Robust Multi-Objective Load Dispatch in Microgrid Involving Unstable Renewable Generation

Ran Wang,Tingli Xu,Hu Xu,Guanyu Gao,Yang Zhang,Kun Zhu
DOI: https://doi.org/10.1016/j.ijepes.2023.108991
IF: 5.659
2023-01-01
International Journal of Electrical Power & Energy Systems
Abstract:The microgrid improve productivity and performance by rational load balancing and intelligent energy management schemes. This paper addresses a robust multi-objective load dispatch problem to tactically schedule various resources for an microgrid equipped with controllable distributed generators and unstable renewable energy (RE) sources. Power generation cost and emissions are considered as the key indicators for the load dispatching. In such smart grid system, the indeterminate renewable energy generation (REG) imposes intractable dispatching challenges. To properly cope with the uncertain RE, a model with uncertain distribution is developed to depict the fluctuation in a flexible yet detailed manner. Chance constraint (CC) approximations and robust optimization (RO) methods are involved to change the prime issue to a more solvable form. We further explore a framework based on deep reinforcement learning (DRL) to deal with multi-objective optimization which is generally computational intractable. Real-world data-based simulation has verified the superiority of the proposed uncertain model in handling unstable REG. Furthermore, the DRL based framework is time-efficient in solving the multi-objective load dispatch problem and surpasses the state-of-art methods by having 56.74% and 15.62% better efficiencies on average.
What problem does this paper attempt to address?