Muhammad ABDUL JALIL, M. F. Miskon, Mohd Bazli Bahar
{"title":"HIGH ACCURACY HUMAN MOTION TRAJECTORY GENERATION FOR EXOSKELETON ROBOT USING CURVE FITTING TECHNIQUE","authors":"Muhammad ABDUL JALIL, M. F. Miskon, Mohd Bazli Bahar","doi":"10.31436/iiumej.v24i2.2296","DOIUrl":null,"url":null,"abstract":"Robotic systems often require trajectory planning algorithms that can generate natural human-like movements for tasks such as grasping and manipulation. However, conventional trajectory planning methods may not accurately capture the complex movement patterns observed in humans. In this paper, we present a trajectory planning algorithm based on polynomial curve fitting that aims to address this issue. The algorithm determines the polynomial coefficient values that accurately match the natural human trajectory profile and is evaluated using MATLAB simulations. We compare the proposed algorithm to the conventional quintic polynomial trajectory method, analysing the accuracy, precision, and via-point continuity. The result shows that the algorithm has the ability to generate a trajectory profile with accuracy of 99.8% and a precision of 0.002°. However, the result for via-point continuity shows an error on every sub-phase transition, with the lowest error of 0.0031 between the transition of sub-phases 1 and 2. The result also shows that the lowest fitting error recorded is 0.00014°. The results demonstrate that our algorithm can generate trajectory profiles with higher accuracy and naturalness, potentially improving the performance and usability of robotic systems.\nABSTRAK: Sistem robotik sering memerlukan algoritma perancangan trajektori yang dapat menghasilkan gerakan semulajadi seperti manusia bagi tugas seperti memegang dan memanipulasi objek. Walau bagaimanapun, kaedah perancangan trajektori konvensional mungkin tidak dapat merekodkan pola gerakan kompleks seperti yang dihasilkan manusia secara tepat. Kajian ini adalah berkenaan algoritma perancangan lintasan berdasarkan penyepaduan lengkung polinomial bagi menyelesaikan masalah ini. Algoritma ini menentukan nilai pekali polinomial yang sepadan dengan profil gerakan semulajadi manusia dan dinilai menggunakan simulasi MATLAB. Algoritma yang dicadangkan ini telah dibandingkan dengan kaedah perancangan lintasan polinomial kuintik konvensional, dianalisis kejituan, ketepatan, dan keberterusan titik lalu. Keputusan menunjukkan bahawa algoritma tersebut mampu menghasilkan profil lintasan dengan kejituan sebanyak 99.8% dan ketepatan sebanyak 0.002°. Walau bagaimanapun, dapatan kajian mengenai keberterusan titik lalu menunjukkan ralat pada setiap peralihan fasa-sub dengan ralat terendah sebanyak 0.0031 pada peralihan antara fasa-sub 1 dan fasa-sub 2. Dapatan kajian juga menunjukkan bahawa ralat penyepaduan terendah yang direkodkan adalah sebanyak 0.00014°. Keputusan ini menunjukkan bahawa algoritma ini mampu menghasilkan profil lintasan dengan ketepatan dan sifat semula jadi yang lebih tinggi, berpotensi meningkatkan prestasi dan kegunaan sistem robotik.","PeriodicalId":13439,"journal":{"name":"IIUM Engineering Journal","volume":"30 1","pages":""},"PeriodicalIF":0.6000,"publicationDate":"2023-07-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"IIUM Engineering Journal","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.31436/iiumej.v24i2.2296","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q3","JCRName":"ENGINEERING, MULTIDISCIPLINARY","Score":null,"Total":0}
引用次数: 0
Abstract
Robotic systems often require trajectory planning algorithms that can generate natural human-like movements for tasks such as grasping and manipulation. However, conventional trajectory planning methods may not accurately capture the complex movement patterns observed in humans. In this paper, we present a trajectory planning algorithm based on polynomial curve fitting that aims to address this issue. The algorithm determines the polynomial coefficient values that accurately match the natural human trajectory profile and is evaluated using MATLAB simulations. We compare the proposed algorithm to the conventional quintic polynomial trajectory method, analysing the accuracy, precision, and via-point continuity. The result shows that the algorithm has the ability to generate a trajectory profile with accuracy of 99.8% and a precision of 0.002°. However, the result for via-point continuity shows an error on every sub-phase transition, with the lowest error of 0.0031 between the transition of sub-phases 1 and 2. The result also shows that the lowest fitting error recorded is 0.00014°. The results demonstrate that our algorithm can generate trajectory profiles with higher accuracy and naturalness, potentially improving the performance and usability of robotic systems.
ABSTRAK: Sistem robotik sering memerlukan algoritma perancangan trajektori yang dapat menghasilkan gerakan semulajadi seperti manusia bagi tugas seperti memegang dan memanipulasi objek. Walau bagaimanapun, kaedah perancangan trajektori konvensional mungkin tidak dapat merekodkan pola gerakan kompleks seperti yang dihasilkan manusia secara tepat. Kajian ini adalah berkenaan algoritma perancangan lintasan berdasarkan penyepaduan lengkung polinomial bagi menyelesaikan masalah ini. Algoritma ini menentukan nilai pekali polinomial yang sepadan dengan profil gerakan semulajadi manusia dan dinilai menggunakan simulasi MATLAB. Algoritma yang dicadangkan ini telah dibandingkan dengan kaedah perancangan lintasan polinomial kuintik konvensional, dianalisis kejituan, ketepatan, dan keberterusan titik lalu. Keputusan menunjukkan bahawa algoritma tersebut mampu menghasilkan profil lintasan dengan kejituan sebanyak 99.8% dan ketepatan sebanyak 0.002°. Walau bagaimanapun, dapatan kajian mengenai keberterusan titik lalu menunjukkan ralat pada setiap peralihan fasa-sub dengan ralat terendah sebanyak 0.0031 pada peralihan antara fasa-sub 1 dan fasa-sub 2. Dapatan kajian juga menunjukkan bahawa ralat penyepaduan terendah yang direkodkan adalah sebanyak 0.00014°. Keputusan ini menunjukkan bahawa algoritma ini mampu menghasilkan profil lintasan dengan ketepatan dan sifat semula jadi yang lebih tinggi, berpotensi meningkatkan prestasi dan kegunaan sistem robotik.
机器人系统通常需要轨迹规划算法,这些算法可以为抓取和操纵等任务生成类似人类的自然运动。然而,传统的轨迹规划方法可能无法准确捕捉到人类观察到的复杂运动模式。本文提出了一种基于多项式曲线拟合的轨迹规划算法来解决这一问题。该算法确定了精确匹配人体自然轨迹轮廓的多项式系数值,并通过MATLAB仿真进行了评估。我们将提出的算法与传统的五次多项式轨迹方法进行了比较,分析了精度、精度和过点连续性。结果表明,该算法能够生成精度为99.8%、精度为0.002°的弹道轮廓。而通过点连续的结果在每个子相转变上都存在误差,其中子相1和子相2的转变误差最小,为0.0031。结果还表明,记录的最小拟合误差为0.00014°。结果表明,我们的算法能够以更高的精度和自然度生成轨迹轮廓,有可能提高机器人系统的性能和可用性。摘要:系统机器人服务的memerlukan算法是一种基于memerkan算法的机器人学算法。Walau bagaimanapun, kaedah perancangan, trajektori, konkonk, konpleks, perkokkan, pola, gerakan, kompleks, perkoka, dihasilkan, manusia, secara, tepat。加加尼算法与加加尼算法、加加尼算法、加加尼算法、加加尼算法、加加尼算法、加加尼算法、加加尼算法。算法ini menentukan nilai pekali多项式,杨思丹,邓根剖面,gerakan semulajadi, dandinilai, menggunakan simulasmatlab。算法yang dicadangkan ini telah dibandingkan dendenan kaedah perancancan lintasan多项式kuintik konvension, dianalkejituan, ketepatan, dankeberterusan titik lalu。Keputusan menunjukkan bahawa算法,但mampu menghasilkan profile lintasan dengan kejituan sebanyak 99.8%,但ketepatan sebanyak 0.002°。Walau bagaimanapun, dapatan kajian mengenai keberterusan titik lalu menunjukkan ralat pada setiap peralihan fasa-sub dengan ralat terendah sebanyak 0.0031 padperalihan antara fasa-sub 1和fasa-sub 2。Dapatan kajian juga menunjukkan bahawa ralat penyepaduan terendah yang direkodkan adalah sebanyak 0.00014°。可普图桑算法可普图桑算法可普图桑算法可普图桑算法可普图桑算法可普图桑算法可普图桑算法可普图桑算法可普图桑算法可普图桑算法可普图桑算法可普图桑算法可普图桑算法
期刊介绍:
The IIUM Engineering Journal, published biannually (June and December), is a peer-reviewed open-access journal of the Faculty of Engineering, International Islamic University Malaysia (IIUM). The IIUM Engineering Journal publishes original research findings as regular papers, review papers (by invitation). The Journal provides a platform for Engineers, Researchers, Academicians, and Practitioners who are highly motivated in contributing to the Engineering disciplines, and Applied Sciences. It also welcomes contributions that address solutions to the specific challenges of the developing world, and address science and technology issues from an Islamic and multidisciplinary perspective. Subject areas suitable for publication are as follows: -Chemical and Biotechnology Engineering -Civil and Environmental Engineering -Computer Science and Information Technology -Electrical, Computer, and Communications Engineering -Engineering Mathematics and Applied Science -Materials and Manufacturing Engineering -Mechanical and Aerospace Engineering -Mechatronics and Automation Engineering