Scholarlink

<p>In this paper, we propose and analyze several Variable Neighborhood Search (VNS) algorithms with different levels of cooperation for solving a novel nonlinear optimization problem known as the Team Generalized Orienteering Problem (TGOP). The TGOP is an extension of the Generalized Orienteering Problemwhere several touristic itineraries must be created that maximize the weighted sum of the scores of the nodes visited by the tourist. Even though they are designed for an extension (the TGOP), the proposed VNS algorithms are first evaluated on a Team Orienteering Problem (TOP) dataset. They proved to be competitive with the state of the art (including 20 different metaheuristics designed specifically for the TOP) in this particular case. The performance of our algorithms is then analyzed in detail on several realistic datasets specifically created for the TGOP, including a total of 750 different instances. This computational experience shows that the algorithms are able to create good solutions for a wide variety of problems and provides interesting insights about the trade-off between score and computational time for the different levels of cooperation. Furthermore, the evaluation of non-cooperative approaches provides insight into the pros and cons of cooperation in this context, which could be used to design other metaheuristics for the TGOP or related problems.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

A variable neighborhood search with different levels of cooperation for the team generalized orienteering problem

  • 期刊论文

  • 发表日期：

    2026年3月16日

查看全文
  • Adolfo Urrutia-Zambrana,
  • Gregorio Tirado,
  • Alfonso Mateos

摘要

In this paper, we propose and analyze several Variable Neighborhood Search (VNS) algorithms with different levels of cooperation for solving a novel nonlinear optimization problem known as the Team Generalized Orienteering Problem (TGOP). The TGOP is an extension of the Generalized Orienteering Problemwhere several touristic itineraries must be created that maximize the weighted sum of the scores of the nodes visited by the tourist. Even though they are designed for an extension (the TGOP), the proposed VNS algorithms are first evaluated on a Team Orienteering Problem (TOP) dataset. They proved to be competitive with the state of the art (including 20 different metaheuristics designed specifically for the TOP) in this particular case. The performance of our algorithms is then analyzed in detail on several realistic datasets specifically created for the TGOP, including a total of 750 different instances. This computational experience shows that the algorithms are able to create good solutions for a wide variety of problems and provides interesting insights about the trade-off between score and computational time for the different levels of cooperation. Furthermore, the evaluation of non-cooperative approaches provides insight into the pros and cons of cooperation in this context, which could be used to design other metaheuristics for the TGOP or related problems.