부산대학교 도서관

An improved fruit fly algorithm(IFOA) is proposed to optimize the size of energy storage in micro-grid system. Firstly, the model of micro-grid considering fuel cost, electricity purchasing cost, energy storage maintenance cost and loss cost is built. Secondly, some improvements are applied in basic fruit fly optimization algorithm. The improved algorithm uses adaptive step size to fully consider both search speed and precision of solution, and records average position to modify position update strategy to avoid falling into local optimum. As for flying speed and direction, a nonlinear coefficient is proposed to improve the performance. To verify the proposed algorithm, two operation mode(isolated mode and grid-connected mode) are selected, and results show that IFOA has a better performance, comparing to FOA, PSOGSA, GSA, PSO and CHGSA.