A comparison of three metaheuristics for the workover rig routing problem |
| |
Authors: | Glaydston Mattos Ribeiro Gilbert Laporte Geraldo Regis Mauri |
| |
Affiliation: | 1. CIRRELT and HEC Montréal, Canada H3T 2A7;2. Federal University of Espírito Santo, São Mateus, ES 29932-540, Brazil;3. Federal University of Espírito Santo, Alegre, ES 29500-000, Brazil |
| |
Abstract: | The workover rig routing problem (WRRP) is a variant of the Vehicle Routing Problem with Time Windows (VRPTW) and arises in the operations of onshore oil fields. In this problem, a set of workover rigs located at different positions must service oil wells requesting maintenance as soon as possible. When a well requires maintenance, its production is reduced or stopped for safety reasons and some workover rig must service it within a given deadline. It is therefore important to service the wells in a timely fashion in order to minimize the production loss. Whereas for classical VRPTWs the objective is to minimize route length, in the WRRP the objective is to minimize the total lost production, equal to the sum of arrival times at the wells, multiplied by production loss rates. The WRRP generalizes the Delivery Man Problem with Time Windows by considering multiple open vehicle routes and multiple depots. This paper compares three metaheuristics for the WRRP: an iterated local search, a clustering search, and an Adaptive Large Neighborhood Search (ALNS). All approaches, in particular ALNS, have yielded good solutions for instances derived from a real-life setting. |
| |
Keywords: | Metaheuristics Iterated local search Clustering search Adaptive large neighborhood search Workover rig routing problem |
本文献已被 ScienceDirect 等数据库收录! |
|