Disturbance-rejection flight control strategy for a heavy-payload agricultural unmanned helicopter

Plant-protection equipment is often required to fully meet the diverse production environments for the high efficiency, accuracy, and adaptability in China. Unmanned aerial vehicles (UAVs) have been widely used to improve pesticide utilization efficiency in an emerging pesticide-application platform. Environmental contamination and labor requirements can also be substantially reduced in agricultural ecosystems. Among them, plant-protection unmanned helicopters are particularly suitable for low-altitude spraying, due to their high payload, long endurance, and adaptability to complex terrain. However, the spray liquid can be continuously depleted to cause time-varying vehicle mass, center-of-gravity location, and inertia parameters during operation. Meanwhile, gusts and stochastic wind disturbances can further degrade attitude stability and trajectory-tracking accuracy. This study aims to solve the flight-control challenges from the coupled effects of continuous payload and wind disturbances. A disturbance-rejection flight control strategy was proposed for a fuel-powered heavy-payload plant-protection unmanned helicopter. A nonlinear rigid-body dynamic model of the unmanned helicopter was developed using physics modeling. A systematic analysis was conducted to determine the forces and moments from the main rotor, tail rotor, fuselage, and payload. A time-varying mass-properties model was then incorporated to represent the continuous variations in the total mass, center-of-gravity location, and inertia parameters, as the spray liquid consumed. The parameters were determined using prototype measurement and identification tests. A flight control system was developed to verify the optimization. A cascaded proportional–integral–derivative (PID) architecture was employed for the inner attitude and the outer position loop. The hierarchical arrangement was used to separate the fast attitude dynamics from the slower translational dynamics, thereby balancing flight stabilization, disturbance attenuation, and trajectory tracking. In the inner loop, the attitude angles and angular rates were regulated to rapidly attenuate high-frequency disturbances and then stabilize the helicopter. Whereas the velocity and position were regulated for trajectory tracking to suppress low-frequency errors in the outer loop. In addition, a proportional–integral (PI) engine-speed controller was designed to maintain a stable rotor speed under varying payload conditions. An artificial-potential-field-based path-planning module was implemented to generate the required flight path. The strategy was evaluated using both local- and full-process simulation of a representative plant-protection mission. The local simulations were used to examine hover stability under gust disturbances, particularly for unloaded and fully loaded conditions. The full-process simulation was used to reproduce the complete sequence, including takeoff at full payload, spraying with continuous liquid depletion, and landing at empty payload, while simultaneously accounting for continuous payload variation and stochastic wind disturbances. The results showed that the maximum deviations in attitude angle and angular rate were below 0.05 rad and 0.02 rad/s, respectively, under either payload condition during gust-disturbed hovering, while the position error remained below 0.25 m. The altitude error was maintained below 0.3 m during the full-process simulation, and the horizontal trajectory-tracking error remained below 0.1 m during steady forward flight. The fluctuation amplitudes of the roll, pitch, and yaw angles were below 0.2, 0.03, and 0.02 rad, respectively. The time-varying mass-properties model was established to represent the effects of continuous spray-liquid consumption on the flight dynamics of the helicopter. The flight control strategy maintained high attitude stability and trajectory-tracking accuracy under the continuous payload and wind disturbances. Thereby the findings can provide a strong reference for the modeling, control, and engineering application of heavy-payload plant-protection unmanned helicopters.