Case data for active distribution network: Optimal dispatching with uncertainties based on affine arithmetic.

<b>Affine Arithmetic (AA) </b><b>is</b><b> </b><b>an</b><b> effective interval analysis method to deal with the uncertainties of the power systems. However, </b><b>pre</b><b>vious research on AA-based optimal problems is challenging to capture the uncertainties of EV and the cumulative effect of energy storage systems. In addition, the reformulated AA model increases massive variables and constraints, resulting in a heavy computational burden. In this paper, an effective AA-based economic dispatch (AAED) method for active distribution network with EVs and ESSs is proposed considering uncertainties. Specifically, an EV Charging load interval (CLI) model is developed to efficiently depict the plug-in/plug-out time and initial/target energy randomness. Confidence level is defined to avoid the </b><b>ov</b><b>erly conservativeness of the CLI model. An</b><b> </b><b>ESS model is formulated in the AA domain to represent the cumulative effect caused by multiple-period </b><b>uncertainties</b><b>, ensuring the accuracy of the state of charge</b><b> bound. Furthermore, a fast-solving strategy is designed to improve the computation efficiency of the AA</b><b>ED</b><b> model without loss of accuracy. Massive state variables and corresponding constraints are eliminated by the derived </b><b>analytical partial deviation formulation, representing the mapping relationship between the state variables and the uncertain resources. The simulation results demonstrate the effectiveness of the proposed model and method.</b>