Multirobot Coverage Path Planning Based on Deep Q-Network in Unknown Environment

Table 5

The average repetition rates in comparison simulations.

	The environment is unknown	The planning is completed in a short time	2 robots in random environments	2 robots in artificial environments	3 robots in random environments	3 robots in artificial environments
Our method	Yes	Yes	17.1%	21.4%	22.1%	17.8%
GBNN	No	Yes	26.8%	7.1%	49.4%	46.4%
CPMPC	No	No	11.6%	14.3%	18.6%	71.4%