Engineering topics
Publications and source records attributed to Kalendra, Eric.
RmultiProcess software is designed to work with multiple correlated sensors. The characteristic that allows data to be filled in or the support to be changed is correlation, the interdependence between observations.
Our purpose is to develop a methodology to quantify the impact of climate change on electric loads in the United States. We perform simple linear regression, assisted by geospatial smoothing, on paired temperature and load time-series to estimate the heating- and coolinginduced sensitivity to temperature across 300 transmission zones and 16 seasonal and diurnal time periods. The estimated load sensitivities can be coupled with climate scenarios to quantify the potential impact of climate change on load, with a primary application being long-term electricity scenarios. The method allows regional and seasonal differences in climate and load response to be reflected in the electricity scenarios. While the immediate product of this analysis was designed to mesh with the spatial and temporal resolution of a specific electricity model to enable climate change scenarios and analysis with that model, we also propose that the process could be applied for other models and purposes.
RmultiProcess software is designed to work with multiple correlated sensors. The characteristic that allows data to be filled in or the support to be changed is correlation, the interdependence between observations.
This poster illustrates methods to substantially improve the understanding of renewable energy data sets and the depth and efficiency of their analysis through the application of statistical learning methods ('data mining') in the intelligent processing of these often large and messy information sources. The six examples apply methods for anomaly detection, data cleansing, and pattern mining to time-series data (measurements from metering points in buildings) and spatiotemporal data (renewable energy resource datasets).
Summary Numerous studies have linked ambient air pollution and adverse health outcomes. Many studies of this nature relate outdoor pollution levels measured at a few monitoring stations with health outcomes. Recently, computational methods have been developed to model the distribution of personal exposures, rather than ambient concentration, and then relate the exposure distribution to the health outcome. Although these methods show great promise, they are limited by the computational demands of the exposure model. We propose a method to alleviate these computational burdens with the eventual goal of implementing a national study of the health effects of air pollution exposure. Our approach is to develop a statistical emulator for the exposure model, i.e. we use Bayesian density estimation to predict the conditional exposure distribution as a function of several variables, such as temperature, human activity and physical characteristics of the pollutant. This poses a challenging statistical problem because there are many predictors of the exposure distribution and density estimation is notoriously difficult in high dimensions. To overcome this challenge, we use stochastic search variable selection to identify a subset of the variables that have more than just additive effects on the mean of the exposure distribution. We apply our method to emulate an ozone exposure model in Philadelphia.