Control of a Robot Arm Using Iterative Learning Algorithm with a Stopping Criterion

Prestasi lasak bagi skema kawalan robot sangat perlu untuk memastikan robot dapat bekerja dengan berkesan seperti yang dikehendaki dalam persekitaran terbatas melibatkan gangguan, perubahan parameter, ketidaktentuan dan kepelbagaian keadaan operasi. Kajian yang dibuat adalah berkaitan dengan satu skema kawalan daya aktif dan algoritma pembelajaran berlelaran (AFCAIL) yang melibatkan satu ciri pembaikan dalam bentuk penggunaan kriteria memberhenti yang sesuai dimuatkan dalam strategi kawalan. Skema tersebut digunapakai terhadap sistem pengolah robotik planar berlengan–dua yang beroperasi secara mendatar. Kriteria memberhenti yang dicadangkan adalah reka bentuk untuk memberhentikan proses pembelajaran berlelaran apabila syarat atau keadaan berkaitan dengan kejituan ketika melakukan tugas serta perolehan matriks inersia anggaran pengolah yang dikehendaki dapat dipenuhi. Dengan cara demikian, robot dikatakan dapat beroperasi dengan baik sebagaimana yang diarahkan. Keberkesanan skema juga dikaji dengan mengambil kira beberapa keadaan bebanan dan operasi. Kata kunci: Robot; kawalan daya aktif; algoritma pembelajaran berlelaran; kriterion memberhenti The robust performance of a robot control scheme is vital to ensure that the robot accomplishes its tasks desirably in a constraint environment involving disturbances, parametric changes, uncertainties and varied operating conditions. The study introduces the Active Force Control and Iterative Learning Algorithm (AFCAIL) scheme with an improved feature in the form of a suitably designed stopping criterion incorporated in the control strategy. The scheme is applied to the control of a horiziontally operated robotic two–link planar manipulator. The proposed stopping criterion is specifically designed to halt the iterative learning process when the conditions related to the accuracy of the performed tasks and the acquisition of appropriate estimated inertia matrix of the robot arm are favourably met. In this way, the robot is said to perform desirably and excellently. The effectiveness of the scheme is also investigated by considering a number different loading and operating conditions. Key words: Robot; active force control; iterative learning algorithm; stopping criterion

[1]  Mohsen Shahinpoor,et al.  Recent trends in robotics: Modeling, control and education , 1986 .

[2]  John F. Gardner,et al.  Applications of neural networks for coordinate transformations in robotics , 1993, J. Intell. Robotic Syst..

[3]  Luca Maria Gambardella,et al.  On the iterative learning control theory for robotic manipulators , 1988, IEEE J. Robotics Autom..

[4]  Spyros G. Tzafestas Engineering Systems with Intelligence , 1991 .

[5]  Suguru Arimoto,et al.  Bettering operation of Robots by learning , 1984, J. Field Robotics.

[6]  Toshiyuki Murakami,et al.  A unified approach to position and force control by fuzzy logic , 1996, IEEE Trans. Ind. Electron..

[7]  Michael J. Grimble,et al.  Iterative Learning Control for Deterministic Systems , 1992 .

[8]  Jean-Jacques E. Slotine,et al.  Robot analysis and control , 1988, Autom..

[9]  S Arimoto,et al.  Convergence, stability and robustness of learning control schemes for robot manipulators , 1986 .

[10]  Maarouf Saad,et al.  Adaptive robot control using neural networks , 1994, IEEE Trans. Ind. Electron..

[11]  John J. Craig,et al.  Hybrid Position/Force Control of Robot Manipulators , 1982 .

[12]  Musa Mailah,et al.  Intelligent active force control of a robot arm using fuzzy logic , 2000, 2000 TENCON Proceedings. Intelligent Systems and Technologies for the New Millennium (Cat. No.00CH37119).

[13]  R. J. Mitchell,et al.  Engineering Systems with Intelligence: Concepts, Tools and Applications , 1993 .

[14]  J. R Hewit,et al.  Fast dynamic decoupled control for robotics, using active force control , 1981 .

[15]  Kevin L. Moore,et al.  Adaptive gain adjustment for a learning control method for robotics , 1990, Proceedings., IEEE International Conference on Robotics and Automation.