2010-06-30 17:05
In the energy domain, one of the most important goals is the integration of more renewable energy sources (RES, e.g., windmills, solar panels). Unfortunately, most RES depend on external factors such as wind speed and the amount of sunlight. Hence, available power from RES cannot be planned as traditional energy sources and thus, there is the need of balancing energy demand and supply. We address this requirement within the MIRACLE (Micro-Request-Based Aggregation, Forecasting and Scheduling of Energy Demand, Supply and Distribution) project with a micro-request-based approach, where acceptable flexibilities (e.g., timeshifts or variable amount) can be specified by consumers and producers along with concrete energy demand and supply, respectively. These flexibilities allow for fine-grained scheduling and thus, balancing of demand and supply. Accurate and efficient forecasts for short-term and long-term horizons of energy consumption and production as well as for requests with timeshifts are a fundamental precondition for dynamic and fine-grained scheduling of energy demand and supply.
In this survey, we give a detailed overview of forecast models for energy demand and supply. First, we reveal typical data characteristics of energy demand and supply. Second, we describe the mathematical background of time series forecasting in general. Furthermore, we review existing domain-specific techniques of forecasting energy demand, supply and prices as well as how these techniques can be integrated into a system architecture of a data management system. Third, we select representative forecast models from the main categories of existing techniques and evaluate their accuracy with regard to different time horizons. Fourth, we identify major challenges and open problems that should be addressed in order to enable accurate and efficient forecasting as a fundamental prerequisite for scheduling energy demand and supply. Finally, the scheduling of energy demand and supply will allow (1) to smoothen cost-extensive peaks, (2) to integrate more renewable energy sources, and (3) to balance energy demand and supply. [read more]
Authors:
Lars Dannecker, SAP; Matthias Boehm, Ulrike Fischer, Frank Rosenthal, TUD; Gregor Hackenbroich, SAP; Wolfgang Lehner, TUD
In this survey, we give a detailed overview of forecast models for energy demand and supply. First, we reveal typical data characteristics of energy demand and supply. Second, we describe the mathematical background of time series forecasting in general. Furthermore, we review existing domain-specific techniques of forecasting energy demand, supply and prices as well as how these techniques can be integrated into a system architecture of a data management system. Third, we select representative forecast models from the main categories of existing techniques and evaluate their accuracy with regard to different time horizons. Fourth, we identify major challenges and open problems that should be addressed in order to enable accurate and efficient forecasting as a fundamental prerequisite for scheduling energy demand and supply. Finally, the scheduling of energy demand and supply will allow (1) to smoothen cost-extensive peaks, (2) to integrate more renewable energy sources, and (3) to balance energy demand and supply. [read more]
Authors:
Lars Dannecker, SAP; Matthias Boehm, Ulrike Fischer, Frank Rosenthal, TUD; Gregor Hackenbroich, SAP; Wolfgang Lehner, TUD
