Consumption modeling based on Markov chains and Bayesian networks for a demand side management design of isolated microgrids

Abstract

Summary This paper proposes a novel simulator of energy consumption patterns that allows designing demand side management (DSM) strategies without economic incentives. The simulator emulates consumers' patterns with and without installed DSM interfaces, based on both actual consumption measurements and surveys applied to the inhabitants of an existing isolated microgrid (Huatacondo, Chile) that has a particular DSM strategy without economic incentives. The simulator uses Markov chains to generate data characterizing consumption patterns without DSM and Bayesian networks for cases in which the users respond to the DSM strategy. Data obtained from the simulator are used to derive a response model of the consumers to the DSM interface, which can be included for the energy management system design. Results show that the implemented strategy can be effective and can generate savings up to 4.45% in diesel consumption for an ideal case where all the dwellings have the interface installed. Copyright © 2016 John Wiley & Sons, Ltd.

Harnessing artificial intelligence for data-driven energy predictive analytics: A systematic survey towards enhancing sustainability

“…Speaking of Bayesian Networks, they are graphs supporting the description of the possibilities of events happening based on the present state (Horný, 2014). In the document, Bayesian Networks have been utilized for user response prediction to demand side management measures (Z. , for detecting prospective variations in electricity markets (Roje et al, 2017), and taking into consideration the uncertainty in energy usage and solar PV energy generation (Sun et al, 2020). It is not hard to see that Bayesian networks are valuable in managing energy because they are capable of quantifying uncertainty, as well as the production of renewable energy might be intermittent, and user schedules can alter.…”

Asymptotic Bayesian Generalization Error in Latent Dirichlet Allocation and Stochastic Matrix Factorization

“…Third, let us study an application of the main result. Markov chains and Bayesian networks are used for many purposes, such webpage ranking [29], marketing [39,30,32], weather forecasting [11,38], operations research [15], computer security [21], control systems [26], and power systems [34]. Here, SMF can be used in inference of a Bayesian network composed of a Markov chain; this Bayesian network is one of the simplest and non-trivial ones.…”

Design and analysis of an aging‐aware energy management system for islanded grids using mixed‐integer quadratic programming

“…Integrated with RES, energy storage systems, distributed power generators, and active loads, microgrids can be utilized in both grid‐connected and standalone (islanded) modes . Although BESS are frequently introduced as part of microgrids, they stand out as one of the costliest parts.…”

“…7 Integrated with RES, energy storage systems, distributed power generators, and active loads, microgrids can be utilized in both grid-connected and standalone (islanded) modes. 8 Although BESS are frequently introduced as part of microgrids, they stand out as one of the costliest parts. 9,10 Consequently, besides addressing the power quality concerns (eg, continuous feeding of the load by staying in strict voltage and frequency limits), a microgrid central controller must also consider being affordable.…”