【摘要】 针对轴流风机叶片的适用翼型选择过少、传统优化设计周期过长的问题,采用类别形状函数变换(CST)参数化方法,拟合出沿叶片径向不同截面处的翼型,并建立轴流风机单流道模型进行数值计算,将全压和全压效率作为目标函数,利用第二代非支配排序遗传算法(NSGA-II)寻优。对遗传算法进化不稳定、报错等问题,采用正态分布交叉算子替代模拟二进制交叉算子,并通过一种单级轴流风机为例来验证该多目标优化算法的有效性。算例结果表明,CST参数化方法拟合效果较好,在一定的流量条件下,通过改进的遗传算法多目标寻优,轴流风机的全压上升35%～37.7%,全压效率提高4.5%～7.6%。更多还原

【Abstract】 To solve the problem of the less airfoil choice apply to axial flow fan blade and long cycle of traditional optimization design, the class-shape function transformation(CST) parameterization method is adopt to fit the airfoils at different section along radial direction. The model of single passage axial flow fan is established, and the second nondominated sorting genetic algorithm(NSGA II) is used to enhance performance of blade taking total pressure and efficiency as objective function. For the problem of genetic algorithm evolution instability, the simulated binary crossover is replaced by normal distribution crossover, and the effectiveness of improved multi-objective optimization algorithm is verified by a singlestage axial flow fan as an example. The results show that the CST parametric method has a good fitting performance. Under the certain flow rate condition, the total pressure and efficiency of axial flow fan increase by 35% ～37.7% and 4.5% ～7.6%respectively through improved multi-objective optimization algorithm.更多还原