UAVs Maneuver Decision-Making Method Based on Transfer Reinforcement Learning

Table 3

Design of the training method for 1vs1 confrontation strategies.

Experiment environment parameters	Parameter values
Task area boundary	[0, 100] × [0, 80] (km)
Number of obstacles,	9
The radius and the number of different kinds of obstacles	4 km (3), 5 km (3), 6 km (3)
Area where obstacles appear randomly	[15, 85] × [15, 65] (km)
Radar detection range of attack UAV,	12 km × 120°
Radar detection range of defense UAV,	8 km × 120°
Discrete number of detection range, m	7
Maximum speed of attack UAV,	340 m/s
Maximum speed of defense UAV,	300 m/s
The upper limit of the acceleration of the attack UAV,	20 m/s^2
The upper limit of the acceleration of the defense UAV,	20 m/s^2
The maximum angular velocity of the attack UAV,	π/15.7
The maximum angular velocity of the defense UAV,	π/22.6
The initial position coordinates and heading angle of the attack UAV	[2.5, 2.5] (km), π/4
The initial position coordinates and heading angle of the defense UAV	[97.5, 77.5] (km), 5π/4
The center point coordinate and radius of the target area	[95, 75] (km), 5 (km)
The attacking radius of the defense UAV,	1 km