Abstract:
The recent developments in the field of industrial automation, especially related to mass customization, have increased the demand for mixed model assembly lines which involve customized production following a particular „product mix‟, i.e., number of models of a base product are jointly processed on a single line, in an increased quantity, quality and conducive environment. Various factors relate to the optimal operating sequence of the operations such as total setup cost, smooth consumption of parts‟ usage, total utility work, etc. In order to provide corrective measures in each case, mixed model assembly lines require the services of evolutionary algorithms.
Genetic Algorithm (binary encoding/decoding, two point crossover and uniform mutation) has been used in this study to address a global problem, i.e., total utility work, by working on associated local problems of manufacturing facility such as part assembly, quality control and supporting staff activities. A methodology has been developed to test and analyze the impact of local problems on the concerned global objectives and defense industry-oriented problems were presented to test the algorithm in real world conditions. The results were critically examined and respective improvement measures were stated along with graphical interpretations.
