Forecasting tools for micro-grid applications

点击翻译按钮获取中文摘要

General Electric (GE) Digital Energy Multilin in Canada, referred to as GE here onward, has partnered with the University of Calgary to develop a Collaborative Research and Development (CRD) program. The program will be focused on price and load forecasting technologies for GE's Grid IQ Micro-grid Control System platform. The platform, developed in Canada, is the first industrial micro-grid controller to enter the world market. The research problem in the CRD program is how to develop forecasting modules required for optimal operation of micro-grids. More specifically, the first task in this proposal is to develop and implement a price forecasting mechanism that is seamlessly adaptable to different market structures. This is to ensure that when installing the platform in different jurisdictions, minimal adjustments are necessary to generate relevant electricity market price forecasts. The second task in this research is to build deterministic and probabilistic micro-grid load forecasting models that are robust against the high volatility of micro-grid loads. The load forecasting models will be designed such that they take into account the specific characteristics of micro-grid load composition. Finally, the third task in this research is to investigate how forecast errors may impact the overall economic efficiency of micro-grids, and how the forecasting models could be adjusted to mitigate any economic deficiency associated with forecast errors. This program will benefit Canada in several ways, including (i) creating economic opportunities from an improved Canadian technology, (ii) facilitating the integration of more renewable resources by improving an industrial micro-grid control platform, and (iii) saving money and avoiding emissions by improving a platform that is an essential part of micro-grid optimization in remote, off-grid communities.