Interval optimization to schedule multi-energy systems considering PV power generation uncertainty

This paper applies interval optimization to schedule a multi-carrier system (MES) while taking into account the uncertainty of photovoltaic (PV) energy generation. The PV energy generation in this study is represented by a predicted interval obtained from a non-parametric distribution, while the quantile function is estimated to predict the distribution. The MES is modeled as mixed-integer linear programming (MILP) and the day-ahead schedule problem is solved by a linear interval optimization. The proposed method is applied on the data set of an industrial site. The simulation results show that the proposed forecasting method is reliable to obtain predicted interval for PV system. Moreover, the sensitivity analysis on the interval optimization solution is done to determine the behavior of the system under different PV uncertainty levels to obtain minimum operational cost.