论文部分内容阅读
本文提出了一种基于Kriging代理模型的风洞应变天平静态校准方法。采用拉丁超立方设计的试验设计(DOE)方法选取校准加载点,以六分量组合加载技术为支撑,采用一阶回归基函数和EXP关联函数对校准数据构建Kriging模型,以电压信号增量和其他条件变量为自变量,通过Kriging模型对天平所受载荷进行预测。校准结果表明,与传统的OFAT(One Factor at A Time)校准方法相比,该方法具有相同的准度,加载点数量极大减少。将该方法应用于某矢量喷管推力测量天平的校准工作中,以通气流量和天平电压信号增量作为自变量,较好预测了喷管产生的推力,解决了传统校准方法难以克服的通气影响问题。
This paper presents a static calibrating method of wind tunnel strain balance based on Kriging agent model. The calibration loading point was selected by the method of Latin Hypercube design (DOE), and the six-component combined loading technique was used as the support. The first order regression basis function and EXP correlation function were used to construct the Kriging model for the calibration data. The voltage signal increment and other The conditional variable is an independent variable, and the load on the balance is predicted by the Kriging model. Calibration results show that this method has the same level of accuracy compared to the traditional OneATAT (At Factor One) calibration method, and the number of loading points is greatly reduced. The method is applied to the calibrating work of a vector nozzle thrust measurement balance. Taking the increase of aeration flow and balance voltage signal as independent variables, the method predicts the thrust produced by the nozzle and solves the influence of ventilation which is difficult to be overcome by traditional calibration methods problem.