Genetic algorithm for integrated model of berth allocation problem and quay crane scheduling with noncrossing safety and distance constraint

Abstract

Berth Allocation and Quay Crane Scheduling are the most important part of container terminal operations since berth and quay cranes are an interface of ocean-side and landside in any port container terminal operation. Their operations significantly influence the efficiency of port container terminals and need to be solved simultaneously. Based on the situation, this study focuses on an integrated model of Continuous Berth Allocation Problem and Quay Crane Scheduling Problem. A comprehensive analysis of safety distance for vessel and non-crossing constraint for quay crane is provided. There are two integrated model involved. For the first integrated model, non-crossing constraints are added wherein quay cranes cannot cross over each other since they are on the same track. The second integrated model is focused on the safety distance between vessels while berthing at the terminal and at the same time, quay crane remains not to cross each other. These two constraints were selected to ensure a realistic model based on the real situation at the port. The objective of this model is to minimise the processing time of vessels. A vessel's processing time is measured between arrival and departure including the waiting time to be berthed and servicing time. A new algorithm is developed to obtain the good solution. Genetic Algorithm is chosen as a method based on flexibility and can apply to any problems. There are three layers of algorithm that provide a wider search to the solution space for vessel list, berth list, and hold list developed in this study. The new Genetic Algorithm produced a better solution than the previous research, where the objective function decreases 5 to 12 percent. Numerical experiments were conducted and the results show that both integrated models are able to minimize the processing time of vessels and can solve problem quickly even involving a large number of vessels. Studies have found that the safety distance set as 5 percent of vessel length gives the best solution. By adding safety distance to the integrated model with non-crossing constraint, the result indicates no improvement in the model objective function due to increasing distance between vessels. The objective function increases in the range of 0.4 to 8.6 percent. However, the safety distance constraint is important for safety and realistic model based on the port’s real situation.