Simulationbased GA Optimization for Production Planning Bioma 2014
Simulation-based GA Optimization for Production Planning Bioma 2014 September 13, 2014 Juan Esteban Díaz Leiva Dr Julia Handl
Production Planning Production levels Production Plan Allocation of resources Business objectives 2
Production Planning Experience & “Sixth sense” Lack of appropriate instrument Inappropriate methods 3
Simulation-based Optimization Simulation Optimization GA DES Aplicable solution 4
Objective Support decision making Production Planning Feasibility Simulation-based optimization Uncertainty & Applicablility Real-life complexity Robustness 5
Simulation-based Optimization Model 6
Simulation-based Optimization Model 7
Simulation-based Optimization Model • 8
Simulation-based Optimization Model • 9
Simulation-based Optimization Model • 10
Simulation-based Optimization Model Ø GA (MI-LXPM) [2] • • • real coded Laplace crossover power mutation tournament selection truncation procedure for integer restrictions parameter free penalty approach [1] K. Deb. An efficient constraint handling method for genetic algorithms. Computer methods in applied mechanics and engineering, 186(2): 311 -338, 2000. [2] K. Deep, K. P. Singh, M. Kansal, and C. Mohan. A real coded genetic algorithm for solving integer and mixed integer optimization problems. Applied Mathematics and Computation, 212(2): 505 -518, 2009. 11
Results Original model Figure 4. Best, mean and worst fitness value of the population at each iteration. 12
Results Model modifications 13
Results Model modifications 14
Results Profit maximization 15 Figure 7. Best, mean and worst fitness value of the population at each iteration (time: 8. 17 h).
Results ILP CDF deterministic Stochastic Simulation-based optimization CDF uncertainty 16
Results Profit maximization 17 Figure 8. CDFs of profit obtained through stochastic simulation.
Conclusions Ø Production plan • production levels and allocation of work centres Ø Process uncertainty • delays Ø Real life complexity • no complete analytic formulation Ø Better performance of solutions • stochastic simulation 18
Post-doc Position Constrained optimization (applied in the area of protein structure prediction) Start date: November 2014 in collaboration between: Computer Sciences (Joshua Knowles), Faculty of Life Sciences (Simon Lovell) and MBS (Julia Handl). Info: j. handl@manchester. ac. uk 19
Q & A 20
Thank you September 13, 2014 Juan Esteban Diaz Leiva Dr Julia Handl 21
- Slides: 21