Research on Dynamic Path Planning of Mobile Robot Based on Improved DDPG Algorithm

Table 1

Algorithm performance comparison.

	Success rate (100%) (success/total)	Training model time (h)
Original DDPG	25/50 = 0.50	14
Add transfer learning	23/50 = 0.46	13
Add prioritized experience replay	35/50 = 0.70	14
Add transfer learning and prioritized experience replay	37/50 = 0.74	13
Add RAdam algorithm	43/50 = 0.86	12
Add curiosity module	45/50 = 0.90	11