Research Article
[Retracted] Reinforcement Learning-Based Continuous Action Space Path Planning Method for Mobile Robots
Table 1
Neural network-related parameter settings.
| Parameters | Value | Meaning |
| Network learning rate | 0.001 | Network learning speed | Reward discount rate | 0.92 | Future rewards at the current value | Soft update parameters | 0.01 | Update parameters of the strategy network and target network | Steps per round | 250 | Maximum number of exploration steps per round | Total number of rounds | 20000 | Maximum number of rounds | Experience pool capacity | 60000 | Experience storage limit | Batch size | 32 | Update network training batch size |
|
|