A multi-unmanned aerial vehicle cluster and obstacle avoidance control method based on skill arrangement

By decomposing multi-UAV missions into cluster and obstacle sub-missions, and constructing low-level skill networks and high-level skill orchestration networks, efficient collaborative control of multiple UAVs in complex environments is achieved. This solves the problems of strategy discontinuity and insufficient environmental adaptability in existing technologies and improves collaborative performance.

CN122411548APending Publication Date: 2026-07-17NANJING UNIV OF AERONAUTICS & ASTRONAUTICS
View PDF 0 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
NANJING UNIV OF AERONAUTICS & ASTRONAUTICS
Filing Date
2026-04-10
Publication Date
2026-07-17

AI Technical Summary

Technical Problem

Existing multi-UAV cooperative control methods struggle to achieve efficient and reliable cooperative control in complex environments, especially in dynamic environments and complex scenarios where they cannot dynamically integrate multiple skills, resulting in strategy discontinuity and insufficient environmental adaptability.

Method used

A skill-based orchestration-based multi-UAV swarm and obstacle avoidance control method is adopted. The task is decomposed into swarm sub-tasks and obstacle sub-tasks. By utilizing a parameter sharing mechanism and a centralized training-distributed execution framework, a low-level skill network and a high-level skill orchestration network are constructed. Multi-UAV collaborative control is achieved through skill weighted fusion.

Benefits of technology

It significantly reduces complexity, improves policy adaptability and robustness, allows multiple skills to work together at the same time, avoids policy discontinuity, and improves collaborative performance under complex tasks.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122411548A_ABST
    Figure CN122411548A_ABST
Patent Text Reader

Abstract

The application discloses a kind of multi-unmanned aerial vehicle cluster and barrier control method based on skill arrangement, it is related to artificial intelligence and unmanned aerial vehicle system field, the method is decomposed into multiple subtasks for collaborative task, and each subtask is modeled as a decentralized partially observable Markov decision process, and then based on parameter sharing mechanism, using multi-agent deep deterministic policy gradient algorithm, for each subtask under each agent is constructed for generating original action low-level skill network, and for generating skill weight high-level skill arrangement network.In execution process, based on overall observation information, dynamically adjust the skill weight, and the original action and skill weight are weighted and fused, to generate the final control instruction of each unmanned aerial vehicle.The application can flexibly adjust the behavior combination according to real-time environmental changes by skill decomposition and collaborative arrangement, significantly improves the learning efficiency, strategy adaptability and task execution robustness of multi-unmanned aerial vehicle in obstacle dense scene.
Need to check novelty before this filing date? Find Prior Art