Motion Planning and Control for Multi-UAV Cooperative Transportation

Apr, 2020 - Sep, 2022

Aims: Design motion planning approaches and formation control schemes for multiple UAVs, enabling them to collaborate in transporting payloads via tethers.

Objects: Multi-UAV Cooperative Transportaion System

Problems:

• Constraints among UAVs and among UAVs and payload exist, which led to the need to maintain a configuration for transportation.
• Offline algorithms cost long time and have low efficiency, and need to avoid obstacles in dynamic environments

Contribution:

• Investigated system dynamics model to reduce planning variables and expand the motion range of each UAV, then proposed a novel path planning algorithm to obtain smooth, collision-avoidance trajectories of UAVs for cooperative transportation.
• Examined reinforcement learning based path planning scheme as a local planner to handle dynamic environmental changes, and proposed an intelligent fault-tolerant control algorithm based on deep deterministic policy gradient algorithm to address the case of actuator failure.