Integrated Optimization of Input/Output Point Assignment and Twin Stackers Scheduling in Multi-Input/Output Points Automated Storage and Retrieval System by Ant Colony Algorithm

Abstract

This paper researches the problem of operational optimization in the multiple-input/output (multi-I/O) points automated storage and retrieval system (AS/RS), where the destination addresses of the retrieval requests are unknown, twin stackers work simultaneously in the same aisle and cannot cross each other, and each stacker can carry two cargo units at the same time. The problem is formalized as an integrated optimization problem with noncrossing constraint for the I/O point assignment and the storage/retrieval (S/R) requests grouping and scheduling. The complexity of the problem is analyzed without considering the I/O point assignment and double units transport (DUT). In order to adapt to the integrated optimization characteristics of the problem, a node structure with multi-I/O points is considered, and the probability selection formula of ant colony algorithm is improved for multiple pheromones. Various numerical experiments show that the average deviation of the objective function value between the proposed method and CPLEX is no more than 0.4%, and the maximum deviation is only 1.22%. Although the phased algorithm avoids the waiting of stacker in operation, the average value of its objective function is 12.14% higher than that of the proposed method, and the maximum deviation is more than 30%.