Air Combat Maneuver Trajectory Prediction Model of Target Based on Chaotic Theory and IGA-VNN

Table 3

Target Y-coordinate multistep prediction performance comparison of four different prediction methods.

Prediction step	Algorithm	RMSE	MAE	MAPE	COR
1	BP	48.0217	3.9851	0.0069	0.9965
	IGA-VNN	15.6534	3.4766	0.0022	0.9987
	GA-VNN	30.7653	5.2296	0.0047	0.9970
	Volterra	40.5759	5.9218	0.0061	0.9971
	RBF	48.8712	6.3457	0.0073	0.9974
2	BP	99.3602	8.6527	0.0139	0.9913
	IGA-VNN	23.0582	4.2662	0.0033	0.9980
	GA-VNN	57.3526	6.6614	0.0078	0.9932
	Volterra	59.8836	7.1030	0.0089	0.9948
	RBF	70.9127	7.8841	0.0109	0.9899
4	BP	123.7001	10.4490	0.0188	0.9851
	IGA-VNN	35.4839	5.5817	0.0056	0.9976
	GA-VNN	119.9039	9.9738	0.0175	0.9752
	Volterra	148.4173	10.9043	0.0221	0.9904
	RBF	119.6835	9.2310	0.0159	0.9798
6	BP	161.1386	11.6505	0.0236	0.9685
	IGA-VNN	44.4239	6.4155	0.0073	0.9952
	GA-VNN	158.5973	11.4905	0.0242	0.9846
	Volterra	189.8012	12.7370	0.0295	0.9880
	RBF	159.7916	10.9816	0.0203	0.9618