Current approaches to forest management must be capable of specifying where, when, and how much can be harvested from the forest to procure wood production while dealing with multiple operational and conservation considerations. To accomplish such a task, in this work, we present a multi-objective optimization model based on the “triad zoning” approach, which proposes dividing a forest under management into three zones: a) conservation, b) wood extensive production, and c) wood intensive production. An instance of functional zoning for the communal forest of San Pedro El Alto, Oaxaca, Mexico was built, which was validated by a panel of experts. This instance was solved through six heuristic techniques which are: local search, local beam search, multi-start, and three hybrids composed by simulated annealing with each one of the heuristic techniques mentioned above. The results obtained by each metaheuristic were compared among themselves and with the current forest management plan carried out by the community of San Pedro. The simulated annealing with simple local search technique (SLS) generated the best results, maximizing economic gains in the planning horizon while respecting the zoning criteria. In general, the results showed that techniques that use guided search respond efficiently to the defined problem.

当前的森林管理方法必须能够指定从森林中采伐的地点，时间和数量来采购木材，同时还要处理多种运营和保护方面的考虑。为了完成这一任务，在这项工作中，我们提出了一个基于“三重区划”方法的多目标优化模型，该模型建议将管理中的森林划分为三个区域：a）保护区，b）木材大量生产和c ）木材集约生产。专家小组对墨西哥瓦哈卡州圣佩德罗·埃尔奥托公共森林的功能区划实例进行了构建。通过六种启发式技术解决了该实例：本地搜索，本地波束搜索，多次启动，以及由上述每种启发式技术通过模拟退火组成的三个混合体。将每种超启发式方法获得的结果相互比较，并与圣佩德罗社区执行的当前森林管理计划进行了比较。使用简单的局部搜索技术（SLS）进行的模拟退火产生了最佳结果，在遵守分区标准的同时，最大化了计划范围内的经济收益。通常，结果表明，使用引导搜索的技术可以有效地响应已定义的问题。在遵守分区标准的同时，在计划范围内最大化经济收益。通常，结果表明，使用引导搜索的技术可以有效地响应已定义的问题。在遵守分区标准的同时，在计划范围内最大化经济收益。通常，结果表明，使用引导搜索的技术可以有效地响应已定义的问题。