Facility location problems have been a major research area of interest in the last several decades. In particular, uncapacitated location problems (ULP) have enormous applications. Variations of ULP often appear, especially as large-scale subproblems in more complex combinatorial optimization problems. Although many researchers have studied different versions of ULP (e.g., uncapacitated facility location problem (UCFLP) and p-Median problem), most of these authors have considered small to moderately sized problems. In this paper, we address the ULP and provide a fast adaptive meta-heuristic for large-scale problems. The approach is based on critical event memory tabu search. For the diversification component of the algorithm, we have chosen a procedure based on a sequencing problem commonly used for traveling salesman-type problems. The efficacy of this approach is evaluated across a diverse range of benchmark problems sourced from the Internet, with a comprehensive comparison against four prominent algorithms in the literature. The proposed adaptive critical event tabu search (ACETS) demonstrates remarkable effectiveness for large-scale problems. The algorithm successfully solved all problems optimally within a short computing time. Notably, ACETS discovered three best new solutions for benchmark problems, specifically for Asymmetric 500A-1, Asymmetric 750A-1, and Symmetric 750B-4, underscoring its innovative and robust nature.

翻译：设施选址问题在过去数十年间一直是重要的研究领域。其中，无容量限制的设施选址问题具有广泛的应用价值。该问题的变体频繁出现，尤其作为复杂组合优化问题中的大规模子问题。尽管已有大量研究关注不同版本的此类问题（如无容量限制设施选址问题和p-中值问题），但多数研究仅针对中小规模问题。本文针对大规模无容量限制设施选址问题，提出一种快速自适应元启发式求解方法。该方法基于关键事件记忆禁忌搜索框架，在算法的多样化环节采用了常用于旅行商类问题的序列优化策略。通过从互联网获取的多样化基准测试问题集，本研究对该方法进行了系统评估，并与文献中四种主流算法进行了全面对比。实验表明，所提出的自适应关键事件禁忌搜索算法在处理大规模问题时表现出卓越性能，能在较短时间内获得所有测试问题的最优解。特别值得指出的是，该算法在Asymmetric 500A-1、Asymmetric 750A-1和Symmetric 750B-4三个基准问题上获得了当前最优的新解，充分证明了其创新性与鲁棒性。