creator Lu, Dan advisor Li, Zuyi The dissertation is composed by four parts, first, load sampling for SCUC based on Principal Component Analysis (PCA) and Kernel Density Estimation (KDE); second, load forecasting based on PCA and Bayesian ridge regression; third, anomalies detection based on Machine Learning methodology; fourth the long-term planning of Battery-based Energy Storage Transportation (BEST) in power system. Mathematical models are constructed to fulfill the research of the three targets, and numerical examples are used to test the models. The first three parts are based on PCA, which reduced the load dimensions. In the first part, a robust power system Unit Commitment (UC) is the aim to fulfil the possible load. In the second part, a novel short-term nodal load forecasting is raised to give better prediction of the next day load to improve the next data UC scheduling. In the third part, anomalies are detected in the reduced power flow space based on the pattern identified in the lower dimensional space. The purpose of the fourth part is to find ways of better utilizing the existing resources from integrating the frontier technology, the mobility of more compact and higher capacity batteries. Mix-integer programming (MIP) is used in the formulation. Submitted by Erma Thomas (thomase@iit.edu) on 2017-11-02T23:12:12Z No. of bitstreams: 1 etdadmin_upload_497473.zip: 4472330 bytes, checksum: f4b31927c5ba77153f6f7ce0a0408983 (MD5) Made available in DSpace on 2017-11-02T23:12:12Z (GMT). No. of bitstreams: 1 etdadmin_upload_497473.zip: 4472330 bytes, checksum: f4b31927c5ba77153f6f7ce0a0408983 (MD5) Previous issue date: 2017-05 Ph.D. in Electrical Engineering, May 2017 2017 2017-05 http://hdl.handle.net/10560/4144 en Anomalies detection Kernel Density Estimation Load Forecasting Load Sampling Principal Component Analysis Security Constrained Unit Commitment Dissertation born digital application/pdf In Copyright http://rightsstatements.org/page/InC/1.0/ Restricted Access ECE / Electrical and Computer Engineering Illinois Institute of Technology Affiliated department