No abstract

An efficient Binary Search-based approach used to determine the sets of human resources allowing one to reach, for an enterprise process, a given performance level is presented. It is based on an enterprise process performance methodology integrating human skill influence. In a first step, this approach is compared with algorithms such as Hybrid Taboo-Descent, Simulated Annealing and Knapsack algorithms for different performance points of view (temporal, financial, quality oriented) on limited and large complexity examples. In a second step, this approach is generalized to integrate simultaneously several viewpoints.

performance des processus industriels. Cet article s'intéresse à la recherche du personnel permettant d'assurer, lors de son affectation, les objectifs de performance recherchés. Il s'appuie sur une méthodologie d'estimation de performance des processus d'entreprise, développée en collaboration avec un partenaire industriel, incluant explicitement l'impact des compétences des individus employés. Cette méthodologie permet de proposer une démarche d'affectation rapide et performante des ressources humaines s'appuyant sur un algorithme dichotomique. Une comparaison avec des approches heuristiques classiques (recuit simulé, etc.) est réalisée sur des exemples de moyenne et grande complexité. ABSTRACT. The human factor becomes a key element of the enterprise. Into a moving environment, it ensures timeless and competitiveness. It is now necessary to model and analyse enterprise processes by taking into account human resources as well as material resources. The proposed work is dedicated to decision making for human resource assignment. Human resources are needed for the execution of process activities and generate performance. A performance estimation methodology which links human capacities and process performance has been developed with an industrial. It underlies the human resource assignment approach proposed. It has been evaluated to determine the possible assignment guaranteeing the satisfaction a performance objective of a process. A fast and performant heuristic binary search algorithm for the human resources assignment is proposed. It is compared to other heuristic algorithms (Simulated Annealing, etc.) on large and medium complexity benchmarks. MOTS-CLÉS : affectation des ressources humaines, estimation de performance, algorithme heuristique.

No abstract