Robust Vehicle Routing Problem in the Presence of a Proactive Attacker

TL;DR

This paper introduces a robust vehicle routing model considering proactive attacks on network arcs, aiming to optimize supply success rate and cost, with algorithms tested on various problem sizes.

Contribution

It develops a novel mathematical model and solution algorithms for vehicle routing under attack scenarios, addressing both small and large-scale problems.

Findings

Success rate is linked to route length and cost.

Shorter routes tend to decrease success rate.

Proposed algorithms effectively solve different problem scales.

Abstract

The vehicle routing problem has great importance and application in transportation and supply chain management. In this case, there are several supply requests in a transportation network. The main goal is to allocate customers to available vehicles and find the sequence of customer visits on each route. It is possible to attack the arcs of the network when two actors compete against each other with conflicting goals. In this case, the attacker may destroy the network arcs or prevent the supply of customers. The defender also intends to fulfill customer demand with the least cost and the most success rate. To this end, we define the robust vehicle routing problem in the presence of a proactive attacker. In this research, we propose a mathematical model with two objective functions to minimize total cost and maximize the total success rate in fulfilling the demand. Also, an exact algorithm for generating feasible routes has been used to solve small-scale problems. To solve large-scale problems, a simulated annealing algorithm is used. We test one numerical example problem and six medium and large size problems to evaluate the proposed algorithms' efficiency. The results demonstrate that the total success rate in supply is directly related to the amount of cost; crossing shorter routes decrease the total of success rate.