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

Table 3

The single-robot repetition rates (%) of different execution methods in 100 simulations.

Simulation number	Experimental group	Control group
1	8.6	8.6
2	11.8	14.7
3	5.7	5.7
4	8.8	32.4
5	0	11.4
6	9.7	22.6
7	2.9	NULL
8	11.4	45.7
9	15.2	27.3
10	14.7	41.2
11	23.5	26.5
12	5.9	26.5
13	2.9	22.9
14	5.7	5.7
15	11.4	11.4
16	8.6	NULL
17	3.1	21.9
18	2.9	2.9
19	11.8	41.2
20	11.8	35.3
21	5.9	11.8
22	15.2	27.3
23	5.9	8.8
24	2.9	2.9
25	5.9	NULL
26	12.1	NULL
27	2.9	2.9
28	0	2.9
29	2.8	13.9
30	0	6.1
31	18.2	27.3
32	0	5.6
33	11.4	8.6
34	8.6	20
35	8.6	37.1
36	8.8	14.7
37	5.9	11.8
38	0	0
39	8.8	14.7
40	8.6	NULL
41	5.6	11.1
42	0	0
43	6.3	NULL
44	12.1	30.3
45	2.9	2.9
46	5.7	17.1
47	0	2.8
48	16.1	NULL
49	0	2.8
50	2.9	NULL
51	12.5	18.8
52	8.8	NULL
53	14.7	NULL
54	0	11.1
55	0	5.6
56	5.7	11.4
57	0	0
58	12.1	18.2
59	0	0
60	15.6	15.6
61	5.7	8.6
62	12.5	NULL
63	0	0
64	5.9	20.6
65	8.8	14.7
66	2.9	8.6
67	9.1	NULL
68	8.6	8.6
69	0	0
70	11.8	20.6
71	17.6	11.8
72	31.3	46.9
73	2.8	8.3
74	0	NULL
75	0	0
76	5.9	5.9
77	2.9	NULL
78	2.9	2.9
79	5.9	14.7
80	5.7	5.7
81	5.6	19.4
82	8.3	2.8
83	5.9	26.5
84	5.6	19.4
85	6.1	NULL
86	2.9	NULL
87	5.9	11.8
88	2.9	11.4
89	2.9	20.6
90	11.8	38.2
91	0	0
92	5.6	19.4
93	5.7	17.1
94	11.8	26.5
95	5.9	29.4
96	8.8	35.3
97	9.4	NULL
98	2.9	8.6
99	0	0
100	14.7	17.6