Solving the team orienteering problem with time windows and mandatory visits using a constraint programming approach

Abstract

This paper presents a constraint programming (CP) approach for solving the team orienteering problem with time windows and mandatory visits (TOPTW-MV), which has many real-world implementations, such as tourist tour planning, routing technicians, and disaster relief routing. In the TOPTW-MV, a set of locations is given; some locations must be visited, while others are optional. For each location, the profit, service time, and service time window information are known. A fleet of homogeneous vehicles is available for visiting locations and collecting the profits. The objective in solving this problem is to create a set of vehicle routes that begin and end at a depot, visit mandatory locations exactly once and optional locations at most once, while observing other restrictions such as time windows and sequence-based travel times. The CP-based approach finds 99 of the best-known solutions and explores 64 new best-known solutions for the benchmark instances.