An Agent Simulation Heuristics for Snowplowing Problems

Abstract

This paper proposes a heuristics method for huge task allocation problem over snowblower tasks. It is difficult to solve the huge task allocation problem. A snowblower problem includes a task allocation problem and agents scheduling problem. Given a directed graph as city map, we should consider some arc-disjoint partitions of the graph as allocation. A graph partition problem is a fundamental problem of combinatorics, there are some effective algorithms for the problem. However, in the case that we consider the task allocation and scheduling, the graph partition problem is too difficult. Because of this fact, we employ an agent simulation for solving the snowblower problem.