Search

This paper considers group scheduling problem in hybrid flexible flow shop with sequence-dependent setup times to minimize makespan. Group scheduling problem consists of two levels, namely scheduling of groups and jobs within each group. In order to solve problems with this context, two new metaheuristics based on simulated annealing (SA) and genetic algorithm (GA) are developed. A design procedure is developed to specify and adjust significant parameters for SA- and GA-based metaheuristics. The proposed procedure is based on the response surface methodology and two types of objective function are considered to develop multiple-objective decision making model. For comparing metaheuristics, makespan and elapsed time to obtain it are considered as two response variables representing effectiveness and efficiency of algorithms. Based on obtained results in the aspect of makespan, GA-based metaheuristic is recommended for solving group scheduling problems in hybrid flexible flow shop in all sizes and for elapsed time SA-based metaheuristic has better results. [ABSTRACT FROM AUTHOR]

In this paper, part-time and mixed-skilled workers scheduling problem have been modeled in which workers' productivity is variable during a day. This paper considers fatigue as a significant human aspect in staff scheduling problem. Considering the effect of fatigue in staff scheduling problem is motivated by human operators who experience fatigue in carrying out tasks in various shifts. The model tries to maximize workers' satisfaction while regarding workers' availability, productivity, priority preference, seniority level, and number of workers required. To solve such an NP-hard problem, simulated annealing (SA) and variable neighborhood search (VNS) have been introduced. In order to improve the efficiency of proposed algorithms, 'time of switching NSSs' is applied as a control parameter. To evaluate validity of the algorithms, efficient lower bound has been provided using a heuristic method. Comparison experiments show that our algorithms are capable of finding near-global optimum solution. Competitiveness of proposed algorithms has been evaluated through solutions of mathematical integer programming. The comparison results indicate that performance of VNS is better than performance of SA in terms of solution quality and computational time. [ABSTRACT FROM AUTHOR]