READ ME File For 'Simulation results'

Dataset DOI: 10.5258/SOTON/D3579

Date that the file was created: May, 2025

-------------------
GENERAL INFORMATION
-------------------

ReadMe Author: Zhipeng Wang, University of Southampton, 0009-0004-1940-1047

This dataset supports the thesis entitled "Unmanned Aerial Vehicle Autonomous Navigation using Semi-Prior Deep Reinforcement Learning"
AWARDED BY: University of Southampton
DATE OF AWARD: 2025

DESCRIPTION OF THE DATA
Simulation results of cumulative reward model, 3D spatial information compression and 3DSIC in POMDP environment and Dual-direction 3DSIC. 

This dataset is recommended to be read using Matlab, and users can use the plot function to plot the learning curves of different types of deep reinforcement learning agents.

This dataset contains:

The learning curves data of DQN, DDPG and FRSVG agents with cumulative reward model and traditional reward model.

The learning curves data of DQN, DDPG and FRSVG agents with and without 3DSIC in MDP environment.

The learning curves data of DQN, DDPG and FRSVG agents with and without 3DSIC in POMDP environment.

Date of data collection: 01.05.2025

--------------------------
SHARING/ACCESS INFORMATION
-------------------------- 

Licence: CC-BY

Recommended citation for the data:

This dataset supports the thesis:
Zhipeng Wang (2025) "Unmanned Aerial Vehicle Autonomous Navigation using Semi-Prior Deep Reinforcement Learning", University of Southampton, name of the 
University Faculty or School or Department, PhD Thesis, pagination. 

Related publication:

[1] Z. Wang, S. X. Ng and M. El-Hajjar, "Deep Reinforcement Learning Assisted UAV Path Planning Relying on Cumulative Reward Mode and Region Segmentation," 
in IEEE Open Journal of Vehicular Technology, vol. 5, pp. 737-751, 2024, doi: 10.1109/OJVT.2024.3402129.

[2] Z. Wang, S. X. Ng and M. El-Hajjar, "A 3D Spatial Information Compression Based Deep Reinforcement Learning Technique for UAV Path Planning in 
Cluttered Environments," in IEEE Open Journal of Vehicular Technology, vol. 6, pp. 647-661, 2025, doi: 10.1109/OJVT.2025.3540174.