In precision agriculture, such as crop spraying, controlling UAVs presents various challenges such as variable payload, inertial coefficient variation, influence of external disturbances such as wind gusts, and uncertainties associated with the dynamics. To address these challenges, this paper proposes a hybrid control technique that combines higher-order integral sliding mode control, fast-terminal sliding mode control, and adaptive law. The objective is to mitigate the effects of variable payload, external disturbances, and uncertainties while maintaining the stability and performance of the UAV during spraying. Initially, a mathematical model is constructed for a coaxial octocopter UAV that is fitted with a spraying tank. This model takes into account the variation in mass and moment of inertia. Then, a two-loop control structure is employed to attain control of both the translational and rotational axis of the UAV. The numerical simulations are performed on a nonlinear model of the agricultural UAV system and compared with neural network based sliding mode control and robust adaptive backstepping control schemes. The robustness of the proposed scheme is tested in wind gusts and sensor measurement error conditions. Finally, hardware-in-loop simulations are performed using the Pixhawk Orange Cube flight controller to validate the real-time capability of the proposed scheme.

中文翻译：

---

稳健的自适应控制律设计，提高农业稳定性农药喷洒无人机


在精准农业中，例如农作物喷洒，控制无人机会带来各种挑战，例如可变有效载荷、惯性系数变化、阵风等外部干扰的影响以及与动力学相关的不确定性。为了应对这些挑战，本文提出了一种混合控制技术，该技术结合了高阶积分滑模控制、快速终端滑模控制和自适应定律。目标是减轻可变有效载荷、外部干扰和不确定性的影响，同时在喷涂过程中保持无人机的稳定性和性能。最初，为装有喷雾罐的同轴八旋翼无人机构建了一个数学模型。该模型考虑了质量和转动惯量的变化。然后，采用双回路控制结构来实现对无人机平移轴和旋转轴的控制。在农用无人机系统的非线性模型上进行了数值模拟，并与基于神经网络的滑模控制和鲁棒的自适应反步控制方案进行了比较。在阵风和传感器测量误差条件下测试了所提方案的鲁棒性。最后，使用 Pixhawk Orange Cube 飞行控制器进行硬件在环仿真，以验证所提方案的实时能力。