Scheduling of twin automatic stacking cranes and automated guided vehicle considering buffer mode on automated container terminal

Abstract

To improve the efficiency of automatic stacking crane (ASC) and automated guided vehicle (AGV) scheduling in synchronized import and export mode at automated container terminal (ACT), with the objective of minimizing the maximum ASC completion time and AGV wait time, a twin ASCs and AGV scheduling model is established, which considers both the size of handshake area and buffer, and the location constraints of handshake area. Longest transport time priority (LTTP) and traversal priority (TP) heuristics strategies are designed to adjust the execution order of ASCs to reduce the wait time of ASCs. Based on the comprehensive learning particle swarm optimization algorithm (CLPSO), the discrete coding CLPSO algorithm (CLDPSO) is developed by designed permutation coding for ASC execution order and integer coding for AGV numbering, and the convergence speed is improved by dynamically adjusting the weight parameters. The experimental results show that increasing the handshake area and buffer size can reduce the fitness value by 3.64 % and 5.58 % on average respectively, the selected optimal handshake area location can reduce the fitness value by 3.51 % on average compared with the conventional location, the TP strategy can reduce the waiting time of ASC by up to 16.89 %. Compared to the PSO and DPSO algorithms, the fitness value of CLDPSO can reduce the fitness value by 13.83 % and 7.80 %, and the computation time is reduced by 10.67 % and 7.43 % on average, respectively.