ISCP-Data: A Vehicle-to-grid Dataset For Commercial Center And Its Machine Learning Application

Vehicle-to-grid (V2G) is an emerging technology for grid stability and environmental protection. However, predicting the unknown data accurately in V2G optimization is difficult. And utilizing the machine learning and big data method for V2G operation becomes a feasible method. Therefore, a vehicle-to-grid dataset for commercial center and its machine learning application are conducted in this paper. Firstly, three categories of datasets, i.e., conventional load data, PV output data and PEV charging data, are selected as the trend of power grid operation. Secondly, a V2G scenario of commercial center is designed, and the maximum values of three parameters were calculated. Next, both the centralized method and the distributed method are used for data annotation. In case study, the dataset is applied through a variety of machine learning methods. The results show that the accuracy of V2G prediction is up to 94%.