Loading...

Autonomous Search of Forests with Minimum Number of Drones for Early Wildfire Detection

Pordal, Reza | 2024

0 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 57418 (45)
  4. University: Sharif University of Technology
  5. Department: Aerospace Engineering
  6. Advisor(s): Banazadeh, Afshin
  7. Abstract:
  8. This study focuses on path planning for a team of Unmanned Aerial Vehicles (UAVs) involved in efficiently searching a region for early wildfire detection. The primary goal of the path planning algorithm is to identify optimized search paths that maximize the likelihood of detecting wildfires. This optimization considers various factors, including UAV flight time constraints and the locations of the final bases. To improve the algorithm's performance, an initial phase incorporates the utilization of density-based clustering to eliminate noise within the probability map. Furthermore, centroid-based clustering is applied to mitigate the complexity of the optimization problem. Subsequently, an optimization technique inspired by ant colonies is utilized to minimize the objective function. This is achieved by introducing ant subgroups to determine an optimal sequence of cluster centers as waypoints for the UAV team. The generated paths are refined using a self-organizing map neural network to improve path smoothness and meet maneuverability requirements. Finally, the effectiveness of the proposed approach is evaluated using spatial datasets of probabilistic wildfire risk components for the United States
  9. Keywords:
  10. Path Planning ; Clustering ; Ant Colony Optimization (ACO) ; Self-Organizing Map (SOM) ; Wildfire ; Self-Organizing Map (SOM)Nueral Networks ; Unmanned Aerial Vehicles (UAV) ; Fire Probability Map

 Digital Object List

 Bookmark

No TOC