Parallel learning based adaptive control method for quadrotor unmanned aerial vehicle hoisting system
By employing parallel learning technology, the quadcopter UAV hoisting system has achieved accurate identification of load mass and stable attitude control during routine flight missions. This solves the problem of the dependence of traditional adaptive control algorithms on continuous excitation conditions and improves the stability and trajectory accuracy of the hoisting process.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- CHONGQING UNIV OF POSTS & TELECOMM
- Filing Date
- 2026-05-29
- Publication Date
- 2026-07-17
AI Technical Summary
In hoisting missions, the load mass of quadcopter drones is unknown and the model parameters are highly uncertain. Traditional adaptive control algorithms rely too heavily on continuous excitation conditions, which makes it impossible to accurately identify the system's physical parameters in regular flight missions. Furthermore, load swaying and external wind interference during hoisting affect the accuracy of the flight trajectory.
An adaptive control method based on parallel learning is adopted. By establishing a dynamic model of the UAV hoisting system, a position subsystem controller and an attitude subsystem controller are designed. By combining a nonlinear disturbance observer and a historical data stack, the system can accurately identify the load quality and stably control the attitude. The parallel learning technology is used to reuse historical data to meet the continuous excitation conditions.
Achieving high-precision parameter identification under non-continuous excitation conditions reduces the impact of load swing and external wind field interference, improves the stability and trajectory accuracy of the hoisting process, and significantly reduces height drop and position overshoot.
Smart Images
Figure CN122411052A_ABST