Multi-objective vehicle loading and routing problem for fresh fruit and vegetable transportation

Abstract

Multi-compartment refrigerated vehicles are recently utilized in the cold chain industry, due largely to their flexibility in storage capacities with different temperature settings. To better comprehend the costs and benefits of this vehicle type in fresh fruits and vegetable transportation, a multi-compartment vehicle routing and loading problem (MCVRLP) with three different objectives – namely (i) minimizing total transportation cost, (ii) minimizing CO2 emissions, and (iii) minimizing weight loss of fresh fruits and vegetable – is herein explored and solved by mathematical formulation and genetic-based evolutionary algorithm approaches. Based on our computational results, we find that large and complicated MCVRLP instances are less likely to be solved to optimality by CPLEX solver within a reasonable computational time period, due to its complexity. However, the proposed genetic-based evolutionary algorithm seems to work well under all MCVRLP settings, as it could provide solutions that match the optimal solutions to small MCVRLP instances and those that outperform CPLEX solutions in larger ones. We also find that, with the same input information, slight differences in loading and routing may lead to solutions with totally different quality; and, interestingly, solutions with fewer vehicles might be worse off in terms of cost under the same routing, due to different temperature settings, which are results from loading decisions.