Energy-Efficient Online Path Planning of Multiple Drones Using Reinforcement Learning

Hong, Dooyoung; Lee, Seonhoon; Cho, Young Hoo; Baek, Donkyu; Kim, Jaemin; Chang, Naehyuck

doi:10.1109/tvt.2021.3102589

Cited by 43 publications

(6 citation statements)

References 18 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…If it is not, the last feasible point of the segment connecting nΓ(wn) to wn (i.e., xc) is used as the central point to perform a random extraction from a gaussian distribution (lines [20][21]. The random point xr becomes the next candidate point to add to the tree (i.e., xw) if the segment going from nΓ(xr) to xr is collision free (lines [22][23][24]. If there is a suitable candidate to be added to the tree (line 26), xw is added to the tree and connected to xn which is either its nearest neighbor, i.e., nΓ(xw) or its preceding point in the tree if its path to xw is feasible (line 27).…”

Section: B Strategic Deconflictionmentioning

confidence: 99%

Multiobjective Modular Strategic Planning Framework for Low-Altitude Missions Within the Urban Air Mobility Ecosystem

Causa,

Fasano

2024

IEEE Trans. Aerosp. Electron. Syst.

View full text Add to dashboard Cite

This paper introduces a multi-metric multiconstraint strategic path planning framework applicable to unstructured urban airspace. The planner is based on a modular and scalable approach to handle several information sources and aspects characterizing urban flight scenarios, such as risk and weather maps, landing site locations, navigation requirements, and mobile and fixed obstacle characteristics. This information is coupled with dynamic constraints and UAV specifications to derive a flyable and safe path connecting a start position and a destination. Strategies for data gathering and synthesis, used to keep a reduced computational burden, are described along with the path planner algorithm. The latter consists in three steps specifically developed to handle both static and time-varying information. A multi-objective cost function with variable weighting coefficients has been implemented so that the most relevant factors for the considered applications can be selected in an adaptive fashion. The performance of the developed algorithms is tested by investigating the planner behavior when changing its inputs as well as the cost function weighting coefficients, demonstrating the ability of the planner in returning an efficient and safe trajectory.

show abstract

Section: B Strategic Deconflictionmentioning

confidence: 99%

Multiobjective Modular Strategic Planning Framework for Low-Altitude Missions Within the Urban Air Mobility Ecosystem

Causa,

Fasano

2024

IEEE Trans. Aerosp. Electron. Syst.

View full text Add to dashboard Cite

show abstract

“…In addition, swarming algorithms and optimized connectivity solutions should be designed to fulfill requirements with minimal energy use. There are recent studies done to improve the energy-efficiency such as applying reinforcement learning for path planning [98], using blockchains both to improve security and energy consumption [99], and utilizing wake-up radios to reserve battery for long-term operations [100]. However, there is more work to be done to achieve long-term operational goals e.g., in many search-andrescue missions.…”

Section: Energy-efficient Drone Swarmsmentioning

confidence: 99%

Situational Awareness for Industry 5.0 and Mobile Robotics: A Review

Höyhtyä¹

2023

Preprint

View full text Add to dashboard Cite

show abstract

“…While UAVs' potential mobility offers exciting possibilities, they also introduce new technological hurdles. UAV pathways must be optimized regarding the essential performance parameters such as throughput, energy efficiency, and latency [24], [25]. In addition, route optimization challenges for the dynamic characteristics and types of drones should be described.…”

Section: Mobile Dbssmentioning

confidence: 99%

Joint position and trajectory optimization of flying base station in 5G cellular networks, based on users' current and predicted location

Sookhak¹,

Mohajerzadeh²

2022

Preprint

View full text Add to dashboard Cite

Nowadays, Unmanned Aerial Vehicles (UAVs) have been significantly improved, and one of their most important applications is to provide temporary coverage for cellular users. Static Base Station cannot service all users due to temporary crashes because of temporary events such as ground BS breakdowns, bad weather conditions, natural disasters, transmission errors, etc., drones equipped with small cellular BS. The Drone Base Station is immediately sent to the target location and establishes the necessary communication links without requiring any predetermined infrastructure and covers that area. Finding the optimal location and the appropriate number (DBS) of drone-BS in this area is a challenge. Therefore, in this paper, the optimal location and optimal number of DBSs are distributed in the current state of the users and the subsequent user states determined by the prediction. Finally, the DBS transition is optimized from the current state to the predicted future locations. The simulation results show that the proposed method can provide acceptable coverage on the network.

show abstract

Energy-Efficient Online Path Planning of Multiple Drones Using Reinforcement Learning

Cited by 43 publications

References 18 publications

Multiobjective Modular Strategic Planning Framework for Low-Altitude Missions Within the Urban Air Mobility Ecosystem

Multiobjective Modular Strategic Planning Framework for Low-Altitude Missions Within the Urban Air Mobility Ecosystem

Situational Awareness for Industry 5.0 and Mobile Robotics: A Review

Joint position and trajectory optimization of flying base station in 5G cellular networks, based on users' current and predicted location

Contact Info

Product

Resources

About