Multi-neighborhood simulated annealing for personalized user project planning

• Published in: Applied soft computing Vol. 119; p. 108566
• Main Authors: Li, Yinuo, Hao, Jin-Kao
• Format: Journal Article
• Language: English
• Published: Elsevier B.V 01.04.2022; Elsevier
• Subjects: Computer Science; Heuristics; Operations Research; Project planning; Resource assignment; Simulated annealing; Resource assignment; Project planning; Heuristics; Simulated annealing
• ISSN: 1568-4946; 1872-9681
• DOI: 10.1016/j.asoc.2022.108566

Effective decision support systems are very useful management tools in many applied domains. However, such systems are still scarce or even missing in social and medico-social establishments. This study investigates the personalized user project planning problem in French social and medico-social establishments, whose purpose is to optimize the assignment of multi-featured activities and resources to a group of residents or users subject to complex imperative constraints. We focus on the design and implementation of an innovative multi-neighborhood local optimization algorithm that serves as the key component of a decision support system for these establishments. We assess the effectiveness of the proposed approach on realistic data and show comparisons with other approaches including mathematical programming and greedy search. •We consider a complex planning problem in social and medico-social sector.•We present a multi-neighborhood simulated annealing algorithm for solving the problem.•We show computational results on realistic test instances.•The algorithm serves as the basis for a practical decision-making system.