Title :
A General Variable Neighborhood Search heuristic for short term load forecasting in Smart Grids environment
Author :
Almassalkhi, Mads ; Simon, Bernd ; Gupta, Arpan
Author_Institution :
Root3 Technol. Inc., Chicago, IL, USA
Abstract :
Energy plants represent large energy consumers with a wide array of energy needs, assets (e.g. boilers, chillers, storage, on-site generation), and constraints on operations. An innovative energy management system for energy plants is presented in this paper. Through predictive optimization of plant assets, energy analytics, pricing signals, and historical and realtime data, the online energy management system supplies energy plants with salient hourly, real-time recommendations and enables “what-if” analysis to achieve improved economic efficiency. Within a systems context, the paper draws upon ideas from power systems and highlights technical issues related to plant optimization. The paper also describes actual implementations of the energy management solution at two energy plants in the US, providing economic details and an analysis of the savings achieved.
Keywords :
energy management systems; optimisation; power generation economics; thermal power stations; energy analytics; energy management system; energy plants; historical data; online energy management; plant asset optimization; power systems; predictive optimization; pricing signals; realtime data; Boilers; Cooling; Economics; Natural gas; Optimization; Turbines; General Variable Neighborhood Search; Load Forecasting; Multi-Start; OptFrame Framework; Power Engineering Computing;
Conference_Titel :
Power Systems Conference (PSC), 2014 Clemson University
Conference_Location :
Clemson, SC
DOI :
10.1109/PSC.2014.6808095