最优预测元模型在小型离心式鼓风机性能优化中的应用  

Application of metamodel of optimal prognosis in performance optimization of small centrifugal blowers

作　　者：李科钦 周忠宁[1] 李意民[1] 纪子宁 周志恒 包昊森 LI Keqin;ZHOU Zhongning;LI Yimin;JI Zining;ZHOU Zhiheng;BAO Haosen(School of Low-Carbon Energy and Power Engineering,China University of Mining and Technology,Xuzhou 221116,China)

机构地区：[1]中国矿业大学低碳能源与动力工程学院,江苏徐州221116

出　　处：《流体机械》2025年第1期105-112,共8页Fluid Machinery

基　　金：国家自然科学基金项目(51776217)。

摘　　要：针对传统的叶片气动设计方法存在依赖于设计者的经验、研发周期长、设计精度难以保证等问题,将最优预测元模型优化方法应用到离心鼓风机叶轮优化中。通过将某小型离心式鼓风机进行参数化造型,使用高级拉丁超立方(ALHS)方法对叶片子午面、进、出口安放角、叶片流面扭转角3个部分的36个参数采样并进行数值模拟计算,构建了200组鼓风机性能数据;在此基础上,采用最优预测元模型(MOP)筛选出了对鼓风机性能影响最大的9个参数,并以最大等熵效率为目标,进、出口压差为约束条件,运用进化算法(EA)对9个参数进行迭代计算,得到了性能最优的鼓风机参数,对优化前、后的鼓风机流场进行了对比分析。结果表明,优化后的鼓风机抑制了叶顶泄漏流,叶顶泄漏流量相较于原模型减少了10%;在叶轮中段(55%子午流线位置处)流体沿叶高优化前、后的最大速度由101.66 m/s减少到82.06 m/s,速度变化?V由53.52 m/s减少到40.74 m/s,流动损失降低;叶轮出口的最大压力也从115.37 kPa增加到116.19 kPa;最优预测元模型能够过滤掉优化过程中不重要的输入参数,从而减少优化成本与时间,研究结果可为离心式鼓风机的优化设计提供参考。Addressing the issues inherent in traditional aerodynamic design methods for blades—such as reliance on designer experience,long development cycles,and difficulty in ensuring design precision—this study applies an optimal surrogate model optimization method to the impeller optimization of a centrifugal blower.By parameterizing the design of a small centrifugal blower and using the Advanced Latin Hypercube Sampling(ALHS)method to sample 36 parameters from the meridional profile,inlet and outlet blade angles,and blade surface twist angle,200 sets of blower performance data were constructed through numerical simulations.Based on this data,the optimal surrogate model(MOP)was employed to identify the 9 parameters that most significantly impact blower performance.Subsequently,an evolutionary algorithm(EA)was utilized to iteratively compute the optimal values of these 9 parameters,with the objective of maximizing isentropic efficiency and constraining the inlet-outlet pressure difference.A comparative analysis of the flow fields before and after optimization was conducted.The results demonstrate that the optimized blower suppressed tip leakage flow,reducing the tip leakage flow rate by 10%compared to the original model.At the mid-span of the impeller(at 55%of the meridional streamline),the maximum fluid velocity along the blade height decreased from 101.66 m/s to 82.06 m/s after optimization,and the velocity variationΔV reduced from 53.52 m/s to 40.74 m/s,indicating reduced flow losses.Additionally,the maximum pressure at the impeller outlet increased from 115.37 kPa to 116.19 kPa.The optimal surrogate model effectively filters out non-critical input parameters during the optimization process,thereby reducing optimization costs and time.The findings of this research provide a reference for the optimal design of centrifugal blowers.

关 键 词：小型离心式鼓风机 参数优化 最优预测元模型 数值模拟 

分 类 号：TH43[机械工程—机械制造及自动化]