Optimal routing and scheduling in a time-dependent and directed multi-graph: a multi-agent temporal rural postman problem

Time-dependent (or temporal) properties may arise in many practical static network search problems, for example routing and scheduling of railway track inspection where availability of track for inspection depends on the rail-schedule. In the absence of side constraints this routing and scheduling problem is a multi-agent rural postman problem on a temporal directed network; having temporal attributes motivated from railway track unavailabilities. This work presents a novel Mixed Integer Linear Programming (MILP) formulation model for a Multi-agent Temporal Rural Postman Problem (MTRPP), and proposes theory for temporal graphs to support the model. A modified Benders’ decomposition algorithm is also proposed to address the non-convex unavailability constraints occurring in its sub-problems. Furthermore, a polynomial time algorithm is designed to address this sub-problem, while restraining the NP-Hard part ar X iv :2 10 1. 04 95 0v 1 [ m at h. O C ] 1 3 Ja n 20 21 to the master problem. A simulation study showing application in rail-track inspection using track-basedagents is presented, alongside comparison study showing improved results in smaller and sparser graphs compared to Time-Dependent Rural Postman Problem (TDRPP) which is developed over temporal attributes arising due to congestion in roadway networks.