离心泵多目标多变量优化设计模型实用化研究

A Practical Study of Multi-objective and Multi-variable Optimization Model for Centrifugal Pump Design

从离心泵整体损失功率最小、抗气蚀性能和无驼峰扬程曲线作为设计目标,在考虑叶片入口预旋的条件下建立了一个实用化离心泵的多目标多变量优化数学模型。应用MATLAB遗传算法工具箱对该多目标多变量优化模型进行数值求解,得到了叶轮进出口直径、叶片进出口安放角、叶片进口位置、叶轮出口宽度、叶片数及包角等关键设计参数的优化结果。最后采用CFD数值模拟的方法来进行验证,表明该优化方法是实用可行的。

By taking into account the minimum overall power loss of centrifugal pump, cavitation resistance and stable head curve requirements, a practical mathematical model of multi-obiective and multi-variable optimization is developed in this paper for centrifugal pumps under the condition of inlet blade pre-recirculation. By using genetic algorithm toolbox in MATLAB, the multivariate multiobjective optimization model is solved numerically. The results of the optimization include most primary design parameters such as the inlet and outlet diameters o{ the impeller, the inlet and outlet angles of blades, the inlet position of the blade, impeller outlet width, numbers of blades and wrap angle of blades. The results of simulation calculation based on CFD indicate that the optimization method is practical and feasible.