An Adaptive Assembly Line Balancing Method Based On Variation Of Operational Time

In the manual-manufacturing industry, all the processes need to be allocated tovarious workstations reasonably so that each workstation in assembly line hasapproximately the same production cycle time and the manufacture capacity could bemade full use of, and this is called assembly line balancing problem(ALBP). In thepast, ALBP was often researched only considering part of the production site factors,and does not take the difficulty of measuring standard operational time in manual lineinto account, as well as the data update, which led to the conclusion may have greatedeviation from the actual production. This method is not robust.This thesis researches on assembly line balancing problem considering bothvariation of operational time and operator skill level based on previous achievements.Because of the high uncertainty of manual operation time and high cost of measuringoperational time, we need to give a feasible method to estimate the distribution ofoperational time from small amount of measured data. Then we can put forward anadaptive line balancing methods which is more robust. It could update the systemparameters and operator efficiency factor database while getting real-time data, inorder to ensure the synchronization with production site.We can find manual work time is normally subject to the Weibull distributionwith three parameters through the analysis of available data. In this basic assumption,in order to avoid the complexity of estimating Weibull distribution parameters, wecan generate an empirical formula to directly estimate the operating time distributioncurve from a certain value and is defined as the standard operating time. So that wecan change the original problem to a determine-type line balancing problem, thenconsider the constraints and different operator skill levels, select the appropriatealgorithm and solve the problem. The empirical parameters in the empirical formulaneed to be constantly renewed by the measured data, to enable the system robust.This study has contact with a cooperation project, which has the specific needsof developing an automatic assembly line balancing procedures base on the abovemethod. We have developed the program and simplified the assigning work ofindustrial engineer, and it turned to be satisfactory results.