《2016 An efficient algorithm for solving the felxible job shop scheduling problem》.pdf

AN EFFICIENT ALGORITHM FOR SOLVING THE FLEXIBLE JOB SHOP SCHEDULING PROBLEM Karin Thörnblad *,1 Ann-Brith Strömberg **,3 Michael Patriksson **,4 Torgny Almgren*,2 *) GKN Aerospace Engine Systems, Dept. of Logistics Development, 461 81 Trollhättan, Sweden 1 ) E-mail: karin.thornblad@ , Tel: +46 520 29 22 66 2 ) E-mail: torgny.almgren@ , Tel: +46 520 29 22 62 **) Chalmers University of Technology and University of Gothenburg, Dept. of Mathematical Sciences, 421 96 Göteborg, Sweden 3 ) E-mail: anstr@chalmers.se, Tel: +46 31 772 53 78 4 ) E-mail: mipat@chalmers.se, Tel: +46 31 772 35 29 ABSTRACT Purpose To investigate the efficiency of a discretization procedure utilizing a time-indexed mathe- matical optimization model for finding accurate solutions to flexible job shop scheduling problems considering objectives comprising makespan and tardiness, respectively. Design/methodology/approach A time-indexed mixed integer programming model is used to find solutions by iteratively employing time steps of decreasing length. The solutions and computation times are compared with results from a known benchmark formulation and an alternative model. Findings The proposed method finds significantly better solutions for the largest instances within the same time frame. Both the other models are better choices for some smaller instances, which is expected since the new method is designed for larger problems. Only our alter- native model is able to solve two of