基于分布式遗传算法的多段翼型优化设计

为了提高采用遗传算法的气动外型优化设计的效率，文中探讨了将分布式计算引入到优化设计过程中，实现了基于分布式遗传算法的多段翼型优化设计，进行了多段翼型的缝隙、重叠量和偏转角度等量的优化设计。设计实践表明，该方法是可行的。

On Using Distributed GA (Genetic Algorithm) to Raise Efficiency in Optimizing Multi-Element Airfoil Configuration

Using GA for optimization of multi element airfoil configuration is generally considered to be quite advantageous but the amount of computation becomes almost prohibitive. We propose a distributed GA to reduce significantly the amount of computation. The distributed GA we developed uses microcomputers in LAN (Local area Network) to optimize multi element airfoil high lift configurations (such as gap, overlap and flap deflection). Our distributed GA adopts the client server technology based on TCP/IP protocol. High lift coefficient is an important design objective; Fig.3 shows the comparison of variations of lift coefficient with the number of evolution generations when our distributed GA and ordinary GA are used respectively for a certain multi element airfoil configuration. The comparison shows that our distributed GA is much more effective. Moreover, the size of client group used by our distributed GA is smaller than that used by ordinary GA.