Path Planning For Unmanned Surface Vehicles Based On Modified Artificial Fish Swarm Algorithm With Local Optimizer

Table 3

Performance comparison between different methods for 8 scenarios.

	Evaluation criteria	Proposed	Standard AFSA	lite	Hybrid
Scenario 1 (complex map)	AVG travel distance (pixels)	13.900	16.240	13.890	15.070
	Time cost (s)	0.099	0.137	0.135	0.135
	Time SD (s)	0.012	0.023	0.028	0.026
Scenario 2 (complex map)	AVG travel distance (pixels)	13.900	16.240	15.660	13.890
	Time cost (s)	0.120	0.146	0.180	0.131
	Time SD (s)	0.014	0.023	0.024	0.026
Scenario 3 (complex map)	AVG travel distance (pixels)	30.380	35.070	30.970	31.800
	Time cost (s)	0.331	0.536	0.588	0.630
	Time SD (s)	0.021	0.43	0.047	0.037
Scenario 4 (complex map)	AVG travel distance (pixels)	27.790	31.800	29.210	28.630
	Time cost (s)	0.306	0.554	0.576	0.623
	Time SD (s)	0.011	0.115	0.064	0.055
Scenario 5 (open water)	AVG travel distance (pixels)	244.000	258.800	248.200	249.300
	Time cost (s)	1.245	3.609	8.115	2.578
	Time SD (s)	0.210	0.700	0.640	0.560
Scenario 6 (open water)	AVG travel distance (pixels)	395.600	415.880	408.960	403.810
	Time cost (s)	1.352	4.506	8.470	2.610
	Time SD (s)	0.312	0.441	0.652	0.760
Scenario 7 (open water)	AVG travel distance (pixels)	318.350	344.180	324.570	326.330
	Time cost (s)	1.124	4.124	8.761	2.523
	Time SD (s)	0.212	0.471	0.572	0.778
Scenario 8 (open water)	AVG travel distance (pixels)	359.000	370.800	360.010	360.970
	Time cost (s)	1.235	5.264	9.021	1.812
	Time SD (s)	0.348	0.473	0.642	0.678