Computers and Electrical Engineering 86 (2020) 106684 Contents lists available at ScienceDirect Computers and Electrical Engineering journal homepage: www.elsevier.com/locate/compeleceng Microgrid protection under weather uncertainty using joint probabilistic modeling of solar irradiance and wind speed  Murli Manohar, Ebha Koley, Subhojit Ghosh ∗ Department of Electrical Engineering, National Institute of Technology, GE Road, Raipur (C.G) 492010, India a r t i c l e i n f o Article history: Received 8 December 2018 Revised 16 April 2020 Accepted 20 April 2020 Available online xxx Keywords: Microgrid protection Distributed energy resources Solar and wind intermittency Joint probability distribution function Optimal random forest Real-time validation a b s t r a c t The present power scenario has witnessed an unprecedented rise in the penetration level of renewable distributed energy resources (DERs). The wider acceptance of photovoltaic (PV) and wind-based DERs in microgrids poses severe protection challenges due to the uncertain variation in solar irradiance level and wind speed. The sporadic variation often results in underreach and overreach of overcurrent relays. With the aim of imparting ro- bustness to the protection scheme against weather intermittency, a technique based on meteorological data-dependent joint probabilistic modeling of the stochastic variation in solar irradiance and wind speed has been integrated with a discrete wavelet transform (DWT) based feature extraction module and an optimal random forest (ORF) based clas- sifier to perform fault detection/classification and zone identification. The performance of the proposed protection scheme has been examined for diverse test cases and compared with other computational intelligence techniques using numerical and real-time simula- tions. © 2020 Elsevier Ltd. All rights reserved. 1. Introduction With the rapid adoption of renewable technology for minimizing the burden on the utility grid, the present power gener- ation scenario has witnessed a gradual transition from centralized power generation to the distributed generation. To ensure sustainable and reliable power generation with low carbon emission and increased efficiency at the distribution level, the deployment of distributed generation with the integration of distributed energy resources commonly referred to as micro- grid has gained wide acceptance in recent times [1]. Among the DERs, the penetration of PV and wind-based sources are highly promoted in the microgrid due to their sustainability and economic operation. Microgrids carry the ability to operate under dual modes i.e. grid-connected and islanded mode. The grid-connected mode enables bi-directional power flow, whereas the islanded mode facilitates the feeding of critical loads in case of faults or disturbance in the grid. Despite the various advantages associated with microgrid, the increased penetration level of DERs poses severe protection challenges due to the dependency of renewable DERs (PV and wind in particular) on the weather condition. Microgrids integrated with PV and wind-based DERs are characterized by intermittent behavior, resulting from the random variation in solar irradiance levels and wind speed [2]. The sporadic nature of PV and wind power generation  This paper is for CAEE special section SI-aires. Reviews processed and recommended for publication to the Editor-in-Chief by Guest Editor Dr. Mustapha Hatti. ∗ Corresponding author. E-mail address: sghosh.ele@nitrr.ac.in (S. Ghosh). https://doi.org/10.1016/j.compeleceng.2020.106684 0045-7906/© 2020 Elsevier Ltd. All rights reserved.