Deep Learning Based Model-Free Robust Load Restoration to Enhance Bulk System Resilience With Wind Power Penetration

This paper proposes a new deep learning (DL) based model-free robust method for bulk system on-line load restoration with high penetration of wind power. Inspired by the iterative calculation two-stage model, neural network (DNN) and convolutional (CNN) are respectively designed to find worst-case condition pickup decision evaluate corresponding security. In order optimal result within limited number checks, checklist generation (LPCG) algorithm is developed ensure optimality. Then, fast strategy acquisition achieved on one-line (OSG) algorithm. The proposed finds in way, holds robustness handle uncertainties, provides real-time computation. It can completely replace conventional optimization supports which better satisfies changeable process. effectiveness validated using IEEE 30-bus 118-bus system, showing computational efficiency considerable accuracy.