Abstract:

In order to reduce the weight of impeller, a multi-objective genetic optimization method and process of the impeller of rail axial fan was established by analyzing the impeller of the numerical calculation method and constructing Kriging model. The numerical calculation method was established by the reverse analysis model and mesh density accuracy. The sampling space of impeller was obtained by Latin hypercube experimental design model. Kriging model of impeller with sufficient accuracy could replace the numerical analysis. The sensitivity of the main technical parameters of the force on the impeller structure was analyzed by the method and the most influential parameters on the impeller structural stress were identified. The weight of impeller is reduced by 31.7% by optimization and save the cost of materials. The maximum stress value is also changed from the initial 21.5MPa to 16.5MPa, which effectively improve the mechanical performance of the impeller. After optimization, dynamic and static experiments were used to the impeller by the slip ring electric and dynamic and static strain testing system, and the impeller rotating machinery equipment and the technology of testing plan was designed. Experimental results and calculation results are very close and the dynamic stress is relatively small, so the impeller has a good mechanics and vibration performance, and has high value in engineering applications.

Key words: impeller, multi-objective genetic optimization, numerical analysis, Kriging model, axial fan

摘要：

为最大限度地减小叶轮质量，通过反求模型和网格密度精度的分析，对叶轮结构进行数值计算，并构造了拉丁超立方试验设计模型和Kriging近似模型，得到叶轮结构的采样空间和高精度的近似模型用来代替数值分析，提出了轨道轴流风机叶轮结构多目标遗传优化方法。采用该方法分析了主要技术参数对叶轮结构受力的灵敏性，并找出了对叶轮结构应力影响最大的参数。叶轮结构通过优化后，质量减小了31.7%，大大节省了材料的成本，最大应力值也由初始方案的21.5MPa变为16.5MPa，有效地提高了叶轮的力学性能。优化后通过滑环引电器和动静态应变测试系统对风机叶轮进行动静态试验，并设计了叶轮旋转机械的工艺工装和试验方案，试验结果与计算结果十分吻合，且动应力相对很小，因此，叶轮具有良好的力学和振动性能，在工程应用中具有较高的价值。

关键词: 叶轮, 多目标遗传优化, 数值分析, Kriging模型, 轴流风机