Abstract

This research tries to solve the dynamic stacking problem that utilizes a stacking environment with uncertainty and multiple time limitations, imitating the complexity of the environment's constraints in real life instead of solving the stacking problem in a stationary setting. In order to build a crane schedule with a minimum number of relocations and high efficiency while sticking to the simulation's restrictions, this research proposed a way to handle the stacking problem using a scoring-based strategy, which involved a weighted sum of criteria scores and penalties to avoid a specific situation in an unpredictable environment. To explore the weight and penalty solutions, this method employs an existing genetic algorithm approach known as Noisy Restricted Tournament Selection (N-RTS) that helps tackle the noisy optimization problem in various scenario settings. The set of weights and penalties outcomes is then used to evaluates each of the various jobs that possibly allocated to the crane, and then produce a schedule for performing the stacking and dispatching job. When a schedule is completed, the solver repeats the evaluation process, allowing it to construct the best schedule for the crane while adapting to the current state of the uncertain environment.