The invention discloses an unknown environment agent collaborative
planning method and
system based on sequential single auction, and belongs to the field of task allocation and path planning. A global tag Petri
network model is constructed according to environment information and task information; performing task point
library reduction on the global tag
Petri net model by using a task point abstraction method, and combining redundant grid libraries to obtain an abstracted global tag
Petri net model and a task Boolean specification; according to the abstracted global tag Petri
network model and a task Boolean specification, performing task allocation on each
robot by adopting an improved task
allocation method based on an auction
algorithm to obtain a task allocation result of each
robot and an initial path set of each
robot; constructing a local tag
Petri net model of each robot, and generating an initial path strategy; and the decision state of the initial path is expanded, a binary sensor is used for simulating environment
cognition, and a complete
motion strategy of each robot is obtained according to sensor signals.