【摘要】 将叶型参数化、试验设计方法、神经网络算法与Pareto类遗传算法相结合,发展了一种平面叶栅多目标优化设计方法.在该算法中,为提高优化效率,提出了并行神经网络算法和改进了NSGA-Ⅱ算法.以极小化三个工况点的总压损失系数为目标函数,将该优化方法应用于某平面叶栅多工况优化设计.与初始叶栅相比,优化后叶栅的总压损失系数在三个工况均有一定的减小,说明该优化方法是有效的。更多还原

【Abstract】 An optimization approach to cascade,incorporating blade parameterization,design of experiments,artificial neural networks(ANN)and Pareto-GA,is presented.To improve optimization efficiency,the idea of parallel artificial neural networks(PANN)is proposed and an improved NSGA-Ⅱalgorithm is developed in this approach.The minimum total pressure loss coefficient is selected as objective function.After optimization,the loss coefficients of the three conditions are decreased,and the results show that the proposed approach is effective for cascade multi-point design.更多还原