基于时间序列分析的雾滴叶面动态接触角预测与建模

Dynamic Contact Angle Forecasting and Modeling for Drop-Leaf Interface Based on Time Series Analysis

摘要:

提出了一种基于图像数据采集的时间序列分析与空间趋势面相结合的雾滴叶面动态接触角预测和建模方法。利用CCD数字摄像技术，测定了质量分数为16.08%～82.14%的8种草甘膦助剂(EF8108-B)水溶液液滴在黄瓜叶面上的动态接触角θ。实验结果表明，接触角在0～2 s内急剧下降，2～10 s内下降平缓并渐趋于稳定；各质量分数EF8108-B液滴在黄瓜叶面的接触角没有随着质量分数的增大呈现接触角降幅增加的趋势。采用时间序列分析建立了黄瓜叶片界面2 s内液滴的动态接触角二阶自回归模型AR(2)、动态液滴尺寸滑动平均模型MA(3)以及动态液滴形状指数平滑模型ES(6)。结合趋势面原理，模拟铺展直径Φ、液滴高度H、液滴面积A和液滴体积V在空间上的分布规律，构建出黄瓜叶面动态接触角的三维趋势面模型。数值模拟检验和实验验证结果表明，模型拟合度达到94.72%，具有较高的模拟精度。

关键词:雾滴;动态接触角;时间序列分析;叶片润湿;趋势面模型

Abstract:

A prediction and modeling method for dynamic contact angle of spraying agricultural drop on cucumber leaf surface based on image data collected was presented with combination of time series analysis and applying spatial trends theory. The dynamic contact angle θ of different mass fraction of glyphosate additives (EF8108-B) aqueous solution range from 16.08% to 82.14% on the cucumber leaf surface were measured by applying high-speed digital camera technology. Experimental results showed that the contact angle declined rapidly within the first 2 s, and then decreased slowly and gradually to reach its stabilized in the following 8 s. The contact angle didn’t show increase of a downward trend with increasing EF8108-B solution concentration. The self-relevant and self-regression model (AR(2)), the mobile average model (MA(3)) and exponential smoothing model (ES(6)) were established according to time series analysis, respectively. Moreover, the spatial distribution of spreading diameter, drop height, drop area and drop volume were simulated by applying trend surface theory, then the 3-D dynamic contact angle trend surface model were constructed. The numerical simulation and experimental verification results indicated that the goodness of fit was up to 94.72%.

Key words:Drop;Dynamic contact angle;Time series analysis;Leaf wetting;Trend surface model