Deep Reinforcement Learning for UAV Intelligent Mission Planning

Table 2

Algorithm: PPO (CLIP).

---

Proximal policy optimization algorithm (PPO)

---

1. For i = 1 to N do

2. Run policy for T timesteps, collecting ;

3. Estimate return and advantage ;

4. For k = 1 to K do

5. Sample minibatch from the trajectory, calculate policy loss and value loss;

6. Optimize surrogate loss function ;

7. Update policy parameter ;

8. End for

9. End for

---