Two-Level Forecast-Based Energy and Load Management for Grid-Connected Local Systems Using General Load and Storage Models

摘要

The ongoing changes in power supply systems cause, besides all technical challenges, increasing costs for electrical energy. Households and small or medium-sized companies and farms with high energy demand are interested in decreasing their expenses by installing renewable generation devices, e.g. solar panels or wind turbines, together with energy storages. Flexible energy management systems (EMS) can optimize their operation in different kinds of systems. A two-level method is proposed that is able to determine cost-efficient operation strategies in 1-minute-steps in real-time, based on forecasts of power demand and renewable generation for systems including renewable energy generation, shiftable loads and storages. On the upper level, a combination of Simulated Annealing together with non-linear optimization is used to determine look-ahead schedules for shiftable loads and setpoints for storages' states of charge over a forecasting time horizon of 24 hours. On the lower level, on a fast timescale of 1-minute steps, controls of storages are determined based on the current system state by non-linear optimization. General load and storage models can be applied. A simulation based on measurement data shows that the approach successfully reduces the overall cost with low computation time.