Resilience-Constrained Hourly Unit Commitment in Electricity Grids

Yifei Wang,Liping Huang,Mohammad Shahidehpour,Loi Lei Lai,Haoliang Yuan,Fang Yuan Xu

DOI: https://doi.org/10.1109/tpwrs.2018.2817929

IF: 7.326

2018-01-01

IEEE Transactions on Power Systems

Abstract:This paper proposes a resilience-constrained unit commitment (RCUC) framework in which the system operation constraints, heterogeneity of power flow distribution and lines forced outages are simultaneously addressed. The proportional hazard model (PHM) is adopted to shape line forced outage rates with the cumulative effect of weather conditions and line loading rates. A sequential and Monte Carlo-based framework is established for RCUC using the recursive sampling process. The RCUC solution in each period affects the previously accumulated outage probabilities, thus it would have an impact on the outage process. In turn, the sampled line outages affect the RCUC solution in the next period and the loading of remaining lines. This is a sequential process in which the RCUC and the outage sampling of power system components are computed alternatively until the scheduling horizon is culminated. During each period, two penalty terms are introduced into the RCUC model to seek a tradeoff between operation cost and the homogeneity of flow distribution in power network and the loading rates of local lines affected by extreme weather. Since the penalty terms are modeled by absolute value functions, a general linearization method for absolute value functions in the mix-integer programming is presented. The overall convergence of the Monte Carlo technique is quantified by the coefficient of variation of total costs over multiple simulations. If the convergence criterion is not satisfied, the Monte Carlo solution approach will be continued. The case studies illustrate the effectiveness of the proposed RCUC model.