基于CFD与BFGS的四加热湿度传感器设计

Design of four heating humidity sensor based on CFD and BFGS

针对湿度传感器在高空气象探测时因表面覆盖水凝物引起沾湿误差的问题,提出了一种具有加热冷却功能的四加热湿度传感器。为实现对加热策略的优化设计,将传感器的工作周期分为4个阶段:加热阶段、稳定阶段、冷却阶段和测量阶段。首先,采用计算流体动力学(CFD)方法计算出4个阶段对应的时间;接着,利用拟牛顿法中使用较多的BFGS算法进行数据拟合,得到加热时间和冷却时间的修正公式;最后,以低气压风洞作为实验平台,开展外场喷雾实验进行对比验证。实验结果表明:四加热湿度传感器加热时间的修正值与实验值差值绝对值的平均值为0.112 s,均方根误差为0.133 s;冷却时间的修正值与实验值差值绝对值的平均值为0.801 s,均方根误差为0.959 s。开启加热时,传感器可以减小10%以上的测量误差,由此验证了该方法的可行性。

Aiming at the problem of wetness error caused by the surface covered with hydrometeor when the humidity sensor detects high-altitude meteorology,a four-heating humidity sensor with heating and cooling functions is proposed.In order to achieve the optimal design of the heating strategy,the working cycle of the sensor is divided into four phases,which are heating phase,stable phase,cooling phase and measurement phase.Firstly,the computational fluid dynamics(CFD)method is used to calculate the time corresponding to the four stages.Next,Broyden Fletcher Goldfarb Shanno(BFGS)algorithm which is widely used in quasi-Newton method,is adopted for data fitting,and the correction formulas for heating time and cooling time are obtained.Finally,the low-pressure wind tunnel is used as an experimental platform to carry out out-field spray experiments for comparison and verification.Experiments show that the average value of the absolute difference between the correction value of the heating time and the experimental value of the four-heating humidity sensor is 0.112 s,and the root mean square error is 0.133 s;the average of the absolute value of the difference between the correction value of the cooling time and the experimental value is 0.801 s,the root mean square error is 0.959s.When the heating is turned on,the sensor can reduce the measurement error by more than 10%,which verifies the feasibility of this method.