arXiv subjects

Sahin Albayrak

Publications and source records attributed to Sahin Albayrak.

22 records · Page 2Linked to original sources

arXiv2020

Kullback-Leibler Divergence-Based Distributionally Robust Unit Commitment Under Net Load Uncertainty

The deepening penetration of renewable resources into power systems entails great difficulties that have not been surmounted satisfactorily. An issue that merits special attention is the short-term planning of power systems under net load uncertainty. To this end, we work out a distributionally robust unit commitment methodology that expressly assesses the uncertainty associated with net load. The principal strength of the proposed methodology lies in its ability to represent the probabilistic nature of net load without having to set forth its probability distribution. This strength is brought about by the notion of ambiguity set, for the construction of which the Kullback-Leibler divergence is employed in this paper. We demonstrate the effectiveness of the proposed methodology on real-world data using representative studies. The sensitivity analyses performed provide quantitative answers to a broad array of what if questions on the influence of divergence tolerance and dataset size on optimal solutions.

math.OC↗

arXiv2020

Forecasting Daily Primary Three-Hour Net Load Ramps in the CAISO System

The deepening penetration of variable energy resources creates unprecedented challenges for system operators (SOs). An issue that merits special attention is the precipitous net load ramps, which require SOs to have flexible capacity at their disposal so as to maintain the supply-demand balance at all times. In the judicious procurement and deployment of flexible capacity, a tool that forecasts net load ramps may be of great assistance to SOs. To this end, we propose a methodology to forecast the magnitude and start time of daily primary three-hour net load ramps. We perform an extensive analysis so as to identify the factors that influence net load and draw on the identified factors to develop a forecasting methodology that harnesses the long short-term memory model. We demonstrate the effectiveness of the proposed methodology on the CAISO system using comparative assessments with selected benchmarks based on various evaluation metrics.

eess.SP↗

arXiv2020

Quantification of the Impact of GHG Emissions on Unit Commitment in Microgrids

The global climate change creates a dire need to mitigate greenhouse gas (GHG) emissions from thermal generation resources (TGRs). While microgrids are instrumental in enabling the deeper penetration of renewable resources, the short-term planning of microgrids needs to explicitly assess the full range of impact of GHG emissions. To this end, we propose a novel unit commitment (UC) approach, which enables the representation of GHG emissions from TGRs, the specification of GHG emission constraints, and the ex-ante evaluation of carbon tax payment with all other costs and benefits. We quantify the relative merits of the proposed approach vis-à-vis the classical UC approach via representative studies. The results indicate that the proposed UC approach yields lower costs than does the classical UC approach and achieves a greater reduction in costs as carbon tax rate increases. Further, increasing carbon tax rates can markedly disincentivize TGR generation under the proposed approach.

math.OC↗

arXiv2010

The Link Prediction Problem in Bipartite Networks

We define and study the link prediction problem in bipartite networks, specializing general link prediction algorithms to the bipartite case. In a graph, a link prediction function of two vertices denotes the similarity or proximity of the vertices. Common link prediction functions for general graphs are defined using paths of length two between two nodes. Since in a bipartite graph adjacency vertices can only be connected by paths of odd lengths, these functions do not apply to bipartite graphs. Instead, a certain class of graph kernels (spectral transformation kernels) can be generalized to bipartite graphs when the positive-semidefinite kernel constraint is relaxed. This generalization is realized by the odd component of the underlying spectral transformation. This construction leads to several new link prediction pseudokernels such as the matrix hyperbolic sine, which we examine for rating graphs, authorship graphs, folksonomies, document--feature networks and other types of bipartite networks.

cs.LG↗