Volume 36, Number 2, April 2018
|Page(s)||323 - 331|
|Published online||03 July 2018|
Decision Modeling of UAV On-Line Path Planning Based on IMM
School Electronics and Information, Northwestern Polytechnical University, Xi'an 710072, China
2 Air Force Military Representatire Office in Jiangxi Area, Nanchang 330024, China
In order to enhance the capability of tracking targets autonomously of UAV, a model for UAV on-line path planning is established based on the theoretical framework of partially observable markov decision process(POMDP). The elements of the POMDP model are analyzed and described. According to the diversity of the target motion in real world, the law of state transition in POMDP model is described by the method of Interactive Multiple Model(IMM) To adapt to the target maneuvering changes. The action strategy of the UAV is calculated through nominal belief-state optimization(NBO) algorithm which is designed to search optimal action policy to minimize the cumulative cost of action. The generated action strategy controls the UAV flight. The simulation results show that the established POMDP model can achieve autonomous planning for UAV route, and it can control the UAV to effectively track target. The planning path is more reasonable and efficient than the result of using single state transition law.
Key words: partially observable markov decision process / interactive multiple model / UAV / target tracking / MATLAB / cost reduction
关键字 : 部分可观测马尔科夫决策过程(POMDP) / 交互多模型(IMM) / 路径规划 / 目标跟踪 / 名义信念状态优化 /
© 2018 Journal of Northwestern Polytechnical University. All rights reserved.
This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.
Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.
Initial download of the metrics may take a while.