基于响应面法和正交试验的涡轮流量计优化设计

为降低流体黏度对涡轮流量计测量精度的影响，将涡轮流量计仪表系数线性度误差最小值作为目标函数，在运用计算流体力学（CFD）仿真的基础上，先通过Plackett-Burman设计筛选结构参数，并根据几何结构对目标函数的影响将其划分为两个等级，即显著影响因素和次显著影响因素；再通过Box-Behnken设计及响应面法对显著影响因素进行优化设计，分析结构参数间的交互作用，得到参数的最优设计点；最后在响应面分析基础上通过正交试验对次显著影响因素进行优化设计，得到最优参数组合。对最优参数组合的涡轮流量计进行试验研究，试验结果与CFD计算值吻合，仪表系数线性度误差由1.71%下降至1.59%，表明优化后的涡轮流量计测量精度得到了显著提高，基于响应面法和正交试验的优化方法可以用于涡轮流量计的结构设计。

Optimization of the design of a turbine flowmeter based on the response surface method and orthogonal tests

In order to reduce the errors in a turbine flowmeter influenced by fluid viscosity, the minimum turbine meter instrumental coefficient linearity error was employed as the target function. Firstly, Plackett-Burman design was used to screen out structure parameters by CFD. The influence of the geometrical structure on the target function can be divided into two levels:significant influencing factors and time-significant influencing factors. Secondly, Box-Behnken design and a response surface method were employed to optimize the design of significant influencing factors. The interactions between structural parameters were analyzed in order to obtain the optimal design parameters. Finally, by combining the response surface analysis with orthogonal tests, the sub-significant influencing factors were optimized and the optimal parameter combination was obtained. In an experimental investigation of the optimal combination of parameters for a turbine flowmeter, the instrumental coefficient linearity error in the test results and CFD calculated values, decreased from 1.71% to 1.59%. The optimized turbine flowmeter thus gave significantly higher measuring precision. Our optimization method based on the response surface method combined with orthogonal tests can be used to improve the structural design of turbine flowmeters.