Multi-criteria Intermodal Vehicle Routing Problem for Maintenance in Manufacturing Systems

Abstract

In this paper, a multi-objective optimization problem of delivering and picking up crews for maintenance and servicing of machines in production systems is considered. The problem is formulated as an intermodal variant of the Vehicle Routing Problem with Pickup and Delivery. Two optimization objectives of total distance traveled by all vehicles and total weighted tardiness of all servicing jobs are considered. A solution in the form of a modified Delimiters Representation paired with a disjunctive graph is proposed. The methods for checking solution feasibility and determining the objective function value are discussed and proven as theorems. Three solution methods for the considered problem are proposed: two metaheuristic methods Tabu Search and Genetic Algorithm and a greedy heuristic serving as a reference. 9765 problem instances of up to 40 servicing jobs are prepared as a benchmark and computer experiments are performed on that instances to test the effectiveness of the proposed solution methods. The results show that Tabu Search vastly outperforms the Genetic Algorithm and the greedy heuristic, providing better Pareto front approximations measured through the Hypervolume Indicator. The research also shows that combining both metaheuristics slightly improves the result. Experiments are also conducted to show the influence of problem parameters on the resulting solution.