工程科学学报，第39卷.第8期：1261-1267,2017年8月 Chinese Journal of Engineering,Vol.39,No.8:1261-1267,August 2017 DOI:10.13374/j.issn2095-9389.2017.08.017;http://journals.ustb.edu.cn 基于半自主导航与运动想象的多旋翼飞行器二维空 间目标搜索 史添玮)，王宏2)，崔文华)，任玲) 1)辽宁科技大学国际金融与银行学院，鞍山1140512)东北大学机械工程与自动化学院，沈阳110819 3)辽宁聚龙金融设备股份有限公司，鞍山114051 ☒通信作者，E-mail:tianweiabbec@163.com 摘要提出一种脑-机接口系统实现多旋冀飞行器室内二维空间目标搜索.系统由半自主导航与决策子系统组成.半自主 导航子系统用于为决策子系统提供可行飞行方向并实现多旋翼飞行器半自主避障.决策子系统采用互相关方法与逻辑回归 方法完成运动想象的脑电特征提取与分类.实际的室内目标搜索实验验证了使用该系统是可行且有效的.相比其他方法，减 少了被试者负担，降低控制难度，控制精度约提高±10cm. 关键词脑-机接口；运动想象；半自主导航；互相关；逻辑回归 分类号R318;V249.1 Two-dimensional space target searching based on semi-autonomous navigation and motor imagery for multi-rotor aircraft SHI Tian-wei,WANG Hong),CUl Wen-hua),REN Ling) 1)School of Intemational Finance and Banking,University of Science and Technology Liaoning,Anshan 114051,China 2)School of Mechanical Engineering and Automation,Northeastern University,Shenyang 110819.China 3)Liaoning Julong Financial Equipment Co.,Ltd.,Anshan 114051,China Corresponding author,E-mail:tianweiabbcc@163.com ABSTRACT A brain computer interface (BCI)system was proposed to realize the two-dimensional indoor space target searching for multi-rotor aircraft.This system consists of semi-autonomous navigation and decision subsystems.The semi-autonomous navigation subsystem is employed to provide feasible directions for the decision subsystem and avoid obstacles semi-automatically for multi-rotor aircraft.The decision subsystem utilizes the cross-correlation (CC)and logistic regression (LR)methods to implement motor imagery (MI)electroencephalogram(EEG)feature extraction and classification,respectively.The actual indoor target searching experiment validates the feasibility and effectiveness of this BCI system.Compared to similar methods,the proposed BCI system reduces the bur- den of the subjects and the control difficulties.The control precision increases by approximately +10cm. KEY WORDS brain computer interface;motor imagery;semi-autonomous navigation;cross-correlation;logistic regression 近十几年来，多旋翼飞行器在军事与民用领域的有预先规划的前提下立即识别障碍物，且无法在路口 应用愈加广泛.尽管自主性在不断增强，但是，多旋翼处做出瞬时决策.通常，人使用双手操控飞行器，这导 飞行器仍然存在诸多应用瓶颈，例如：在目标搜索中，致其无法同时完成更多的控制任务.此外，由于实际 基于视觉与惯性导航等]全自主导航系统无法在没 应用中存在无法预测的复杂性，有时需要人为控制的 收稿日期：2016-09-06 基金项目：国家自然科学基金资助项目(51405073)：辽宁省高校创新团队资助项目(LT2014006):辽宁省教育厅资助项目(2016HZZD05)工程科学学报,第 39 卷,第 8 期:1261鄄鄄1267,2017 年 8 月 Chinese Journal of Engineering, Vol. 39, No. 8: 1261鄄鄄1267, August 2017 DOI: 10. 13374 / j. issn2095鄄鄄9389. 2017. 08. 017; http: / / journals. ustb. edu. cn 基于半自主导航与运动想象的多旋翼飞行器二维空 间目标搜索 史添玮1) 苣 , 王 宏2) , 崔文华1) , 任 玲3) 1) 辽宁科技大学国际金融与银行学院, 鞍山 114051 2) 东北大学机械工程与自动化学院, 沈阳 110819 3) 辽宁聚龙金融设备股份有限公司, 鞍山 114051 苣 通信作者, E鄄mail: tianweiabbcc@ 163. com 摘 要 提出一种脑鄄鄄机接口系统实现多旋翼飞行器室内二维空间目标搜索. 系统由半自主导航与决策子系统组成. 半自主 导航子系统用于为决策子系统提供可行飞行方向并实现多旋翼飞行器半自主避障. 决策子系统采用互相关方法与逻辑回归 方法完成运动想象的脑电特征提取与分类. 实际的室内目标搜索实验验证了使用该系统是可行且有效的. 相比其他方法,减 少了被试者负担,降低控制难度,控制精度约提高 依 10 cm. 关键词 脑鄄鄄机接口; 运动想象; 半自主导航; 互相关; 逻辑回归 分类号 R318; V249郾 1 Two鄄dimensional space target searching based on semi鄄autonomous navigation and motor imagery for multi鄄rotor aircraft SHI Tian鄄wei 1) 苣 , WANG Hong 2) , CUI Wen鄄hua 1) , REN Ling 3) 1) School of International Finance and Banking, University of Science and Technology Liaoning, Anshan 114051, China 2) School of Mechanical Engineering and Automation, Northeastern University, Shenyang 110819, China 3) Liaoning Julong Financial Equipment Co. , Ltd. , Anshan 114051, China 苣 Corresponding author, E鄄mail: tianweiabbcc@ 163. com ABSTRACT A brain computer interface (BCI) system was proposed to realize the two鄄dimensional indoor space target searching for multi鄄rotor aircraft. This system consists of semi鄄autonomous navigation and decision subsystems. The semi鄄autonomous navigation subsystem is employed to provide feasible directions for the decision subsystem and avoid obstacles semi鄄automatically for multi鄄rotor aircraft. The decision subsystem utilizes the cross鄄correlation (CC) and logistic regression (LR) methods to implement motor imagery (MI) electroencephalogram (EEG) feature extraction and classification, respectively. The actual indoor target searching experiment validates the feasibility and effectiveness of this BCI system. Compared to similar methods, the proposed BCI system reduces the bur鄄 den of the subjects and the control difficulties. The control precision increases by approximately 依 10 cm. KEY WORDS brain computer interface; motor imagery; semi鄄autonomous navigation; cross鄄correlation; logistic regression 收稿日期: 2016鄄鄄09鄄鄄06 基金项目: 国家自然科学基金资助项目(51405073); 辽宁省高校创新团队资助项目(LT2014006); 辽宁省教育厅资助项目(2016HZZD05) 近十几年来,多旋翼飞行器在军事与民用领域的 应用愈加广泛. 尽管自主性在不断增强,但是,多旋翼 飞行器仍然存在诸多应用瓶颈,例如:在目标搜索中, 基于视觉与惯性导航等[1鄄鄄2] 全自主导航系统无法在没 有预先规划的前提下立即识别障碍物,且无法在路口 处做出瞬时决策. 通常,人使用双手操控飞行器,这导 致其无法同时完成更多的控制任务. 此外,由于实际 应用中存在无法预测的复杂性,有时需要人为控制的