Development of a software for simulating active force control schemes of a two–link planar manipulator

Abstract

The paper describes the development of a software in the form of an interactive computer program that integrates a number of robotic control schemes with the active force control (AFC) strategy as the key element of the robotic system that assumes a rigid two–link planar configuration. The various AFC schemes are employed in conjunction with a number of conventional and intelligent techniques embedded in the main control loop to approximate the estimated inertia matrix of the robot arm. The schemes have been individually developed and rigorously experimented through simulation studies. The results of these studies clearly indicate that the AFC technique provides a practical solution to enhance the robustness of the robotic system even in the wake of uncertainties, disturbances and varied loading conditions. Thus, it is deemed useful to develop software that can integrate a number of individual AFC schemes into a single program using a graphic user interface (GUI) technique. In this manner, the user can effectively select and execute any scheme by the manipulation of a few keystrokes or buttons of the input devices. This resulted in a program that is user friendly, readily accessible, flexible and proved very convenient. On top of that, the graphical results can be observed and analysed on–line while the program is running. By using MATLAB and its GUI facility, all the AFC schemes already described in the previous works such as the AFC with crude approximation method, AFC and Iterative Learning (AFCAIL), AFC and Neural Network (AFCANN), AFC and Fuzzy Logic (AFCAFL), and AFC and Genetic Algorithm (AFCAGA) schemes were linked into a single menu–driven program where each of the scheme can be easily selected and executed by the user. A classic proportional–derivative (PD) control scheme was also included in the program for the purpose of benchmarking.