Title :
Robot kinematics computations using GMDH learning strategy
Author :
Chen, C.L.P. ; McAulay, A.D.
Author_Institution :
Dept. of Comput. Sci. & Eng., Wright State Univ., Dayton, OH
Abstract :
Summary form only given, as follows. The group method of data handling (GMDH), a data analysis technique for identification of nonlinear complex systems, is a feature-based mapping neural network. It is also an example of a polynomial neural network (PNN). The PNN can be successfully trained to interpolate an unknown function by observing a few samples. The PNN is used to interpolate robot forward and inverse kinematics computations (FKC and IKC). The FKC and IKC are, respectively, designed to find the mapping from the joint space to the Cartesian space, and the mapping from the Cartesian space to the joint space. A PNN simulation software package has been developed for solving both the FKC and the IKC. The simulation was performed in a two-degree-of-freedom manipulator. The solutions of the built FKC and IKC networks were compared with the analytic equations. The PNN successfully learns the indicated path
Keywords :
control engineering computing; digital simulation; identification; kinematics; learning systems; neural nets; robots; Cartesian space; GMDH learning strategy; control engineering computing; data analysis; feature-based mapping neural network; forward kinematics computation; group method of data handling; identification; inverse kinematics computations; joint space; mapping; nonlinear complex systems; polynomial neural network; robot; Application software; Computational modeling; Computer networks; Computer science; Data analysis; Data handling; Neural networks; Orbital robotics; Polynomials; Robot kinematics;
Conference_Titel :
Neural Networks, 1991., IJCNN-91-Seattle International Joint Conference on
Conference_Location :
Seattle, WA
Print_ISBN :
0-7803-0164-1
DOI :
10.1109/IJCNN.1991.155673