Issues of patronage : the Denver Civic Center sculptures

Copyright held by the author. DU faculty and students may access digitized MRPs by logging in with their DU ID in the Library's Catalog. Those not affiliated with DU may request digitized or microfilmed MRPs through interlibrary loan. Hard copies of MRPs are available to use by appointment in the Art Office in the Shwayder Art Building on campus.

Short-term Load Forecasting Using Neural Network For Future Smart Grid Application

Short-term load forecasting of power system has been a classic problem for a long time. Not merely it has been researched extensively and intensively, but also a variety of forecasting methods has been raised. This thesis outlines some aspects and functions of smart meter. It also presents different policies and current statuses as well as future projects and objectives of SG development in several countries. Then the thesis compares main aspects about latest products of smart meter from different companies. Lastly, three types of prediction models are established in MATLAB to emulate the functions of smart grid in the short-term load forecasting, and then their results are compared and analyzed in terms of accuracy. For this thesis, more variables such as dew point temperature are used in the Neural Network model to achieve more accuracy for better short-term load forecasting results.