Simulation Methods for Stochastic Storage Problems: A Statistical\n Learning Perspective

2018/03/29 by Michael Ludkovski, Ludkovski, Michael, Aditya V. Maheshwari +1 · 1 citation
Computer Science · Decision Sciences · #Advanced Multi-Objective Optimization Algorithms #Simulation Techniques and Applications #Advanced Data Storage Technologies

paper · pdf · doi:10.48550/arxiv.1803.11309

Abstract

We consider solution of stochastic storage problems through regression Monte\nCarlo (RMC) methods. Taking a statistical learning perspective, we develop the\ndynamic emulation algorithm (DEA) that unifies the different existing\napproaches in a single modular template. We then investigate the two central\naspects of regression architecture and experimental design that constitute DEA.\nFor the regression piece, we discuss various non-parametric approaches, in\nparticular introducing the use of Gaussian process regression in the context of\nstochastic storage. For simulation design, we compare the performance of\ntraditional design (grid discretization), against space-filling, and several\nadaptive alternatives. The overall DEA template is illustrated with multiple\nexamples drawing from natural gas storage valuation and optimal control of\nback-up generator in a microgrid.\n

Cited by

• On the Effectiveness of Classical Regression Methods for Optimal Switching Problems

Related

• Reinforcement Learning via Parametric Cost Function Approximation for Multistage Stochastic Programming
• Stochastic Search for a Parametric Cost Function Approximation: Energy storage with rolling forecasts
• A Generalization of a Bin-Based Modification to the Stochastic\n Simulation Algorithm
• Regress-Later Monte Carlo for Optimal Inventory Control with applications in energy
• A deep learning model for gas storage optimization