Distribution law of rice pollen in the wind field of small UAV

The wind field produced by the rotor-wing UAV has a significant impact on the distribution of rice pollen, which directly influences hybrid rice breeding. This research aimed to explore the distribution law of rice pollen in the wind field of small UAV. Aviation Beidou Positioning System UB351 positions all sampling nodes for precise corresponding coordinates and spacing information, and draws UVA's flying trajectories, thus providing accurate data for tests. Wireless sensor network measurement system was used to study the three-direction wind field produced by the rotor-wing UAV under various experimental factors and acquires wind field width and wind speed, and the data were compared with the area ratio and width of pollen distribution. Test of univariate normality was conducted through Shapiro-Wilk test and Kolmogorov-Smimov test. In order to figure out the appropriate flight speed for UAV's pollination under pollen distribution law, it is also necessary to perform analysis of variances on regression model. The comparison of wind speed in longitudinal (X) and lateral (Y) direction show that as the major force of the horizontal wind field produced by the rotor-wing UAV, the wind from Y-direction forms the widest wind field. Moreover, flight speed mainly influences wind field width. To be specific, the width of horizontal wind field decreases as flight speed increases. Meanwhile, UAV flight speed also exerts significant impact on vertical wind field. Both the pollen distribution width of more than 5 pollen grains and the area ratio reached the maximum when the UAV flight was at 4.53 m/s, which was the most favorable speed to pollination. In addition, pollen quantity is closely associated with both horizontal and vertical wind field. With comparison of the pollen quantity of sampling nodes, it was found that the wind field produced by the rotor-wing UAV exerted asymmetrical impact on pollen distribution. Q-Q plot of SPSS verifies that pollen distribution is against normal distribution. The establishment of a multiple linear regression model of pollen distribution and wind speed in three directions indicates that pollen distribution quantity only shares positive linear relation with the wind field in X-direction. These findings provide a theoretical guidance for rice pollination by using agricultural UAVs.