TY - CHAP U1 - Konferenzveröffentlichung A1 - da Costa Fernandes, Jesus Salazar A1 - Schmidt, Michael A1 - Feldmann, Thomas A1 - Bollin, Elmar T1 - Use of load profiles to optimize microgrids of non-residential environments T2 - 33rd EU PVSEC Conference Proceedings N2 - Three real-lab trigeneration microgrids are investigated in non-residential environments (educational, office/administrational, companies/production) with a special focus on domain-specific load characteristics. For accurate load forecasting on such a local level, à priori information on scheduled events have been combined with statistical insight from historical load data (capturing information on not explicitly-known consumer behavior). The load forecasts are then used as data input for (predictive) energy management systems that are implemented in the trigeneration microgrids. In real-world applications, these energy management systems must especially be able to carry out a number of safety and maintenance operations on components such as the battery (e.g. gassing) or CHP unit (e.g. regular test runs). Therefore, energy management systems should combine heuristics with advanced predictive optimization methods. Reducing the effort in IT infrastructure the main and safety relevant management process steps are done on site using a Smart & Local Energy Controller (SLEC) assisted by locally measured signals or operator given information as default and external inputs for any advanced optimization. Heuristic aspects for local fine adjustment of energy flows are presented. KW - Microgrids KW - Trigeneration KW - Smart Grid Operation Y1 - 2017 UR - www.eupvsec-proceedings.com SN - 2196-100X SS - 2196-100X SN - 3-936338-47-7 SB - 3-936338-47-7 U6 - https://doi.org/10.4229/EUPVSEC20172017-6BV.3.90 DO - https://doi.org/10.4229/EUPVSEC20172017-6BV.3.90 SP - 2739 EP - 2746 CY - München ER -