Virtual engineering: optimal cell layout method for improving productivity for industrial robot

Abstract

This paper presents a method to optimize the positions of robot tasks in a robotic work cell in terms of minimum cycle time. In order to improve the efficiency of the overall algorithm, the method is proposed to be decomposed into 3 stages. Firstly, a good position for each task is derived by putting such robot task to a preferred region; secondly, based on the results from the first stage, a switched method is designed to seek the best space sorting of the tasks; with the benefit from these two stages which can reduce the search space greatly in optimization, in the third stage, the positions of all robot tasks are adjusted simultaneously by means of Simulated AnnealingMethod. Several test cases verified the effectiveness of the proposed method