Fine-Grained Cooperative Instruction Architecture for Heterogeneous Unmanned Swarms in Dynamic Task Scenarios

Aiming at the new requirements and challenges faced by task planning and collaborative control in the dynamic scenario of heterogeneous unmanned swarms, such as scalability, task diversification, command and control interoperability, as well as the characteristics and difficulties of swarm system control under heterogeneous units, complex space, difficult collaboration and diverse constraints, this paper defines the concept and connotation of fine-grained collaborative instruction architecture for task planning and collaborative control in heterogeneous unmanned swarms, a three-tier instruction architecture consisting of task instruction layer, action instruction layer and load instruction layer is proposed, the directory and collection of constraint and effect instructions are constructed, presents the instruction layer generation model. The instruction architecture can comprehensively consider the complexity and constraints of heterogeneous unmanned swarms' collaborative task execution, and plan a reasonable and feasible task sequence, action sequence and load sequence in the dynamic task scenario, it provides theoretical support and design reference for improving the information processing and instant decision-making ability of swarm command.