2015/12/22 by Thomas Browne, Bertrand Iooss, Browne, Thomas +7 · 1 citation
Engineering · #FOS: Computer and information sciences #FOS: Mathematics #Manufacturing Process and Optimization #Methodology (stat.ME) #Statistics Theory (math.ST) #Technology Assessment and Management
paper · pdf · doi:10.48550/arxiv.1512.07060
openalex publication_date 2015/12/22 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28
This paper deals with the optimization of industrial asset management\nstrategies, whose profitability is characterized by the Net Present Value (NPV)\nindicator which is assessed by a Monte Carlo simulator. The developed method\nconsists in building a metamodel of this stochastic simulator, allowing to get,\nfor a given model input, the NPV probability distribution without running the\nsimulator. The present work is concentrated on the emulation of the quantile\nfunction of the stochastic simulator by interpolating well chosen basis\nfunctions and metamodeling their coefficients (using the Gaussian process\nmetamodel). This quantile function metamodel is then used to treat a problem of\nstrategy maintenance optimization (four systems installed on different plants),\nin order to optimize an NPV quantile. Using the Gaussian process framework, an\nadaptive design method (called QFEI) is defined by extending in our case the\nwell known EGO algorithm. This allows to obtain an "optimal" solution using a\nsmall number of simulator runs.\n