LAUSR

Significant penetration of renewable generations (RGs) and mass roll-out of plug-in electric vehicles (PEVs) will pay a vital role in delivering the low carbon energy future and low emissions of greenhouse gas (GHG) that are responsible for the global climate change. However, it is of considerable difficulties to precisely forecast the undispatchable and intermittent wind and solar power generations. The uncoordinated charging of PEVs imposes further challenges on the unit commitment in modern grid operations. In this paper, all these factors are comprehensively investigated for the first time within a novel hybrid unit commitment framework, namely UCsRP, which considers a wide range of scenarios in renewable generations and demand side management of dispatchable PEVs load. UCsRP is however an extremely challenging optimisation problem not only due to the large scale, mixed integer and nonlinearity, but also due to the double uncertainties relating to the renewable generations and PEV charging and discharging. In this paper, a meta-heuristic solving tool is introduced for solving the UCsRP problem. A key to improve the reliability of the unit commitment is to generate a range of scenarios based on multiple distributions of renewable generations under different prediction errors and extreme predicted value conditions. This is achieved by introducing a novel multi-zone sampling method. A comprehensive study considering four different cases of unit commitment problems with various weather and season scenarios using real power system data are conducted and solved, and smart management of charging and discharging of PEVs are incorporated into the problem. Test results confirm the efficacy of the proposed framework and new solving tool for UCsRP problem. The economic effects of various scenarios are comprehensively evaluated and compared based on the average economic cost index, and several important findings are revealed.