基于强化学习的改进三维A
*
算法在线航迹规划
任智, 张栋, 唐硕
Improved three-dimensional A
*
algorithm of real-time path planning based on reinforcement learning
Zhi REN, Dong ZHANG, Shuo TANG
表6
各场景仿真结果表
Table 6
Simulation results table of different scenarios
场景
DDPG训练策略
指数衰减方法
仿真时间/s
节点
仿真时间/s
节点
1
1.373
52
9.612
69
2
2.651
61
38.347
76
3
0.894
54
2.811
54
4
1.081
46
4.692
50
5
1.670
52
19.793
74
6
0.612
47
6.185
48