Alternatif altgrafik montaj hattı dengeleme problemi için bir çözüm yaklaşımı

Assembly line balancing problem (ALBP) is the process of assigning a set of tasks to a group of workstations, considering the precedence relations between the assembly tasks. Precedence relationships between tasks are shown with help of a graph diagram. In more complex versions of this problem, tasks may have alternative precedence relationships. This situation has led to the emergence of the Alternative Subgraph Assembly Line Balancing Problems (ASALBP). In this thesis, we worked on the Alternative Subgraph Assembly Line Balancing Problem of type 1 (ASALBP-1). ASALBP-1 aims to assign the tasks to a minimum number of workstations under precedence relations and cycle time constraints after the alternative subgraph is selected for a line balancing problem with different mounting alternatives. In this study, ASALBP-1 is solved with firefly, bat and proposed hybrid firefly-bat metaheuristic algorithms. At the beginning of proposed hybrid firefly-bat algorithm, the alternative subgraph selections of the problem are made, and then the initial solutions are generated by using the ranked positional weight method and randomly. After the initial solutions are formed, the firefly algorithm is used to generate new solutions initially. Later, the hybrid algorithm is terminated by adding the operators of the bat algorithm to the solution. Benchmark problems are solved with firefly and bat algorithms in addition to the hybrid algorithm. These three algorithms are used to solve twelve problem sets with different sizes, cycle times and subgraphs. Performance evaluations are made by comparing the hybrid algorithm against some heuristics taken from the literature for firefly, bat, and proposed hybrid algorithms. Comparative conclusions of algorithms give that the proposed hybrid firefly - bat algorithm is capable of producing promising results for ASALBP-1.