Seasonal business operations hire workers depending on environmental conditions and market prices. For example, during the growing and harvest seasons, agricultural businesses employ multiple workers to perform activities such as tilling soil, sowing seed, spreading fertiliser, spraying pesticides, removing weeds, and threshing crops. This study proposes two mixed-integer programming (MIP) models with an effective heuristic to solve the problem of simultaneously assigning multiple multi-skilled workers to the numerous tasks that require different skill sets during single-and multiple-period operations. The multi-skilled workforce management (MSWM) problem is NP hard in the strong sense, and it seems unlikely that large-sized realistic instances could be solved efficiently by exact algorithms directly except for some instances with very sparse tasks and skill sets. Thus, this study presents a heuristic algorithm using k-Opt as a diversification strategy embedded within the Tabu search for this complex problem. To assess the solution quality of the k-Opt heuristic, we solved two sets of instances with different sizes by running the exact solver Gurobi and the proposed heuristic algorithm with a single processor as well as running Gurobi with multiple processors. This heuristic is applicable to other multitasking situations where many workers with multiple capabilities are deployed.

季节性业务运营根据环境条件和市场价格雇佣工人。例如，在生长和收获季节，农业企业雇佣多名工人从事诸如翻土、播种、施肥、喷洒农药、除草和脱粒等活动。本研究提出了两种具有有效启发式的混合整数编程 (MIP) 模型，以解决在单周期和多周期操作中同时将多个多技能工人分配给需要不同技能集的众多任务的问题。多技能劳动力管理 (MSWM) 问题在强烈意义上是 NP 难的，并且似乎不太可能通过精确算法直接有效地解决大型现实实例，除了一些具有非常稀疏的任务和技能集的实例。因此，本研究提出了一种启发式算法，使用 k-选择作为禁忌搜索中嵌入的多样化策略来解决这个复杂的问题。为了评估 k- Opt启发式算法的解决方案质量，我们通过使用单个处理器运行精确求解器 Gurobi 和建议的启发式算法以及使用多个处理器运行 Gurobi 来求解具有不同大小的两组实例。这种启发式方法适用于部署了许多具有多种功能的工人的其他多任务处理情况。