Decentralized Cost-Optimized Fuzzy Control of DC Microgrids

Abstract

This paper presents a novel method for designing a decentralized fuzzy controller for DC microgrids, aiming to reduce operating costs. Therefore, a classical fuzzy control is created for each power supplying grid user based on the voltage droop control. The control of the active rectifier is then extended with respect to the electric costs as an additional input. The input membership functions of this controller are in a next step optimized with a genetic algorithm, whereas two different approaches were used: first, only the membership function of the rectifier was optimized, secondly, the energy storage system was added to the optimization. The optimization was performed in terms of minimizing the operative costs of the DC microgrid. This method and its results were in the end compared with the optimized characteristic diagrams presented in [1]. These diagrams were an extension of the voltage droop curves and the optimization was also based on the operating costs of the grid. Results achieved with the new concept are as good as in the previously presented characteristic diagrams approach. In addition optimization time is reduced significantly (up to 50 times) and the definition of membership functions is more handy.

Load Profile Cycle Recognition for Industrial DC Microgrids with Energy Storage Systems

“…For model validation, a DC microgrid test rig (Fig. 7) was set up which has already been used in [29]- [32]. The test rig contains mains supply (sup) with adaptive droop curve control which is the main supplier to the DC microgrid.…”