Developing a Fuzzy Knowledge Based Optimisation System for Storage and Retrieval Operations of Long Stay Containers

Abstract

Owing to the many uncertainties involved, the management of container yard operations is very challenging. The storage of containers is one of those operations that require proper management to achieve efficient utilisation of the yard, short handling time and a minimum number of re-handlings. The aim of this study is to develop a fuzzy knowledge based optimisation system based on genetic algorithm named ‘<em>F</em>KBGA' for the management of container yard operations that take into consideration factors and constraints of long stay containers that exist in real-life situations. One of these factors is the duration of stay of a container in each stack. Because the duration of stay of containers stored with pre-existing containers varies dynamically over time, an ‘ON/OFF' strategy is proposed to activate or deactivate the duration of stay factor in the estimation of departure time if the topmost containers for each stack have been stored for a similar time period. A genetic algorithm module based Multi-Layer concept is developed which identifies the optimal fuzzy rules required for each set of fired rules to achieve a minimum number of container re-handlings when selecting a stack. An industrial case study is used to demonstrate the applicability and practicability of the developed system. The proposed system has the potential to produce more effective storage and retrieval strategies, by reducing the number of re-handlings of containers. The performance of the proposed system is assessed by comparing with other Constrained-Probabilistic Stack Allocation “CPSA" and Constrained-Neighbourhood Stack Allocation “CNSA" storage and retrieval techniques.