Simultaneously line balancing and buffer allocation for multi model assembly line

Assembly lines are important part of many production systems. One of the most basic problems about the design of the assembly lines is the assembly line balancing problem. The assembly line balancing problem is to determine the number of work stations (workers) for a given throughput rate, and to assign the tasks to the stations taking into account the precedence relations of tasks.The most common goal in this problem is to reduce the number of work stations. Different versions of the problem have been studied in the literature. In our thesis, we consider the assembly line balancing problem for a real-life assembly line of a dishwasher manufacturing plant. In our problem, task times are considered as random values, which is the case in most of the assembly lines. Our problem distinguishes from the classical assembly line problems by recognizing that the number of inter-station buffers is limited, and how these buffers are assingned is an important factor affecting the throughput rate. Because the buffer assignment affects the blockage probability of stations due to the random task times. Thus, assembly line balancing and the buffer allocation problem should be considered simultaneously in order to find the best solution under random task times. This thesis suggests a heuristic approach that assigns tasks to stations and allocates buffers between stations simultaneously for a given target throughput rate. The suggested heuristic involves an optimisation model for assigning tasks to the stations, a search heuristic for allocating buffers, and a simulation model to evaluate the performance of candidate solutions. We showed the benefits of the suggested approach using both a real life case and a set of random problems based on a problem from the literature.