Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Effective Underwater Glider Path Planning in Dynamic 3D Environments Using Multi-Point Potential Fields

Hanzhi Yang, Nina Mahmoudian

TL;DR

This work tackles real-time 3D path planning for underwater gliders operating in dynamic environments with unknown obstacles and flow fields. It extends the Multi-Point Potential Field (MPPF) approach by adding velocity-based repulsion and a flow-field potential, along with a UG-aware sawtooth pre-plan and vertical local-minima avoidance, to produce a robust total potential $U_{tot}$ guiding navigation. Validated on the ROUGHIE prototype through extensive simulations, the enhanced MPPF demonstrates improved obstacle avoidance (static and dynamic), effective local minima escape, and better time efficiency under currents, including complex flow fields and multi-depth trajectories. The results support practical deployment of real-time, flow-aware, 3D path planning for UGs in near-shore, fjord-like, and harbor-adjacent environments, paving the way for reliable autonomous underwater exploration.

Abstract

Underwater gliders (UGs) have emerged as highly effective unmanned vehicles for ocean exploration. However, their operation in dynamic and complex underwater environments necessitates robust path-planning strategies. Previous studies have primarily focused on global energy or time-efficient path planning in explored environments, overlooking challenges posed by unpredictable flow conditions and unknown obstacles in varying and dynamic areas like fjords and near-harbor waters. This paper introduces and improves a real-time path planning method, Multi-Point Potential Field (MPPF), tailored for UGs operating in 3D space as they are constrained by buoyancy propulsion and internal actuation. The proposed MPPF method addresses obstacles, flow fields, and local minima, enhancing the efficiency and robustness of UG path planning. A low-cost prototype, the Research Oriented Underwater Glider for Hands-on Investigative Engineering (ROUGHIE), is utilized for validation. Through case studies and simulations, the efficacy of the enhanced MPPF method is demonstrated, highlighting its potential for real-world applications in underwater exploration.

Effective Underwater Glider Path Planning in Dynamic 3D Environments Using Multi-Point Potential Fields

TL;DR

This work tackles real-time 3D path planning for underwater gliders operating in dynamic environments with unknown obstacles and flow fields. It extends the Multi-Point Potential Field (MPPF) approach by adding velocity-based repulsion and a flow-field potential, along with a UG-aware sawtooth pre-plan and vertical local-minima avoidance, to produce a robust total potential guiding navigation. Validated on the ROUGHIE prototype through extensive simulations, the enhanced MPPF demonstrates improved obstacle avoidance (static and dynamic), effective local minima escape, and better time efficiency under currents, including complex flow fields and multi-depth trajectories. The results support practical deployment of real-time, flow-aware, 3D path planning for UGs in near-shore, fjord-like, and harbor-adjacent environments, paving the way for reliable autonomous underwater exploration.

Abstract

Underwater gliders (UGs) have emerged as highly effective unmanned vehicles for ocean exploration. However, their operation in dynamic and complex underwater environments necessitates robust path-planning strategies. Previous studies have primarily focused on global energy or time-efficient path planning in explored environments, overlooking challenges posed by unpredictable flow conditions and unknown obstacles in varying and dynamic areas like fjords and near-harbor waters. This paper introduces and improves a real-time path planning method, Multi-Point Potential Field (MPPF), tailored for UGs operating in 3D space as they are constrained by buoyancy propulsion and internal actuation. The proposed MPPF method addresses obstacles, flow fields, and local minima, enhancing the efficiency and robustness of UG path planning. A low-cost prototype, the Research Oriented Underwater Glider for Hands-on Investigative Engineering (ROUGHIE), is utilized for validation. Through case studies and simulations, the efficacy of the enhanced MPPF method is demonstrated, highlighting its potential for real-world applications in underwater exploration.
Paper Structure (15 sections, 16 equations, 12 figures)

This paper contains 15 sections, 16 equations, 12 figures.

Table of Contents

  1. INTRODUCTION
  2. METHODOLOGY
  3. Multi-Point Potential Field (MPPF) review
  4. MPPF for underwater gliders
  5. Case 1 - start point and end point at different depths
  6. Case 2 - start point and end point at the same depth, $Z_{max} \leq d_{max}$
  7. Case 3 - start point and end point at the same depth, $Z_{max} > d_{max}$
  8. Local minima problem
  9. CASE STUDY
  10. SIMULATION RESULTS AND DISCUSSION
  11. Sawtooth path following
  12. Static obstacles and local minima
  13. Dynamic obstacles
  14. Flow field
  15. CONCLUSION AND FUTURE WORK

Figures (12)

  • Figure 1: Custom-made underwater glider ROUGHIE was used for validation studies.
  • Figure 2: 3D illustration of ROUGHIE with MPPF. Attractive potential generates sampling waypoints around the glider and calculates the distance to the target point and repulsive potential generates points on the obstacles and calculates the distance to the points.
  • Figure 3: Schematic of closed-loop control system implemented on ROUGHIE.
  • Figure 4: Pre-planned sawtooth trajectory tracking using attractive potential, shown in (a) 2D top view and (b) 3D view.
  • Figure 5: Static obstacle avoidance by repulsive potential, shown in (a) 2D top view and (b) 3D view.
  • ...and 7 more figures