WEKO3
アイテム
人工ポテンシャル場と特徴抽出法に基づく屋内環境における移動ロボットの経路計画
https://doi.org/10.18997/00004488
https://doi.org/10.18997/00004488d789e1c3-b6b5-43bb-92c9-a1a6681bdd4f
| 名前 / ファイル | ライセンス | アクション |
|---|---|---|
sei_k_267.pdf (5.2 MB)
|
|
| アイテムタイプ | 学位論文 = Thesis or Dissertation(1) | |||||||
|---|---|---|---|---|---|---|---|---|
| 公開日 | 2016-08-03 | |||||||
| 資源タイプ | ||||||||
| 資源タイプ識別子 | http://purl.org/coar/resource_type/c_db06 | |||||||
| 資源タイプ | doctoral thesis | |||||||
| タイトル | ||||||||
| タイトル | Artificial Potential Field and Feature Extraction Method for Mobile Robot Path Planning in Structured Environments | |||||||
| 言語 | en | |||||||
| タイトル | ||||||||
| タイトル | 人工ポテンシャル場と特徴抽出法に基づく屋内環境における移動ロボットの経路計画 | |||||||
| 言語 | ja | |||||||
| 言語 | ||||||||
| 言語 | eng | |||||||
| 著者 |
W.M.M., Tharindu Weerakoon
× W.M.M., Tharindu Weerakoon
|
|||||||
| 抄録 | ||||||||
| 内容記述タイプ | Abstract | |||||||
| 内容記述 | Mobile robots are widely used in many applications in industrial fields as well as in academic and research fields. The robot path planning problem is a key problem in making truly autonomous robots. It is one of the most important aspects in mobile robot research and plays a major role in their applications but is a complex problem. The role of path planning of mobile robot can be described as finding a collision free path in a working environmentenriched with obstacles from a specified starting point to a desired destination position called the goal. Additional characteristic of path planning in known environments is to satisfy some certain optimization criteria. Most of the traditional path planning approaches such as visibility graphs, cell decomposition, voronoi diagram, etc. are designed and functioning well in static known environments. However the real environments are consisting of both the stationary and moving obstacles. Artificial potential field based methods can be applicable for both the static and dynamic environments as well as for the known or unknownenvironments. Because of the analytical complexity of the dynamic environments, researchon path planning in dynamic environments is limited but there are hundreds of research workhave been reported on path planning in static known environments.Because of the mathematical simplicity, easy implementation and real time applicability ofartificial potential field, it has become popular in robot path planning. However, the potential field based path planning shows some inherit shortcomings such as dead-lock. Recently, in the field mobile robotics, some different techniques have been proposed to overcome the dead-lock issue associated with the artificial potential field based path planning. Most of these research work targeted only a specified situation where the dead-lock can happen.In this research, we proposed a method for avoiding robot from dead-lock caused in differentsituations of mobile robot path planning using artificial potential field. In the proposedmethod we have introduced a new repulsive force component which is depended on therobot’s heading direction. The proposed method is evaluated for different conditions whichcreate dead-lock for traditional artificial potential field method. The simulations of theproposed approach have indicated that it has a capability of avoiding dead-locking associatedwith the traditional method, and is simpler and easier to implement. However in realimplementation it is required to extract the geometric features of the environment such aswalls and corners for our consideration in structured environments. Consequently, we havediscussed a segmentation and feature extraction adaptive algorithm for structuredenvironments. In this study, several adaptive techniques proposed in literature forsegmentation of laser range data have been implemented and tested in different environmentsto compare the performances of them with the proposed technique. The experimental resultshave shown that the proposed method is superior to other adaptive techniques. Furtherdiscussion is continued to analyze the implementation issues of the artificial potential field approach in geometrical structured environments. The segmented features of the walls are used to generate the potential force for robot navigation. These segmented features arematched with the pre-observed features to extend or merge them together to generate a mapof the environment and this map is used in potential force generation process. Combining thesegmentation and representation of geometrical obstacles for artificial potential fieldgeneration in robot path planning, simulation experiments were done and performances arecompared for the traditional and the proposed approach.Based on the simulation results from various case studies, we have concluded that theproposed artificial potential field method for mobile robot path planning is able to solve the dead-lock problems that are with traditional method. The segmentation and feature extraction algorithm proposed in this thesis has shown better performances than the existing methods by experimental results. Geometrical representation of the structured environment is used to implement the artificial potential field based path planning on the robot and implementation barriers are discussed. | |||||||
| 目次 | ||||||||
| 内容記述タイプ | TableOfContents | |||||||
| 内容記述 | 1: Introduction||2: Path Planning Background and Literature Review||3: Proposed Artificial Potential Filed Based Algorithm||4: Feature Extraction and Landmark Detection||5: Representation of Geometric Environment and Implementation of P-APF||6: Conclusions and Future Work | |||||||
| 備考 | ||||||||
| 内容記述タイプ | Other | |||||||
| 内容記述 | 九州工業大学博士学位論文 学位記番号:生工博甲第267号 学位授与年月日:平成28年3月25日 | |||||||
| キーワード | ||||||||
| 主題Scheme | Other | |||||||
| 主題 | Artificial potential field | |||||||
| キーワード | ||||||||
| 主題Scheme | Other | |||||||
| 主題 | Local minima | |||||||
| キーワード | ||||||||
| 主題Scheme | Other | |||||||
| 主題 | Deadlock | |||||||
| キーワード | ||||||||
| 主題Scheme | Other | |||||||
| 主題 | Path planning | |||||||
| キーワード | ||||||||
| 主題Scheme | Other | |||||||
| 主題 | Segmentation | |||||||
| キーワード | ||||||||
| 主題Scheme | Other | |||||||
| 主題 | Feature extraction | |||||||
| アドバイザー | ||||||||
| 石井, 和男 | ||||||||
| 学位授与番号 | ||||||||
| 学位授与番号 | 甲第267号 | |||||||
| 学位名 | ||||||||
| 学位名 | 博士（工学） | |||||||
| 学位授与年月日 | ||||||||
| 学位授与年月日 | 2016-03-25 | |||||||
| 学位授与機関 | ||||||||
| 学位授与機関識別子Scheme | kakenhi | |||||||
| 学位授与機関識別子 | 17104 | |||||||
| 学位授与機関名 | 九州工業大学 | |||||||
| 学位授与年度 | ||||||||
| 内容記述タイプ | Other | |||||||
| 内容記述 | 平成27年度 | |||||||
| 出版タイプ | ||||||||
| 出版タイプ | VoR | |||||||
| 出版タイプResource | http://purl.org/coar/version/c_970fb48d4fbd8a85 | |||||||
| アクセス権 | ||||||||
| アクセス権 | open access | |||||||
| アクセス権URI | http://purl.org/coar/access_right/c_abf2 | |||||||
| ID登録 | ||||||||
| ID登録 | 10.18997/00004488 | |||||||
| ID登録タイプ | JaLC | |||||||
