Reconfigurable kinematics, dynamics and control process for industrial robots.

This work, aims at developing a highly reconfigurable control system which intelligently unifies the reconfiguration and manages the interaction of individual robotics control systems within a reconfigurable manufacturing system (RMS). For performing any reconfigurable control process, a reconfigurable plant model that represents different robotic systems was developed. Instead of modeling and creating new reconfigurable systems, new modular robots and machines, the existing systems, robots and machines were defined as reconfigurable systems, modular robots and machines according to their reconfigurable aspects. From the study of existing robotic software and reviewing the literature, the idea of grouping robots according to their kinematic similarities was conceived and the Reconfigurable PUMA-Fanuc (RPF) model was developed. A generic solution module called the Unified Kinematic Modeler and Solver (UKMS) implements the geometric approach for solving the inverse kinematic problem for the (RPF) model. A Reconfigurable PUMA-Fanuc Jacobian Matrix (RPFJM) and reduced Reconfigurable PUMA-Fanuc Singularity Matrix (RPFSM) were developed. The Reconfigurable Robot Workspace (RRW) was developed using the Filtering Boundary Points (FBP) method. For the symbolic calculation of the RPF dynamics equations, named Reconfigurable PUMA-Fanuc Dynamic Model (RPFDM), the recursive NewtonEuler algorithm was employed, using the symbolic algebra package MAPLE 10®. The simplification of the model was done using the Automatic Separation Method (ASM). The significance of the RPFDM is that it automatically generates each element of the inertia matrix A, Coriolis torques matrix B, centrifugal torques iii Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. matrix C, and gravity torques vectors G. This model was extended to include the robot actuator dynamics and the complete electro-mechanical model named RPFDM+ is presented. The Reconfigurable Control Platform (RCP) was developed using Matlab/Simulink® software. As a case study, the PUMA 560 robot was selected and the reconfigurable “PI” controller was designed as a function of the motor parameters. All eight modules (RPF, UKMS, RPGJM, RPFSM, RRW, RPFDM, RPFDM+, and RCP) can be reconfigured by changing the parameters K l ,K 2,K 2,K 4,K 5, and K 6. These parameters represent the sinus and cosines of the robot twist angles. Using the ABB, Fanuc and PUMA 560 robots, the examples were performed.