[Paper Review] Practical Mission Planning for Optimized UAV-Sensor Wireless Recharging

Qiuchen Qian, James OKeeffe+2 more|arXiv (Cornell University)|Mar 9, 2022

UAV Applications and Optimization4 citations

TL;DR

This paper proposes ROMP, a hybrid offline/online mission planner for UAVs in Wireless Rechargeable Sensor Networks, combining Guided Local Search with a Context-aware Black Hole Algorithm to optimize energy recharging under dynamic wind and battery constraints. FastROMP, an enhanced version, reduces computation time by up to 93.3% through real-time route adjustment, improving recharged energy by 21.47% with minimal energy cost increase.

ABSTRACT

Optimal maintenance of sensor nodes in a Wireless Rechargeable Sensor Network (WRSN) requires effective scheduling of power delivery vehicles by solving the Charging Scheduling Problem (CSP). Deploying Unmanned Aerial Vehicles (UAVs) as mobile chargers has emerged as a promising solution due to their mobility and flexibility. The CSP can be formulated as a Mixed-Integer Non-Linear Programming problem whose optimization objective is maximizing the recharged energy of sensor nodes within the UAV battery constraint. While many studies have demonstrated satisfactory performance of heuristic algorithms in addressing specific routing problems, few studies explore online updating (i.e., mission re-planning `on the fly') in the CSP context. Here we present a new offline and online mission planner leveraging a first-principles power consumption model that uses real-time state information and environmental information. The planner, namely Rapid Online Metaheuristic-based Planner (ROMP), supplements solutions from a Guided Local Search (GLS) with our Context-aware Black Hole Algorithm. Our results demonstrate that ROMP outperforms GLS in most cases tested. We developed and proposed FastROMP to speed up the online mission (re-)planning algorithm by introducing a new online adjustment operator that uses the latest state information as input, eliminating the need for re-initialization. FastROMP not only provides a better quality route, but it also significantly reduces computational time. The reduction ranges from 39.57% in sparse deployment to 93.3% in denser deployments.

Motivation & Objective

• Address the dynamic Charging Scheduling Problem (CSP) in Wireless Rechargeable Sensor Networks (WRSNs) where UAVs must recharge sensor nodes under energy and environmental constraints.
• Overcome limitations of static offline planners by enabling real-time mission re-planning in response to changing environmental conditions such as wind and battery levels.
• Improve solution quality and computational efficiency for UAV path planning in NP-hard TSP and Orienteering Problem (OP) scenarios under energy constraints.
• Develop a lightweight, scalable planner suitable for real-world deployment on embedded UAV flight computers.
• Demonstrate that online re-planning using real-time state data significantly enhances energy recharging performance and reduces computational overhead.

Proposed method

• Formulate the UAV charging scheduling problem as a mixed-integer non-linear program maximizing recharged energy under UAV battery and wind-induced power consumption constraints.
• Develop a first-principles power consumption model for multi-rotor UAVs (DJI M100) incorporating wind effects and inductive power transfer (IPT) efficiency.
• Implement an offline planner, ROMP, using a two-stage approach: an Initial Solver (GLS) for initial route generation and an Optimization Solver (Context-aware Black Hole Algorithm) to refine solutions.
• Introduce FastROMP, a novel online adjustment operator that reuses prior routes and updates them using real-time state data (e.g., remaining battery, wind speed) without re-initialization.
• Apply parallelization across multiple cores to accelerate the offline solver, achieving 99.8% solution quality improvement over GLS at 5000 population size.
• Use a hybrid metaheuristic framework combining single-solution (GLS) and population-based (CBHA) optimization to balance exploration and exploitation in dynamic environments.

Experimental results

Research questions

• RQ1 Can a hybrid offline/online mission planner significantly improve recharged energy in WRSNs while reducing computational time compared to static offline methods?
• RQ2 How effective is the integration of a context-aware Black Hole Algorithm with Guided Local Search in enhancing solution quality for UAV charging scheduling under dynamic wind conditions?
• RQ3 To what extent does the online adjustment operator in FastROMP reduce computation time without degrading solution quality in dynamic CSP scenarios?
• RQ4 What is the impact of real-time state information (e.g., battery level, wind speed) on the performance and adaptability of UAV mission re-planning?
• RQ5 Can a lightweight, real-time re-planning algorithm be efficiently deployed on embedded UAV flight computers for practical WRSN applications?

Key findings

• ROMP outperforms GLS in solution quality, achieving a 21.47% gain in recharged energy with only a 1.89% increase in UAV energy consumption in dynamic wind conditions.
• FastROMP reduces computational time by 39.57% in sparse deployments and up to 93.3% in denser deployments compared to ROMP, due to efficient online route adjustment.
• The distributed implementation of ROMP with four processing cores achieves 99.8% solution quality improvement over GLS across 500 executions at a 5000 population size.
• The worst-case time complexity of FastROMP is O(n²), making it computationally efficient for real-time deployment, especially in high-density sensor networks.
• The context-aware Black Hole Algorithm significantly enhances route optimization by adapting to environmental dynamics such as wind speed and battery state.
• Field and wind tunnel experiments are planned to validate the planner’s robustness and to refine model parameters like drag coefficient and fluid density for real-world accuracy.