Matheurística dinâmica e pro-ativa integrada a inversores híbridos para o gerenciamento robusto de recursos energéticos distribuídos

Abstract

The growing adhesion of residential consumers to renewable sources of Distributed Generation (DG), through small generation systems, combined with the perspective of Energy Storage Systems (ESS) in integration with energy distribution systems electricity by means of hybrid inverters, makes it necessary to use methods that manage these energy resources. Battery Energy Storage System (BESS) and Photovoltaic generation systems are examples of Distributed Energy Resources (DER) that enable consumers to actively interact with current electricity systems. A Smart Home Energy Management (SHEM) integrates DER for efficient and economical operation in dynamic pricing scenarios for purchase and sale energy. In this thesis, a dynamic and proactive SHEM matheuristic approach is presented to address the problem of planning and operating in real time of a residential storage system with distributed photovoltaic generation. The integration of a hybrid inverter to the SHEM provides maximum comfort to the user because it does not require a demand control system, additionally allowing the reduction of energy purchase costs with meeting the load at the time when it allows the commercialization of energy surplus. To deal with the dynamic problem related to SHEM and the uncertainties regarding the load, renewable energy and purchase and sales energy tariffs, a Mixed Integer Linear Programming model was developed that encompasses a robust optimization responsive to uncertainties (Robust-MILP, also involving technical restrictions and criteria to increase the life cycle of the storage system. The dynamic aspect gives the system a real applicability from the permanent correction of the Robust-MILP solution with a heuristic that acts proactively in the planning, allowing a quick solution of each operation step and a structure based on rules to adjust the robust optimization considering the operation in real time and the information collected from the previous steps. The case studies presented demonstrates that the mathematical heuristic was able to offer benefits to the energy planning and operation in real time of the SHEM while presenting a computational complexity compatible with its use in real systems.