Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

SKYSURF: A Self-learning Framework for Persistent Surveillance using Cooperative Aerial Gliders

Houssem Eddine Mohamadi, Nadjia Kara

TL;DR

SKYSURF addresses persistent aerial surveillance by enabling a flock of soaring-capable UAVs to harvest energy from atmospheric thermals. It introduces a self-learning framework where UAVs act as rational agents with local decision-making and a global manager, coupled with an $H{{_{ ext{w}}}{H}}$ path-planning module and a DLnT-based PID controller tuning. Key contributions include a modular cooperative architecture, a lift-map with decay for shared energy sources, a task-planning and routing system, and comprehensive simulations showing increased endurance and target detection (two times better than some baselines) with around 6% battery consumption over six hours. The work demonstrates practical potential for long-duration surveillance and points to future extensions to other lift sources and moving-obstacle scenarios.

Abstract

The success of surveillance applications involving small unmanned aerial vehicles (UAVs) depends on how long the limited on-board power would persist. To cope with this challenge, alternative renewable sources of lift are sought. One promising solution is to extract energy from rising masses of buoyant air. This paper proposes a local-global behavioral management and decision-making approach for the autonomous deployment of soaring-capable UAVs. The cooperative UAVs are modeled as non-deterministic finite state-based rational agents. In addition to a mission planning module for assigning tasks and issuing dynamic navigation waypoints for a new path planning scheme, in which the concepts of visibility and prediction are applied to avoid the collisions. Moreover, a delayed learning and tuning strategy is employed optimize the gains of the path tracking controller. Rigorous comparative analyses carried out with three benchmarking baselines and 15 evolutionary algorithms highlight the adequacy of the proposed approach for maintaining the surveillance persistency (staying aloft for longer periods without landing) and maximizing the detection of targets (two times better than non-cooperative and semi-cooperative approaches) with less power consumption (almost 6% of battery consumed in six hours).

SKYSURF: A Self-learning Framework for Persistent Surveillance using Cooperative Aerial Gliders

TL;DR

SKYSURF addresses persistent aerial surveillance by enabling a flock of soaring-capable UAVs to harvest energy from atmospheric thermals. It introduces a self-learning framework where UAVs act as rational agents with local decision-making and a global manager, coupled with an path-planning module and a DLnT-based PID controller tuning. Key contributions include a modular cooperative architecture, a lift-map with decay for shared energy sources, a task-planning and routing system, and comprehensive simulations showing increased endurance and target detection (two times better than some baselines) with around 6% battery consumption over six hours. The work demonstrates practical potential for long-duration surveillance and points to future extensions to other lift sources and moving-obstacle scenarios.

Abstract

The success of surveillance applications involving small unmanned aerial vehicles (UAVs) depends on how long the limited on-board power would persist. To cope with this challenge, alternative renewable sources of lift are sought. One promising solution is to extract energy from rising masses of buoyant air. This paper proposes a local-global behavioral management and decision-making approach for the autonomous deployment of soaring-capable UAVs. The cooperative UAVs are modeled as non-deterministic finite state-based rational agents. In addition to a mission planning module for assigning tasks and issuing dynamic navigation waypoints for a new path planning scheme, in which the concepts of visibility and prediction are applied to avoid the collisions. Moreover, a delayed learning and tuning strategy is employed optimize the gains of the path tracking controller. Rigorous comparative analyses carried out with three benchmarking baselines and 15 evolutionary algorithms highlight the adequacy of the proposed approach for maintaining the surveillance persistency (staying aloft for longer periods without landing) and maximizing the detection of targets (two times better than non-cooperative and semi-cooperative approaches) with less power consumption (almost 6% of battery consumed in six hours).
Paper Structure (19 sections, 16 equations, 19 figures, 3 tables, 3 algorithms)

This paper contains 19 sections, 16 equations, 19 figures, 3 tables, 3 algorithms.

Table of Contents

  1. Introduction
  2. Related works
  3. Autonomous soaring algorithms
  4. Persistent surveillance
  5. Cooperative autonomous soaring
  6. Problem formulation
  7. Proposed approach
  8. Cooperative surveillance and soaring
  9. Decision-making policy
  10. Path planning
  11. Control strategy
  12. Results and analysis
  13. Soaring flight and duration analysis
  14. Updrafts detection and exploitation analysis
  15. Targets detection analysis
  16. ...and 4 more sections

Figures (19)

  • Figure 1: Velocity profile of a sample updraft in different development phases
  • Figure 2: High-level depiction of the components in the framework (at UAV and flock levels)
  • Figure 3: Illustration of the discrete behaviors (states /transitions) that make up the soaring surveillance system
  • Figure 4: Step-wise overview of the delayed learning and tuning (DLnT) algorithm
  • Figure 5: 3D view of flight paths for three UAVs after 1 hour
  • ...and 14 more figures