No abstract

No abstract

At container terminals, many cargo handling processes are interconnected and occur in parallel. Within short time windows, many operational decisions need to be made and should consider both time efficiency and equipment utilization. During operation, many sources of disturbance and, thus, uncertainty exist. For these reasons, perfectly coordinated processes can potentially unravel. This study analyzes simulation-based optimization, an approach that considers uncertainty by means of simulation while optimizing a given objective. The developed procedure simultaneously scales the amount of utilized equipment and adjusts the selection and tuning of operational policies. Thus, the benefits of a simulation study and an integrated optimization framework are combined in a new way. Four meta-heuristics—Tree-structured Parzen Estimator, Bayesian Optimization, Simulated Annealing, and Random Search—guide the simulation-based optimization process. Thus, this study aims to determine a favorable configuration of equipment quantity and operational policies for container terminals using a small number of experiments and, simultaneously, to empirically compare the chosen meta-heuristics including the reproducibility of the optimization runs. The results show that simulation-based optimization is suitable for identifying the amount of required equipment and well-performing policies. Among the presented scenarios, no clear ranking between meta-heuristics regarding the solution quality exists. The approximated optima suggest that pooling yard trucks and a yard block assignment that is close to the quay crane are preferable.

In this work, the potentials of data-driven optimization for the well-known berth allocation problem are studied. The aim of robust berth scheduling is to derive conflict-free vessel assignments at the quay of a terminal, taking into account uncertainty regarding the actual vessel arrival times which may result from external influences as, e.g., cross wind and sea current. In order to achieve robustness, four different Machine Learning methods-from linear regression to an artificial neural network-are employed for vessel arrival time prediction in this work. The different Machine Learning methods are analysed and evaluated with respect to their forecast quality. The calculation and use of so-called dynamic time buffers (DTBs), which are derived from the different AIS-based forecasts and whose length depends on the estimated forecast reliability, in the berth scheduling model enhance the robustness of the resulting schedules considerably, as is shown in an extensive numerical study. Furthermore, the results show that also rather simple Machine Learning approaches are able to reach high forecast accuracy. The optimization model does not only lead to more robust solutions, but also to less actual waiting times for the vessels and hence to an enhanced service quality, as can be shown by studying the resulting schedules for real vessel data. Moreover, it turns out that the accuracy of the resulting berthing schedules, measured as the deviation of planned and actually realisable schedules, exceeds the accuracy of all forecasts which underlines the usefulness of the DTB approach.

At container terminals, many cargo handling processes are interconnected and take place in parallel. Within short time windows, many operational decisions need to be taken considering both time and equipment efficiency. During operation, many sources for disturbance exist. These are the reason why perfectly coordinated processes are possibly unraveled. An approach that considers disturbance factors while optimizing a given objective is simulation-based optimization. This study analyses simulation-based optimization as a procedure to simultaneously scale the number of utilized equipment and to adjust the choice and tuning of operational policies. The four meta-heuristics Tree-structured Parzen Estimator, Bayesian Optimization, Simulated Annealing, and Random Search guide the simulation-based optimization process. The results show that simulation-based optimization is suitable to identify the amount of required equipment and well-performing policies. Thereby, there is no clear ranking which of the meta-heuristics finds the best approximation of the optimum. The approximated optima suggest that pooling terminal trucks as well as a yard block assignment close to the quay crane is preferable. With an increasing number of quay cranes, the number of optimal terminal trucks for each quay crane decreases as well as the range of truck utilization within one experiment.