Winner determination problem in multiple automated guided vehicle considering cost and flexibility

This study solves the integration difficulty between scheduling and routing aspects of the multiple Automated Guided Vehicle (AGV) problem through the winner determination problem (WDP). Our model reduces the cost of using multiple AGV for logistics and warehousing applications. We show that solving the WDP by the mixed integer linear programming (MILP) is inefficient as the assignment of routes is made complicated by the combination of large number of AGVs. We illustrate an efficient approach through our proposed genetic algorithm using knowledge based operators, which decomposes a non-linear combinatorial auction model into a linear model. Simulation results showed the efficacy of MILP, the conventional GA and our proposed GA-based method. However, MILP only works well for small scale data. When the number of routes and AGV increases, the proposed GA method supersedes the other methods as indicated by cost and route flexibilities.