Abstract: To deal with the problems of cargo loading/unloading sequence constraint and impending collision between rail guided vehicles (RGVs) in automated storage and retrieval system (AS/RS), the concepts of the conflicts of cargo loading/unloading sequence and collision are introduced. And a mathematical programming model considering the characteristics of eliminating the conflicts is established to minimize the total complete time. CPLEX is employed to optimally solve the small-scale problems, and a modified harmony search(MHS) algorithm is constructed for medium to large scale problems. In the algorithm design, the harmony memory is repeatedly divided and regrouped to accelerate the searching process, and four variable neighborhood search strategies and a mutation operation are integrated to overcome the deficiencies of the basic harmony search algorithm, such as limited search depth and tendencies to trap into local optimum. Simulation experiments of the MHS algorithm are compared with other evolution algorithms, and the simulation results show that the algorithm has high solution quality and good convergence performance, and can improve the running speed and the practicality of the AS/RS.

Key words: automated storage and retrieval system (AS/RS); collision avoidance scheduling; modified harmony search (MHS) algorithm; loading/unloading sequence constraint