Local Search for a Planning Problem with Resource Constraints in Health Simulation Center

Simon Caillard (1, 2) , Laure Brisoux Devendeville , Corinne Lucet

1 UPJV - Université de Picardie Jules Verne
2 MIS - Modélisation, Information et Systèmes - UR UPJV 4290

The Health Simulation Center SimUSanté performs training sessions for several different healthcare actors. This paper presents in detail the planning problem concerning time and resources encountered by SimUSanté, proposes 0-1 linear program modelization and the combination of greedy algorithm SimU G with a local search algorithm SimU LS, to solve it. New instances stemmed from the Curriculum-Based Courses Timetabling Problem (CB-CTT) are generated, in order to test SimU G and SimU LS on representative instances. The computational results show respectively that SimU G and SimU LS schedule on average 93.55% of the activities on the tested instances. Moreover, SimU LS obtains results with an average gap of 9.97% from the known optimal solution.

Domaines: Informatique [cs]; Recherche opérationnelle [math.OC]

Simon Caillard, Laure Brisoux Devendeville, Corinne Lucet. Local Search for a Planning Problem with Resource Constraints in Health Simulation Center. 2021. ⟨hal-03162224⟩