A hybrid genetic algorithm for scheduling of the panel block assembly shop in shipbuilding

This paper describes a scheduling problem of the panel block assembly shop in a shipbuilding industry. Because the shipbuilding is a labor intensive industry, the most important consideration in a panel block assembly shop is the workload balancing, which balances man-hour, weight and welding length and so on. We need to determine assembly schedule and workstation considering a daily load balancing and a workstation load balancing simultaneously. To solve the problem, we develop a hybrid genetic algorithm. Hybrid genetic algorithm proposed in this paper consists of two phases. The first phase uses the heuristic method to find a initial feasible solution which provides a useful information about optimal solution. The second phase proposes the genetic algorithm to derive the optimal solution with the initial population consisting of feasible solutions based on the initial solution. Finally, we carried out computational experiments for this load balancing problem, which indicate that developed method is effective for finding good solutions.