Automatic Control Optimization for Large-Load Plant-Protection Quadrotor

In this article, methods for the attitude control optimization of large-load plant-protection quadrotor unmanned aerial vehicles (UAVs) are presented. Large-load quadrotors can be defined as equipped with sprayers and a tank containing large amount water or pesticide, allowing to plants spray pesticide during flight. Compared common small quadrotors, two main points need considered in quadrotors—first, gradually diminishes flight physical parameters change process. Second, size mass rotors especially large, which greatly slows response rate rotors. We present an extended-state reinforcement learning (RL) algorithm solve these problems. The moment inertia (MOI) three axes dynamic constant included state list training process, so that controller learn changes models. controlling laws automatically generated optimized, simplifies tuning process compared those traditional algorithms. article is tested on 10 kg class quadrotor, performance verifies effectiveness our work.