SIA OpenIR  > 海洋机器人卓越创新中心
水下机器人操作脑电控制技术研究
Alternative TitleResearch on Electroencephalogram(EEG) Control Technology for Underwater Vehicle Operations
张进1,2
Department海洋机器人卓越创新中心
Thesis Advisor李伟 ; 俞建成
ClassificationTP242
Keyword水下机械手 脑-机接口 事件相关电位 脑电信号分类 组合分类器
Call NumberTP242/Z32/2017
Pages90页
Degree Discipline模式识别与智能系统
Degree Name博士
2017-11-30
Degree Grantor中国科学院沈阳自动化研究所
Place of Conferral沈阳
Abstract针对任务复杂的水下机器人作业中操作人员由于双手被束缚无法同时手动操作其它设备问题,首次尝试了将脑-机接口技术引入水下机器人作业中,通过解析脑电信号并将其映射为具体指令从而控制机械手完成水下作业,为操作人员提供一种独立于手动操作之外的控制方式来控制机械手,从而使其能够同时完成多项任务。然而,将脑电控技术应用到水下机器人作业中面临着一些问题:现阶段脑电控技术还不太成熟,直接使用真实的水下机器人进行实验具有风险;现阶段脑电控制机械手所执行的都是简单、不连贯的动作,实时性、准确性方面仍有不足;现阶段脑电控制是整个系统中唯一控制外设的手段,将脑电控制技术应用到水下作业中时手臂操作带来的干扰会使识别操作人员意图的准确率降低。为此,本文在总结和分析国内外现有研究工作的基础上,针对上述问题,对水下机器人脑电控技术进行了研究。将脑电控技术应用到水下机器人作业时,水下机械手采用脑电控制方式,为了实现对机械手的有效控制,建立了机械手的运动学和动力学模型。在前期方法验证阶段,为了减少风险和方便实验,在Webots环境下建立了机械手和载体的虚拟模型。为了使操作人员熟悉水下作业过程,为脑电控制水下机器人做好前期准备,搭建了水下机器人作业半物理平台,操作人员可以针对水下基本任务使用主手在该平台上进行操作训练。训练的过程中采用相对位置增量式控制方式并考虑了来自水下环境的外部扰动和视觉影响。为了说明平台的可行性和有效性,设计了两个典型的水下操作任务:抓取海洋生物样本和到达一个给定位置。文中给出了两个比较结果:遥控操作虚拟机械手和真实机械手的表现比较;3位操作人员使用虚拟平台训练前后的表现比较。为了使操作人员熟悉ROV作业时水下机械手的不同控制方式,以快速适应脑电控制水下机械手作业,本文搭建了ROV半物理作业平台。操作人员使用水下机械手不同控制方式在平台上执行任务,包括两种基于开关的操作、一种基于主从的操作和结合遥控操作和自主操作的操作。为了说明平台的可行性和有效性,3位操作人员被邀请参与实验,并给出了他们的作业结果。为了解放水下机械手操作人员的双手,提出了基于视觉诱发模式的ERP(事件相关电位)脑电信号来控制水下机械手的策略。通过融合脑电控制与水下机械手作业的各自特点和优化ERP视觉诱发界面,使操作人员能够快速地完成给定任务。8位被试者被邀请在建立的实验平台上进行控制实验,最终得到的辨识操作人员意图平均准确率、系统信息传输率与完成任务平均控制时间分别为91.5%、27.7bits/min与90.1s。与同类系统相比,所提控制策略系统性能更好,且作业效率满足实际作业要求。当手臂操作与脑电控制被同时应用到水下机器人作业中且操作人员处于不同作业状态时,针对使用单一脑电信号分类器无法获得较为理想的控制意图识别准确率问题,提出了使用组合分类器选取分类结果和根据实际作业情况的特殊性修正分类结果的方法来提升识别准确率。该方法首先使用Fisher判别方法分别对无手臂操作与存在手臂操作产生的数据进行训练得到两种作业状态下的分类器,其次将两分类器进行组合并使用曲线拟合的方式确定用来判定分类结果的基准距离差值(该差值的选取考虑了个体差异),然后根据实际作业情况的特殊性使用距离修正函数对距离差值进行修正,最后通过比较基准距离差值与修正后距离差值的大小来确定最终分类结果。实验结果显示,在设计的在线实验中,该方法相对于其它三种方法在识别准确率上分别提升了13.42%、5.55%和5.55%,说明该方法是可行且有效的。
Other AbstractTo overcome the limitations of cannot manipulating additional equipment simultaneously for the operator when he/she is doing a complicated underwater operational tasks while both hands of the operator have to be occupied by operating the manipulator, this thesis works at controlling the manipulator by the instructions transformed from the brain signals received by electroencephalograph (EEG), so that the operator is able to control the underwater manipulator without need for hands. This study is the first to combine the BCI (brain-computer interface) technology with underwater operational tasks, which provides the operator an effective way of doing multiple tasks simultaneously. However, applying the EEG control technology to the operation of underwater vehicle faces some problems: It is risk of directly using the real underwater vehicle to conduct experiments because the EEG control technology is immature at present; The manipulator actions performed by EEG control are simple and inconsistent and the real-time and accuracy are still insufficient at present; The EEG control is the only way to control peripheral equipment in the whole system at present, while the interference of arm operation will reduce the accuracy of identifying the operator's intentions when applying the EEG control technology to underwater operations. Therefore, the EEG control technology of underwater vehicle is studied in view of the above problems after summarizing and analyzing the existing research work at home and abroad. The underwater manipulator is controlled by EEG signals when applying the EEG control technology to the operation of underwater vehicle. In order to control the manipulator effectively, the kinematic and dynamic models of the manipulator are established. The virtual models of the manipulator and the carrier are established in Webots environment for reducing the risk and making the experiment convenient at the early stage of the method validation. A semi physical platform about underwater vehicle operations is established for making operators familiar with the underwater operation process and making a preparation for controlling underwater vehicles by EEG signals. The operator can use the master arm to perform the training process about the underwater basic task on the platform. The relative position incremental control method is adopted in the training process, and the external disturbance and visual influence from the underwater environment are taken into account. In order to demonstrate the feasibility and effectiveness of the virtual platform, two typical underwater operational tasks are designed: grasping a marine organism sample, and reaching at a given position. This thesis presents the comparative studies: 1. The performances demonstrated by remotely controlling the virtual manipulator and the real manipulator; 2. The operating performances delivered by three operators before and after training when using the platform. In order to make the operator familiar with the different control methods of the underwater manipulator under the ROV operation, a semi physical platform about ROV operations is established. The platform enables operators to quickly adapt to the EEG control of underwater manipulator operation. The operator uses different control methods of the underwater manipulator to conduct tasks on the platform, including two switch-based and one master arm-based operation, and combined by remote and autonomous operations. In order to demonstrate the feasibility and effectiveness of the platform, three operators were invited for the training operations, and their operating results are presented. To free two hands of the underwater manipulator operator, a control strategy is proposed for controlling the underwater manipulator via event-related potentials (ERP) of brainwaves (evoked by visual stimuli). The operator can quickly complete a given task after optimizing the visual evoked ERP interface and combining the characteristics of the EEG control and the underwater manipulator operation. Eight subjects were invited to conduct experiments on a self-developed experimental platform. The average accuracy of identifying an operator’s intention, the system information transfer rate and the average control time of completing the task are 91.5%, 27.7bits/min and 90.1s, respectively. Comparing with similar systems, system performance of the proposed control strategy is better, and the operational efficiency satisfies the practical operational requirements. It is difficult to use a single EEG classifier to achieve an ideal recognition accuracy of control intention when the operator is in different operating states using both arm operation and EEG control in the underwater vehicle operation. An algorithm is proposed in this paper to improve the recognition accuracy by selecting the classification result by using the combination classifier and correcting the classification result according to the specific situation of the actual operation. Firstly, the Fisher discriminant method is used to train the data generated by the armless operation and the arm operation to get the classifier in two operation states. Secondly, the two classifiers are combined and the curve fitting is used to determine the reference distance difference which is used to determine the classification result (the selection of the difference takes into account individual differences). Thirdly, the distance difference is corrected by the distance correction function according to the particularity of the actual operation situation. Finally, the final classification result is determined by comparing the difference between the reference distance difference and the corrected distance difference. The experimental results show that the proposed algorithm improves the recognition accuracy by 13.42%, 5.55% and 5.55% respectively compared with the other three methods in the designed online experiment, which demonstrates that the algorithm is feasible and effective.
Language中文
Contribution Rank1
Document Type学位论文
Identifierhttp://ir.sia.cn/handle/173321/21275
Collection海洋机器人卓越创新中心
Affiliation1.中国科学院沈阳自动化研究所
2.中国科学院大学
Recommended Citation
GB/T 7714
张进. 水下机器人操作脑电控制技术研究[D]. 沈阳. 中国科学院沈阳自动化研究所,2017.
Files in This Item:
File Name/Size DocType Version Access License
水下机器人操作脑电控制技术研究.pdf(4243KB)学位论文 开放获取CC BY-NC-SAApplication Full Text
Related Services
Recommend this item
Bookmark
Usage statistics
Export to Endnote
Google Scholar
Similar articles in Google Scholar
[张进]'s Articles
Baidu academic
Similar articles in Baidu academic
[张进]'s Articles
Bing Scholar
Similar articles in Bing Scholar
[张进]'s Articles
Terms of Use
No data!
Social Bookmark/Share
All comments (0)
No comment.
 

Items in the repository are protected by copyright, with all rights reserved, unless otherwise indicated.