Abstract: It firstly introduces the working principle of agricultural UAV and establishes its dynamic model. Then, it deeply studied the trajectory planning of agricultural quadrotor UAV based on the deep learning algorithm of artificial bee colony. Finally, it adopted the linear active disturbance rejection control method to realize the trajectory control of UAV. The experimental results show that the agricultural quadrotor UAV can run smoothly from the starting point to the end point, and avoid the obstacles on the way, which proves the reliability and stability of the system.