Conference article

Modeling and Portfolio Optimization of Stochastic Discrete-Event System through Markovian Approximation: an Open-Pit Mine Study

Roberto G. Ribeiro
Graduate Program in Electrical Engineering, Federal University of Minas Gerais - UFMG, Brazil

Rodney R. Saldanha
Dep. of Electrical Engineering, Federal University of Minas Gerais - UFMG, Brazil

Carlos A. Maia
Dep. of Electrical Engineering, Federal University of Minas Gerais - UFMG, Brazil

Download articlehttp://dx.doi.org/10.3384/ecp17142214

Published in: Proceedings of The 9th EUROSIM Congress on Modelling and Simulation, EUROSIM 2016, The 57th SIMS Conference on Simulation and Modelling SIMS 2016

Linköping Electronic Conference Proceedings 142:31, p. 214-220

Show more +

Published: 2018-12-19

ISBN: 978-91-7685-399-3

ISSN: 1650-3686 (print), 1650-3740 (online)

Abstract

Operation in an open-pit mining is a complex task with stochastic nature. Usually, this kind of system is analyzed by means of DES (Discrete Event System) simulation. This work considers optimizing the investment in new projects in such a way to reach maximum production of an open-pit mine. When a DES model is associated with an optimization problem, the time taken to run such model is a crucial aspect. In order to analyze the project impacts in a reasonable time, this work presents a DES markovian model which represents a load-haulage cycle. The results obtained were compared with the results acquired from validated simulation models which represent the same system. In the optimization context, the complexity is exponential. Therefore, this work proposes a formulation that considers the inter-relationship between projects, which aims to help decision makers. Instead of trying all the possible projects combinations, the proposed method searches for identifying the set of projects that produce good feasible solutions based on performance measure from designed DES model.

Keywords

Markov chain, closed queuing network, project portfolio, open-pit mining

References

No references available

Citations in Crossref