TY - GEN
T1 - Data-driven power generation design and operation under demand uncertainty
AU - Alkatheri, Mohammed
AU - Rizwan, Muhammad
AU - Alhameli, Falah
AU - ElKamel, Ali
AU - Almansoori, Ali
AU - Douglas, Peter
N1 - Publisher Copyright:
© 2019, IEOM Society International.
PY - 2019
Y1 - 2019
N2 - A Data-driven stochastic optimization framework that leverages big data in design and operation of power generation units is proposed. A k-means clustering algorithm is adopted to generate uncertainty scenarios for the stochastic optimization framework. In order to do this, the power generating design and operation problem is formulated as a two-stage stochastic programming model. The first stage variables are associated with design decisions, whereas the second stage variables are associated with unit commitment operation (i.e. scheduling). The historical demand data was first collected and reprocessed. After that, the processed electrical demand (uncertain parameter) is processed and recognized using unsupervised machine learning. K-means clustering algorithm is used to produce electrical demand scenario profiles. These scenarios are used as inputs to the stochastic model. The proposed model is formulated as a mixed integer linear programming (MILP) and solved using GAMS. The stochastic data driven method enjoys the following features: it is based on information derived from real data without explicitly knowing the data distribution and it applies the recent advances of data analysis tools (e.g. machine learning) to generate a reduced size data set (i.e. clusters) integrated into mathematical model (i.e. design and planning model) that leads to a computationally tractable problem.
AB - A Data-driven stochastic optimization framework that leverages big data in design and operation of power generation units is proposed. A k-means clustering algorithm is adopted to generate uncertainty scenarios for the stochastic optimization framework. In order to do this, the power generating design and operation problem is formulated as a two-stage stochastic programming model. The first stage variables are associated with design decisions, whereas the second stage variables are associated with unit commitment operation (i.e. scheduling). The historical demand data was first collected and reprocessed. After that, the processed electrical demand (uncertain parameter) is processed and recognized using unsupervised machine learning. K-means clustering algorithm is used to produce electrical demand scenario profiles. These scenarios are used as inputs to the stochastic model. The proposed model is formulated as a mixed integer linear programming (MILP) and solved using GAMS. The stochastic data driven method enjoys the following features: it is based on information derived from real data without explicitly knowing the data distribution and it applies the recent advances of data analysis tools (e.g. machine learning) to generate a reduced size data set (i.e. clusters) integrated into mathematical model (i.e. design and planning model) that leads to a computationally tractable problem.
UR - http://www.scopus.com/inward/record.url?scp=85079291218&partnerID=8YFLogxK
M3 - Conference contribution
AN - SCOPUS:85079291218
SN - 9781532359507
T3 - Proceedings of the International Conference on Industrial Engineering and Operations Management
SP - 345
EP - 355
BT - 4th North American IEOM Conference. IEOM 2019
T2 - 4th North American IEOM Conference. IEOM 2019
Y2 - 23 October 2019 through 25 October 2019
ER -