|本期目录/Table of Contents|

[1]谷湘煜,彭志远,鲜开义,等.基于深度强化学习的轨道式巡检机器人巡检驻点规划方法[J].工业仪表与自动化装置,2021,(01):73-76.[doi:10.3969/j.issn.1000-0682.2021.01.017]
 GU Xiangyu,Peng Zhiyuan,XIAN Kaiyi,et al.The planning method of orbiting inspection robot’s inspection station based on deep reinforcement learning[J].Industrial Instrumentation & Automation,2021,(01):73-76.[doi:10.3969/j.issn.1000-0682.2021.01.017]
点击复制

基于深度强化学习的轨道式巡检机器人巡检驻点规划方法

《工业仪表与自动化装置》[ISSN:1000-0682/CN:61-1121/TH]

卷:
期数:
2021年01期
页码:
73-76
栏目:
出版日期:
2021-02-15

文章信息/Info

Title:
The planning method of orbiting inspection robot’s inspection station based on deep reinforcement learning
作者:
谷湘煜1彭志远1鲜开义1杨利萍1陈 伟1邹 娟1刘满禄2
1.深圳市朗驰欣创科技股份有限公司,广州 深圳 518000;
2.西南科技大学,四川 绵阳 621000
Author(s):
GU XiangyuPeng ZhiyuanXIAN KaiyiYANG LipingCHEN WeiZOU JuanLIU Manlu
1.Shenzhen Langchi Xinchuang Technology Co., LTD.Guangzhou Shenzhen 518000, China;
2. Southwest University of Science and Technology, Sichuan Mianyang 621000, China
关键词:
轨道式巡检机器人驻点规划强化学习
Keywords:
orbital inspection robot stagnation planning reinforcement learning
分类号:
TH142.71
DOI:
10.3969/j.issn.1000-0682.2021.01.017
文献标志码:
B
摘要:
针对三峡水电站中轨道式巡检机器人的实际巡检需求,对巡检驻点规划展开研究。机器人的巡检驻点一般为人为根据经验及场景预设,因此难以规划最优驻点,影响检测效果。该文通过机器人本体搭载的各种传感设备信息构建综合环境感知函数,作为无模型强化学习算法的环境信息,引导机器人找到最优检测驻点。最后设计实验评估了该方法的有效性。
Abstract:
Aiming at the actual inspection requirements of the track-type inspection robots in the Three Gorges Hydropower Station, the patrol stagnation point planning was studied. The patrol station of the robot is generally artificially preset according to experience and scene, so it is difficult to plan the optimal stagnation point and affect the detection effect. In this paper, the comprehensive environment perception function is built by the various sensor equipments carried by the robot body and used as the environment information of the model-free reinforcement learning algorithm to guide the robot to find the optimal detection stagnation point. The final design experiment evaluated the effectiveness of the method.

参考文献/References:

[1]吕小文. 悬挂轨道式智能巡检机器人研究[D].杭州:浙江大学, 2019.

[2]鲜开义. 变电站轨道式巡检机器人控制系统设计[D].成都:西南交通大学,2018.
[3]郭峰. 轨道式变电站设备巡检机器人的设计与应用[D]. 鞍山:辽宁科技大学,2016.
[4]张营,鲁守银,李建祥,等.基于模糊认知图的变电站巡检机器人的行为规划研究[J].黑龙江大学自然科学学报,2017,34(03):345-351.
[5]宋丹一. 改进A*算法在钢铁企业安全巡检机器人路径规划的研究[D]. 鞍山:辽宁科技大学,2017.
[6]周立辉,张永生,孙勇,等.智能变电站巡检机器人研制及应用[J].电力系统自动化,2011,35(19):85-88+96.
[7]李炜,李全龙,刘政怡.基于加权的K近邻线性混合显著性目标检测[J].电子与信息学报,2019,41(10):2442-2449.
[8]唐红梅,王碧莹,韩力英,等.基于目标紧密性与区域同质性策略的图像显著性检测[J].电子与信息学报,2019,41(10):2532-2540.
[9]汤荣志. 数据归一化方法对提升SVM训练效率的研究[D]. 济南:山东师范大学,2017.
[10]金硕,阮江军,朱琳,等.基于臭氧检测法的电晕起始电压测量方法[J].高电压技术,2017,43(12):4020-4025.
[11]MNIH V, KAVUKCUOGLU K, SILVER D,et al. Human-level control through deep reinforcement learning[J]. Nature, 2015, 518(7540):529-533.
[12]胡华,张强,胡海洋,等.基于Q-learning的移动群智感知任务分配算法[J].计算机集成制造系统,2018,24(07):1774-1783.

相似文献/References:

备注/Memo

备注/Memo:
收稿日期:2020-07-06
作者简介:
谷湘煜(1972),男,工程师,硕士,从事电子信息技术行研发及研发管理工作。
更新日期/Last Update: 1900-01-01