Departments View all

Total Impact Points
Total Impact Points
Total Impact Points

Recent Publications View all

  • [Show abstract] [Hide abstract]
    ABSTRACT: Clustering methods are often used to model energy consumption for two reasons. First, clustering is often used to process data and to improve the predictive accuracy of subsequent energy models. Second, stable clusters that are reproducible with respect to non-essential changes can be used to group, target, and interpret observed subjects. However, it is well known that clustering methods are highly sensitive to the choice of algorithms and variables. This can lead to misleading assessments of predictive accuracy and mis-interpretation of clusters in policymaking. This paper therefore introduces two methods to the modeling of energy consumption in buildings: clusterwise regression, also known as latent class regression, which integrates clustering and regression simultaneously; and cluster validation methods to measure stability. Using a large dataset of multifamily buildings in New York City, clusterwise regression is compared to common two-stage algorithms that use K-means and model-based clustering with linear regression. Predictive accuracy is evaluated using 20-fold cross validation, and the stability of the perturbed clusters is measured using the Jaccard coefficient. These results show that there seems to be an inherent tradeoff between prediction accuracy and cluster stability. This paper concludes by discussing which clustering methods may be appropriate for different analytical purposes.
    Applied Energy 12/2015; 160:153-163. DOI:10.1016/j.apenergy.2015.08.126
  • [Show abstract] [Hide abstract]
    ABSTRACT: Producing electricity from wind is attractive because it provides a clean, low-maintenance power supply. However, wind resource is intermittent on various timescales, thus occasionally introducing large and sudden changes in power supply. A better understanding of this variability can greatly benefit power grid planning. In the following study, wind resource is characterized using metrics that highlight these intermittency issues; therefore identifying areas of high and low wind power reliability in southern Africa and Kenya at different time-scales. After developing a wind speed profile, these metrics are applied at various heights in order to assess the added benefit of raising the wind turbine hub. Furthermore, since the interconnection of wind farms can aid in reducing the overall intermittency, the value of interconnecting near-by sites is mapped using two distinct methods. Of the countries in this region, the Republic of South Africa has shown the most interest in wind power investment. For this reason, we focus parts of the study on wind reliability in the country. The study finds that, although mean Wind Power Density is high in South Africa compared to its neighboring countries, wind power resource tends to be less reliable than in other parts of southern Africa—namely central Tanzania. We also find that South Africa’s potential varies over different timescales, with higher reliability in the summer than winter, and higher reliability during the day than at night. This study is concluded by introducing two methods and measures to characterize the value of interconnection, including the use of principal component analysis to identify areas with a common signal.
    Applied Energy 08/2015; DOI:10.1016/j.apenergy.2015.08.069
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: We present a detailed model of the integrated power system coupled to hydrodynamics that allows us to study global sensitivities in the All-Electric Ship. A novel element of our formulation is the stochastic mod-eling of the coupled system to account for uncertainty in the parameters or operating conditions. This new com-putational framework is applied to a model of the DDG-51 destroyer that involves a 19 MW 15-phase induc-tion machine and an indirect field oriented controller. In particular, we simulate extreme events correspond-ing to propeller emergence and firing of pulsed power weapons.


  • Address
    77 Massachusetts Avenue, 02139, Cambridge, MA, United States
  • Head of Institution
    Rafael Reif
  • Website
Information provided on this web page is aggregated encyclopedic and bibliographical information relating to the named institution. Information provided is not approved by the institution itself. The institution’s logo (and/or other graphical identification, such as a coat of arms) is used only to identify the institution in a nominal way. Under certain jurisdictions it may be property of the institution.

6728 Members View all

View all

Top publications last week by reads

Journal of Biological Physics and Chemistry 01/2015; 15(3):121-159. DOI:10.4024/11SA15R.jbpc.15.03
576 Reads
Harvard business review 03/2007; 85(2):92-100, 156. DOI:10.1109/EMR.2009.5235483
294 Reads

Top Collaborating Institutions


This map visualizes which other institutions researchers from Massachusetts Institute of Technology have collaborated with.

Rg score distribution

See how the RG Scores of researchers from Massachusetts Institute of Technology are distributed.