TY - JOUR
T1 - Simulation of the energy efficiency auction prices via the markov chain monte carlo method
AU - López-Gonzales, Javier Linkolk
AU - Souza, Reinaldo Castro
AU - Da Silva, Felipe Leite Coelho
AU - Carbo-Bustinza, Natalí
AU - Ibacache-Pulgar, Germán
AU - Calili, Rodrigo Flora
N1 - Publisher Copyright:
© 2020 by the authors.
PY - 2020/9
Y1 - 2020/9
N2 - Over the years, electricity consumption behavior in Brazil has been analyzed due to financial and social problems. In this context, it is important to simulate energy prices of the energy efficiency auctions in the Brazilian electricitymarket. TheMarkov ChainMonte Carlo (MCMC)method generated simulations; thus, several samples were generated with different sizes. It is possible to say that the larger the sample, the better the approximation to the original data. Then, the Kernel method and the Gaussian mixture model used to estimate the density distribution of energy price, and the MCMC method were crucial in providing approximations of the original data and clearly analyzing its impact. Next, the behavior of the data in each histogram was observed with 500, 1000, 5000 and 10,000 samples, considering only one scenario. The samplewhich best approximates the original data in accordancewith the generated histograms is the 10,000th sample, which consistently follows the behavior of the data. Therefore, this paper presents an approach to generate samples of auction energy prices in the energy efficiency market, using theMCMC method through theMetropolis-Hastings algorithm. The results show that this approach can be used to generate energy price samples.
AB - Over the years, electricity consumption behavior in Brazil has been analyzed due to financial and social problems. In this context, it is important to simulate energy prices of the energy efficiency auctions in the Brazilian electricitymarket. TheMarkov ChainMonte Carlo (MCMC)method generated simulations; thus, several samples were generated with different sizes. It is possible to say that the larger the sample, the better the approximation to the original data. Then, the Kernel method and the Gaussian mixture model used to estimate the density distribution of energy price, and the MCMC method were crucial in providing approximations of the original data and clearly analyzing its impact. Next, the behavior of the data in each histogram was observed with 500, 1000, 5000 and 10,000 samples, considering only one scenario. The samplewhich best approximates the original data in accordancewith the generated histograms is the 10,000th sample, which consistently follows the behavior of the data. Therefore, this paper presents an approach to generate samples of auction energy prices in the energy efficiency market, using theMCMC method through theMetropolis-Hastings algorithm. The results show that this approach can be used to generate energy price samples.
KW - Demand side bidding
KW - Energy
KW - Energy efficiency
KW - Gaussian mixture model
KW - MCMC
UR - http://www.scopus.com/inward/record.url?scp=85090706761&partnerID=8YFLogxK
U2 - 10.3390/en13174544
DO - 10.3390/en13174544
M3 - Article
AN - SCOPUS:85090706761
SN - 1996-1073
VL - 13
JO - Energies
JF - Energies
IS - 17
M1 - en13174544
ER -