Smoothed Particle Hydrodynamics (SPH) is a fully Lagrangian method that does not require the use of any mesh. Due to its advantages, SPH method is usually adopted to simulate flows with free surfaces. Based on traditional SPH, non-inertial SPH as a new method is proposed recently to deal with cases when the liquid is subjected to inertial force for the motion of reference frame. This study finds that the transport velocity of liquid particles has additional effects on the solution of momentum equation in Navier-Stokes equations. Comparisons and validations are conducted to quantificationally investigate this influence in this paper. Furthermore, the modified non-inertial SPH method is applied to simulate the attitude-liquid-control coupled dynamics of a liquid-filled quadrotor unmanned aerial vehicle (UAV). Parallelization technology and particle searching method based on KD-tree algorithm are used to speed up the simulation. Coupled model of the liquid-filled quadrotor is built firstly, and the liquid sloshing effect on the position & attitude of UAV and the control output of power system are then simulated and discussed. Results prove that the liquid sloshing will deteriorate the trajectory tracking performance, especially the tracking error and rotor output.

平滑粒子流体动力学 (SPH) 是一种完全拉格朗日方法，不需要使用任何网格。由于其优点，通常采用SPH方法来模拟具有自由表面的流动。在传统SPH的基础上，最近提出了一种新的非惯性SPH方法来处理液体受到惯性力引起参考系运动的情况。本研究发现，液体粒子的输运速度对 Navier-Stokes 方程中动量方程的解具有附加影响。本文进行了比较和验证，以量化研究这种影响。此外，应用改进的非惯性 SPH 方法来模拟充液四旋翼无人机 (UAV) 的姿态-液体-控制耦合动力学。采用并行化技术和基于KD-tree算法的粒子搜索方法来加速仿真。首先建立了充液四旋翼飞行器的耦合模型，然后仿真讨论了液体晃动对无人机位置姿态和动力系统控制输出的影响。结果证明液体晃动会恶化轨迹跟踪性能，特别是跟踪误差和转子输出。