A hierarchical deep learning approach to optimizing voltage and frequency control in networked microgrid systems

• Published in: Applied energy Vol. 377; p. 124313
• Main Authors: Khosravi, Nima, Dowlatabadi, Masrour, Sabzevari, Kiomars
• Format: Journal Article
• Language: English
• Published: Elsevier Ltd 01-01-2025
• Subjects: Distributed energy resources; Hierarchical deep-learning; Microgrids operation management; Voltage and frequency control; MPC; DG; DNN; HiDeNN; IPFC; PV; Iact; RNN; RCNN; MGs; ADMM; DNNC; AC; NMGs; P2P; Microgrids operation management; EHs; Voltage and frequency control; RL; DERs; FC; VS; ANN; CNN; HDL-RCNN; FN; FP; V/F; FS; FOMCON; PDC; UAVs; DQN-LS; RDeNN; SE; QoS; RESs; DQN; Hierarchical deep-learning; LSTM; TS-ESV; Distributed energy resources; BMS; WCE; IDPP; PI; PMU; TN; FOPID; DC; DRL
• ISSN: 0306-2619
• DOI: 10.1016/j.apenergy.2024.124313

Distributed energy sources (DERs) and microgrids (MGs) will play an important role in improving the resilience, reliability and sustainability of the grid through dedicated generation, load management and additional capacity struggling to cope with challenges. This study addresses the challenges faced by MG systems, especially in monitoring voltage-frequency operation (V/F) using the proposed two-layer operation scheme that aims to improve MG performance. A pioneering approach is to determine controller coefficients with information from the system components using hierarchical deep-learning-based recurrent convolutional neural network (HDL-RCNN)-excluded attributes have enabled these distributions themselves to determine the optimal conditions for optimal V/F control. Further, the fractional order proportional integral derivative (FOPID) approach, along with the root of the proposed technique, will serve as comparative methods to assess the performance of the HDL-CNN approach. The effectiveness of the proposed method is demonstrated through implementation and validation using the MATLAB/SIMULINK platform.
•The study focuses on optimal management of voltage and frequency operations in microgrid systems.•A dual-layered strategy aims to efficiently optimize microgrid operations.•A pioneering approach utilizes a hierarchical deep-learning-based RCNN for optimizing microgrid operations management.