An MILP-Based Planning Model of a Photovoltaic/Diesel/Battery Stand-Alone Microgrid Considering the Reliability.

Xiong Wu,Wencheng Zhao,Xiuli Wang,Haoyu Li
DOI: https://doi.org/10.1109/tsg.2021.3084935
IF: 10.275
2021-01-01
IEEE Transactions on Smart Grid
Abstract:For a stand-alone microgrid, the reliability criterion is often taken into account to guarantee the supply adequacy. This paper proposes a mixed integer linear programming (MILP) based model to determine the optimal sizing of a photovoltaic/diesel/battery stand-alone microgrid considering the reliability. Firstly, an economic-oriented microgrid planning model is developed. Subsequently, a rolling horizon based Sequential Monte Carlo method is developed to evaluate the system reliability. In particular, a multidimensional piecewise linearization method is proposed to approximate the nonlinear and implicit reliability evaluation function. Therefore the reliability criterion can be safely integrated into the microgrid planning model as an MILP problem which can be solved easily. The advantage of this method is that the costly reliability evaluation is calculated offline. Consequently, the time consumption of the online implement of the model becomes much less. The conventional techniques, however, integrate the reliability evaluation into the procedure. Thus a lot of iterations are needed to search for a solution that satisfies the reliability criterion. Finally, numerical experiments based on one microgrid test system are performed. The results indicate that larger sized DERs are deployed if a higher reliability criterion is required. Additionally, the proposed method is much faster than conventional evolutionary algorithms.
What problem does this paper attempt to address?