Task Assignment of Heterogeneous Robots Based on Large Model Prompt Learning

In order to ensure that the heterogeneous robot clusters performing the task can complete the target search at the specified location of the target object that the user needs to search according to their own field of view capabilities, we propose a task assignment algorithm for heterogeneous indoor robot clusters based on the robot’s own field of view constraints. In particular, the heterogeneous robot clusters need to be parsed by linguistic commands to obtain the assignment results. Therefore, we solve the task assignment of heterogeneous robot clusters by performing cue learning on a large model to achieve the maximum utilization of heterogeneous robots while satisfying the field-of-view constraint; the simulation verifies the effectiveness of the task assignment of this method.