Simulation-based optimisation using local search and neural network metamodels

This paper presents a new algorithm for enhancing the efficiency of simulation-based optimisation using local search and neural network metamodels. The local search strategy is based on steepest ascent Hill Climbing. In contrast to many other approaches that use a metamodel for simulation optimisation, this algorithm alternates between the metamodel and its underlying simulation model, rather than using them sequentially. On-line learning of the metamodel is applied to improve its accuracy in the current region of the search space. The proposed algorithm is applied to a theoretical benchmark problem as well as a real-world manufacturing optimisation problem and initial results show good performance when compared to a standard Hill Climbing strategy.

[1]  Fred W. Glover,et al.  New advances and applications for marrying simulation and optimization , 2004, Proceedings of the 2004 Winter Simulation Conference, 2004..

[2]  Henri Pierreval,et al.  An investigation on neural network capabilities as simulation metamodels , 1992 .

[3]  Fred Glover,et al.  Future of simulation optimization , 2001, Proceeding of the 2001 Winter Simulation Conference (Cat. No.01CH37304).

[4]  Z. K. Zhang,et al.  Global convergence of unconstrained and bound constrained surrogate-assisted evolutionary search in aerodynamic shape design , 2003, The 2003 Congress on Evolutionary Computation, 2003. CEC '03..

[5]  Daniel J. Fonseca,et al.  Simulation metamodeling through artificial neural networks , 2003 .

[6]  Anu Maria,et al.  Introduction to modeling and simulation , 1997, WSC '97.

[7]  László Monostori,et al.  Design and real-time reconfiguration of robust manufacturing systems by using design of experiments and artificial neural networks , 1997 .

[8]  Ken R. McNaught,et al.  A comparison of experimental designs in the development of a neural network simulation metamodel , 2004, Simul. Model. Pract. Theory.

[9]  Taho Yang,et al.  Design of manufacturing systems by a hybrid approach with neural network metamodelling and stochastic local search , 2002 .

[10]  Paul A. Fishwick,et al.  Neural Network Models In Simulation: A Comparison With Traditional Modeling Approaches , 1989, 1989 Winter Simulation Conference Proceedings.

[11]  F. Azadivar.  Simulation optimization methodologies , 1999, WSC'99. 1999 Winter Simulation Conference Proceedings. 'Simulation - A Bridge to the Future' (Cat. No.99CH37038).

[12]  László Monostori,et al.  Hybrid, AI- and simulation-supported optimisation of process chains and production plants , 2001 .

[13]  Mary Lou Padgett,et al.  Neural networks and simulation: Modeling for applications , 1992, Simul..

[14]  H. Pierreval.  Training A Neural Network By Simulation For Dispatching Problems , 1992, Proceedings of the Third International Conference on Computer Integrated Manufacturing,.

[15]  Farhad Azadivar,et al.  Simulation optimization methodologies , 1999, WSC '99.

[16]  Mary Lou Padgett.  Neural Networks Simulation: Modeling for Applications , 1992, Simul..

[17]  Ihsan Sabuncuoglu,et al.  Simulation metamodelling with neural networks: An experimental investigation , 2002 .

[18]  Robert W. Blanning,et al.  The construction and implementation of metamodels , 1975 .