例表明，本文建立的方法具有良好的鲁棒性，有一定的工程 应用价值。 关键词：遗传算法，OpenMP，有限体积法，多重网格技术，CST 参数化方法，气动优化设计 Abstract An efficient two dimensional aerodynamic optimization method is developed, using an optimization algorithm based on Genetic Algorithms (GAs). Several aerofoil optimization calculations are implemented using the developed method. A binary coded GAs with elitist strategy is used as a fundamental optimization method. Fitness calculations of the population are parallelized using OpenMP. A finite volume solver, based on Jameson method, is integrated with implicit residual smooth and agglomerate multigrid method to enhance the speed of convergence to the largest extent. To improve the accuracy of flow calculation, a vortex correction skill is utilized. Domain mesh is generated with a spring analogy method to maximize the optimization efficiency. Aerofoil geometry is parameterized using a newly developed CST (Class-Shape function Transformation) method. A NACA0012 surface pressure reconstruction problem is solved to validate the feasibility of the developed method used in aerofoil optimization. Then a shock control bump optimization on RAE2822 aerofoil under transonic flow conditions is implemented with the developed method. After that an aerofoil lift optimization with constraints under transonic flow conditions is implemented. Results show that the developed method is robust and efficient when used in aerofoil aerodynamic o