STUDY OF TWO SIDED ASSEMBLY LINE BALANCING USING GENETIC ALGORITHM

Abstract

Two-sided assembly lines are common industrial practice in the assembly of large-sized products. It provides shorter line length, reduced throughput time, lower cost of tools and fixtures, and less material handling. In a two-sided assembly line, the products wait during the cycle time at each mated-station where there are two operators working at the opposite sides of the line simultaneously performing the different tasks on the same individual product. Genetic algorithms have received an increasing attention from the researchers since it provides an alternative to traditional optimization techniques by using directed random searches to locate optimum solutions in complex landscapes. In this thesis various features of genetic algorithms proposed for balancing of two sided assembly lines are studied and discussed. Initial population is one of the key feature in the genetic algorithm. Performance of algorithm and early convergence is dependent on the initial population. The diverse initial population helps to check all areas of solution space and prevents the trapping in local optimum. The effect of direction of generation over diversity in initial population is studied. A new method is developed to generate initial population to increase the diversity. The proposed method is applied on small sized problems to measure the performance. It is evident from the results that though diversity in population largely depends upon the precedence relations of the tasks in problem, it can vary to a great extent with the use of random direction of generation.