A Novel Searching Method Using Reinforcement Learning Scheme for Multi-UAVs in Unknown Environments

Yue Yue,Wang Wang,Guan Guan

doi:10.3390/app9224964

Abstract

In this paper, the important topic of cooperative searches for multi-dynamic targets in unknown sea areas by unmanned aerial vehicles (UAVs) is studied based on a reinforcement learning (RL) algorithm. A novel multi-UAV sea area search map is established, in which models of the environment, UAV dynamics, target dynamics, and sensor detection are involved. Then, the search map is updated and extended using the concept of the territory awareness information map. Finally, according to the search efficiency function, a reward and punishment function is designed, and an RL method is used to generate a multi-UAV cooperative search path online. The simulation results show that the proposed algorithm could effectively perform the search task in the sea area with no prior information.

Highlights

With the rapid development of sensors, wireless communication, intelligent control, and other technologies, the functions and the application fields of unmanned group systems are increasing day by day
In order to solve the above problems, this paper proposes a multi-unmanned aerial vehicles (UAVs) cooperative search method based on reinforcement learning (RL), which fully considers the characteristics of an unknown sea area
In order to verify the effectiveness of the algorithm, a multi-UAV cooperative search simulation environment was established in MATLAB

Summary

Introduction

With the rapid development of sensors, wireless communication, intelligent control, and other technologies, the functions and the application fields of unmanned group systems are increasing day by day. Because of their expansibility, strong cooperation, and low loss, the cooperative theory and applied research on unmanned group systems received increasing attention in the fields of academia, industry, and national defense [1]. Multi-UAV cooperative sea area searching is one of the important research directions of unmanned group systems [2]. In References [6,7], a search map model was established according to the existence probability of target, and distributed model predictive control was used to solve the problem, which effectively reduced the solution scale of the search decision problem

Methods

Results

Conclusion