Using a Neural Network instead of IKM in 2R Planar Robot to follow rectangular path
الموضوعات :
Ali Ali
1
(Mechatronics , Mechanical and Electrical Engineering , Tishreen University , Lattakia , Syria)
الکلمات المفتاحية: Direct Kinematic Model (DKM), Denavit Hartenberg (DH), Neural Network (NN), Inverse Kinematic Model (IKM), Back-Propagation(BP),
ملخص المقالة :
Abstract— An educational platform is presented here for the beginner students in the Simulation and Artificial Intelligence sciences. It provides with a start point of building and simulation of the manipulators, especially of 2R planar Robot. It also displays a method to replace the inverse kinematic model (IKM) of the Robot with a simpler one, by using a Multi-Layer Perceptron Neural Network (MLP-NN), to make the end-effector able to track a specific path, which has a rectangular shape (in this article), and allocated in the robot's workspace. The method is based on Back-Propagation Levenberg Marquardt algorithm. This paper also suggests a good strategy for the simulation of the robot's motion in Matlab to tell the beginners how the operation could be done quite closely to the built-in Matlab functions. The control part was ignored here for the simplicity. So we can classify this paper as a manual in the robotic world.
1. Dorsey, N. Top 5 Robotic Systems to Watch in Agriculture. 2018 [cited 2019 19 September ]; Available from: https://www.precisionag.com/in-field-technologies/top-5-robotic-systems-to-watch-in-agriculture/.
2. Bouchard, S. Industrial robots: What are the different types? 2014 [cited 2019 19 September]; Available from: https://blog.robotiq.com/bid/63528/what-are-the-different-types-of-industrial-robots.
3. Household robots. [cited 2019 19 September ]; Available from: http://www.allonrobots.com/household-robots.html.
4. [cited 2019 19 September]; Available from: https://www.youtube.com/watch?v=ZWZGfWU8_p0.
5. Zhao, J., et al., A learning-based multiscale modelling approach to real-time serial manipulator kinematics simulation. Neurocomputing, 2019; Available from: https://www.sciencedirect.com/science/article/abs/pii/S0925231219314456.
6. Kieffer, S., V. Morellas, and M. Donath. Neural network learning of the inverse kinematic relationships for a robot arm. in Proceedings. 1991 IEEE International Conference on Robotics and Automation. 1991. IEEE.
7. Banga, V., Y. Singh, and R. Kumar, Simulation of robotic arm using genetic algorithm & AHP. World Academy of Science, Engineering and Technology, 2007. 25(1): p. 95-101; Available from: https://pdfs.semanticscholar.org/cbef/ee29e571ec3a50637e8de3c6f6ce56d00338.pdf.
8. Hao, W.G., Y.Y. Leck, and L.C. Hun. 6-DOF PC-Based Robotic Arm (PC-ROBOARM) with efficient trajectory planning and speed control. in 2011 4th International Conference on Mechatronics (ICOM). 2011. IEEE.
9. Spong, M.W., S. Hutchinson, and M. Vidyasagar, Robot modeling and control. 2020: John Wiley & Sons.
10. Beale, H.D., H.B. Demuth, and M. Hagan, Neural network design. Pws, Boston, 1996; Available from: https://pdfs.semanticscholar.org/6a81/22861b80842ee6f406acdeec35aec913f1b8.pdf.
